Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Building networks for change: how ed-tech coaches broker information to lead instructional reform
(USC Thesis Other)
Building networks for change: how ed-tech coaches broker information to lead instructional reform
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Running head: BUILDING NETWORKS FOR CHANGE
Building Networks for Change:
How Ed-Tech Coaches Broker Information to Lead Instructional Reform
by
Ayesha K. Hashim
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(EDUCATION)
August 2017
Copyright 2017 Ayesha K. Hashim
BUILDING NETWORKS FOR CHANGE
i
DEDICATION
To everyone working to make lives better through education.
BUILDING NETWORKS FOR CHANGE
ii
ACKNOLWEDGEMENTS
It takes a village to pursue your dreams and earn the status of “expert” in a field that you
care so passionately about. This dissertation has been seven years in the making and could not
have been finished without the support of many generous people.
I would like to thank the district that I studied for my dissertation, which I refer to using
the pseudonym “Lexington Unified School District” (LUSD). The strategic planning of LUSD’s
Board of Education, superintendent, assistant superintendents, and other central office
administrators made it possible for me to examine the role of ed-tech coaches in supporting a
one-to-one computing program for students. Whereas many districts have struggled to roll out
digital devices to students, facing numerous technical challenges, LUSD supplied these devices
without a hitch and could thus focus on the complex work of improving teacher practice. By
investing in a group of innovative teachers to lead this process of instructional change, district
leaders put their faith in the professional capacity of their own teachers, providing a rich setting
for me to explore how distributed models of leadership work in practice. Thank you to LUSD’s
district leaders, ed-tech coaches, principals and teachers who joined forces to implement this
program and participated in this study.
I would also like to thank the University of Southern California (USC) for giving me the
opportunity to pursue a doctoral degree in education policy. It is because of USC’s generosity,
tremendous resources, and the Trojan network that I could find meaningful work in Los Angeles
while immigrating from Bangkok, Thailand to build a life with my husband. Through
coursework as a Master’s in Public Policy student I met my Ph.D. faculty advisor, Dr. Katharine
O. Strunk, who encouraged me to apply to the Ph.D. program at the Rossier School of Education
and became instrumental in shaping my career as an education researcher. At Rossier, I was
BUILDING NETWORKS FOR CHANGE
iii
fortunate to be paid a stipend while developing my research skills and studying issues in
education policy that I care deeply about. Thank you to Laura Romero, my Ph.D. program
director, for always being a student advocate and going out of your way to support me in my
studies. Thank you to the Rossier faculty, especially Drs. Bill Tierney and Gale Sinatra, whose
coursework during my first year as a Ph.D. student helped me focus on education technology and
professional development as key areas of interest.
I would also like to thank the classmates who pushed me to think critically in our
coursework and gave me endless encouragement during my Ph.D. studies. I would especially
like to thank Tracey Weinstein, who was my “big sister” in the program and who I still look up
to for support and guidance in my research and in making important career and life decisions.
Cheryl Ching, I could not have made it through these last three years without your guidance and
support. You are an incredible scholar, one of the most intelligent people I know, and one of my
closest and dearest friends. Thank you for patiently reading and commenting on my dissertation
writing. I would like to thank Alice Huguet, Stephani Wrabel, Susan Bush, Federick Ngo,
Matthew Duque, Laura Mulfinger, and Bradley Marianno, who have all played a part in
supporting my dissertation research whether it was talking through my research questions,
reading drafts, or providing a shoulder for me to lean on when things got tough. It would also be
very remiss of me to not acknowledge the CANA group at USC’s Department of Preventive
Medicine for helping me master the social network chops needed to complete this dissertation,
with special thanks to Timothy Hayes for showing me how to code and analyze social network
data in R.
I would also like to acknowledge the support of my dissertation committee. To my
dissertation chair and Ph.D. faculty advisor Katharine Strunk, thank you for seeing my potential
BUILDING NETWORKS FOR CHANGE
iv
as a researcher so many years ago. I always dreamed of earning a Ph.D. and becoming an
academic, but did not believe it was possible to achieve these aspirations until I met you. You
have been my role model in so many ways, from showing me how to form meaningful research
partnerships with districts, to conceiving of new and interesting research projects, to teaching
and caring for you students, to balancing work with life. As I leave Rossier, I am comforted
knowing that you will continue to be a central person in my career and a close and dear friend.
To my other committee members, Drs. Julie Marsh and Thomas Valente, thank you for
lending your expertise to me along this journey. Julie – in many ways, I see my dissertation as
representing the joint influence that you and Katharine have had on my development as a
scholar. I am thrilled that I could study a topic that was in your wheelhouse (coaching) and hope
to collaborate more with you on my research. Thank you for patiently teaching me your process
for conducting rigorous and in-depth qualitative research. I know that I still have more to learn
on this methodology, but appreciate you expanding my horizons and guiding me to fold in a
qualitative component to my dissertation. Tom – thank you for teaching me the ropes of social
network analysis and introducing me to the CANA group. I remember feeling very intimidated
about using social network analysis (and ERGMs) for my dissertation, but your openness and
encouragement kept me going.
Finally, I would like to thank my family for believing in me. To my husband, Nizar,
thank you for being my most loyal friend and supporter. For cooking me dinner, washing the
dishes late at night, walking the dog, looking after our daughter so that I could have dedicated
time to read and write, and tolerating my grumpy moods when I felt down about work. You
helped me keep mind and soul together through these long years and I am forever grateful. I love
you. To my daughter, Noor Claire, thank you for making me smile every day and reminding me
BUILDING NETWORKS FOR CHANGE
v
that there is more to life than work. I have new dreams and aspirations now that you are in my
life. To Lulu (our maltipoo), thank you for being the best writing partner a girl could ask for. I
will always cherish those long days where you sat patiently by my side as I drafted my
qualifying exam, dissertation proposal, and these final dissertation chapters. To my mom and
dad, thank you for educating your children and always encouraging us to pursue our passions in
life. I could not be where I am today without your love and support. Dad – special thanks to you
for supporting me to leave behind my miserable job in investment banking and opening the doors
for me to work at the Sakhumvit Trust in Bangalore, India, showing me the vast inequities that
exist in education and the important role of philanthropy, policy, and research in rectifying these
injustices. To the Dream School Foundation where I spent a lot of my time during these
formative years in India – Maitreyee, Sabu, and my students – thank you for helping me find my
place in the education field and inspiring me to do good for others. As I continue this work in the
future, I will carry with me memories, lessons and inspiration from everyone that has helped
along the way.
BUILDING NETWORKS FOR CHANGE
vi
TABLE OF CONTENTS
CHAPTER ONE -- Introducing Ed-Tech Coaching ............................................................ 0
A Brief Review of Policy Context Surrounding Ed-Tech Coaching ..................................... 0
Overview of the Dissertation ................................................................................................. 3
CHAPTER TWO -- Exploring Ed-Tech Coach Leadership and Brokering ................... 10
Introduction .......................................................................................................................... 10
Instructional Leadership in Systemic School Reform ......................................................... 11
Instructional Leadership of Ed-Tech Coaches ................................................................. 14
Brokering Practices Related to Instructional Leadership ..................................................... 16
Central Office Coordination ............................................................................................ 18
Top-Down Communication ............................................................................................. 20
School Coordination ........................................................................................................ 23
Bottom-Up Communication ............................................................................................. 25
Organizational Conditions Related to Brokering ................................................................. 26
Power and Influence ........................................................................................................ 27
Access to Expertise .......................................................................................................... 28
Informal Networks of Communication ............................................................................ 29
Contributions to the Extant Literature ................................................................................. 32
CHAPTER THREE -- Social Network Theory as a Lens for Brokering ......................... 36
Introduction .......................................................................................................................... 36
Review of Social Network Theory ...................................................................................... 36
Brokering Properties from Social Network Theory ......................................................... 39
Social Structure Properties from Social Network Theory ............................................... 42
Dynamics of Social Structures ......................................................................................... 45
Limitations of Social Network Theory ............................................................................ 47
CHAPTER FOUR --LUSD’s Digital Coaching Program .................................................. 49
Introduction .......................................................................................................................... 49
Theory of Action for LUSD’s Technology Rollout ............................................................. 50
Digital Coaching Program ............................................................................................... 53
CHAPTER FIVE -- Social Network Data and Methods .................................................... 60
Data Collection .................................................................................................................... 60
Administrative Data ......................................................................................................... 60
Social Network Data ........................................................................................................ 61
Descriptive Analysis ............................................................................................................ 64
Exponential Random Graph Modeling ................................................................................ 67
Tie-Sender and Receiver Parameters ............................................................................... 68
Dyadic Parameters ........................................................................................................... 70
Network Parameters ......................................................................................................... 71
Model Fit .......................................................................................................................... 72
Methodological Limitations ................................................................................................. 72
CHAPTER SIX -- Social Network Findings ........................................................................ 76
Brokering Practices of DLCs ............................................................................................... 76
Hypothesis 1..................................................................................................................... 76
Summary of Brokering Practices of DLCs ...................................................................... 84
Hypothesis 2..................................................................................................................... 88
BUILDING NETWORKS FOR CHANGE
vii
Hypothesis 3..................................................................................................................... 91
Hypothesis 4..................................................................................................................... 93
Hypothesis 5..................................................................................................................... 96
Summary of Findings on Organizational Context ........................................................... 98
CHAPTER SEVEN -- Multiple-Case Study on the Social Positioning of DLCs ........... 101
Introduction ........................................................................................................................ 101
Conceptual Framework ...................................................................................................... 103
Individual Attributes of Ed-Tech Coaches .................................................................... 104
Brokering Practices of Ed-Tech Coaches ...................................................................... 104
Social Positioning of Ed-Tech Coaches ......................................................................... 106
Contextual Factors: Social and Other School Conditions and District Context ............ 107
Contributions to the Extant Literature ........................................................................... 109
Case Study Approach ......................................................................................................... 110
Case Site and Participant Selection ................................................................................ 110
Data Sources .................................................................................................................. 114
Data Analysis ................................................................................................................. 117
Representation of Data ................................................................................................... 119
Trustworthiness .............................................................................................................. 121
Methodological Limitations ............................................................................................... 122
CHAPTER EIGHT -- Qualitative Findings on Social Positioning of DLCs .................. 124
Introduction ........................................................................................................................ 124
Weak Social Positioning of DLCs as CCSS Experts ......................................................... 125
Formal Responsibilities of DLCs Distanced from Balanced Literacy ........................... 126
Process versus Product-Oriented Coaching ................................................................... 128
Minimal Focus of Instructional Support on Balanced Literacy ..................................... 131
Individual and School Conditions Informed DLC Brokering Practices ........................ 133
Different Collaborative Practices of DLCs for Technology-Enabled Instruction ............. 144
Formal Responsibilities of DLCs Prioritized DFs Over Other Teachers ...................... 145
Collaborative versus Non-Collaborative Brokering Practices ....................................... 146
Individual and School Conditions Informing DLC Brokering Practices ....................... 149
Cross-Case Conclusions..................................................................................................... 160
CHAPTER NINE -- Implications for Policy, Practice, and Research ............................ 164
Review of Dissertation and Findings ................................................................................. 164
Contributions to Research .................................................................................................. 168
Implications for Policy and Practice .................................................................................. 170
Future Research ................................................................................................................. 175
Summary ............................................................................................................................ 176
REFERENCES ..................................................................................................................... 178
TABLES AND FIGURES ................................................................................................... 198
APPENDIX A: Goodness of Fit Tests for ERG Models ................................................... 228
APPENDIX B: Complete ERG Model Results ................................................................. 230
APPENDIX C: Interview Questions linked to Conceptual Framework ......................... 232
APPENDIX D: Data Matrix Supporting Qualitative Findings ....................................... 243
BUILDING NETWORKS FOR CHANGE
0
CHAPTER ONE -- Introducing Ed-Tech Coaching
A Brief Review of Policy Context Surrounding Ed-Tech Coaching
In the past decade, school districts have faced increasing pressure from federal and state
policies to use technology to innovate teaching and learning. A prominent example is the
adoption of the Common Core State Standards (CCSS), now in 42 states, which require teachers
to use technology to teach rigorous content, core literacy and math skills, and higher-order skills
such as creativity, collaboration, and critical thinking, with assessments offered online (CCSS,
2014a, 2014b).
1
In addition, philanthropic funding (e.g., The Bill and Melinda Gates Foundation
Next Generation Learning Challenge, The Carnegie Corporation of New York’s Opportunity by
Design initiative) and federal initiatives such as the Federal Communication Commission’s E-
Rate program, Race to the Top and, more recently, the Every Student Succeeds Act, have
incentivized districts to experiment with personalized learning initiatives that use technology to
meet individual student needs. In response to these policies, thousands of districts have rolled out
one-to-one computing initiatives in which teachers and students have individual (or one-to-one)
access to digital devices for teaching and learning. These initiatives are costly, with districts
spending billions of dollars on digital devices and content, software, and related teacher
professional development (PD) (Herold, 2015; Miller et al., 2016).
Ultimately, district leaders expect these large investments in school technology to
improve students’ academic performance, higher-order thinking skills, and college and
workforce readiness. However, these expectations are rooted in assumptions about district and
school capacity that may not hold in practice. Some of the more crucial assumptions concern the
ability of teachers to transition from traditional classroom routines, such as whole-class lecturing
1
While originally intended as nationwide initiative, 42 states have fully or partially adopted the CCSS to date
(National Center on Standards & Assessment Implementation, 2016).
BUILDING NETWORKS FOR CHANGE
1
and drill-and-practice, to constructivist approaches that support student engagement and critical
thinking (Lazonder & Harmsen, 2016; Pane, Griffin, McCaffrey, & Karam, 2013; Zhang et al.,
2016). These constructivist teaching practices involve teachers using technology to tailor
instruction to students’ needs and interests (“personalized instruction”) (Bingham, Pane, Steiner,
& Hamilton, 2016; Tanenbaum, Le Floch, & Boyle, 2013), and to support student-led inquiry of
content standards (“inquiry-based instruction”) (Aubusson, Burke, Schuck, Kearney, &
Frischknecht, 2014).
2
Unlike more prescriptive instructional approaches, personalized and
inquiry-based teaching practices place large information search burdens on teachers, requiring
them to constantly research and adapt new digital resources and teaching practices to meet
changing student needs (Bingham, 2016; Frank, Zhao, Penuel, Ellefson, & Porter, 2011).
However, despite increased access to technology in schools and urgent calls for change,
teachers have been slow to make these pedagogical shifts. In fact, both national reports (Gray,
Thomas, & Lewis, 2010) and local studies (Bingham, 2016; Blanchard, LePrevost, Tolin, &
Gutierrez, 2016) show that teachers use technology to supplement traditional instruction rather
than transform their instructional approach. Given these minimal shifts, it is perhaps not
surprising that researchers have also found mixed effects of education technology interventions
on student achievement (e.g., Barrow, Markman, & Rouse, 2009; Means, Toyama, Murphy, &
Baki, 2013; Rouse & Krueger, 2004).
Because of these challenges, federal, state, and local governments are investing heavily in
teacher PD as a critical leverage point for change (Lawless & Pellegrino, 2007). In particular, a
growing number of districts are recruiting expert teachers from schools to work as education
technology coaches or “ed-tech coaches” to guide teachers in using technology creatively and
2
I refer to these personalized and inquiry-based teaching practices as “technology-enabled” instruction throughout
this dissertation.
BUILDING NETWORKS FOR CHANGE
2
effectively in the classroom (Flanigan, 2016). These coaches are early to middle-stage career
teachers with strong self-efficacy for instruction, taking it upon themselves to learn how to use
technology for instruction (e.g., experimenting with new software apps, earning a master’s
degree in technology and instruction) and willing to wear multiple hats to meet the different
needs of schools and teachers (Cavanagh, 2015; Flanigan, 2016).
3
Accordingly, ed-tech coaches
fulfill many demanding roles, including advising teachers on technical aspects of technology
(e.g., how to use software, resolving technical glitches), as well as demonstrating how
technology can be used to improve lesson content and pedagogy, student assessment, classroom
management, and student digital citizenship (International Society for Technology in Education,
2011).
4
Outside of these classroom tasks, ed-tech coaches are expected to serve as instructional
leaders, facilitating district-wide improvements in teaching and learning. This includes helping
central office and school leaders develop a joint vision of how technology should support
instruction; promoting new digital tools and teaching practices across schools; and developing
PD and other supports that are responsive to changing school needs (International Society for
Technology in Education, 2011).
While ed-tech coaching programs are gaining prominence among districts (Flanigan,
2016; Herold, 2015), these programs make strong assumptions about the capacity of ed-tech
coaches to shift teacher practice that are grounded in little evidence (Kopcha, 2012; Lawless &
Pellegrino, 2007). To investigate these assumptions, education scholars have primarily focused
on the one-on-one instructional support that ed-tech coaches provide teachers (Glazer, Hannafin,
Polly, & Rich, 2009; Kopcha, 2012; Swan & Dixon, 2006). Much less attention has been
3
I define self-efficacy for instruction as teachers’ perceived abilities and confidence in delivering instruction (or
certain kinds of instruction such as technology-enabled instruction) to achieve their goals for student learning.
4
Student digital citizenship refers to norms of appropriate and responsible use of technology for both educational
and personal purposes.
BUILDING NETWORKS FOR CHANGE
3
directed toward understanding how ed-tech coaches work as instructional leaders, which
involves understanding how these coaches circulate information among central office leaders and
staff, school principals, and teachers to inform district-wide procedures and practices for
instruction (Swinnerton, 2007). This perspective is important given the “knowledge-intensive”
nature of education technology reforms, which require districts to engage in a continuous cycle
of searching, experimenting, and disseminating best practices so that central office and school
decision-making and individual teacher practice can adapt to new insights on how to use
technology to improve student learning (Bingham et al., 2016; Ho & Ng, 2017; Honig &
Ikemoto, 2008).
5
Overview of the Dissertation
To support these knowledge-intensive demands, many districts have embraced a
distributed leadership approach (Spillane & Diamond, 2007; Spillane, Halverson, & Diamond,
2004) in which central office and school actors collaborate and engage expertise throughout their
school systems to inform instructional procedures and practice (Durand, Lawson, Wilcox, &
Schiller, 2016; Ho & Ng, 2017; Hopkins, Ozimek, & Sweet, 2016). Ed-tech coaches are central
to this distributed process of change, brokering information between central office and school
actors to support instructional coordination, coherence, and alignment, which I discuss below as
being key processes for systemic instructional change.
Brokering is a relational process that involves connecting individuals or groups to
provide access to social capital for achieving individual or collective goals (Burt, 2000). The idea
of coaches brokering information is not new to policy and practice (Coburn & Stein, 2006;
Coburn & Woulfin, 2012; Lave & Wenger, 1991). In fact, most instructional coaching studies
5
These best practices are not just focused on improving how teachers teach, but also on improving how central
office administrators and school leaders create district and school conditions conducive to instructional change.
BUILDING NETWORKS FOR CHANGE
4
assume that coaches broker information from the central office into schools to align teacher
practice with central office goals and resources for instructional change. I break from this
tradition by arguing that ed-tech coaches are also expected to broker information laterally in their
school systems to coordinate instructional improvement on a system-wide basis (Coburn, Toure,
& Yamashita, 2009; Hopkins et al., 2016; Mayer, Woulfin, & Warhol, 2014), and in a bottom-up
direction from schools to the central office to build coherence between central office directives
and school goals and strategies for improving student learning (Daly & Finnigan, 2016; Honig,
2004).
As research has shown (Bridwell-Mitchell & Cooc, 2016; Daly, 2010; Daly & Finnigan,
2016; Moolenaar, 2012), this relationship building is complex in nature, with ed-tech coaches
having to navigate complex organizational structures, norms, and informal networks of relations
to direct information flows in their school systems. In this dissertation, I draw on social network
theory and methods to observe these organizational conditions as they manifest in the social
structure of districts and schools. Social structures are networks of relationships (i.e., social ties)
that facilitate and/or constrain access to social capital (Daly, 2010; Lin, 2001; Scott, 2000).
According to social network theory, access to social capital depends on the qualities of the
relationships that individuals share with others (e.g., the strength and/or reciprocal nature of
these relationships), as well as structural features of networks such as the positioning of ed-tech
coaches (relative to other actors) to direct information flows in their district and the overall
configuration of network relationships (Bridwell-Mitchell, forthcoming; Daly, 2010; Moolenaar,
2012). I use these ideas to examine how organizational conditions influence the brokering
practices of ed-tech coaches, including how ed-tech coaches are positioned to have power and
influence over instructional practice (Moolenaar, Daly, & Sleegers, 2010) and to access expertise
BUILDING NETWORKS FOR CHANGE
5
of other central office and school actors to inform their brokering (Liou, 2016; Penuel, Sun,
Frank, & Gallagher, 2012). I also investigate how the brokering practices of ed-tech coaches are
influenced by informal communication patterns such as dense ties, reciprocal exchange, and
fragmented social networks that feature inward-looking among actor groups as opposed to
outward-looking ties (Bridwell-Mitchell & Cooc, 2016; Daly & Finnigan, 2012).
I situate my study in a mid-sized, diverse urban school district – Lexington Unified
School District
6
– that is ahead of most districts in the country in completing a one-to-one
technology rollout of iPads (elementary and middle school) and laptops (high school) to support
implementation of the CCSS. To help teachers use these digital tools to fulfill instructional and
assessment objectives of the CCSS, district leaders implemented a digital coaching program
(“DCP”) as a central mechanism for generating and disseminating best practices for technology-
enabled instruction of these content standards. In the context of this reform, I ask three research
questions:
1. How, if at all, are ed-tech coaches in LUSD brokering information on (a) integrating
technology with instruction and (b) teaching the CCSS?
2. How does the organizational context of LUSD, as mediated by social structure, inform the
brokering practices of these ed-tech coaches?
3. Are there similarities and/or differences in the brokering practices of ed-tech coaches for
these reform efforts, and if so, why?
I answer my research questions using social network theory as a lens for observing the brokering
practices and organizational context of ed-tech coaches, along with social network analysis and
complementary qualitative methods to inform my analysis. I ground my analysis in
6
All names of individuals, schools, and the district in this dissertation are pseudonyms to preserve the anonymity of
my research site and individual research participants.
BUILDING NETWORKS FOR CHANGE
6
administrative, survey, and social network data collected from central office administrators,
school leaders, ed-tech coaches, and teachers in the 2014-15 school year, as well as interviews
with a purposive sample of these study participants.
I make several contributions to the extant literature on ed-tech coaching. First, I provide
some of the first empirical evidence on the instructional leadership of ed-tech coaches by
describing their brokering practices, examining the organizational context surrounding their
brokering ties, and attending to the implications of their brokering for teacher learning and
systemic change. I also explore how ed-tech coaches’ leadership varies across instructional
reforms (i.e., integrating technology with instruction and teaching the CCSS) that district leaders
increasingly expect to be linked in practice with the concurrent rollout of one-to-one computing
programs and the CCSS standards.
Beyond ed-tech coaching, I contribute to research on instructional coaching and the work
of district leaders who are increasingly relying on instructional coaches to support knowledge-
intensive reform. First, I shed light on the ways in which ed-tech coaches – as a sub-group of
instructional coaches – demonstrate instructional leadership to support district-wide
improvements in teaching and learning. In particular, I add to the largely qualitative evidence-
base on instructional coaching, which suggests that instructional leadership is an important yet
understudied aspect of coaching (Firestone & Martinez, 2007; Swinnerton, 2007), by using
social network analysis to quantify the extent to which ed-tech coaches broker information and
demonstrate this instructional leadership: (a) across different levels of school systems (i.e.,
within and between the district central office and schools), and (b) in support of interrelated
instructional reforms. As mentioned earlier, most of the extant literature has focused on how
coaches broker information in a top-down manner from the central office into schools (through
BUILDING NETWORKS FOR CHANGE
7
the provision of teacher PD) instead of examining how coaches broker information laterally and
in a bottom-up direction from schools to the central office. Moreover, these prior studies have
focused primarily on the role of coaches in supporting singular reforms (e.g., a new literacy or
mathematics curriculum) instead of multiple, interrelated reforms.
I also offer new insights to the role of organizational context in shaping brokering and
related processes of instructional change (e.g., Hopkins et al., 2016; Marsh, Bertrand, & Huguet,
2015; Matsumura & Wang, 2014). I describe how organizational conditions such power and
influence, access to expertise, and informal communication patterns manifest in a social structure
that facilitates and/or constrains brokering across an entire school district. This perspective
represents a substantial shift from prior studies, which have so far examined how these
organizational conditions influence communication between central office and school actors in
isolated pockets of school systems (e.g., the influence of teacher trust on coach interactions with
teachers in schools) rather than on a systemic basis (e.g., Anderson, Feldman, & Minstrell, 2014;
Atteberry & Bryk, 2011; Matsumura & Wang, 2014). In addition, I demonstrate how the social
structures of districts and schools can be dynamic and affect brokering in different ways based
on the topic of the instructional reform being discussed (i.e., integrating technology with
instruction or teaching the CCSS).
The remainder of this dissertation proceeds as follows. In chapter two, I draw on the
extant literature to develop a conceptual framework of the brokering practices that support
coordination, coherence, and alignment in education technology reforms. I also review findings
from this literature on organizational conditions that facilitate and/or constrain these brokering
practices. Because this perspective on brokering is new to research on ed-tech coaching and
instructional coaching in general, I motivate my dissertation as being one of the first studies to
BUILDING NETWORKS FOR CHANGE
8
investigate the role of instructional coaches in leading district-wide improvements in teaching
and learning in knowledge-intensive reform.
In chapter three, I draw on social network theory to develop a framework for observing
the brokering practices of ed-tech coaches. A social network perspective affords several
advantages to my research. First, social network theory offers a set of brokering properties for
observing the extent to which different brokering practices expected of ed-tech coaches
contribute to information flows in districts and schools. Second, social network theory provides a
lens for translating organizational conditions that prior research has shown to influence brokering
and instructional change into a coherent set of properties for describing the social structures of
districts and schools. These properties, in turn, allow me to observe how organizational context
shapes information flows and consequently the brokering practices of ed-tech coaches. To guide
my analysis, I draw on the above brokering and social network properties to develop a series of
hypotheses on the expected brokering practices of ed-tech coaches and how the social structures
of districts affect these brokering practices. Finally, because social network theory assumes that
social ties and structures vary based on the purpose of relationships, I argue that the brokering
practices and organizational context of ed-tech coaches can change across reform topics and that
social network theory provides a lens for observing these differences.
In chapter four, I introduce LUSD and its DCP as a case for investigating my research
questions and hypotheses. In chapter five, I outline my social network data collection and
analytic methods for answering my research questions, and present my results from this analysis
in chapter six. I then review my qualitative data and multiple-case study research design to build
on my social network findings and investigate the brokering practices of two ed-tech coaches in
LUSD in closer detail. I present the findings from this case study analysis in chapter eight.
BUILDING NETWORKS FOR CHANGE
9
Finally, I conclude with a discussion of the implications and limitations of my dissertation
research in chapter nine.
BUILDING NETWORKS FOR CHANGE
10
CHAPTER TWO -- Exploring Ed-Tech Coach Leadership and Brokering
Introduction
As noted above, education technology reforms are growing in popularity among school
districts, but are grounded in little empirical evidence on the effectiveness of these reforms in
transforming classroom instruction and student learning (Bingham, 2016; Blanchard et al., 2016;
Cuban, 2012; Zheng, Warschauer, Lin, & Chang, 2016). Because these reforms demand
substantial shifts in teacher practice and how districts and schools are organized to support
classroom instruction, researchers have started focusing on early-stage implementation activities
in these reforms to get a better sense of what is working and not working as intended (e.g.,
Bingham et al., 2016; Lake, Hill, & Maas, 2015; Miller et al., 2016; Tanenbaum et al., 2013).
These studies have generally found a range of implementation challenges that detract from the
efficacy of education technology reforms, including inadequate school infrastructure
(Tanenbaum et al., 2013), incoherent teacher professional development (Lawless & Pellegrino,
2007), and misaligned expectations for student learning and school success among policymakers,
educators, and students (Bingham et al., 2016).
While there are many underlying causes of these implementation challenges, an
emerging finding from the extant literature is need for instructional leadership that can bring
about creative, coordinated, and responsive solutions for supporting teachers to improve
instruction (DeArmond & Gross, 2016; Lake et al., 2015; Tanenbaum et al., 2013). Districts are
increasingly relying on ed-tech coaches to demonstrate this leadership in terms of supporting
teachers to change their instructional approach (Flanigan, 2016), and adapting district-wide
instructional procedures in response to how schools are experimenting with and using
technology for instruction (Cavanagh, 2014b; International Society for Technology in Education,
BUILDING NETWORKS FOR CHANGE
11
2011). In this chapter I develop a conceptual framework (Figure 1) that translates the
instructional leadership practices of ed-tech coaches into brokering practices that support
district-wide change. I start by reviewing literature on instructional leadership in systemic school
reform to describe three core leadership practices – instruction coordination, coherence, and
alignment (Box 1) –expected of frontline staff such as ed-tech coaches who interface between
the central office and schools. I then draw parallels between these leadership practices and the
responsibilities of ed-tech coaches as currently voiced by districts, the media, advocates for
technology-enabled instruction (e.g., the International Society for Technology in Education) and
the research community.
Since these leadership practices require ed-tech coaches to build relationships and
circulate information and resources on a district-wide basis, I use brokering as a lens for
observing these leadership practices in action (Box 2). And because brokering requires ed-tech
coaches to forge new relationships and facilitate new channels of communication in districts,
which is work that depends on prevailing structures, norms, and informal networks of
communication, I review findings on organizational conditions (Box 3) that can influence the
ability of ed-tech coaches to broker information on a district-wide basis.
Instructional Leadership in Systemic School Reform
There is a strong research base to support the notion that instructional leadership matters
for supporting systemic changes in teacher practice and student learning (e.g., Hitt & Tucker,
2016; Wenner & Campbell, 2016). While most of this research has focused on instructional
leadership as exercised within schools by school leaders (Honig & Coburn, 2006), there is a
growing body of evidence on the instructional leadership practices of central office “mid-level”
administrators and “frontline” staff (Honig, 2003, p. 299) who interface between the central
BUILDING NETWORKS FOR CHANGE
12
office and schools and exercise instructional leadership on a district-wide basis (e.g., Coburn &
Talbert, 2006; Daly & Finnigan, 2016; Ho & Ng, 2017; Honig & Hatch, 2004; Hopkins et al.,
2016; Spillane & Burch, 2004).
7,8
This literature describes frontline staff as supporting systemic reform through three core
leadership practices (Box 1): instructional coordination, coherence and alignment. As shown by
the double headed arrow in Box 1, these practices are iterative and interconnected, building on
one another to improve teaching and learning at scale (Honig & Hatch, 2004). Instructional
coordination involves establishing a vision for instructional change that cuts across all
instructional programs and services in a district (1A.1), and developing the necessary procedures,
structures, and resources to bring this vision to fruition (1A.2) (Coburn et al., 2009; Durand et
al., 2016; Lake et al., 2015). Because frontline staff are responsible for supervising and
implementing instructional reforms in schools, they are strategically positioned to glean insights
on how central offices should be coordinating decision-making across educational programs and
services to affect instructional change (Honig, 2004; Spillane & Burch, 2004). This coordinated
decision-making, in turn, can help school leaders and teachers make sense of and implement
instructional reform as intended, identifying connections between these reforms and their day-to-
day responsibilities for educating students and leveraging resources and supports available to
them to achieve reform goals (Coburn et al., 2009). As part of their work in schools, frontline
staff are also expected to coordinate instructional decision-making and practices across school
7
This body of literature has examined a range of polices such as school and community partnerships (Honig, 2003),
data-driven decision-making (Coburn, Toure, & Yamashita, 2009), and high-stakes accountability regimes (Daly,
Finnigan, Moolenaar, & Che, 2014).
8
Mid-level central office staff are defined as “program managers, content area directors, budget specialists, and
others who administer or manage programs and services but are not in top-cabinet positions (Spillane & Burch,
2004, p. 1). Frontline staff such as curriculum specialists, instructional coaches, special education coordinator are
also responsible for administering and supervising instructional programs, but have less authority in the
organizational hierarchy of the central office. Given my interest in ed-tech coaches, I focus on the work of these
frontline staff in this literature review and use the terms “frontline staff” and “coaches” interchangeably in my
writing.
BUILDING NETWORKS FOR CHANGE
13
sites, exposing school leaders and teachers to best practices that can further teacher learning and
change and expand the footprint of instructional change in their district (Hopkins et al., 2016).
To ensure that school efforts in instructional innovation are authentic and sustained,
frontline staff are responsible for making central office directives for instructional change
coherent with the instructional goals of school leaders and teachers (e.g., Daly & Finnigan, 2012,
2016; Honig, 2004). Honig and Hatch (2004) break down coherence as consisting of two
ongoing tasks, namely: translating central office directives for instructional change into goals
and strategies that are specific to school goals and needs (which I refer to as “school coherence”)
(1B.1), and adjusting central office procedures and resources to foster district conditions that are
more responsive to school goals and needs (which I refer to as “system coherence”) (1B.2).
Through building school and system coherence, frontline staff help school leaders and teachers
adapt instructional reforms to meet the specific school needs and ensure that schools are
continuously supported by the central office to pursue these localized improvement efforts.
Third, to ensure that central office and school goals for instruction translate into tangible
shifts in teacher practice, frontline staff are expected to work closely with teachers to align their
instructional practice with central office and school goals (1C) (Coburn & Woulfin, 2012;
Woulfin, 2014). In many ways, this work in instructional alignment is an extension of how
frontline staff build school coherence (1B.1) and coordinate instructional improvement within
and across schools (1A.3). However, because shifting teacher practice requires shifting teacher
knowledge, beliefs, and motivation for instructional change, this process requires more intense
communication in the form of sustained, situated, content-specific, and collaborative teacher
learning experiences (Darling-Hammond, Wei, Andree, Richardson, & Orphanos, 2009;
Desimone, 2009; Garet, Porter, Desimone, Birman, & Yoon, 2001). Through this joint work,
BUILDING NETWORKS FOR CHANGE
14
frontline staff bridge teacher practice to central office and school goals for instructional change,
ensuring that instructional reforms affect classroom instruction and ultimately student learning as
intended (Woulfin, 2014).
Instructional Leadership of Ed-Tech Coaches
As expert teachers who have been selected by district leaders to supervise and support
technology-enabled and CCSS instruction, ed-tech coaches are part of a larger, nationwide trend
of districts promoting expert teachers to the ranks of frontline staff to support district-wide,
knowledge-intensive instructional change. Not surprisingly, there are many parallel between the
core leadership practices of frontline staff as described above and how districts, the media,
advocates for technology-enabled instruction, and the research community have described the
work of ed-tech coaches in leading education technology reform.
In terms of instructional coordination, ed-tech coaches are expected to work with district
leaders and central office administrators to position technology as a central resource for teaching
content standards, preparing students for standards-based assessments, and supporting other core
programs for student learning and school improvement (International Society for Technology in
Education, 2011). This involves developing district procedures (e.g., adoption of new content
standards and assessments) and resources (e.g., purchasing educational software, designing
professional development) that provide school leaders and teachers with comprehensive support
for integrating technology with instruction (Cavanagh, 2014b; Superville, 2015). Through their
work with teachers, ed-tech coaches are supposed to coordinate information and resource
exchange within and across schools, exposing school leaders and teachers to best practices for
technology-enabled instruction (Herold, 2015; Superville, 2015). In theory, these combined
efforts are supposed to coordinate district expectations, procedures, and resources for improving
BUILDING NETWORKS FOR CHANGE
15
teacher practice, student outcomes and school performance and therefore mitigate many of the
implementation challenges plaguing education technology reform so far (e.g., incoherent teacher
PD, misaligned expectations for school success).
The success of education technology reform also depends on schools experimenting with
instruction to develop personalized and inquiry-based teaching practices that enhance student
learning (Cavanagh, 2014a; Herold, 2015). Ed-tech coaches are expected to kick-start this school
experimentation through building school coherence, translating central office goals, procedures,
structures, and resources for instruction into school goals and strategies that use technology to
address specific student needs (Glazer et al., 2009; Ho & Ng, 2017; Lawless & Pellegrino, 2007).
To sustain these school efforts in the long-term, ed-tech coaches are also supposed to build
system coherence, providing feedback to central office leaders and administrators on how to
adjust district procedures and resources for instruction in response to changing school needs,
insights, and plans for improvement (International Society for Technology in Education, 2011).
Through these ongoing efforts, ed-tech coaches are supposed to transform districts and schools
into self-correcting, adaptive systems that are constantly responding to student needs and latest
developments in the field of education technology (Cavanagh, 2014b; Ho & Ng, 2017; Molnar,
2017).
Ed-tech coaches also play a central role in instructional alignment in terms of supporting
teachers to shift their instructional practice to reflect central office and school goals for
instructional change (Herold, 2015; International Society for Technology in Education, 2011).
This requires ed-tech coaches to engage teachers in an ongoing trial-and-error process for
incorporating technology into classroom lessons; modeling instruction, observing teachers,
providing feedback on their instructional practice, and engaging teachers in reflective
BUILDING NETWORKS FOR CHANGE
16
conversations on instructional change (Glazer & Hannafin, 2008; Kopcha, 2012; Sugar, 2005).
Through this joint work, ed-tech coaches are expected to shift teachers’ knowledge, beliefs, and
motivations for instructional change. This includes building teachers’ technical knowledge for
operating digital devices and related software applications or apps, as well as teachers’ content
and pedagogical knowledge for using technology to fulfill specific learning objectives (Lawless
& Pellegrino, 2007; Mishra & Koehler, 2006). Ed-tech coaches are also supposed to improve
teacher self-efficacy for using technology to accomplish classroom goals (Wang, Ertmer, &
Newby, 2004), shift teacher beliefs toward a more student-centric view of teaching and learning
(Ertmer, Ottenbreit-Leftwich, Sadik, Sendurur, & Sendurur, 2012), and help teachers overcome
perceived barriers against using technology in their classrooms (Kopcha, 2012).
9
Through
facilitating opportunities for teachers to collaborate with colleagues, ed-tech coaches are
expected to develop social networks that reinforce teacher learning (Ertmer, 2005; Swan &
Dixon, 2006) and sustain instructional innovation (Coburn, Russell, Kaufman, & Stein, 2012;
Frank et al., 2011).
Brokering Practices Related to Instructional Leadership
Given the importance of instructional coordination, coherence, and alignment in
education technology reform, along with the central role of ed-tech coaches in enacting these
leadership practices, ed-tech coach leadership is a critical piece in the implementation of
education technology reform that warrants further investigation. This then begs the question of
how to observe ed-tech coach leadership in action? There is a growing body of literature that
suggests instructional leadership is a distributed phenomenon, meaning that districts and school
leaders must socially engage one another and pool expertise throughout their organizations to
9
Examples of these barriers include external pressure to conform to traditional instructional styles, unclear
procedures for accessing technology in schools, and technical glitches when using technology (Kopcha, 2012).
BUILDING NETWORKS FOR CHANGE
17
improve teaching and learning (Gronn, 2000; Harris, 2004; Spillane & Diamond, 2007; Spillane
et al., 2004). This distributed approach has received special emphasis in the context of
knowledge-intensive reform where district leaders do not have a prescribed approach for how
teachers should be teaching in their classrooms and instead have had to leverage expertise
throughout their school systems to develop and disseminate best practices for instruction
(Bingham et al., 2016; Durand et al., 2016; Ho & Ng, 2017; Liou, 2016). In the context of ed-
tech coaching, a distributed leadership perspective implies that these coaches need to build
relationships with central office and school actors to coordinate, build coherence, and align
instructional practice.
To describe this relationship building, education researchers have used theoretical and
empirical approaches that acknowledge the social dimension of educational organizations (e.g.,
Coburn & Stein, 2006; Little, 2002; Louis, Marks, & Kruse, 1996; McLaughlin & Talbert,
2003). One of the main findings from this research is that districts consist of complex networks
of social relationships for circulating information, resources, and other forms of social capital in
support of organizational change (Daly, 2010; Moolenaar, 2012). I take a similar approach by
focusing on brokering as a social practice that enables the instructional leadership of ed-tech
coaches (research question 1). Researchers have long argued that brokering can support
instructional coordination, coherence, and alignment (Coburn & Stein, 2006; Daly & Finnigan,
2016; Honig & Hatch, 2004). That said, a cohesive discussion of what brokering looks like and
how it maps onto these leadership practices is missing from the extant literature. This is because
researchers have so far associated brokering with processes such as partnership building, conflict
resolution, and stakeholder coordination that bridge districts to external resources for
instructional change rather than exploring the potential for brokering to facilitate knowledge
BUILDING NETWORKS FOR CHANGE
18
sharing and exploitation within districts (e.g., Durand et al., 2016). Moreover, the studies that
have described brokering exchanges within districts have largely done so from a top-down
perspective, assuming that brokering is supposed to align teacher practice with central office
goals for instructional change (e.g., Coburn & Stein, 2006; Woulfin, 2014) rather than to
coordinate instructional practices within the central office and across schools or to build
coherence between the central office and schools (Daly & Finnigan, 2016).
To address these limitations, I draw on research on central office instructional leadership,
instructional coaching, and professional learning to build a more complete framework for
understanding how brokering enables instructional leadership. Specifically, I connect leadership
practices of instructional coordination, coherence, and alignment to brokering practices (Box 2)
that describe how ed-tech coaches circulate information in all directions of their school systems.
These brokering practices include coordinating information within the central office (2A),
engaging in top-down communication from the central office into schools (2B), coordinating
information within and among schools (2C), and engaging in bottom-up communication
channels from schools to the central office (2D).
Central Office Coordination
Research on central office instructional leadership suggests that communication among
central office administrators and staff is necessary to develop a shared vision for instructional
change that cuts across all educational programs offered in a district and that is appropriately
supported (Coburn et al., 2009; Durand et al., 2016; Liou, 2016). To achieve this goal, frontline
staff are tasked with breaking down barriers between the different hierarchical levels and
divisions housed within the central office. This involves middle-up communication, defined as
frontline staff sharing information with senior leaders and administrators in the central office
BUILDING NETWORKS FOR CHANGE
19
(Honig, 2004; Spillane & Burch, 2004), and lateral communication, defined as frontline staff
exchanging information with similarly-ranked colleagues within and across divisions of the
central office (Coburn et al., 2009; Firestone & Martinez, 2007).
Despite the importance of central office coordination, there is surprisingly little research
on the extent to which frontline staff engage in middle-up and lateral communication (Liou,
2016). Most studies, which pre-date knowledge-intensive reforms such as education technology
and CCSS instruction, suggest that information exchanges to coordinate policy in the central
office are sparse and infrequent (e.g., Coburn et al., 2009; Finnigan & Daly, 2010; Honig, 2003).
For instance, some studies show that, in the absence of coordinated decision-making across
central office divisions, schools receive multiple and conflicting messages for instructional
change that make reforms more challenging to implement (Coburn et al., 2009; Lake et al.,
2015). Recent studies suggest that, with the onset of more complex knowledge-intensive
reforms, district leaders are imbuing central offices with a collaborative culture geared toward
coordinating central office instructional goals, procedures, and resources (Durand et al., 2016;
Liou, 2016). This collaborative work is driven by an adaptive leadership approach of
redistributing tasks to individuals situated lower in the organizational hierarchy of central office
and schools to foster system-wide communication and collaboration (Hargreaves & Fink, 2005;
Heifetz, Grashow, & Linsky, 2009; Spillane, 2013).
For example, Durand and colleagues (2016) find that “odds-beating” districts (i.e., those
that achieved beyond predicted outcomes on CCSS assessments) proactively fostered
collaboration within central offices relative to “typical performing” districts (i.e., those that
achieved at predicted performance levels). The authors argue that leaders in these odds-beating
districts engaged in adaptive leadership practices such as: (a) establishing collaborative models
BUILDING NETWORKS FOR CHANGE
20
of decision-making in which senior central office administrators and staff would engage school
leaders and teachers in substantive dialogue related to instructional reform; (b) creating formal
routines for joint decision-making between different central office divisions; (c) redefining the
formal roles of central office administrators to include the management and coordination of
multiple, interrelated programs related to student achievement; and (d) developing a shared
language for discussing reform efforts that emphasizes coordination across instructional
programs and services. These adaptive leadership practices consistent with the leadership
responsibilities of ed-tech coaches, who are expected to collaborate with central office
administrators and staff to develop district-wide procedures and resources that position
technology as a central resource for supporting standards-based instruction and assessment along
with other instructional programs. Yet, while it is assumed that districts are empowering ed-tech
coaches through these adaptive leadership practices, there has been little research to show if this
is the case and the extent to which ed-tech coaches are informing central office decision-making.
Top-Down Communication
As mentioned earlier, research on instructional coaching has mainly focused on how
coaches, as frontline staff, work in a top-down manner to align teacher practice with central
office directives for change. Researchers have further characterized these top-down brokering
practices in terms of buffering and/or bridging teachers to/from messages for instructional
change (Coburn & Woulfin, 2012) and in terms of social learning routines that inform teacher
practice (Coburn & Russell, 2008; Marsh et al., 2015; Mudzimiri, Burroughs, Luebeck, Sutton,
& Yopp, 2014).
Bridging and buffering. Buffering, defined as sheltering schools from external demands
for instructional change, can involve a range of actions from ignoring or blocking central office
BUILDING NETWORKS FOR CHANGE
21
demands for reform, to reducing the complexity of instructional change expected from teachers,
to providing more space and time for teachers to experiment with a subset of instructional ideas
(Honig & Hatch, 2004). In contrast, bridging, defined as bringing external demands for
instructional change closer to schools, includes creating a sense of urgency for reform, actively
negotiating the meaning of instructional reforms with school leaders and teachers, and providing
resources to support teacher learning (Coburn & Woulfin, 2012; Honig & Hatch, 2004).
Research on instructional coaching suggests that teachers demonstrate more authentic shifts in
classroom instruction in response to bridging practices such as persuasion, and more superficial,
symbolic, and compliance-oriented changes in response to buffering practices or bridging
practices such as pressure (Coburn & Woulfin, 2012; DiPaola & Tschannen-Moran, 2005; Honig
& Hatch, 2004). That said, because top-processes for achieving instructional alignment depend
on how teachers interpret reform messages for instructional change and the extent to which these
messages fit with a school’s existing mission, culture, operations, and political and normative
environment (Honig & Hatch, 2004), coaches often use a mix of buffering and bridging
strategies when working with teachers (Coburn & Woulfin, 2012; Swinnerton, 2007).
Social learning routines. Because coaches often work with teachers in the context of
providing teacher PD, researchers have further characterized their top-down communication in
terms of social learning routines that inform teacher learning and change. Social learning
routines refer to collaborative activities or artifacts that deepen the professional learning of
education practitioners in authentic work settings (Honig, 2012). The first of these social
learning routines involves initiating and sustaining social engagement with teachers on
instructional matters through dialogue and norms for instructional change (Coburn & Russell,
2008; Marsh et al., 2015). For instance, coaches can engage teachers in dialogue on certain
BUILDING NETWORKS FOR CHANGE
22
aspects of teacher practice (e.g., classroom management, student behavior) and can initiate these
conversations in a one-to-one or group setting, depending on established norms for discussing
teacher practice.
Second, coaches can integrate new teaching practices with joint work that is already of
value to teachers, whether it be accomplishing central office directives that are important to
teachers or working with teachers in professional learning communities (PLCs) focused on
achieving specific school goals (Marsh et al., 2015; Mayer et al., 2014). Finally, coaches can
help teachers develop a shared repertoire of conceptual (e.g., principles, frameworks, ideas) and
practical (e.g., practices, strategies, and resources) tools that, in turn, can support personal
experiences of success with new teaching practices, and the continued negotiation of
instructional change with other teachers (Grossman, Smagorinsky, & Valencia, 1999; Marsh,
McCombs, & Martorell, 2010). Examples of these tools include regular cycles of observing and
debriefing on teacher practice, as well as sharing or co-developing lesson plans, assessment data,
and observation rubrics for guiding classroom instruction.
While the above social learning routines are generally thought of as improving teacher
learning and practice, these routines can be implemented in ways that lead to varying degrees of
instructional change. For example, studies have shown that dialogue focused explicitly on
instructional tasks, student thinking, and teacher practice are more effective in aligning teacher
practice with reform goals, as compared to dialogue focused on non-instructional matters such as
technology, assessment data, and curricular content (Coburn & Russell, 2008; Marsh et al., 2015;
Swan & Dixon, 2006). As this example suggests, variation in teacher-coach dialogue offers
another perspective for understanding how coaches might bridge (i.e., instructional dialogue)
and/or buffer (i.e., non-instructional dialogue) teachers to/from reform messages for instructional
BUILDING NETWORKS FOR CHANGE
23
change.
Despite the detailed schema for describing how coaches engage teachers in top-down
communication to support instructional alignment, this critical lens has not been applied to the
work of ed-tech coaches. The few studies that have investigated interactions between ed-tech
coaches and teachers suggest that coaches use a combination of buffering and bridging practices
to gradually build teacher self-efficacy for making complex shifts in their instructional routines
(e.g., Glazer et al., 2009; Kopcha, 2012; Sugar, 2005; Swan & Dixon, 2006) (I review these
studies in more detail in chapter 7). However, given that many of these studies pre-date CCSS
instruction and the recent push for one-to-one computing to support standards-based instruction
and assessments, it is possible that ed-tech coaches are brokering information in entirely different
ways that are not well-documented or known.
School Coordination
Coordinating information within and across schools is another brokering practice that
frontline staff use to disseminate best practices for instruction at scale, build school coherence
for implementing instructional reform, and align teacher practice with instructional reform goals.
There are several reasons why forming these lateral ties within and among schools is important.
First, school leaders are a critical mediating factor in the implementation of instructional reforms
(Atteberry & Bryk, 2011; Marsh et al., 2010; Matsumura, Garnier, & Resnick, 2010). Because
school leaders set school goals for teaching and learning, influence teacher sense-making of
instructional reforms, and inform teacher perceptions of the instructional responsibilities of
coaches (Bean, Draper, Hall, Vandermolen, & Zigmond, 2010; Matsumura & Wang, 2014),
frontline staff need to coordinate information between school leaders and teaching staff when
building school coherence and aligning teacher practice with reform goals. Second research
BUILDING NETWORKS FOR CHANGE
24
suggests that, due to the novel resource-intensive nature of technology-enabled instruction,
teachers often depend on conditions outside of their immediate control to use technology in their
classrooms (e.g., access to digital tools, technical assistance, extended instructional time, school
leader and parent acceptance of new ways to teach and assess student learning) (Bingham et al.,
2016; Zhao, Pugh, Sheldon, & Byers, 2002). To manage this dependence on external factors, ed-
tech coaches need to coordinate resources on a school-wide basis so that teachers face minimal
barriers against instructional change.
Third, bridging practices and social learning routines that encourage teacher collaboration
are commonplace in the PD that coaches provide teachers (Hopkins et al., 2016; Marsh et al.,
2015). These collaborative opportunities, in turn, can coordinate instruction across classrooms so
that best practices are implemented at scale (Frank et al., 2011), build collective efficacy among
teachers for achieving instructional reform goals (a form of school coherence) (Moolenaar,
Sleegers, & Daly, 2011), and reinforce teacher learning and instructional alignment (Coburn &
Russell, 2008; Sun, Penuel, Frank, Gallagher, & Youngs, 2013). Studies suggest that
collaborative interactions between school leaders and teachers across schools also contribute to
instructional coordination, coherence, and alignment in the same way (Daly & Finnigan, 2012;
Marsh, Bush-Mecenas, & Hough, 2016; Spillane, Hopkins, & Sweet, 2015). However, these
between-school ties are more sparse and take time to develop, requiring deliberate and concerted
action from frontline staff to build interschool networks for instructional change (Daly &
Finnigan, 2016; Hopkins et al., 2016).
Despite the many reasons why frontline staff should be coordinating information within
and across schools as a central brokering practice of their instructional leadership, there is
surprisingly little research on their efforts to do so. This is because, as noted earlier, researchers
BUILDING NETWORKS FOR CHANGE
25
have mainly studied the brokering practices of frontline staff from a top-down perspective,
focusing on the one-on-one interactions between these frontline staff and teachers rather than
examining how these frontline staff engage in more inclusive and collaborative routines to build
school and district-wide networks for instructional change (Hopkins et al., 2016; Marsh et al.,
2015; Mayer et al., 2014)
Bottom-Up Communication
Building system coherence is a central practice for how frontline staff support district-
wide improvements in teaching and learning. This work involves facilitating bottom-up
communication channels from schools to the central office to make district procedures and
resources more responsive to school improvement goals and needs (Honig, 2004; Honig &
Hatch, 2004). Because this communication often features complex information exchange and
joint problem solving to differentiate district procedures and resources according to local school
circumstance, these bottom-up relations need to take place regularly and involve reciprocal
dialogue between central office administrators and school leaders (Daly & Finnigan, 2012). In
the case of education technology reform, these bottom-up communication channels are especially
important because the shifts being demanded of school instructional programs and teacher
practice are demanding and need to be adapted to specific context of schools in order to be
sustained in the long-term (Ertmer et al., 2012; Frank et al., 2011).
Despite the importance of bottom-up communication, research suggests that there is little
communication between central office administrators and school leaders and that these ties are
often top-down rather than reciprocated. These studies have also found low-performing schools
to be among the most isolated in district school reform networks, suggesting that central office
administrators are lacking the information needed to support these schools with more targeted
BUILDING NETWORKS FOR CHANGE
26
resources (Daly & Finnigan, 2012; Finnigan & Daly, 2010). While these findings suggest a
troubling absence of direct ties emanating from school leaders to the central office, what is less
known is whether frontline staff are acting as intermediaries who broker information from school
leaders and teachers to other central office leaders, administrators, and frontline staff. Earlier
studies suggest that even frontline staff are prone to buffering senior central administrators and
district leadership from school sites, preventing school demands from being addressed and
keeping innovative school practices at the margin of school systems (Honig, 2003, 2004;
Spillane & Burch, 2004). That said, these cases pre-date knowledge-intensive reform that
increasingly require central office administrators to incorporate insights from school
experimentation with instruction into their decision-making.
Organizational Conditions Related to Brokering
In sum, the extant literature has mainly focused on how frontline staff facilitate top-down
communication channels to align teacher practice with central office goals for instructional
change, and less so on their efforts to coordinate information within the central office, coordinate
information within and among schools, or engage in bottom-up communication from schools to
the central office. Moreover, the few studies that have explored these alternative brokering
practices point to discouraging results, implying that districts are failing to achieve instructional
coordination, coherence, and alignment on a system-wide basis. However, with knowledge-
intensive reform now placing heightened demands on districts to innovate teaching and learning,
frontline staff could be brokering information in ways that are different from what has been
documented in the extant literature.
While my first research question is concerned with describing how ed-tech coaches
broker information in this new policy context, my second and third research questions are
BUILDING NETWORKS FOR CHANGE
27
focused on understanding the organizational conditions (Box 3) that affect these brokering
practices and might lead to different brokering practices across reform topics (i.e., integrating
technology with instruction and teaching the CCSS). These questions have also been an
important area of inquiry for prior research, which suggests that power and influence, access to
expertise, and informal communication patterns such as dense ties, reciprocal exchange,
homophily, and closure are important organizational conditions that influence brokering.
Power and Influence
As frontline staff who occupy an ambiguous role in the top-down organizational
hierarchy of districts, ed-tech coaches need power and influence to broker information on
instructional reform (Coburn & Woulfin, 2012; Ho & Ng, 2017; Hopkins et al., 2016). Research
shows that central office and school actors are more likely to seek advice from colleagues in
more senior formal positions (e.g., a senior central office administrator relative to a mid-level
administrator, a school leader relative to a coach or teacher) (e.g., Daly, Moolenaar, Liou,
Tuytens, & Del Fresno, 2015; Liou, 2016; Spillane et al., 2015). This is because seniority serves
as a signal for expertise, leadership, and strong interpersonal skills (Wilhelm, Chen, Smith, &
Frank, 2016), and because districts and schools operate within an entrenched top-down
organizational culture that privileges individuals in formal positions of authority (Finnigan &
Daly, 2010; Liou, 2016).
As boundary spanners who work between the central office and schools, ed-tech coaches
do not occupy a formal position that gives them direct authority over others, instead drawing
heavily on their instructional background (e.g., personal attributes such as instructional expertise,
interpersonal skills, prior experience) and social environment (e.g., support from district and
school leadership, school norms for instructional change) to negotiate for influence over
BUILDING NETWORKS FOR CHANGE
28
instructional practice (Atteberry & Bryk, 2011; Bean et al., 2010; Ho & Ng, 2017; Marsh et al.,
2010; Matsumura et al., 2010). While there is growing evidence on individual and social
conditions that position frontline staff such as ed-tech coaches with influence over teacher
practice (which I discuss in greater length in chapter 7), researchers have mainly studied these
conditions from a top-down perspective in terms of how frontline staff facilitate instructional
coherence and alignment in schools. In contrast, there has been less research on how frontline
staff are positioned in districts to engage in other brokering practices to coordinate instruction on
a district-wide basis (i.e., coordinate information within the central office and among schools)
and build system coherence (i.e., bottom-up communication). Given that the success of
knowledge-intensive reform hinges on frontline staff engaging in a complete array of brokering
practices, it is important to understand the individual and social conditions that position these
staff to pursue this work in its entirety.
Access to Expertise
In addition to determining access to power and influence, the social positioning of ed-
tech coaches can also influence their access to expertise from other central office and school
actors in their district. Recent studies suggest that central office and school actors acquire
instructional expertise through interaction with colleagues on reform-related matters and that this
exposure is positively associated with improvements in teacher practice (Coburn et al., 2012;
Daly, 2010; Liou, 2016; Sun et al., 2013). Some of these studies even show that this informal
exposure to expertise is perhaps more valuable for sustaining instructional innovation than
conventional measures of expertise such as participation in formal professional development
(Coburn et al., 2012; Frank et al., 2011).
The idea of expertise residing in personal networks is highly relevant to brokering in
BUILDING NETWORKS FOR CHANGE
29
knowledge-intensive reform, where central office and school actors lack a priori expertise for
prescribing how teachers should be teaching in the classroom or how the central office and
schools should be supporting instructional redesign, and instead must experiment and actively
learn from one another (Durand et al., 2016; Herold, 2015; Superville, 2015). As discussed
earlier, one of the main responsibilities of frontline staff in these reform efforts, as brokers, is to
develop personal networks for curating expertise throughout their school systems and to then
share the insights learned from these networks with others. While access to expertise is an
important condition for brokering information in support of instructional change, researchers
have not examined the advice-seeking practices of frontline staff much less ed-tech coaches, and
how this is related to their power and influence in school systems.
Informal Networks of Communication
The informal organizational context of ed-tech coaches can also influence their
communication with central office and school actors. While researchers have described this
organizational context as informal communities of educators who share a common set of norms
and values for teaching and learning, another way of characterizing these communities is in
terms of how these norms and values are expressed in the structure of relationships between
educators (Bridwell-Mitchell & Cooc, 2016; Moolenaar, 2012). From this perspective,
communities are informal not because of the content of the norms and values that they share, but
because these norms and values result in educators engaging one another based on interpersonal
preferences and organization-wide communication patterns that are independent of formal role
and structure. This perspective is useful in studies aiming to describe the district-wide brokering
practices of frontline staff, where it might not be feasible to observe the norms and values of
different communities of educators in the district with nuance but where it can still be possible to
BUILDING NETWORKS FOR CHANGE
30
map out the structure of the social ties within and between these communities. In particular,
researchers have shown density of ties, reciprocal exchange, homophily and closure to be
relevant informal conditions that influence brokering and information exchange.
Density of ties. Density of ties refers to the proportion of all possible relationships that
exist between individuals in an organization (Scott, 2000), and is used in education research to
observe the level of social engagement in instructional reform (Bridwell-Mitchell & Cooc, 2016;
Daly, Moolenaar, Bolivar, & Burke, 2010). Dense networks move resources quickly (Adler &
Kwon, 2002; Lin, 2001; Nahapiet & Ghoshal, 1998), help to coordinate action between actors
(Obstfeld, 2005), support higher levels of organizational performance (Reagans & Zuckerman,
2001), and by virtue of doing so, are more likely to have individuals broker information than
sparse networks where these brokering opportunities are less utilized (Burt, 1992; Granovetter,
1973). In general, research suggests that dense district and school networks are associated with
professional trust, a shared understanding of reform goals, and support for educational change
(Daly et al., 2010; Finnigan & Daly, 2010; Moolenaar, 2012).
Reciprocal ties. While there is some evidence that dense networks are conducive for
brokering information, there is less empirical evidence on other informal conditions linked to
these dense structures, including reciprocal ties and closure (Scott, 2000; Valente, 2010). For
instance, researchers have long claimed that organizational norms that engender trust are
essential for instructional change (Bryk & Schneider, 2002; Bryk, Sebring, Allensworth,
Luppesco, & Easton, 2010). To see how trust shapes informal relations in districts and schools,
researchers have focused on the formation of reciprocal ties, defined as relationships in which
both actors engage one another as a source of support (Daly & Finnigan, 2012; Daly et al.,
2010). Reciprocated ties allow for predictability in social relations, help reduce individual
BUILDING NETWORKS FOR CHANGE
31
vulnerability, and ultimately deepen professional trust (Adler & Kwon, 2002; Nahapiet &
Ghoshal, 1998; Tsai & Ghoshal, 1998). In so doing, reciprocal ties also increase the depth of
exchange between actors and are ideal for communicating complex information and coordinating
action (Albrecht & Bach, 1997; Larson, 1992; Uzzi, 1997).
Based on this evidence, researchers have claimed that frontline staff are more capable of
brokering information in support of instructional reforms in districts and schools with high levels
of reciprocal exchange (Anderson et al., 2014; Atteberry & Bryk, 2011; Liou, 2016). However,
in taking a closer look at how trust influences information sharing on a district-wide basis, a
handful of studies show that these conditions can shield school communities from central office
leadership and reform efforts, leading to an absence of trust between the central office and
schools that can make it more challenging to broker information into or among schools
(Anderson et al., 2014; Matsumura et al., 2010).
10
This same tension applies to the reverse
scenario in terms of frontline staff shielding themselves and central office decision-making
processes from school insights and needs (Daly & Finnigan, 2012; Honig, 2004; Spillane &
Burch, 2004). Together, these findings suggest that researchers need to explore how trust and
reciprocal ties form on a district-wide basis, as well as in the space between the central office and
schools, to fully understand the brokering practices of frontline staff (Daly & Finnigan, 2012).
Fragmented networks: Homophily and closure. Research on social networks in
educational settings shows that central office administrators, school leaders, and teachers are
more likely to interact with colleagues who are like them – who match on gender or race, who
are physically proximate, and who share similar work roles and routines (e.g., working in the
same central office division, school type, or grade-level). Homophily engenders reliability, trust,
10
For instance, while most studies suggest that teacher collaboration is positively related to teacher participation in
coaching, Matsumara and colleagues (2010) find a negative relationship due to the presence of strong school culture
against the instructional reform associated with the work of coaches in schools.
BUILDING NETWORKS FOR CHANGE
32
and common understanding that, in turn, facilitates the sharing of tacit, non-routine, and complex
information (Borgatti & Foster, 2003; Ibarra, 1993). At the same time, however, homophilous
ties restrict information sharing to cliques of similar actors that can be challenging for brokers to
gain access to (Reagans & McEvily, 2003).
In larger groups, actor cliques develop through the formation of closed triad where actors
who are indirectly connected by a third party are also directly connected to each other (Bridwell-
Mitchell & Cooc, 2016; Valente, 2010). The accumulation of closed triads can create an efficient
social infrastructure for sharing and coordinating resources on instructional change, providing
the same advantages for instructional change as noted above for dense, reciprocal, and
homophilous ties (Hansen, 1999; Nahapiet & Ghoshal, 1998; Reagans & McEvily, 2003). At the
same time, however, closure can create a fragmented context with cliques of actors that reinforce
existing norms for instructional change, circulate redundant and over-exploited information
within these cliques, and shield actors from outside sources of influence (Coleman, 1988;
Hansen, 1999). This, in turn, can challenge frontline staff to broker information as an external
actor within and between these cliques.
Because homophily and closure are both imperative for instructional change but can give
rise to fragmented social structures that limit the reach of brokering ties and information
exchange, it is important to understand how frontline staff such as ed-tech coaches navigate this
organizational tension as part of their instructional leadership. Yet there has been little empirical
evidence on this topic to date, leaving researchers and policymakers in the dark on the extent to
which fragmented organizational settings facilitate and/or constrain the brokering practices of
frontline staff.
Contributions to the Extant Literature
BUILDING NETWORKS FOR CHANGE
33
As the above review of the literature suggests, ed-tech coach leadership can be enacted
through a range of brokering practices that fulfill different purposes in support of instructional
change (i.e., coordination, coherence, alignment). There are also a range of organizational
conditions that influence how these brokering practices take place. While informative, the extant
literature faces some limitations in terms of understanding how frontline staff such as ed-tech
coaches broker information in knowledge-intensive reform. Most notably, this literature is
somewhat outdated with regards to understanding the new policy context of knowledge-intensive
reform where frontline staff are expected to circulate information in all directions in school
systems to support systemic change. This systemic approach to brokering also requires
researchers to go beyond documenting if and how brokering takes place to assessing the extent to
which frontline staff are engaging in different brokering practices and the implications of this
variation for teacher learning, instructional change, and district-wide improvement. For example,
while the extant literature suggests that top-down communication is commonplace in districts
(Coburn & Woulfin, 2012; Daly & Finnigan, 2011), we do not know the full extent to which
these top-down communication channels are shaping information flows relative to other
brokering ties (e.g., within school or central office coordination and bottom-up communication).
As such, we cannot fully grasp the implications of these brokering practices for knowledge-
intensive reforms that hinge on system-wide information sharing.
Moreover, while the capacity of ed-tech coaches to broker information on a district-wide
basis clearly depends on their organizational context, researchers do not have a sophisticated
understanding of these contextual factors (Daly & Finnigan, 2016). Specifically, research
suggests that power and influence are important (Moolenaar et al., 2010), but we do not know if
ed-tech coaches possess the power and influence needed to influence instructional practice in
BUILDING NETWORKS FOR CHANGE
34
education technology reform, how this influence is shaped by the instructional background of
these coaches and their social environment, and how this influence enables and/or constraints the
full range of brokering practices that these coaches are expected to undertake. Similarly, while
access to expertise is a critical aspect of brokering (Gould & Fernandez, 1989) and an important
mechanism for supporting instructional change (Frank et al., 2011; Sun et al., 2013), researchers
have not studied the advice-seeking practices of ed-tech coaches and how this advice-seeking
relates to their influence as brokers. Finally, while studies suggest that information sharing in
districts and schools depends on informal communication patterns such as the density of ties,
reciprocal exchange, homophily, and closure (Bridwell-Mitchell & Cooc, 2016; Moolenaar,
2012), these informal conditions can have both positive or negative influences on brokering that
have not been examined in close detail.
Drawing on social network and qualitative methods, I build on the extant literature by
examining the systemic brokering practices of ed-tech coaches for supporting technology-
enabled and CCSS instruction and the organizational context surrounding these brokering
practices. In answering my first research question, I use social network methods to provide
evidence on the extent to which ed-tech coaches are engaging in different brokering practices
that contribute to instructional change (e.g., within central office or school coordination, top-
down or bottom-up communication). This allows me to observe the extent to which ed-tech
coaches are brokering information in different directions in their school systems and, based on
this evidence, to make inferences on the emphasis of their instructional leadership (i.e.,
coordination, coherence, and/or alignment) and the implications of this leadership for teacher
learning and systemic change.
My second research question then focuses on how these brokering practices are informed
BUILDING NETWORKS FOR CHANGE
35
by the organizational context of districts. Specifically, I use social network methods to describe
the organizational context of ed-tech coaches in terms of their power and influence, access to
expertise, and exposure to informal communication patterns such as dense relations, reciprocal
exchange, homophily, and closure, and to assess how these organizational conditions inform the
system-wide brokering practices of these coaches. My third research question then examines if
the brokering practices of ed-tech coaches and their organizational context, as apparent in my
data and analysis for research questions 1 and 2, differ across reform efforts and the reasons for
these differences. In so doing, I shed light on how the relationship between organizational
context and brokering might differ across knowledge-intensive reforms that district leaders
increasingly expect to be interrelated in practice. For both research questions 2 and 3, I draw on
qualitative data and methods to further unpack the organizational surroundings of ed-tech
coaches and substantive differences between technology-enabled and CCSS instruction that
might lead to different brokering practices across these reforms.
BUILDING NETWORKS FOR CHANGE
36
CHAPTER THREE -- Social Network Theory as a Lens for Brokering
Introduction
I draw on social network theory to translate the brokering practices of ed-tech coaches
and their organizational conditions into a social structure that can be readily observed and
studied (Figure 1, Box 4). Social network theory provides a comprehensive set of properties for
observing the extent to which brokering practices contribute to information flows on
instructional reform in districts and schools. Social network theory also offers a coherent set of
properties for observing the organizational conditions that influence brokering in the social
structure of districts and schools Finally, social network theory assumes that social structures can
vary based on the purpose of relationships and hence the topic of instructional reform. This
perspective adds more nuance to our understanding of how organizational context can influence
brokering by allowing researchers to observe how social structures vary across different albeit
interrelated reform efforts (i.e., from integrating technology with instruction to teaching the
CCSS).
In what follows, I introduce social network theory and discuss these theoretical
advantages in more detail. As part of this discussion, I outline several hypotheses on the extent to
which ed-tech coaches are engaging in different brokering practices to support instructional
change (research question 1) and how their organizational context might be influencing these
brokering practices (research question 2).
Review of Social Network Theory
Social network theory draws on the concept of social capital, which suggests that
individuals access social resources through relationships with others to achieve individual and/or
collective outcomes (Lin, 2001). While many theorists have foregrounded the cultural,
BUILDING NETWORKS FOR CHANGE
37
psychological, and network aspects of social capital (Bourdieu, 1986; Coleman, 1988; Portes,
2000; Putnam, 1993), social network theorists examine how properties of individual
relationships or ties (“dyads”) and the broader structure (“networks”) that these ties create,
facilitate social capital transactions and access to resources (Daly, 2010; Lin, 2001).
When describing these social structures, social network theorists focus on properties
pertaining to both formal organizational structures (e.g., the formal position, department
assignment, physical location of actors) and informal organizational arrangements (e.g.,
individual personal preferences to interact with colleagues, organizational norms for
communication) (Borgatti & Ofem, 2010; Spillane, Kim, & Frank, 2012). As such, social
network theorists do not make a priori assumptions on the importance of formal positions,
routines, and structures in shaping communication patterns in organizations. Rather, social
network theorists map out all possible relations and then examine the properties of these ties and
social structures to identify both formal and informal determinants of social capital. Moreover,
social network theory does not presume that these properties have a uniform relationship with
processes of organizational change, but rather keeps an open perspective with regards to
supportive and/or constraining conditions associated with these properties (Daly, 2010; Nahapiet
& Ghoshal, 1998; Penuel, Riel, Krause, & Frank, 2009).
Another central concept of social network theory is that of “social embeddedness”
(Nahapiet & Ghoshal, 1998; Scott, 2000), which suggests that individuals and their immediate
relationships are nested in larger, nested structures that determine their access to social resources
(Daly & Finnigan, 2012). Social embeddedness implies that changes in individual relationships
can lead to higher-order changes in social structures (Scott, 2000). In other words, the
interpersonal conditions that influence whether two actors socially engage one another both
BUILDING NETWORKS FOR CHANGE
38
inform, and are informed by, their broader social structure. This logic implies that dyadic
relations represent more than just two actors connected to one another, but rather an entire
system of relationships (Burt, 2000).
Finally, social network theory assumes that social structures are dynamic and vary based
on the purpose of relationships and the content being exchanged (Daly, 2010; Ibarra, 1993;
Moolenaar, 2012). Education researchers have used social network theory to distinguish
between: (a) social structures consisting of instrumental relationships focused on work-related
issues and achieving district and school goals, and (b) social structures consisting of expressive
relationships focused on the personal interests of school actors (e.g., networks of friendship or
personal guidance). Within these relationship categories, education researchers have mapped out
social structures across different types of instrumental or expressive relationships such as social
networks focused on different aspects of teacher practice (e.g., lesson planning, instructional
reform knowledge) (Daly et al., 2010) and social networks focused on different aspects of school
restructuring in high-stakes accountability regimes (Finnigan & Daly, 2010).
The principles of social network theory offer several advantages for studying the
brokering practices of ed-tech coaches. First, by giving equal importance to all possible
relationships in school systems, social network theory makes it possible to observe if social ties
exist between all possible combinations of central office and school actors and the extent to
which these ties contribute to prominent brokering patterns in the social structure of districts and
schools (e.g., coordination, top and bottom-up communication). Second, based on ideas of social
embeddedness and that both formal and informal conditions contribute to information flows in
organizations, social network theory offers a cohesive set of properties for observing the
complex structure of these conditions and their influence on tie formation. Third, in considering
BUILDING NETWORKS FOR CHANGE
39
how social structures vary based on the topic and related goals of content that is being
exchanged, social network theory provides a lens for observing how brokering and
organizational context might vary across reforms where central office and school actors are
engaging one another for different purposes. I discuss each of these theoretical advantages in
more detail below.
Brokering Properties from Social Network Theory
Social network theory offers a detailed schema for describing brokering relations based
on actors’ group affiliation and their respective goals and interests (Gould & Fernandez, 1989).
These brokering properties are rooted in the idea that brokers bridge structural holes to unleash
social capital in networks (Burt, 2000). Structural holes refer to the absence of communication
between groups or sub-groups of actors in a network. Because actors on either side of these
structural holes circulate in different flows of information and have access to different social
capital, bridging these structural holes can provide access to new, non-redundant sources of
information and in so doing, improve the overall capacity of individuals and organizations.
Social network theory conceptualizes brokering as a two-path relationship in which a broker
mediates communication between two disconnected actors by seeking advice from one of them
and being approached for advice from the other.
The benefits that arise from bridging structural holes varies based on the group affiliation
of actors involved. Given the pertinence of formal roles and structure in shaping communication
in education settings (Finnigan & Daly, 2010; Liou, 2016; Spillane et al., 2012), I distinguish
actors based on their relative positioning in the organizational hierarchy of districts. Specifically,
I distinguish between administrators and frontline staff who work in or from the central office
(i.e., senior and mid-level administrators, ed-tech coaches, and other frontline staff) from
BUILDING NETWORKS FOR CHANGE
40
principals and teachers who work in schools. Based on this classification scheme, I use social
network theory to observe six distinct brokering properties that map onto brokering practices for
sharing information within and among the central office and schools as reviewed earlier. These
properties can be used to describe communication patterns across an entire social network, as
well as the specific communication patterns of specific groups of actors. For my research, I
formulate a series of research hypotheses on these brokering properties as observed for ed-tech
coaches in their district’s social network.
Table 1 summarizes the six brokering properties I observe in my data (Gould &
Fernandez, 1989). Betweenness, measures the total number of two-path ties where an actor is
situated between two other actors, indicating the total brokering activity of an actor in a
network.
11
The remaining brokering properties deconstruct betweenness into specific brokering
relations. Coordination involves brokering information between actors with the same group
affiliation to support the goals of the group. This property is equivalent to ed-tech coaches
coordinating information within the central office to develop a comprehensive vision for
technology-enabled and CCSS instruction and the appropriate procedures and resources for
supporting this vision. Itinerant coordination involves brokering information between actors
within the same group. The broker, however, does not share in this group identity and instead
works as an outsider who mediates information sharing between two group insiders to support
their goals. This property is equivalent to ed-tech coaches coordinating information within
schools to make central office directives coherent with school goals and strategies for
instructional improvement and to align teacher practice with joint vision for instructional change
as shared between the central office and schools. Liasing involves a scenario where the broker,
11
Gould and Fernandez’s (1989) betweenness measure is the maximum score for betweenness centrality, which is
the total number of times a node lies on the shortest path connecting two actors (Freeman, 1978). In my dissertation,
I refer to this measure as betweenness to distinguish it from betweenness centrality.
BUILDING NETWORKS FOR CHANGE
41
as well as the two actors that he/she is mediating between, are all affiliated with different groups
and are working to build connections between groups that do not have any prior allegiance to one
another. This property captures the expected role of ed-tech coaches to build interschool
networks that share best practices across schools in their district, further enabling processes of
instructional alignment and school coherence.
The last two properties describe brokering practices that support vertical channels of
communication between the central office and schools in districts. Representation, refers to when
the broker and the actor who is sharing information are from the same group, while the actor who
is receiving this information is an outsider from another group. In this case, the broker acts as a
representative who must negotiate how insights from his/her group are shared with outsiders.
Representation can be understood as ed-tech coaches engaging in top-down communication
channels to inform school actors about central office expectations for instructional improvement,
develop school programs that are coherent with these expectations, and to align teacher practice
with these goals. Gatekeeping involves a scenario where the broker and the actor who is
receiving information are from the same group while the actor who is sharing information is an
outsider from another group. In this case, the broker acts as a gatekeeper who can decide whether
to grant an outsider access to his/her group, as well has how to present information from this
outsider to his/her group members. Gatekeeping is equivalent to ed-tech coaches acting as
intermediaries who facilitate bottom-up communication channels from schools to the central
office to make central office procedures and resources more responsive to school needs.
The above properties provide a comprehensive framework for observing essential
brokering practices for instructional change and assessing the extent to which these brokering
practices contribute to information flows in districts. Because ed-tech coaches are expected to
BUILDING NETWORKS FOR CHANGE
42
engage in all these brokering practices to support systemic improvements in teaching and
learning, I hypothesize that:
Hypothesis 1: Ed-tech coaches are prominent brokers (i.e., have high-levels
of betweenness and brokering relations for coordination, itinerant
coordination, liasing, representation and gatekeeping) for (a) integrating
technology with instruction and (b) teaching the CCSS.
Next, I describe how social network theory can be used to observe organizational conditions that
facilitate and/or constrain the brokering practices of ed-tech coaches as they move throughout
their school systems.
Social Structure Properties from Social Network Theory
Social network theory provides a coherent set of properties for observing the broader
social structure in which brokering ties are embedded. As mentioned earlier, these properties are
based on the idea that informal communication patterns can be just as influential as formal role
and structure in providing access to social capital. More importantly, these properties are
grounded in the idea of social embeddedness, that individuals’ access to social capital is
determined by their immediate relationships with others and the broader structure that these
relationships create. The properties of this broader social structure include how actors are
positioned in their network to advise (in-degree centrality) and seek advice from others (out-
degree centrality), and more diffuse, network-wide communication patterns such as overall
levels of communication and social engagement in networks (network density), norms for
reciprocating advice and support (network reciprocity), and the tendency for actors to
communicate in cliques (homophily and network closure).
In-degree centrality. Opinion leadership refers to an individual’s ability to exert
considerable power or influence over how information is circulated within his/her organization
regardless of formal position. Accordingly, opinion leaders have the power to create (or hinder)
BUILDING NETWORKS FOR CHANGE
43
new linkages and access to resources that, in turn, may enhance (or detract from) social capital,
organizational capacity for change, and the uptake of new innovations (Rogers, 2003; Stuart,
1998; Tsai, 2000). Social network theorists measure opinion leadership in terms of in-degree
centrality, which is the total number of ties that an individual receives from other actors in
his/her network (Freeman, 1978; Valente, 2010).
12
Analysis of social networks further suggests
that in-degree centrality is strongly correlated with betweenness centrality as a measure of
individual control of information flows (Freeman, 1978). In light of this evidence, the in-degree
centrality of frontline staff can be interpreted as their ability to redistribute authority over
instruction within the top-down organizational hierarchy of districts to broker information for
instructional change. Based on this idea, I hypothesize that:
Hypothesis 2: Ed-tech coaches with power and influence (i.e., high-levels of in-
degree centrality) are more capable at brokering information on (a) integrating
technology with instruction, and (b) teaching the CCSS.
Out-degree centrality. Out-degree centrality, which counts the total number of actors
that an individual nominates in his/her network as a source of information or advice, is a social
network measure that captures advice-seeking interactions and opportunities for individuals to
leverage expertise in their immediate surroundings (Scott, 2000). While out-degree is
conventionally thought of as signaling an individual’s dependency on others and lack of control
of information flows in his/her network (Burt, 2000; Daly & Finnigan, 2012; Snijders, van de
Bunt, & Steglich, 2010), applications of out-degree centrality in education research suggest that
it can measure an individual’s efforts to accumulate expertise and social status in districts and
schools (Coburn & Russell, 2008; Frank et al., 2011; Jabbar, 2015; Penuel et al., 2012; Sun et al.,
2013). These findings are also consistent with the fundamental idea of brokering, which involves
12
Social network theorists have also observed opinion leadership based on a measure of betweenness, which refers
to the total number of times that an individual is positioned in between two actors as observed using brokering
properties described earlier (e.g., coordination, itinerant coordination, representation, gatekeeping).
BUILDING NETWORKS FOR CHANGE
44
frontline staff developing personal networks of their own and then sharing insights learned from
these personal networks with others (Gould & Fernandez, 1989). Moreover, seeing as frontline
staff are expected to circulate information on a district-wide basis in knowledge-intensive
reform, it is critical that their personal networks include multiple actors from diverse positions
and perspectives in their school systems to build credibility as a source of expertise. I
hypothesize that:
Hypothesis 3: Ed-tech coaches with access to expertise (i.e., high-levels of out-
degree centrality) are more capable of brokering information on (a) integrating
technology with instruction, and (b) teaching the CCSS.
Network density and reciprocity. Social network theory offers several properties to
observe norms of trust and collaboration in organizational settings that are important for
instructional innovation. The first of these properties, network density, refers to the proportion of
ties that exist among all possible connections between actors in a network (Scott, 2000). As
noted earlier, research suggests that organizations with dense networks are more effective at
spreading and coordinating information (Adler & Kwon, 2002; Lin, 2001; Nahapiet & Ghoshal,
1998; Obstfeld, 2005) and, by virtue of doing so, should have fewer structural holes and enabling
of brokering (Burt, 1992; Granovetter, 1973). Research also suggests that dense networks are
associated with more reciprocal ties which, as noted earlier, create a sense of interdependence,
trust, and flexibility in organizational settings that helps minimize risk and support complex
processes of change (Albrecht & Bach, 1997; Larson, 1992; Uzzi, 1997). Network reciprocity,
the proportion of all ties in a network that are reciprocated, accounts for the prevalence of these
reciprocal ties in organizational settings. Based on these properties, I hypothesize that:
Hypothesis 4: Ed-tech coaches are more capable of brokering information on
(a) integrating technology with instruction and (b) teaching the CCSS in dense
and reciprocated networks.
BUILDING NETWORKS FOR CHANGE
45
While most studies suggest that dense and reciprocated networks are positive conditions
for organizational change, there is some evidence to suggest that these properties work against
information sharing and brokering specifically. As noted earlier, research in both education (e.g.,
Moolenaar, 2012; Penuel et al., 2009) and non-education settings (e.g., Coleman, 1988; Reagans
& McEvily, 2003) show that dense and/or reciprocated networks can consist of cliques that
shield actors from external sources of influence, reinforce existing norms of change, and
circulate redundant and over-exploited information (Hansen, 1999). These contrary findings
suggest that, in addition to the overall density and reciprocity of networks, it is important to
account for the formation of cliques within networks.
Homophily and closure. As discussed in chapter 2, fragmented social structures,
consisting of homophilous ties and closure, can support communication within actor groups but
hinder communication between groups (Hansen, 1999; Reagans & McEvily, 2003). Social
network theorists observe fragmented communication through dyadic measures of homophily
(i.e., pair-wise indicators for whether two actors are more likely to send/receive social ties
to/from one another based on sharing the same attributes) and network clustering, which refers to
the proportion of all triads in a network that are closed through transitive ties between all three
actors. Given that fragmented communication can make it challenging for frontline staff to
engage in brokering ties that bridge distinct actor groups (e.g., itinerant coordination, liasing,
representation, and gatekeeping), I hypothesize that:
Hypothesis 5: Ed-tech coaches are less capable of brokering information on (a)
integrating technology with instruction and (b) teaching the CCSS in fragmented
networks with actor cliques (i.e., homophilous ties and network clustering).
Dynamics of Social Structures
Social network theory adds further nuance to our understanding of organizational context
BUILDING NETWORKS FOR CHANGE
46
and brokering by recognizing that social structures vary based on the content exchanged and
overall purpose of relationships. This is an advantage that only a few education researchers have
exploited when describing processes of instructional reform. For example, in examining the role
of teachers’ social networks in a district-wide reform focused on literacy instruction, Daly and
colleagues (2010) find significant variation both within and between schools across different
reform-related networks (i.e., networks focused on lesson planning, knowledge of reading
comprehension, and implementation of reform efforts). Based on this evidence, the authors argue
that the differences across reform-related networks can have varied implications for the uptake,
depth, and spread of change efforts in schools. In an analysis of leadership networks of school
restructuring under the No Child Left Behind Act (NCLB), Finnigan and Daly (2010) also find
similarities and differences in the flow of organizational resources among and between central
office and school leaders for different aspects of school restructuring (e.g., program
improvement, sharing best practices, instructional innovation). The authors argue that, to support
system-wide gains in school performance, district leaders need to be aware of how information is
being communicated across interrelated aspects of school restructuring and improvement.
13
As the above findings suggest, education research on instructional reform, especially
studies focused on how frontline staff broker information to support interrelated knowledge-
intensive reforms, can benefit from understanding how the social structures of districts vary
across reform efforts. By providing a lens to observe these differences, social network theory
offers a more nuanced understanding of how frontline staff broker information in knowledge-
13
The authors find sparse and non-reciprocated networks across all topics of school restructuring as well as minimal
evidence of vertical connections between the central office and schools, all of which make it more challenging for
restructuring schools to access the expertise needed to improve school performance. The authors also find evidence
of school leaders who are opinion leaders for innovating school practices as being marginal actors in leadership
networks focused on other aspects of school restructuring, thus suggesting that most central office and school
leaders are not leveraging the advice of these innovative school leaders to transform other aspects of their work
related to school restructuring.
BUILDING NETWORKS FOR CHANGE
47
intensive reform and the influence of organizational context on these brokering practices.
Limitations of Social Network Theory
In this chapter, I argue that social network theory offers several advantages for
understanding how ed-tech coaches broker information in knowledge-intensive reform. These
contributions include providing: (a) brokering properties to assess the extent to which the
different brokering practices expected of ed-tech coaches shape information flows in districts; (b)
cohesive social network properties for observing the organizational context of ed-tech coaches
and how these conditions facilitate and/or constrain their brokering behavior across school
systems; and (c) a lens for observing how social structures and hence brokering patterns and
organizational context vary across interrelated, knowledge-intensive reforms.
While these contributions are substantial, there are limitations to social network theory
that make this theoretical approach non-ideal for advancing other aspects of the extant literature
on brokering and instructional change. First, because social network theory places a strong
emphasis on understanding the influence of social structure on access to social capital and
organizational change, this theory is not well-suited to examine individual attributes and sense-
making processes that affect brokering (Daly, 2010). In the case of ed-tech coaching, social
network theory does not elucidate how the instructional background of coaches, or the individual
attributes of other district and school actors, position coaches with power and influence to affect
instructional change on a district-wide basis. In addition, social network theory does not provide
a framework for observing the content, activities, and underlying dynamics of relationships in
social structures (Coburn et al., 2012; Daly, 2010). As such, a social network perspective does
not describe these social exchanges with enough nuance to capture norms, values, and other
aspects of social engagement that can influence instructional change. For similar reasons, social
BUILDING NETWORKS FOR CHANGE
48
network theory does not shed light on bridging, buffering, and social learning routines that ed-
tech coaches might use to facilitate instructional coordination, coherence, and alignment.
Given these limitations, I use both social network theory and findings from the extant
literature to guide my analysis of the brokering practice of ed-tech coaches. As shown in Figure
1, my conceptual framework makes specific connections between the instructional leadership
(Box 1) and brokering practices (Box 2) of ed-tech coaches. I then situate these leadership and
brokering practices in a broader organizational context that includes conditions such as power
and influence, access to expertise, and informal communication patterns such as density of ties,
reciprocal exchange, homophily, and closure (Box 3). I also show that the social structure of
districts (Box 4) can mediate the effects of these organizational conditions on brokering,
representing a critical focal point of my study. Based on this framework, I develop a series of
hypotheses on: (a) the extent to which ed-tech coaches engage in brokering practices that are
central to their instructional leadership (research question 1), and (b) how the above
organizational conditions facilitate and/or constrain these brokering practices as mediated by
social structure (research question 2). As I discuss in chapter 5, I use social network data and
methods to investigate these hypotheses and answer my research questions. However, because a
focus on social structure alone does not capture all aspects of brokering and the organizational
context of districts, I also draw on empirical findings from the extant literature to inform my
analysis. In chapter 7, I expand on certain elements of this conceptual framework (highlighted in
red) to guide my qualitative data collection and analysis for research questions 2 and 3.
BUILDING NETWORKS FOR CHANGE
49
CHAPTER FOUR --LUSD’s Digital Coaching Program
Introduction
I investigate my research questions in LUSD, a mid-sized district that passed a
multimillion dollar bond measure to modernize schools with iPads, laptops, software
applications, and an online content management system. LUSD provides an ideal setting for
studying the instructional leadership of ed-tech coaches. First, LUSD is generalizable to other
districts in the nation. It is a mid-sized school district with a diverse student population,
including a sizeable share of low-income (approximately 45% of students) and English language
learner students (approximately 23% of students), located in the southwestern region of the
United States. Like other districts nationwide, LUSD has had to reinvent its educational practices
to better address the educational needs of its changing student population. Moreover, LUSD has
pursued this transformation within a relatively thin and simple organizational structure (unlike
the complex bureaucracies of larger urban districts), making it feasible for ed-tech coaches to
engage individuals throughout the district, build networks for instructional change, and broker
information as needed.
Second, LUSD is ahead of most districts in completing a technology rollout to fulfill
teaching and learning objectives of the CCSS. Looking to capitalize on a gap year in standards-
based testing and school accountability arising from its state’s transition to the CCSS, LUSD
leaders invested in a sweep of technology tools to modernize instruction and assessment
practices. In 2012, local voters approved a bond measure that raised $135 million over a 30-year
period to upgrade the district’s technological infrastructure to support STEM education, college
and career preparation, and CCSS instruction. In the 2013-14 school year, the district provided
students three-to-one access to iPads from kindergarten to third grade and one-to-one access to
BUILDING NETWORKS FOR CHANGE
50
iPads from fourth to eight-grade. In the 2014-15 school year, the district provided students one-
to-one access to Toshiba laptops in high school. These digital devices were all pre-loaded with
educational software apps.
Theory of Action for LUSD’s Technology Rollout
The main goal for this investment in school technology was to modernize educator
practices throughout the district. In his messaging to schools, the district Superintendent
described this modernization as teachers “teaching rigorous standards through engaging
strategies and support from technology.” As this phrase implies, district leaders expected
teachers to embed technology throughout the curriculum of instruction to support student
learning. My conversations with district leaders also revealed a teaching philosophy grounded in
principles of personalized and inquiry-based instruction, including expectations for teachers to
tailor instruction to individual student needs and facilitate authentic learning experiences for
students to practice 21
st
century learning skills such as critical thinking, collaboration,
communication and creativity (referred to as the “4Cs”). As shown in Figure 2, district leaders
expected these student-centric teaching practices to improve student engagement in learning,
content knowledge and 21
st
century learning skills (i.e., the 4Cs). District leaders expected to
measure these improvements in student learning using student performance on the CCSS
assessments, which district leaders viewed as a valid assessment for evaluating student’s college
and career readiness, and through monitoring the achievement gap between Title 1 (i.e., low-
income, predominantly English language learner schools) and non-Title 1 schools.
To achieve these goals, central office administrators in LUSD established certain
structures, procedures, and resources, including modernizing school technology so that teachers
and students could use new digital tools and apps, improved Internet connectivity, ad high-
BUILDING NETWORKS FOR CHANGE
51
quality technical support in the classroom. The district also rolled out an online learning content
management system (called “Haiku”) that principals, teachers, and students were expected to use
for all facets of school management and instruction, from communicating with parents, to
assigning student work and logging student grades, to assigning student work, to looking up
resources for lesson planning. Alongside these investments in school technology, the district
adopted “signature practices” for guiding teacher instruction. These signature practices were
expected to guide teachers’ use of technology in the classroom (the SAMR scale), and
instruction in core subject areas of the Common Core such as English language arts (Balanced
Literacy, adopted district-wide in 2014-15 for elementary schools) and math (Cognitively
Guided instruction or “CGI”, adopted district-wide in 2015-16), all focusing on different aspects
of personalized and inquiry-based teaching.
14
Balanced Literacy is a differentiated instructional program for developing the reading and
writing skills for each individual student. The program is implemented through the Reader’s and
Writer’s workshop models, which include a combination of teacher modeling, small group
instruction, and independent student work to build student proficiency in vocabulary, reading,
and writing. Teachers are supposed to use the Reader’s and Writer’s workshop models to guide
student learning in all subject areas. Balanced Literacy does not prescribe reading content,
curriculum scripts, or pacing guides for teachers, placing teachers in charge of selecting content
and developing lesson plans according to classroom needs.
15
14
While LUSD planned to adopt signature practices to guide instruction across all school-levels, I focus on the
signature practices adopted at the elementary school level since this is where the district was focusing these efforts
at first.
15
CGI incorporates similar principles of personalized and inquiry-based instruction. However, because LUSD only
adopted CGI as a signature practice in the 2015-16 school year, I do not go into much detail on this signature
practice.
BUILDING NETWORKS FOR CHANGE
52
While Balanced Literacy focuses on personalizing instruction to student needs, the
SAMR scale focuses more on inquiry-based teaching practices. The scale rates teachers’ use of
technology according to a four-level rubric, ranging from Substitution to Augmentation to
Modification to Redefinition. Each level of this scale describes teachers’ progress in using
technology to create more autonomous, interest-driven learning experiences in which students
create original products of knowledge and practice 21
st
century skills. As teachers progress along
the SAMR scale, they are providing students with more enriched and engaged learning
experiences. For instance, substitution refers to teachers using technology to replace pen-and-
paper tasks that do not build on student engagement in learning, whereas redefinition refers to
teachers using technology to facilitate transformative learning experiences that could not be
possible without technology and that support student self-expression, collaboration, and
authenticity in learning.
As Figure 2 shows, district leaders expected improved technology access in schools and
adoption of these signature practices to enable teachers to engage students in personalized and
inquiry-based instruction that improves student learning. However, to ensure that teachers made
effective use of these technology resources and signature practices, district leaders also invested
in teacher PD as a critical leverage point for instructional change. While the district pursued
several PD strategies including district-wide and school trainings on technology-enabled and
CCSS instruction, as well as using Haiku to provide principals and teachers with instructional
resources, LUSD’s Digital Coaching Program (DCP) was arguably most central to these efforts.
Instead of being a one-shot effort to improve teacher practice, district leaders intended for the
DCP to build a social infrastructure that would allow the central office and schools to constantly
improve teaching and learning on a district-wide basis. Below I describe the DCP in more detail,
BUILDING NETWORKS FOR CHANGE
53
showing that the responsibilities of ed-tech coaches for developing this social infrastructure map
onto the same instructional leadership and brokering practices that have been used to
characterize the work of ed-tech coaches nationally.
Digital Coaching Program
In the 2014-15 school year, the district recruited 15 expert teachers (digital learning
coaches or “DLCs”) to work as ed-tech coaches and provide content-focused, situated, sustained,
and collaborative PD to teachers across schools in the district.
16,17
Teachers were recruited to
work as DLCs based on their prior track-record in teaching, use of technology for instruction,
and demonstrated interpersonal skills for working in adult education. District leaders inferred
these qualities and made their final selection of DLCs through a formal application process in
which interested teachers described their strengths across these areas and delivered a sample
lesson in front of a panel of interviewers. Based on opinions of central office administrators and
school principals as to who might be qualified candidates, district leaders also invited certain
teachers to apply for DLC positions and participate in the interview process.
Background of DLCs. As shown in Table 3, LUSD recruited DLCs with a range of
grade-level and subject area expertise to match the demographics and instructional assignments
of teachers in the district. On average, DLCs were comparable to other teachers in terms of
16
In my analysis of LUSD’s implementation of its technology rollout, I refer to these ed-tech coaches as DLCs
since this is how they were identified in the district.
17
While LUSD first distributed iPads to elementary and middle schools in the 2013-14 school year and had
assigned DLCs to schools to support this rollout, the district experienced major technical difficulties and delays, as
is common with most district one-to-one computing programs (e.g., Bingham, Pane, Steiner, & Hamilton, 2016;
Zheng, Warschauer, Lin, & Chang, 2016), such that teachers and students did not receive these devices until the
spring of 2014. I focus on 2014-15 as the first time when teachers and students had access to technology throughout
the school year and when the district had resolved most technical issues related with providing one-to-one
computing access in schools (e.g., sufficient broadband capacity, efficient processes for distributing and collecting
devices). All DLCs (except for one coach at the high school level) were hired as teachers on special assignment and
did not continue teaching students in the classroom. Five DLCs worked with teachers across subject areas at the
elementary school-level. The remaining 10 DLCs worked at the middle and high school-levels, each specializing in
a subject area of instruction.
BUILDING NETWORKS FOR CHANGE
54
gender and even more racially diverse than LUSD’s teaching population. While fewer DLCs had
teaching experience in high school grade-levels or in secondary social science courses than
LUSD’s teaching staff, these coaches were otherwise comparable to LUSD teachers in all other
grade-levels and subject areas of instruction. DLCs also had fewer years of teaching experience
and higher levels of educational attainment than other teachers in the district, which is consistent
with how ed-tech coaches across the country have been described (Flanigan, 2016).
While district leaders ultimately intended for DLCs to help teachers use technology to
support CCSS instruction, district leaders did not select DLCs based on their prior knowledge of
these content standards and the district’s signature practices. This is because, as one senior
administrator explained, LUSD was in the process of adopting its signature practices and
developing curriculum frameworks aligned to the CCSS at the same time of its technology
rollout. Given these shifting pieces, district leaders prioritized selecting DLCs whom they
perceived as having strong self-efficacy for instruction and who could learn on their feet to make
technology relevant across different content standards, curriculum materials, and pedagogical
frameworks.
18
This assumption once again is consistent with how other districts nationwide have
described their selection and recruitment strategies for ed-tech coaches, given that these coaches
are often expected to work in a range of instructional settings (Flanigan, 2016; Herold, 2015).
Instructional leadership and brokering of DLCs. As shown in Figure 2, DLCs were
expected to support instructional coordination, coherence, and alignment (Box 1) through their
provision of teacher PD and to engage in brokering practices related to these leadership
18
The one exception is that district leaders had decided to rollout Balanced Literacy and Readers and Writer’s
Workshop as CCSS-aligned curricular frameworks for English Language Arts instruction in elementary schools for
the 2014-15 school year. However, district leaders still did not recruit DLCs based on their expertise in these
curriculum frameworks and rather assumed that, because these frameworks had been piloted in schools across the
district, that most DLCs would have had some exposure to these frameworks based on their prior teaching
experience.
BUILDING NETWORKS FOR CHANGE
55
practices. Specifically, DLCs were expected to coordinate information among themselves and
other central office administrators and staff (Box 2A) to position technology as a central resource
for supporting signature practices for CCSS instruction (Box 3A). This communication was
supposed to be reciprocated and collaborative in nature, meaning that DLCs were expected to
seek guidance to inform their work with teachers, and to offer their expertise to guide other
DLCs, senior administrators, and staff working in other instructional divisions from the central
office. Because of the close link between technology-enabled and CCSS instruction, district
leaders focused explicitly on building connections between DLCs and LUSD’s Curriculum and
Instruction Administrator (CIA) and curriculum coordinators (CCs) who, at the time, were
developing curriculum frameworks and teaching guides to help teachers implement signature
practices such as Balanced Literacy.
19
To facilitate this collaboration, DLCs participated in
weekly meetings with LUSD’s Coordinator of Education Technology (CET), a senior central
office administrator responsible for supervising DLCs and other instructional procedures and
resources involving technology, and in monthly meetings with LUSD’s CIA and CCs.
DLCs were also responsible for brokering information in a top-down manner (Box 2B) to
provide one-on-one instructional support to a group of eight to ten voluntary teachers per coach
(these teachers were referred to as “Digital Fellows” or DFs).
20
In total, LUSD trained 102 DFs
in elementary and middle schools in the 2013-14 school year, and another 126 DFs across all
school levels in the 2014-15 school year. DLCs met with DFs in weekly coaching cycles focused
on classroom work and reflective discussions on teacher practice. These weekly routines were
adopted from another coaching program that the district had implemented for several years
19
For instance, CCs were responsible for linking specific standards from the CCSS to grade-level or subject area
curricular resources and developing teacher PD on how to use these resources as requested by schools.
20
DFs were recruited through a voluntary process. First, teachers applied outlining their teaching goals and prior
experience with education technology. LUSD’s CET then reviewed these applications and made the final selection
of DFs while also seeking input from school principals and DLCs.
BUILDING NETWORKS FOR CHANGE
56
through a partnership with an education-focused philanthropic foundation.
21
Not surprisingly,
DLC weekly meetings with their DFs included many of the bridging strategies and social
learning routines as highlighted in the literature on instructional coaching (e.g., Huguet, Marsh,
& Farrell, 2014; Kopcha, 2012; Mudzimiri et al., 2014), which I discuss in greater detail below.
DLCs and DFs met at the beginning of the school year to discuss each DF’s instructional
goals and interests in using technology to support these goals. Under the direction of this initial
conversation, DLCs then met with DFs on a weekly basis to guide and observe DFs in
implementing instructional tasks related to their goals. These weekly meetings involved: (a) pre-
briefings at the start of the school day to discuss the immediate tasks that DFs would be focusing
on and the role of the DLC in supporting this work; (b) joint work in the classroom that included
a range of activities, from DLCs observing DFs to co-teaching with DFs to modeling instruction
for DFs to observe to directly working with students; and (c) post-briefings in which DLCs and
DFs would reflect on the day’s work and plan for their next meeting. Through these weekly
coaching cycles, DLCs were expected to build the knowledge, beliefs and motivation of their
DFs for using technology to teach the CCSS in ways that would address the specific needs of
their students (Box 3B), thereby aligning teacher practice with central office goals for
instructional change and building coherence between these central office goals and school goals
and strategies for improving student learning. For example, district leaders expected DFs to
develop technological, pedagogical, and content knowledge for embedding technology
throughout the curriculum of instruction; to embrace student-centric views of teaching and
learning that advance their practice further along the SAMR scale and allow them to improve
21
I do not name this foundation to preserve the anonymity of my research site. However, the coaching program
supported by this foundation was intended to help talented educators excel in their instructional practice through
intensive mentoring that took place in weekly coaching cycles. This mentoring program is available to all schools in
LUSD, with individual teachers having to apply to participate.
BUILDING NETWORKS FOR CHANGE
57
student learning in their classrooms; and to build self-efficacy for researching and experimenting
with new technology tools for improving teaching and learning in their school context.
In addition to improving and aligning the instructional practice of DFs, district leaders
expected DLCs to support school-wide improvements in teaching and learning. This involved
coordinating information within schools (Box 2C), with DLCs drawing on advice and support
from school principals and DFs to inform their work with other teachers. For instance, DLCs
were expected to collaborate with school principals to develop a school vision for using
technology to achieve specific student learning goals (Box 3C). DLCs met with principals at the
beginning of the school year to understand their goals for using technology and supported
principals in bringing these goals to fruition through their work with DFs and through organizing
PD for other teaching staff. As part of DF training, DLCs were also supposed to coordinate
teaching practices among a broader group of teachers (Box 2C), arranging opportunities for DFs
to showcase their instructional work and collaborate with colleagues on instruction (Box 3D) and
developing informal teacher networks for sustaining instructional innovation and change (Box
3G). Together, these activities were supposed to build school coherence, allowing schools to
translate central office directives for instructional change into goals and strategies that were
tailored to their local context, and align teacher practice with this school vision for instructional
improvement.
22
Eventually, through ongoing school-wide efforts to integrate technology with instruction,
district leaders also expected principals and DFs to become experts for supporting district-wide
improvements in teaching and learning. Through regular exchanges with school principals and
DFs, DLCs were expected to learn from school efforts to experiment with technology and share
22
To support sharing of best practices in schools, DFs committed to sharing their use of technology with colleagues
at least once a month in department or staff meetings and in a limited number of after-school events organized by
the central office.
BUILDING NETWORKS FOR CHANGE
58
these insights across schools, building inter-school networks to further enhance school coherence
and instructional alignment (Box 3E). In addition, principals and DFs were supposed to forge
bottom-up channels of communication to advise central office administrators and staff on how to
continue supporting localized school improvement efforts (Box 3F), thereby contributing to
system coherence for instructional change. Ultimately, these interschool networks and bottom-up
communication channels were supposed to further build out LUSD’s social infrastructure for
disseminating and sustaining best practices for instruction (Box 3G).
For instance, many DFs also served on LUSD’s Technology Taskforce Committee
(TTC), a committee comprised of LUSD’s CET and teacher representatives from schools to
discuss and coordinate the district’s ongoing efforts in technology-enabled instruction. Similarly,
through monthly meetings with LUSD’s Assistant Superintendent of Instruction, principals were
supposed to learn and share best practices for integrating technology with other signature
practices and instructional programs. As intermediaries who regularly interface between the
central office and schools, DLCs were expected to mediate these bottom-up communication
channels, learning from DFs’ efforts to experiment with technology and sharing these insights
with central office administrators and CCs.
Assumptions about organizational context. As discussed earlier, district leaders
perceived the DCP as being central to improving teacher practice at scale and building self-
correcting mechanisms within schools and the central office to allow for ongoing improvements
in teaching and learning. Because the instructional leadership and brokering practices of DLCs
were central to building an adaptive social infrastructure for instructional change, district leaders
implicitly assumed that DLCs could leverage favorable conditions in their district’s social
structure to build relationships and broker information as needed (Box 4 in Figure 2). Moreover,
BUILDING NETWORKS FOR CHANGE
59
these assumptions directly relate to organizational conditions that I reviewed earlier in chapter 2.
The first of these assumptions was that DLCs could claim power and influence over instructional
practice (i.e., in-degree centrality) from other formal leaders such as central office
administrators, instructional support staff (e.g., CCs) and school principals, while also
redistributing authority to DF, TTC representatives, and other teachers to lead instructional
change. District leaders also assumed that DLCs could cultivate insights on technology-enabled
and CCSS instruction from central office and school actors throughout the district (i.e., out-
degree centrality) to inform central office, school, and teacher practices for improving teaching
and learning. It was also assumed that DLCs could leverage informal communication patterns,
such as high levels of social engagement and mutual ties on instructional reform (i.e., dense and
reciprocal ties), to broker information on a system-wide basis. In addition, district leaders
assumed that DLCs could navigate actor cliques (i.e., homophilous ties and network closure) to
share new information with actor groups that otherwise shield themselves from external
intervention.
Overall, the theory of action for LUSD’s technology rollout and the DCP provides a rich
setting for examining the extent to which ed-tech coaches engage in brokering practices central
to instructional leadership in education technology reforms and exploring how LUSD’s
organizational context influences these brokering practices. My dissertation investigates these
questions to identify successes and shortcomings in the district’s underlying theory of change
and discuss implications for systemic instructional change.
BUILDING NETWORKS FOR CHANGE
60
CHAPTER FIVE -- Social Network Data and Methods
Data Collection
I investigate my research questions and hypotheses using administrative and social
network data on central office leaders and administrators (n=18)
23
, CCs (n=12)
24
, DLCs (n=15),
school principals (n=43), DFs (n=186), TTC representatives (n=35), and other classroom
teachers (n=504) in LUSD.
25
I collected these data in the 2014-15 school year.
Administrative Data
I accessed district records on the employment location (i.e., school name or central
office), demographics (race/ethnicity, gender), educational attainment, job position, and years of
experience for all district employees.
26
These administrative data also include information on the
grade-level, course of instruction, and classroom demographics of teachers such as the percent of
students who are under-represented minorities, English language learners, and eligible for free
and reduced-price lunch (FRL) and special education. I supplemented these administrative data
with publicly-available school data on school level (e.g., elementary, middle, high or alternative
23
I limited my sample of central office leaders and administrators to those working in an instructional capacity in
this district. This includes LUSD’s chief academic officer and assistant superintendents for instruction, as well as
administrators overseeing LUSD’s gifted and talented, bilingual, elementary, secondary, special education,
education technology, and curricular development programs. I did not include central office staff working in
specialized pupil services (e.g., speech pathologists, IEP coordinators), music instruction, health and psychological
services, or those responsible for program evaluation and school attendance.
24
This number includes CCs (n=7) as well as Balanced Literacy coaches (n=5) who worked at a limited number of
schools. Schools had to use their own budgets to hire these coaches.
25
I limited my sample of DFs, TTC representations, and other classroom teachers to those teaching core academic
subjects with an assigned classroom of students (i.e., elementary multiple subjects, ELA, math, social science, and
science instruction).
26
I collected complete administrative data for all district employees in LUSD, except for some who were missing
data on their education status. Because social network analysis requires complete data on all actors in a network, I
assume all district employees who responded to my social network survey but who were missing data on these
demographic variables were like most employees in the district (i.e., they have a Master’s degree or higher). These
missing data imputations apply to 2.67% of survey respondents in Sample 1 and 1.53% of survey respondents in
Sample 2. In my analyses, I run alternative models where I assume that respondents with missing demographics data
do not have a Master’s degree and observe similar results.
BUILDING NETWORKS FOR CHANGE
61
schools), school enrollment, student demographics, and academic performance on CCSS-aligned
assessments in ELA and math for the 2014-15 school year.
Social Network Data
I administered a social network survey at the end of the fall semester of the 2014-15
school year. Because social network studies are sensitive to missing data, especially in the case
of directed networks (i.e., where a tie from actor i to j is different from a tie from actor j to i), I
provided a small financial incentive to increase my survey response rate. I used these survey
data to map out two social networks in LUSD. First, I administered this survey to all district
leaders, central office administrators, CCs, principals, DFs, and TTC representatives to map out
communication among all personnel working in an instructional leadership capacity in the
district (herein referred to as “Sample 1”). Sample 1 is my main sample for observing the
brokering practices of DLCs given that district leaders put in place formal routines to build social
relations among these actors in the district (e.g., scheduled meetings between DLCs, CCs, and
central office administrators, the assignment of DFs to DLCs).
Seventy-eight percent of study participants in this sample responded to my survey. Table
4 compares respondents to non-respondents in this sample, showing that non-respondents were
more likely to be DFs who, as teachers, also had lower levels of educational attainment relative
to central office actors and school administrators who responded to the survey. Non-respondents
also had slightly more years of district experience. I do not find any other differences between
respondents and non-respondents in terms of demographic characteristics (gender and race),
location of work in the district (i.e., central office versus schools), school level (elementary,
middle or high school), or school demographics and academic performance.
BUILDING NETWORKS FOR CHANGE
62
Because Sample 1 does not allow me to observe how DLCs are interacting with teachers
who are not working in a formal leadership role (i.e., other classroom teachers who are not DFs
or TTC representatives), I also collected social network data from 15 school sites in LUSD. This
involved administering the same survey instrument to other classroom teachers (n=211) in nine
out of the 18 elementary schools and all six middle schools in the district.
27
I then combined
these data with social network data collected from district leaders, central office administrators,
CCs, DLCs, and from the school principals (n=19), DFs (n=103), and TTC representatives
(n=20) working at these schools (herein referred to as “Sample 2”).
28
In providing evidence on
how DLCs interact with the entire teaching staff at a sample of schools, Sample 2 provides more
complete evidence on brokering practices such as itinerant coordination (i.e., within-school
coordination), liasing (i.e., inter-school coordination), representation (i.e., top-down
communication), and gate-keeping (i.e., bottom-up communication) that depend on DLCs
forming social ties with school actors.
29
I purposively sampled schools in Sample 2 to maximize variation in school size, student
and staff demographics, and school performance. As shown in Table 5, I was successful in
achieving this variation, which is important for understanding how the organizational context of
schools facilitates and/or constrains interactions between DLCs and school staff on technology-
27
While I originally planned to collect social network data from all teachers in LUSD schools, I was not able to
obtain response rates to my survey that were sufficient for the execution of network analyses. As a result, I focused
my efforts on collecting these data from a sample of elementary and all middle schools in the district instead.
28
Ten of the TTC representatives in Sample 2 were also trained as DFs.
29
Because Sample 1 includes social interactions across all school levels in LUSD (elementary, middle, and high)
whereas Sample 2 only includes social interactions at the elementary and middle school levels, a valid concern is
that any difference in findings between these samples could be a function of school level rather than the addition of
social ties involving other classroom teachers. To mitigate this concern, I run all analysis for Sample 1 with a sub-
sample of central office and school actors who are just working at the elementary and school levels (e.g., DLCs,
school principals, DFs, and TTC representatives who are working in elementary and middle schools) and find
identical results as seen with my complete sample. This assures me that any difference in results across Samples 1
and 2 are due to the inclusion of social interactions with other classroom teachers and not due to the exclusion of
high school educators from my sample.
BUILDING NETWORKS FOR CHANGE
63
enabled and CCSS instruction. As shown in Table 5, 68% of school administrators and staff
responded to my survey, with response rates ranging from 55% to 89% by school.
30
Tables 6
compares respondents to non-respondents showing that teachers, especially those with less
education and those who teach middle school math and sixth grade, are under-represented. In
contrast, school principals, DLCs, DFs and teachers who teach fourth-grade elementary are over-
represented in my social network data.
Because I was unsuccessful in collecting social network data from all study participants
in Samples 1 and 2, my data and analysis are vulnerable to response bias. The lower
representation of certain groups of teachers in these networks (DFs in Sample 1 and middle
school math and sixth grade teachers in Sample 2) suggest that I am missing data on teacher
interactions with colleague that could bias my overall findings. The lower representation of these
sub-groups also means that my findings might not be generalizable to the entire population of
actors included in these networks. I address these challenges in my analysis plan and discussion
sections of my findings (chapter 6).
My survey included two socio-metric questions
31
in which respondents could nominate
up to ten colleagues whom they approach (a) for advice on integrating technology with
instruction, and (b) for advice on teaching the CCSS. These questions provide the main social
network data for my analysis. Survey respondents were asked to nominate colleagues who
immediately came to mind as a source of support during the current school year and could search
for these names in a name generator pre-populated with the names of all district employees. I did
not restrict survey respondents to naming contacts in a specific location (e.g., the district central
30
Because DLCs and a select number of CCs were officially assigned to work at school sites, I include these staff
when calculating school response rates to my survey.
31
Socio-metric questions are the questions asked in a survey for collecting social network data. These questions ask
respondents to nominate individuals whom they approach for support and/or advice for a specific topic.
BUILDING NETWORKS FOR CHANGE
64
office or their school) so that I could map social networks both within and across schools in
LUSD. Survey respondents were informed that it was not necessary to fill all ten spaces and
were given the option to write out the names of colleagues whom they could not find in the name
generator. Limiting total nominations to ten responses is consistent with how prior studies have
collected data on social networks in educational settings (e.g., Spillane et al., 2012). While this
numerical limit could bias my network data, less than two percent of respondents in my samples
listed ten contacts for any of the above socio-metric questions.
32
On average, survey respondents
nominated 4.29 individuals as a source of advice for integrating technology with instruction and
3.87 individuals as a source of advice for CCSS instruction in Sample 1. In Sample 2,
respondents nominated an average of 4.00 individuals as a source of advice for integrating
technology with instruction and 3.76 individuals as a source of advice for CCSS instruction.
Descriptive Analysis
I use these data to visualize LUSD’s social networks and conduct descriptive analyses
related to my research hypotheses. First, to generate evidence on the different possible brokering
roles of DLCs in LUSD (hypotheses 1), I calculate brokerage scores for betweenness,
coordination, itinerant coordination, liasing, representation, and gatekeeping in each of LUSD’s
social networks for (a) integrating technology with instruction and (b) teaching the CCSS. As
explained in chapter 3, these brokerage scores are based on the group affiliation of actors, which
for my analysis includes the central office and individual schools represented in my data (29
groups in Sample 1 and 16 groups in Sample 2).
33
32
This option is important because it allows me to observe social networks both within and across schools in LUSD,
which is germane for exploring how DLCs broker information on a district-wide basis.
33
In other words, I classify the actors in my network data as working from the central office (i.e., central office
leaders and administrators, CCs, and DLCs) or in specific schools (school principals, DFs, TTC representatives, and
other classroom teachers). This means there are 29 distinct groups in my network data for Sample 1, and 16 groups
of actors in my network data for Sample 2.
BUILDING NETWORKS FOR CHANGE
65
I calculate these brokerage scores using the statnet package in R, which defines the
brokerage score for actor i as the number of ordered triads (i.e., a transitive path between three
actors) in which actor i is positioned in the middle in one of the six brokering roles listed above.
The brokerage function in statnet computes “global” scores for each brokering role across all
actors in a network, along with corresponding expectations, standard deviations, associated z-
scores, and p-values under the Gould-Fernandez random association model.
34
These global
scores indicate the extent to which each brokering relation is a prominent feature of the social
network and contributes to how information is being shared.
The brokerage function in R also computes standardized “individual” brokerage scores
for each actor in the network. I use these individual brokerage scores to assess the extent to
which DLCs contribute to the brokering patterns observed in my network data relative to other
central office and school actors. Specifically, I compare the distribution (i.e., mean and range) of
individual brokerage scores for DLCs to the distribution of these individual brokerage scores for
other actor groups, including central office administrators, CCs, school principals, DFs and TTC
representatives, and other classroom teachers. Table 7 explains how I interpret these brokerage
scores to investigate my hypotheses.
Next, to explain the brokering patterns in my network data, I run whole-network analyses
to explore the organizational conditions outlined in research hypotheses 2-5. First, I compare
DLCs to other central office and school actors in terms of degree-based measures of centrality
(i.e., in-degree and out-degree), t-testing whether DLCs have significantly higher or lower in-
34
Gould and Fernandez (1989) argue that these brokerage statistics are asymptotically distributed under the null
model for sufficiently large networks (15 actors for global scores and 30 actors for individual scores). The authors
caution that, in networks with very low density rates, it is possible for the distribution of their brokerage statistics to
be skewed. An indicator for this concern would be when the expected value of brokerage scores is less than twice
the magnitude of the standard deviation. In my dissertation, I am working with large social networks that have low
rates of network density, making it possible that the global brokerage scores for these networks are not
asymptotically distributed. However, I confirm in all my results that the expected value of these global brokerage
scores is more than twice the magnitude of the standard deviation.
BUILDING NETWORKS FOR CHANGE
66
degree and out-degree centrality relative to other actor groups.
35
As discussed earlier, to broker
information on instructional change, I expect DLCs to demonstrate significantly higher in-degree
(hypothesis 2) and out-degree (hypothesis 3) centrality than other central office and school actors
in LUSD.
I also calculate whole-network statistics such as network density, reciprocity, and
clustering to examine informal organizational conditions that influence the brokering practices of
DLCs. Network density measures the overall level of communication in a network (i.e., the
proportion of all possible ties that are realized), with low density networks indicating the
presence of structural holes and the absence of brokering ties. Network reciprocity (i.e., the
proportion of social ties that are reciprocated between actors) indicates the amount of
interdependence, trust, and flexibility in a network, which are important conditions for
coordinating and brokering information on a system-wide basis. I expect networks with high
rates of density and reciprocity to support the brokering capacity of DLCs (hypothesis 4).
Network clustering (i.e., the proportion of triads in a network that are closed) provides
evidence of actors communicating in separate cliques within a network. When network
clustering is higher than network density, this indicates that actors are communicating in
densely-connected cliques with sparser regions in between. While network clustering supports
information sharing within actor cliques, it also makes it more challenging for DLCs, as
outsiders, to broker new information into these cliques. As such, I expect networks with high
rates of clustering (relative to network density) to be a more challenging environment for DLCs
to broker information (hypothesis 5).
35
I standardize these degree-based measures so that I can make comparisons between social networks for
integrating technology with instruction and teaching the CCSS.
BUILDING NETWORKS FOR CHANGE
67
Exponential Random Graph Modeling
I then run exponential random graph (ERG) models to obtain more fine-grained evidence
on how the above organizational conditions influence tie formation and brokering in LUSD’s
social networks (see Lusher, Koskinen, & Robins, 2012; Robins, Pattison, Kalish, & Lusher,
2007 for more technical details on these models). The intuition behind ERG models is that local
interactions between individuals at the relational level (i.e., dyadic level) give rise to larger social
networks. Moreover, individual, dyadic, and whole-network characteristics can be modeled
explicitly to describe these larger social networks. While these models are often used to predict
the likelihood of given social network to exist, researchers have used ERG models to isolate the
effects of individual and dyadic attributes on tie formation while controlling for endogenous
network conditions that lead to dependencies between dyads (Jabbar, 2015; Wimmer & Lewis,
2010). For example, ERG models can predict the effect of DLCs receiving and/or sending ties,
while accounting for the general tendency for central office and school actors to reciprocate ties
or for school actors to nominate colleagues with greater out-degree (i.e., out-degree popularity).
This empirical robustness allows me to determine, with greater accuracy, the extent to which
organizational conditions of interest at the individual, dyadic, and network-levels influence tie
formation in LUSD’s social networks while controlling for other relevant factors.
ERG models are similar to ordinary logistic regression models in that they predict the
log-odds that any given edge (or tie) in a network will exist. The difference between ERG
models and ordinary logistic regression models is that the former explicitly controls for network
dependencies between dyads when predicting the state of networks. The general equation for
these models is as follows:
!"# $
%&
= 1 = )
*
+[#(.,0)]
%&
BUILDING NETWORKS FOR CHANGE
68
Where $
%&
represents a social tie or dyad from actor j (tie-sender) to actor i (tie-receiver) and is
directed such that $
%&
does not equal $
&%
. #(.,0) is a vector of measures derived from a given set
of network relations between actors, y, and a matrix of actor attributes, X. These measures
include individual covariates for tie-senders and receivers, dyadic covariates for each pair of tie-
receivers and -senders in the network, and endogenous network parameters. +[#(.,0)]
%&
represents the change in #(.,0) associated with the formation of a social tie between two actors
(i.e., when $
%&
switches from 0 to 1) and ) is a vector of coefficient estimates for these
associated changes.
I run my models using the R package statnet, which uses Markov Chain Monte Carlo
maximum likelihood estimation (Lusher et al., 2012). I estimate two models for each data
sample, one predicting LUSD’s social network for integrating technology with instruction and
another for predicting LUSD’s social network teaching the CCSS. For Sample 1, I observe
52,212 dyads cross-nested within 229 central office and school actors (tie-receivers and senders).
For sample 2, I observe 69,960 dyads cross-nested within 265 central office and school actors
(tie-receivers and senders).
Tie-Sender and Receiver Parameters
Table 8 outlines the tie-sender and receiver, dyadic, and network parameters that I
specify in my ERG models and their connection to my research hypotheses. First, I control for
the formal position of tie-receivers and senders who are CCs, DLCs, school principals, DFs, TTC
representatives and other teachers, leaving central office administrators as the reference group.
36
36
I choose to make central office leaders and administrators the reference group as this will best allow me to assess
the influence of other central office and school actors who are positioned lower in the organizational hierarchy of
LUSD. Given that DFs were trained in two cohorts (78 teachers in the 2013-14 school year and 108 teachers in the
2014-15 school year), I also run a model where I distinguish between these two cohorts and see similar effects for
both groups of DFs.
BUILDING NETWORKS FOR CHANGE
69
These covariates provide evidence on the likelihood of DLCs receiving social ties and sending
social ties in their networks relative to other central office and school actors. A positive tie-
receiver effect for DLCs indicates high in-degree centrality and the ability of DLCs to direct
information flows on instructional change (hypothesis 2), whereas a positive tie-sender effect for
DLCs indicates high out-degree centrality and provides evidence of these coaches seeking advice
to inform their brokering (hypothesis 3).
I control for other tie-receiver and sender attributes that can predict tie formation in
LUSD’s social networks as demonstrated in the extant literature. These attributes include grade-
level and subject area expertise (Spillane et al., 2015), which I specify as indicators for whether
or not tie-receivers and senders have taught (or currently teach in the case of teachers) secondary
ELA/social science, secondary math and secondary science, leaving elementary multiple subjects
as the reference group. I also include tie-receiver and sender covariates for experience in English
language learner and special education instruction, and indicators for whether tie-receivers and
senders are teachers who are assigned to tested grade-levels for CCSS-aligned assessments.
To account for the influence of teaching experience (Penuel et al., 2009; Wilhelm et al.,
2016), I include binary variables for tie-receivers and senders who have two or fewer, three to
five, and six to nine years of teaching experience, leaving as the reference group tie-receivers
and senders with 10 or more years of teaching experience.
37
I also specify tie-receiver and sender
covariates for gender and highest level of educational attainment (Master’s degree or higher) to
control for the effects of these personal attributes on tie receiving and sending (Spillane et al.,
37
I choose these categories for years of teaching experience since research suggests that younger teachers have
greater self-efficacy for technology-enabled instruction (Inan & Lowther, 2009; Snoeyink & Ertmer, 2001) and
since most DLCs in LUSD have fewer than 10 years of teaching experience. I distinguish between novice teachers
and those with three to five and six to nine years of teaching experience since research shows that teachers make the
largest gains in instructional effectiveness during the first few years of their careers followed by smaller (although
persistent) gains after 10 years of experience (Harris, 2011; Papay & Kraft, 2015).
BUILDING NETWORKS FOR CHANGE
70
2015; Wilhelm et al., 2016).
38
To account for the school context of principals and teachers (Daly
& Finnigan, 2012; Finnigan & Daly, 2010), I include tie-receiver and sender covariates for the
percent of students who met or exceeded proficiency standards in ELA for CCSS-aligned
assessments in the 2014-15 school year.
39
Dyadic Parameters
At the dyadic-level, I control for conditions of homophily that contribute to a fragmented
organizational context for brokering (hypothesis 5), but that are not readily apparent in aggregate
measures of network clustering (Liou, 2016; Moolenaar, Daly, Sleegers, & Karsten, 2014;
Spillane et al., 2015; Wilhelm et al., 2016). To control for homophily in formal role and status, I
include a dyadic indicator for whether tie-receivers and senders both work from the central office
or in the same school.
40
In addition, I include a dyadic indicator for whether tie-receivers and
senders both work in the same formal role. To account for homophily in expertise, I include a
dyadic indicator for whether tie-receivers and senders share the same subject area expertise,
teach in the same grade-level (teachers only), and fall into the same categories of years of
teaching experience as outlined earlier. In addition, I control for whether tie-receivers and
senders share the same gender and level of educational attainment. Positive and significant
effects for these dyadic covariates indicate a fragmented organizational environment that should
38
While I originally planned to control for whether tie-receivers and senders are racial/ethnic minorities, there was
very little variation in this attribute which created issues of collinearity when running my models.
39
Because this test performance measure is highly correlated (at 0.65 or higher) with school performance in math,
and the percent of students in schools and teachers’ classrooms who are socioeconomically disadvantaged and
English language learners, I do not control for these other measures of school context. I run separate models where I
substitute these school measures for one another and obtain similar results. For teachers where I have more fine-
grained data on student assignments to their classrooms, I control for classroom demographics rather than school
demographics. For district leaders, central office administrators, and CCs who work from the central office, I assign
these actors districtwide demographics and performance values. For select CCs and some DLCs who work at
multiple school sites, I assign them average school demographics and performance values across their assigned sites.
40
While DLCs and some CCs work in both the central office and schools, I specify these actors as working in the
central office given that this is how I observe brokering patterns in my network data. I run specifications of this
model where I also attribute DLCs and CCs as working in both the central office and their assigned schools and
obtain similar results.
BUILDING NETWORKS FOR CHANGE
71
be more challenging for DLCs to navigate when brokering information. I also include a dyadic
indicator for the absolute difference in school ELA performance between tie-receivers and
senders. A negative and significant effect for this absolute difference suggests that frontline staff
and school actors serving different student populations are less likely to engage one another on
instructional reform, providing further evidence of a fragmented social structure.
Network Parameters
I also include network parameters to investigate my research hypotheses on network
closure and other informal communication patterns. I include out-degree popularity to model the
socially-embedded nature of expertise in knowledge-intensive reform and the advice-seeking
behaviors expected of brokers (hypothesis 3). This measure captures the tendency for actors who
send more ties for advice to also be a central source of advice for others. A positive effect for this
term indicates that influential actors in LUSD’s social networks are valued for their efforts to
build relations and seek-advice from others, which is what I expect to observe of DLCs given the
knowledge-intensive nature of LUSD’s technology rollout.
I also include network parameters to account for conditions of network density
(hypothesis 4), reciprocity (hypothesis 4), and clustering (hypothesis 5). Specifically, I control
for the number of ties (or edges) as a measure for network density. I also control for reciprocal
ties (mutual ties) to predict the likelihood of receiving a social tie when that tie is reciprocated.
Positive and significant effects for both edges and mutual ties should indicate favorable
conditions for brokering. To observe the effects of network closure, I include a parameter for
cyclic closure that predicts the likelihood of an actor receiving a tie when part of a cyclic triad.
41
41
As I show below, LUSD’s reform networks demonstrate low rates of density, making it necessary to specify
network parameters in my ERG models that are better suited for modeling large networks with sparse ties
(Goodreau, 2007). Specifically, to capture the effects of cyclic closure, I use a statistical measure called the
geometrically weighted edgewise shared partner distribution (GWESP). Instead of a census of number of closed
BUILDING NETWORKS FOR CHANGE
72
A positive and significant effect for this parameter indicates a fragmented organizational
environment that should be less favorable to brokering.
Model Fit
ERG models require a stringent goodness-of-fit test, which involves simulating networks
from a predicted model to see how well the model captures network features that are not
explicitly modeled (Hunter, Goodreau, & Handcock, 2008). The most common network features
included in this test are in-degree, out-degree, edge-wise shared partners (a non-parameterized
distribution of shared partners in a network)
42
, and minimum geodesic distance (the pairwise
path distance between two actors).
43
If a model captures much of the variation in the formation
of social ties, it will generate a distribution of network features encompassing those observed in
the actual network. I run this goodness-of-fit test for my main models for each of LUSD’s social
networks. The graphical results from these tests for Sample 1 are shown in Appendix A. These
graphs compare the observed network features (represented by a thick black line) to box plots
indicating the distribution of these features from 100 network simulations. As shown in these
graphs, the black line tracks the box plots very closely, suggesting strong model fit.
Methodological Limitations
A limitation of this study is that survey-based measures of advice-seeking for technology-
triangles in a network, the shared partner count is calculated for each edge in a network, producing a distribution of
shared partner counts across all edges. The GWESP then defines a parametric form for this distribution with a
declining positive effect for each additional shared partner on tie formation (Goodreau, Kitts, & Morris, 2009). This
approach has been shown to be particularly effective for overcoming model degeneracy when estimating ERG
models to fit large social networks (Goodreau, 2007). In my analysis, I specify a geometric rate of decline (t) of
0.25 although my results and model fit are robust to a range of values from 0.00 to 0.50. The process of setting a
value for t entailed testing for model fit for t values beginning at zero and then increasing this value by 0.05 until
my measures of model fit no longer improved. I began at zero and moved upward since lower values of t are less
likely to generate degenerate models (Goodreau et al., 2009).
42
While the GWESP term in my model is guaranteed to predict the mean number of shared partners in a network,
there is no guarantee that my model will fit the full distribution of this shared partner count. Hence, this statistic still
represents a valid measure of assessing model fit (Goodreau et al., 2009).
43
For example, actors who share a tie are distance 1, actors who are connected to the same colleague but who have
no direct connection to one another are distance 2, and so on (Hunter, Goodreau, & Handcock, 2008).
BUILDING NETWORKS FOR CHANGE
73
enabled and CCSS instruction might not reflect the actual advice-seeking behaviors of central
office and school actors. This is in fact a common critique of the extant literature on teacher PD
and instructional improvement in general, which often relies on survey-based measures to
observe teacher experiences in PD (Desimone, 2009; Lawless & Pellegrino, 2007). That said,
researchers have shown that, in the case of collecting social network data, respondents are
capable of reporting stable patterns of social interaction with accuracy than social interactions
that take place in time-specific episodes (e.g., during a PD training) (Marsden, 1990, 2011). I
take advantage of this response behavior by asking respondents to report whom they approach
for instructional advice on both reform topics rather than asking them to describe social
interactions at a specific event or time period.
I also took steps to validate network data. While the primary focus of my dissertation is
the quantitative analysis of social network data to observe the brokering ties of ed-tech coaches
and the influence of social structure on these brokering ties, my initial pilot work and qualitative
interviews helped ensure that survey respondents interpreted my socio-metric questions as
intended. Nevertheless, measurement error is still possible in terms of respondents incorrectly
interpreting these questions or accidently selecting individual names in the online questionnaire.
Given that I administered by social network survey in the middle of the 2014-15 school year, it is
possible that the configuration of LUSD’s social networks might have changed by the end of the
school year. However, as I discuss in greater detail in chapter 7, my interview data suggests that
respondents’ ego-networks were stable over time.
Another limitation of this study concerns the relatively low response rates from teachers
in both samples (DFs in Sample 1 and other classroom teachers in Sample 2). Unlike
econometric analysis where it is possible to run sensitivity analyses to account for missing data
BUILDING NETWORKS FOR CHANGE
74
and non-response bias, there are no robust procedures for imputing missing data in social
network analysis (Huisman, 2014; Moolenaar, 2012; Valente, 2010). This makes non-response
problematic because it could bias my estimation of whole-network properties such as network
density, reciprocity, and clustering in ways that I cannot fully gauge. Moreover, the lower
representation of teachers in my network data limits the generalizability of my findings. To
determine the extent to which my findings are vulnerable to non-response bias, I estimate the in-
degree of non-respondents to determine if they are central actors in LUSD. In general, I find that
survey non-respondents have significantly lower in-degree than survey respondents, suggesting
that most non-respondents were peripheral actors and that I was able to capture most of the social
ties in LUSD’s social networks.
44
As shown in Tables 12 and 13, my estimates for network
density, reciprocity, and clustering are also relatively stable between Samples 1 and 2 even
though I have a substantially higher response rate for Sample 1. This provides further assurance
that these network properties are not severely affected by non-response bias.
My analytic methods also rely on somewhat blunt measures of opinion leadership and
access to expertise to explain the brokering practices of ed-tech coaches. For instance, while
research shows that both individual attributes and the distribution of power and influence in
social structures give rise to opinion leadership (Atteberry & Bryk, 2011; Moolenaar et al., 2010;
Rogers, 2003), I only observe the latter in my network data and cannot account for how the
individual attributes of DLCs imbue these coaches with power and influence. This detailed
44
In LUSD’s reform network for technology, I find that survey respondents in Sample 1 are nominated as a source
of advice by 2.92 individuals on average whereas non-respondents are only nominated by 0.52 individuals on
average (this difference is statistically significant at p<0.05). This trend also holds in Sample 2, where survey
respondents are nominated as a source of advice by 2.95 individuals on average whereas non-respondents are
nominated as a source of advice by only 0.70 individuals on average (this difference is statistically significant at
p<0.001). In LUSD’s reform network for the CCSS, I find that survey respondents in Sample 1 are nominated as a
source of advice by 1.90 individuals on average whereas non-respondents are nominated by 0.45 individuals on
average (this difference is statistically significant at p<0.01). In Sample 2, survey respondents are nominated as a
source of advice by 2.18 individuals on average whereas non-respondents are only nominated by 0.88 individuals on
average (this difference is statistically significant at p<0.001).
BUILDING NETWORKS FOR CHANGE
75
analysis requires collecting data on leadership attributes for all central office and school actors in
LUSD’s reform networks (not just DLCs), which I was unable to collect this dissertation.
Similarly, I rely on a rough proxy for access to expertise of DLCs (out-degree centrality), using
the number of actors these coaches approach for advice as a measure of the expertise residing in
their personal networks rather than looking at the attributes of the individuals with whom they
are in touch. As such, I could be misrepresenting the access to expertise of DLCs in my data if
they are reaching out to a limited number of the most informed actors in their school systems on
instructional reform (e.g., other DLCs who are experts on technology-enabled instruction). On
the other hand, because knowledge-intensive reform requires frontline staff such as DLCs to
interact with a range of central office and school actors to understand their different perspectives
of instructional reform, and not just actors who have the most expertise on instructional change,
it is still informative to use the out-degree centrality of DLCs as a proxy for their access to
expertise. For instance, while principals and teachers might not possess the most expertise on
technology-enabled instruction, DLCs are still expected to engage these school actors and seek
out their perspectives on LUSD’s technology rollout to inform central office procedures for
instructional change. This provides an example where looking at the total number of actors
whom DLCs approach for advice can be more informative than measuring the instructional
expertise of these actors.
Finally, my dissertation is a cross-sectional study of the brokering practices of DLCs and
as such, cannot speak to causal impact of associations seen in my data. However, given the
relative scarcity of research on this topic, my research still provides a valuable first look at
whether DLCs are brokering information on a district-wide basis and the organizational
conditions that might be influencing their brokering behaviors.
BUILDING NETWORKS FOR CHANGE
76
CHAPTER SIX -- Social Network Findings
Brokering Practices of DLCs
Table 9 reports the global brokerage scores for betweenness (B), coordination (C),
itinerant coordination (IC), liasing (L), representation (R), and gatekeeping (G) in each of
LUSD’s social networks. I report these brokerage scores for both Samples 1 and 2, and for a sub-
sample of central office actors in the district (i.e., central office administrators, DLCs and CCs
working from the central office). Positive standardized scores indicate that the total count of
brokering ties exceeds their expected value given the total number of ties and actor groups in the
network, suggesting that brokering ties are a prominent feature of the social network.
Conversely, negative standardized scores indicate that the total count of brokering ties does not
exceed their expected value, suggesting that central office and school actors are avoiding
brokerage relations by pursuing direct relations with one another or forgoing indirect contact
entirely (Gould & Fernandez, 1989). The stars in this table indicate if these global brokerage
scores are statistically significant.
In both Samples 1 and 2 I find that, regardless of reform topic, certain brokering ties are
more prominent than others. Coordination, representation, and gatekeeping ties are prominent
communication channels in LUSD’s reform networks, whereas itinerant coordination and liasing
are less prominent. I find higher levels of coordination, representation, and overall betweenness
in LUSD’s reform network for technology, but higher levels of gatekeeping in LUSD’s reform
network for teaching the CCSS.
Hypothesis 1: Ed-tech Coaches are Prominent Brokers
BUILDING NETWORKS FOR CHANGE
77
To determine how DLCs are contributing to these brokering patterns, I examine who are
the most active brokers in these social networks. Table 10 reports the mean standardized
brokerage scores for central office administrators, CCs, DLCs, school principals, DFs and TTC
representatives, and other classroom teachers. Because these brokerage scores are already
standardized, I compare them to one another in terms of magnitude rather than t-testing to see if
these means are significantly different from one another. I am able to compare these mean scores
across actor groups and brokering ties. For instance, I can compare these mean scores to
determine if DLCs or CCs are more engaged in coordinating information within the central
office, as well as to determine if DLCs engage in coordination ties more than other kinds of
brokering ties (i.e., itinerant coordination, liasing, representation and gatekeeping). Together,
these comparisons allow me to assess which actor group is most actively engaged in brokering
information and the extent to which these actor groups are investing in certain kinds of brokering
relations over others.
In addition to comparing these means scores, I consider the distribution of individual
brokerage scores within each actor group to see if most actors have standardized scores above
1.96 or less than -1.96, which is the threshold z-score value for determining if individual
brokerage scores are statistically significant at p<0.05. Individual brokerage scores above
(below) the threshold value of 1.96 indicates that actors are significantly contributing to (or
avoiding) brokering ties in their social network. I plot the distribution of these individual
brokerage scores for each actor group in LUSD’s reform networks for technology and CCSS in
Figures 3 and 4 respectively (for Sample 1 only). These figures include six boxplots that show
the distribution of individualized brokerage scores for each actor groups in terms of overall
betweenness, coordination, and itinerant coordination (top panel) and for liasing, representation,
BUILDING NETWORKS FOR CHANGE
78
and gatekeeping (bottom panel). The red lines in these boxplots indicate threshold values of 1.96
and -1.96 for determining statistical significance.
DLCs are active brokers in LUSD’s technology reform network. Table 10 (top panel)
suggests that DLCs are the most active brokers in LUSD’s social network for integrating
technology with instruction, with mean betweenness scores (B) of 11.26 and 11.49 standard
deviation units in Samples 1 and 2 respectively (almost four times as large as the global
betweenness scores for the network, see Table 9). As shown in Figure 3, the entire distribution of
brokerage scores for DLCs is concentrated above the threshold value of 1.96, suggesting that all
DLCs are significant brokers in the network. DLCs also have large mean coordination scores
(26.06 and 31.70 standard deviation units in Samples 1 and 2 respectively) and mean
representation scores (34.72 and 37.34 standard deviation units in Samples 1 and 2 respectively).
The distribution of these coordination and representation scores also exists above the threshold
value of 1.96 (see Figure 3), suggesting that most if not all DLCs are engaging in coordination
and representation ties in LUSD. The fact that these brokerage scores are much larger than the
brokerage scores of other actors in the network suggests that DLCs’ efforts to coordinate
information within the central office (i.e., coordination) and to communicate insights from the
central office into schools (i.e., representation) are the most prominent communication channels
in LUSD.
To observe if DLC’s coordination ties within the central office span across hierarchical
levels of authority (middle-up) and instructional divisions (lateral), I next calculate global
brokerage scores for the central office region in LUSD’s reform network for technology – i.e.,
global brokerage scores for social ties among central office administrators, DLCs, and CCs in the
BUILDING NETWORKS FOR CHANGE
79
district.
45
In other words, rather than treating these actor groups as all affiliated to the central
office, I now distinguish between these actors groups in terms of their formal role and estimate
the extent to which their communication within the central office is taking place within formal
role (i.e., coordination) or spanning across formal roles (i.e., itinerant coordination, liasing,
representation and gatekeeping). These global brokerage scores (reported in Table 9) indicate
that brokering within the central office are largely concentrated within rather than across formal
roles. I find a positive and marginally significant coordination score (1.70 standard deviation
units, p<0.10) but a negative and significant gatekeeping score (-2.35 standard deviation units,
p<0.05), suggesting that central office actors are mainly coordinating information on technology-
enabled instruction with others who share the same formal role rather than those who are outside
of their formal designation. Therefore, while DLCs are actively coordinating information within
the central office on technology-enabled instruction, this coordination mainly involves
exchanging information with other DLCs rather engaging in middle-up or lateral ties to
coordinate LUSD’s technology rollout with administrators and frontline staff overseeing other
instructional programs in the district.
DLCs also do not engage in itinerant coordination, liasing, or gatekeeping in LUSD’s
technology reform network. Table 10 shows that DLCs, on average, engage in itinerant
coordination more so than other central office and school actors, but that the magnitude of these
scores is too small to be statistically significant at conventional levels. Similarly, DLCs have
negative liasing and gatekeeping scores in this reform network, but these mean scores are also
too small to be statistically significant. Figure 3 shows that most itinerant coordination, liasing,
45
In this analysis, I classify actors based on their formal role in the central office. This leaves me with three actor
groups to observe brokering ties within the central office, one for central office administrators, another for DLCs,
and the last one for CCs.
BUILDING NETWORKS FOR CHANGE
80
and gatekeeping scores for DLCs fall squarely within the threshold values of -1.96 and 1.96,
suggesting that DLCs are not actively pursuing these brokering ties.
Other brokers in LUSD’s technology reform network. Aside from DLCs, I do find
evidence of other central office and school actors brokering information on technology-enabled
instruction, although the magnitude of these brokerage scores are much smaller than those
observed for DLCs. For instance, central office administrators coordinate information within the
central office, with mean coordination scores of 23.31 and 28.37 standard deviation units in
Samples 1 and 2 respectively, and represent information from the central office in schools, with
mean representation scores of 16.46 and 12.46 standard deviation units in Samples 1 and 2
respectively. As shown in Figures 3, the distribution of coordination and representation scores
for central office administrators surpasses the threshold value of 1.96, suggesting that most of
these administrators are prominent brokers in their network.
School actors also coordinate information within their school sites, with principals, DFs
and TTC representatives, and classroom teachers all demonstrating mean coordination scores
that exceed 1.96 in value (see Table 10). However, perhaps due to the smaller size of school
networks, the mean coordination scores for principals, DFs and TTC representatives, and other
classroom teachers are smaller in magnitude than those seen for DLCs and central office
administrators. The distribution of coordination scores for these school actors also falls within
the threshold range 1.96 to -1.96 (see Figure 3), suggesting that most principals and teachers are
not coordinating information in their schools but that there are several outliers in these school
networks.
I also find that principals are active gatekeepers in LUSD’s reform network for
technology, with gatekeeping scores of 1.12 and 3.22 standard deviation units in Samples 1 and
BUILDING NETWORKS FOR CHANGE
81
2.
46
The boxplot in Figure 3 shows that the distribution of gatekeeping scores for principals just
barely exceeds the threshold value of 1.96, indicating that a select number of principals are
critical gatekeepers in this reform network. Together, these results suggest that principals and
teachers are distilling and coordinating information amongst themselves to inform school-wide
programs for technology-enabled instruction, even though DLCs are not involved in facilitating
these intra-school ties (given their low itinerant coordination scores). That said, this intra-school
communication is still much less prevalent than coordination ties within the central office and
top-down communication from the central office into schools.
DLCs are peripheral actors in LUSD’s CCSS reform network. In contrast to their
central role in LUSD’s technology reform network, I find that DLCs are peripheral actors in
LUSD’s social network for teaching the CCSS. As shown in Table 10 (bottom panel), DLCs
have mean betweenness scores of 1.02 and 0.14 standard deviation units in Samples 1 and 2
respectively, with these scores being too small to be considered statistically significant at
conventional levels. This same trend holds across the different brokering ties that DLCs could be
engaging in. DLCs have slightly larger mean coordination scores, 4.18 and 4.30 in Samples 1
and 2 respectively, and slightly larger mean representation scores, 3.46 and 1.52 in Samples 1
and 2 respectively. As shown in Figure 4, the distribution of these coordination and
representation scores also exceeds the threshold value of 1.96, indicating that certain DLCs are
actively coordinating information within the central office and informing school practices for
teaching the CCSS. That said, these brokerage scores are still much lower in magnitude than
those observed for DLCs in LUSD’s reform network for technology and those observed for other
46
While it is possible that principals are accessing information from other schools (instead of the central office) to
share with teachers, a closer analysis of the actors whom principals nominate as a source of instructional support
confirms that they are mainly seeking advice from central office leaders, CCs, and DLCs.
BUILDING NETWORKS FOR CHANGE
82
central office actors in LUSD’s reform network for teaching the CCSS (e.g., central office
administrators and CCs, discussed later in this section).
I also find little evidence of DLCs engaging in itinerant coordination, liasing, and
gatekeeping ties to support CCSS instruction. As shown in Table 10, DLCs have negative mean
scores for itinerant coordination and liasing ties and positive mean scores for gatekeeping (in
Samples 1 and 2), but the magnitude of these scores are all too small to be significant at
conventional levels. The entire distribution of brokerage scores for DLCs for these three
brokering categories also falls within the range of -1.96 to 1.96 (see Figure 4), suggesting that
DLCs are not significant contributors to these information exchanges.
Overall, the above numbers imply that DLCs are not active brokers in LUSD’s reform
network for teaching the CCSS. I also find that, to the extent that DLCs are coordinating
information within the central office on teaching these content standards, these brokering ties
once again involve communication with other DLCs rather than central office administrators and
CCs. As shown in Table 9, I find a slight positive global coordination score (not statistically
significant) in the central office region of LUSD’s reform network for teaching the CCSS, but
negative and significant global itinerant coordination and gatekeeping scores. This suggests that
central office actors are mainly brokering information within rather than across formal roles to
develop instructional programs and services for teaching the CCSS.
Other brokers in LUSD’s CCSS reform network. If DLCs are not contributing to
brokering ties in LUSD’s reform network for teaching the CCSS, then who is fulfilling this role?
My results suggest that CCs are the most active brokers in this network, with mean betweenness
scores of 4.91 and 3.49 standard deviation units in Samples 1 and 2 respectively. As shown in
Figure 4, the distribution of individual brokerage scores for this actor group clearly surpasses the
BUILDING NETWORKS FOR CHANGE
83
threshold value of 1.96, indicating that most CCs are significant brokers in the network. I also
find some evidence of central office administrators and school principals being active brokers in
this reform network. As shown in Table 10, central office administrators have mean betweenness
scores of 2.97 and 1.10 standard deviation units in Samples 1 and 2 respectively, while principals
have mean betweenness scores of 0.86 and 2.52 in Samples 1 and 2 respectively. As shown in
Figure 4, the distribution of betweenness scores for these actor groups slightly surpasses the
threshold value of 1.96, suggesting that certain central office administrators and school principals
are active brokers in this network.
In terms of specific brokering ties, I find that CCs maintain a prominent role in
coordinating information within the central office on the CCSS, with mean coordination scores
of 18.61 and 19.08 standard deviation units in Samples 1 and 2 respectively. Central office
administrators are also active coordinators within the central office, with mean coordination
scores of 10.09 and 10.34 standard deviation units in Samples 1 and 2 respectively. I also find
positive mean coordination scores for principals, DFs and TTC representatives, and other
teachers that exceed 1.96 in value (see Table 10). Once again, perhaps due to the smaller size of
their school networks, the coordination scores for principals, DFs and TTC representations, and
other classroom teachers are smaller in magnitude than those seen for CCs and central office
administrators. As shown in Figure 4, the distribution of coordination scores for these school
actors falls within the threshold range 1.96 to -1.96, suggesting that most principals and teachers
are not coordinating information on teaching the CCSS in schools and that the large means for
these actor groups are driven by outliers in these school communities.
Unlike LUSD’s reform network for technology, I do not find as strong evidence of
central office actors representing information from the central office into schools on teaching the
BUILDING NETWORKS FOR CHANGE
84
CCSS. Both CCs and central office administrators have positive mean representation scores, 9.29
and 7.68 standard deviation units for CCs in Samples 1 and 2 respectively and 6.10 and 4.22
standard deviation units for central office administrators in Samples 1 and 2 respectively. While
large in magnitude, these representation scores are still much smaller in magnitude than those
observed for DLCs in LUSD’s reform network for technology. Moreover, as shown in Figure 4,
the distribution of representation scores for central office administrators and CCs exceeds the
threshold value of 1.96, but not to the same extent as seen for DLCs in LUSD’s reform network
for technology (Figure 3).
In contrast, it appears that principals are more active mediators between the central office
and schools in LUSD’s reform network for teaching the CCSS, with mean gatekeeping scores of
8.01 and 14.20 standard deviation units in Samples 1 and 2. The boxplot in Figure 4 shows that
the distribution of gatekeeping scores for principals is concentrated above the threshold value of
1.96, indicating that most principals are critical gatekeepers in this reform network. Seeing as
these gatekeeping scores are also much larger in magnitude than those seen for principals in
LUSD’s reform network for technology, this suggests that principals are better equipped to
translate information from the central office on teaching the CCSS into guidance for teachers
than they are for integrating technology with instruction.
Summary of Brokering Practices of DLCs
There are several important takeaways from the above discussion. First, the total
brokering activity (or betweenness) in LUSD’s social network for integrating technology with
instruction far exceeds the brokering activity observed in LUSD’s social network for teaching
the CCSS. While central office administrators, CCs, and principals are more active than DLCs at
brokering information in LUSD’s CCSS reform network, the betweenness scores of these actors
BUILDING NETWORKS FOR CHANGE
85
are much lower overall than the betweenness score of DLCs in LUSD’s technology reform
network. This suggests that LUSD’s social structure is different across reform topics and perhaps
more challenging to navigate for supporting CCSS instruction.
Second, both reform networks demonstrate similar brokering ties, with information being
coordinated within the central office and schools (i.e., coordination) rather than within (i.e.,
itinerant coordination) or across (i.e., liasing) schools. Information in these reform networks is
also communicated in a top-down manner from the central office into schools instead of in a
bottom-up manner from schools to the central office. DLCs seem to be actively contributing to
these coordination and top-down brokering ties in LUSD’s technology reform network, but are
less essential in LUSD’s CCSS reform network where central office administrators, CCs, and are
instead more prominent brokers. I summarize these results in Table 7.
Together these results support only parts of the theory of action of LUSD’s technology
rollout (Figure 2). First, I find little evidence of DLCs brokering information on teaching the
CCSS (Boxes 2A-2D), suggesting that these coaches are not bridging their support for
technology-enabled instruction to the instruction of these content standards. This, in turn, could
take away from the comprehensiveness of instructional support that these coaches are providing
to teachers and the overall quality of best practices that are they are developing and
disseminating to improve teaching and learning in LUSD (Boxes 3A-3G).
Second, I find that DLCs are active brokers in LUSD’s reform network for integrating
technology with instruction. Upon taking a closer look at the brokering relations of DLCs in
LUSD’s technology reform network, I find that these coaches are mainly coordinating
information on technology-enabled instruction amongst themselves rather than coordinating
information with other central office administrators and CCs (Box 2A). This suggests that DLCs
BUILDING NETWORKS FOR CHANGE
86
are not coordinating their support on technology-enabled instruction with central office
administrators and CCs who directly supervise instruction of the CCSS (Box 3A), once again
affirming that these coaches are supporting the integration of technology with instruction in
general but not necessarily with instruction of these content standards.
I do find that DLCs are mainly communicating information on technology-enabled
instruction in a top-down manner from the central office into schools (Box 2B), which suggests
that these coaches are providing teachers with one-on-one instructional support to integrate
technology with teaching and learning (Box 3B) and to understand how the central office’s goals
for technology-enabled instruction relate to more immediate school goals and strategies for
improving student learning (3C). However, I do not find evidence of DLCs coordinating
information within or between schools (Box 2C), suggesting that these coaches are not mediating
collaborative exchanges between school principals and teachers to build out school programs for
instructional change (Box 3C) or to share information on technology-enabled instruction within
and across schools (Boxes 3D and 3E). These results do not mean that school actors are not
collaborating on localized improvement efforts or building inter-school networking to shift
teacher practice, but rather that DLCs are not involved in facilitating these exchanges.
I also find minimal evidence of DLCs facilitating bottom-up communication channels
from schools to the central office (Box 2D). This suggests that, without engaging in school-wide
efforts to integrate technology with instruction, DLCs do not have a rich knowledge base of
school experiences and insights to pull from to guide central office decision-making and
practices for supporting district-wide instructional change (Box 3F). Taken together, these
brokering patterns suggest that DLCs have so far built an incomplete social infrastructure to
facilitate the district-wide adoption of technology-enabled teaching practices in LUSD (Box 3G).
BUILDING NETWORKS FOR CHANGE
87
I next focus on the organizational conditions that could be contributing to the incomplete
brokering practices of DLCs. I explore these organizational conditions in terms of properties of
LUSD’s social structure that could be facilitating and/or constraining social ties and hence
brokering relations in the district. I investigate these structural conditions through a combination
of network visuals, whole-network analyses, and ERG models. Figures 5a and 5b show graphs of
LUSD’s social networks for integrating technology with instruction (Samples 1 and 2
respectively), and Figures 6a and 6b show graphs of LUSD’s social networks for teaching the
CCSS (Samples 1 and 2 respectively). Each node in these network graphs represents an
individual actor, with nodes colored according to the formal role of actors in the district. The
lines between the nodes represent the exchange of social ties, with the arrow indicating the
direction from which advice is being sought (i.e., an arrow pointing toward actor i from to actor j
indicates that actor j is seeking advice from actor i).
Figures 5a-5b and 6a-6b show that, regardless of reform topic, there are more social ties
among central office administrators, CCs, and DLCs who are positioned at the center of these
networks and fewer ties among principals, DFs, TTC representatives, and other classroom
teachers who are positioned at the periphery of these networks. In fact, in LUSD’s social
network for teaching the CCSS (see Figures 6a and 6b), there are quite a few DFs, TTC
representatives, and other classroom teachers who are isolates in the network (i.e., the actors
have no social ties to or from other actors). These visual trends suggest that communication in
LUSD is centralized, with central office actors directing information flows on technology-
enabled and CCSS instruction to school actors who are at the extremes of the network. This
visualization confirms the earlier brokering results in terms of there being more communication
and brokering in LUSD’s technology reform network than in LUSD’s CCSS reform network,
BUILDING NETWORKS FOR CHANGE
88
and with both networks demonstrating prominent top-down channels of communication from the
central office into schools. That said, further quantitative analyses reveal other structural features
that could be shaping the brokering practices of DLCs.
Hypothesis 2: Power and Influence of Ed-tech Coaches Enables Brokering
Table 10 reports the in-degree (influence and power) of central office administrators,
CCs, DLCs, principals, DFs and TTC representatives and classroom teachers in LUSD’s social
network for integrating technology with instruction (top panel). The stars in this table indicate if
mean standardized in-degree and out-degree scores are significantly different between actor
groups. These results show that, in both Samples 1 and 2, DLCs have significantly greater in-
degree (2.40 and 2.61 standard deviation units respectively) relative to other central office and
school actors. As the other active group of brokers in this network, central office administrators
also have significantly greater in-degree relative to other central office and school actors (0.80
and 0.65 standard deviation units respectively), although their in-degree is still much smaller
than that of DLCs. In contrast, principals, DFs and TTC representatives, who have some of the
lowest brokering scores in this reform network, also have lower in-degree than central office
actors. Together, these numbers suggest that influence and power are important enabling
conditions for DLCs to broker information on technology-enabled instruction.
In-degree also appears to be important for brokering information in LUSD’s social
network for teaching the CCSS. This time, however, DLCs have lower in-degree than central
office administrators and CCs who are the most central actors in the network. As shown in Table
10 (bottom panel), while DLCs have significantly greater in-degree than other central office and
school actors in Sample 1 (0.55 standard deviation units), this in-degree is much smaller in
magnitude than the in-degree for central office administrators (1.50 standard deviation units) and
BUILDING NETWORKS FOR CHANGE
89
CCs (1.53 standard deviation units). These same comparisons hold in Sample 2 where, in
addition to central office administrators and CCs, DLCs also have lower in-degree than school
principals (0.77 standard deviation units). Once again, DFs, TTC representatives, and other
classroom teachers have the lowest in-degree relative to other central office and school actors.
ERG model results. The above in-degree trends suggest that DLCs have significant
power and influence in LUSD’s technology reform network and are therefore better positioned to
broker information in this network. In contrast, DLCs have almost no power and influence over
instructional practice in LUSD’s CCSS network, which could be hindering their ability to
facilitate information sharing. While informative, these results do not account for other
individual, dyadic, and network conditions that could be contributing to the in-degree of DLCs. I
rely on my ERG model results to account for some of these endogenous conditions. Table 11
reports my main model results (in odds ratios) predicting the formation of social ties in LUSD’s
social networks as a function of tie-receiver, tie-sender, dyadic, and network conditions. The first
two columns report my model results for LUSD’s social network for integrating technology with
instruction (Samples 1 and 2), while the third and fourth columns report my model results for
LUSD’s social network for teaching the CCSS (Samples 1 and 2). I organize my coefficient
estimates in several panels. The first panel reports the tie-receiver and –sender effects for formal
role. The second panel reports dyadic effects for conditions of homophily. The third panel
reports the effects of network conditions including out-degree popularity, edges (network
density), reciprocated ties, and cyclic closure (network closure or clustering). I include my
complete model results with all tie-receiver, sender, and related dyadic controls in Appendix B.
The ERG model results add further nuance to the in-degree trends reported above. In this
section, I focus on the tie-receiver effects of DLCs and other central office and school actors in
BUILDING NETWORKS FOR CHANGE
90
Table 11 (panel 1). I find that DLCs demonstrate significantly higher odds of receiving social
ties than central office administrators in LUSD’s social network for integrating technology with
instruction (with odds ratios of 3.80 and 2.98 in Samples 1 and 2 respectively), but that CCs and
TTC representatives are now just as likely to receive social ties in this network that central office
administrators. This suggests that central office administrators do not have as much influence in
this reform network as previously suggested. I also find that principals, DFs and other classroom
teachers have much lower odds of receiving social ties in this network than central office leaders
and administrators.
In LUSD’s social network for teaching the CCSS, DLCs are just as likely as other central
office administrators to receive social ties in both Sample 1 (odds ratio of 1.40) and Sample 2
(odds ratio of 1.30), while CCs are significantly more likely to receive social ties in both samples
(with odds ratios of 1.87 and 3.09 respectively) as are school principals in Sample 2 (odds ratio
of 1.62). That said, the tie-receiver effects for CCs and principals are still much smaller than
those observed for DLCs in LUSD’s technology reform network, suggesting that these actors do
not possess the same degree of social influence to affect instructional change regarding the
CCSS. Once again, I continue to see that DFs, TTC representatives, and other classroom teachers
are significantly less likely to receive social ties relative to other central office administrators.
The ERG model results provide strong evidence in support of in-degree influencing the
brokering practices of DLCs (hypothesis 2). First, because DLCs have far more in-degree in
LUSD’s social network for integrating technology with instruction, these actors seem to be better
positioned to broker information on a district-wide basis in this reform network than in LUSD’s
social network for teaching the CCSS. Second, the distribution of social influence in these
networks aligns with the top-down direction of brokering ties reported earlier. More specifically,
BUILDING NETWORKS FOR CHANGE
91
because CCs and DLCs demonstrate significantly higher in-degree in both reform networks,
while principals, DFs, TTC representatives and other teachers demonstrate significantly lower
in-degree in these networks, it follows that information on instructional change is communicated
through top-down communication channels from the central office into schools rather than
bottom-up channels from schools to the central office.
Hypothesis 3: Ed-tech coaches’ Access to Expertise Enables Brokering
Table 10 (top panel) also reports the out-degree (access to expertise) of central office
administrators, CCs, DLCs, principals, DFs and TTC representatives and classroom teachers in
LUSD’s social network for integrating technology with instruction. These results show that
DLCs have significantly greater out-degree (1.45 and 1.42 standard deviation units in Samples 1
and 2 respectively) relative to other central office and school actors. Other than DLCs, CCs also
have significantly higher out-degree in Sample 1 (0.86 standard deviation units) and Sample 2
(1.00 standard deviation units), and principals have significantly higher out-degree in Sample 2
(0.55 standard deviation units). In contrast, DFs, TTC representatives, and other classroom
teachers have significantly lower out-degree in this reform network. Together, these results
support the notion that access to expertise is an enabling condition for DLCs to broker
information on technology-enabled instruction.
I find less consistent evidence of access to expertise enabling brokering in LUSD’s social
network for teaching the CCSS. As shown in Table 10 (bottom panel), DLCs have significantly
higher out-degree (0.92 standard deviation units) than other central office and school actors in
Sample 1. In fact, the out-degree of DLCs is the highest in this network compared to the out-
degree of central office administrators (0.21 standard deviation units), CCs (0.71 standard
deviation units), school principals (0.44 standard deviation units), and DFs and TTC
BUILDING NETWORKS FOR CHANGE
92
representatives (-0.23 standard deviation units). In Sample 2, DLCs also have significantly
higher out-degree (0.68 standard deviation units) that is comparable in magnitude to that of CCs
(0.68 standard deviation units) and slightly lower than that of school principals (0.80 standard
deviation units). However, despite having more outward social ties, DLCs are not brokering
information on teaching the CCSS.
ERG model results. The above trends so far suggest that access to expertise is an
important condition for brokering information on technology-enabled instruction but not
necessarily for teaching the CCSS. However, these descriptive results do not account for the
influence of other individual, dyadic, or network conditions on advice-seeking relations in
LUSD. For instance, as I discuss in the following section, it is possible that DLCs are in a denser
region of LUSD’s reform networks that leads them to have greater out-degree than school actors.
The tie-sender effects reported in Table 11 (panel 1) suggest that controlling for these network
conditions is in fact important. DLCs are now significantly less likely to send social ties in
LUSD’s social network for integrating technology with instruction relative to central office
administrators (odds ratios of 0.40 and 0.45 in Samples 1 and 2 respectively), whereas principals,
DFs, TTC representatives, and other classroom teachers are significantly more likely to send ties
in this social network (with odds ratios substantially larger than one). In LUSD’s social network
for teaching the CCSS, both DLCs and CCs are no more likely to send ties relative to central
office administrators, whereas principals, DFs, TTC representatives, and other classroom
teachers are once again significantly more likely to send ties in this network. The out-degree
popularity effects reported in Table 11 (panel 3) are also less than one and statistically significant
for both reform networks, suggesting a negative association between actor in-degree and out-
degree that is especially pronounced in LUSD’s technology reform network.
BUILDING NETWORKS FOR CHANGE
93
These results suggest that the actors who are the most influential brokers in LUSD’s
social networks are also the least likely to seek advice from others. Rather than showing a
positive relationship between out-degree and brokering, my ERG model results point to a
negative relationship between out-degree and social influence in LUSD’s reform networks. I find
that the actors who are the most influential and active brokers in these social networks to be the
least likely to send advice ties to other actors. The only exception is school principals who are by
far the most likely to send ties in both reform networks and who are active gatekeepers for
translating information from the central office to share in their schools. This is perhaps not
surprising since gatekeepers are supposed to search for new information outside of their
organizations and hence engage in advice-seeking relations (i.e., out-degree). This suggests that
out-degree might be an important structural condition for middle-management such as principals
who are looking to adapt their organization to shifting expectations in their external environment,
but not for central office representatives (i.e., DLCs and CCs) who are more focused on sharing
insights from the central office to guide principals and teachers on this process of school change.
These findings are also consistent with other research showing that, in addition to frontline staff,
principals are critical intermediaries for translating central office goals for instructional change
into school improvement efforts (Finnigan & Daly, 2010; Matsumura & Wang, 2014).
Hypothesis 4: Dense and Reciprocated Networks Enables Ed-tech Coach Brokering
Table 12 reports the size, density, reciprocity, and clustering of LUSD’s social networks.
I report these network statistics for each reform network overall and separately for the central
office and all schools in these networks.
Low density of ties. These numbers indicate that LUSD’s social networks are quite
sparse with an overall density of 0.01, meaning that only one out of 100 potential advice
BUILDING NETWORKS FOR CHANGE
94
relationships between central office and school actors are realized. This finding is consistent with
other studies showing that large networks are less dense (Kadushin, 2012) and that
communication between central office and school actors in district-wide reform efforts is
infrequent (Daly & Finnigan, 2012; Finnigan & Daly, 2010). In examining the density of
relations within the central office versus within and among schools, I find more communication
within the central office, with 12% and 7% of all ties present for technology-enabled instruction
and teaching the CCSS respectively. In contrast, only 1% or fewer ties among school actors are
present for technology-enabled instruction and teaching the CCSS.
These network density statistics provide more detail on how informal communication
patterns in LUSD could be influencing the brokering practices of DLCs. First, the low rates of
density in both social networks suggest that information sharing on technology-enabled and
CCSS instruction is generally constrained. As shown in the ERG model results in Table 11
(panel 3), the network parameter effect for edges is close to zero for both reform networks,
suggesting that the odds of an additional tie forming in these networks is close to none. This
limited communication could make it more challenging to broker information on instructional
change, although DLCs do not seem to be as hindered by the low density of social ties when it
comes to brokering information on technology-enabled instruction. The fact that communication
is denser within the central office than among schools also explains why coordinating
information within the central office is a more prominent brokering practice in both reform
networks than brokering practices that involve initiating ties with school actors (e.g., itinerant
coordination, liasing, representation, and gatekeeping).
Lack of reciprocal ties. Both reform networks in LUSD demonstrate limited reciprocal
exchange. In LUSD’s social network for integrating technology with instruction, only 15% to
BUILDING NETWORKS FOR CHANGE
95
16% of social ties are reciprocated in Samples 1 and 2 respectively. An interesting pattern
emerges when I examine how reciprocal ties are distributed between the central office and
schools. In LUSD’s social network for integrating technology with instruction, I find that 33% of
social ties are reciprocated among central office actors whereas only 25% and 19% of social ties
are reciprocated between school actors in Samples 1 and 2 respectively. These numbers suggest
that there is greater depth of exchange and coordination of resources within the central office to
support technology-enabled instruction than what is taking place within and among schools. The
lack of reciprocal ties within and among schools can explain, in part, why schools are receptive
to top-down communication from the central office on technology-enabled instruction, seeing as
schools lack the in-depth knowledge and support needed to advance teacher practice on their
own. Given that the proportion of reciprocated ties within the central office and within and
among schools is still larger than the overall proportion of reciprocated ties in the district, this
suggests that few ties extending from the central office into schools or vice versa are
reciprocated. The brokering practices of DLCs are consistent with this pattern, with these
coaches representing information from the central office into schools but not facilitating bottom-
up communication from schools to the central office.
In LUSD’s social network for teaching the CCSS, the proportion of reciprocal ties is
higher, with 17% and 23% of social ties reciprocated in Samples 1 and 2 respectively. This time
I find that 21% of social ties are reciprocated among central office actors, whereas 18% and 27%
of social ties are reciprocated between school actors in Samples 1 and 2 respectively. Unlike
LUSD’s social network for integrating technology with instruction, it appears that school actors
are engaging in similar levels (if not more) of back-and-forth dialogue on teaching the CCSS as
central office actors. This could explain why these schools are more closed off to communication
BUILDING NETWORKS FOR CHANGE
96
from the central office on instructional change, with the representation scores of central office
administrators and CCs in this network much lower than those observed for DLCs in LUSD’s
technology reform network. These results also seem to explain why principals, who are internal
to school communities and engaged in reciprocal dialogue on CCSS instruction, are more active
than other central office and school actors at brokering information on the CCSS. As noted
earlier, principals are far more active than other central office and school actors when it comes to
translating (i.e., gatekeeping) information from the central office to share with teachers (see
Table 10). Once again, I find that the proportion of reciprocated ties within the central office and
within and among schools is higher than overall rate of reciprocity for the network, suggesting
that information on teaching the CCSS is still being communicated in a top-down manner from
the central office into schools with less information being shared from schools up to the central
office.
My ERG model results confirm the above findings. As shown in Table 11 (panel 3),
reciprocal ties have positive effects on tie formation in both of LUSD’s reform networks, but
these effects are only significant in Sample 1 of LUSD’s social network for integrating
technology with instruction (odds ratio of 2.22) and in Sample 2 of LUSD’s social network for
teaching the CCSS (odds ratio of 1.96). These coefficient estimates are consistent with the
descriptive trends noted earlier in terms of there being a higher proportion of reciprocal ties
among central office actors in LUSD’s technology reform network (who represent most actors in
Sample 1) and among school actors in LUSD’s CCSS reform network (who represent most
actors in Sample 2).
Hypothesis 5: Fragmented Networks Hinders Ed-Tech Coach Brokering
The above discussion suggests that social ties in LUSD’s reform networks are internally-
BUILDING NETWORKS FOR CHANGE
97
focused, meaning that central office and school actors tend to communicate in cliques with sparer
regions of communication in between. The network clustering results reported in Table 12
confirm this idea. In both reform networks, I find network clustering rates that far exceed
network density. In LUSD’s social network for integrating technology with instruction, I find
network clustering rates of 0.15 and 0.17 in Samples 1 and 2 respectively, and in LUSD’s social
network for teaching the CCSS I find even higher network clustering rates of 0.22 in both
samples. This clustering is also more prevalent among central office actors than among school
actors, which is consistent with earlier results showing that central office actors are more likely
to be coordinating information within rather than across actor groups (Coburn et al., 2009).
ERG model results. My ERG model results in Table 11 (panel 3) also indicate a
consistently large positive effect of cyclic closure on tie formation across both reform networks
and data samples. I find positive and significant dyadic effects (panel 2) for conditions of
homophily in both networks, including effects for when tie-receivers and senders work from the
same location (i.e., both from the central office or in the same school), and share the same formal
role, the same subject area expertise, same grade-level assignment (teachers only), same years of
teaching experience (technology reform network only), and same gender. These results are not
surprising given prior research showing strong homophily effects in social networks in
educational settings (e.g., Liou, 2016; Spillane et al., 2015; Wilhelm et al., 2016). I also find that
central office and school actors are marginally less likely to form social ties for each percentage
point difference in the percent of students who meet or exceed standards in ELA achievement in
their school setting. This finding is again consistent with research showing that, in the context of
high-stakes reforms for improving student achievement and school performance, schools serving
the most disadvantaged and low-performing students are often the most isolated in district reform
BUILDING NETWORKS FOR CHANGE
98
networks for instructional change (Daly & Finnigan, 2012; Finnigan & Daly, 2010).
These results suggest that LUSD’s social structure is fragmented in similar ways across
both reform networks that could make brokering more challenging. In LUSD’s social network
for integrating technology with instruction, this fragmented structure does not seem to be
deterring DLCs from representing information from the central office into schools, but could be
hindering their efforts to coordinate information across actor groups within the central office as
well as within and between schools. In the case of LUSD’s social network for teaching the
CCSS, where DLCs have little influence over instructional practice, this fragmented context
could be deterring these coaches from brokering information entirely.
Summary of Findings on Organizational Context
An analysis of the social structure of LUSD highlight several factors that facilitate and
constrain the brokering practices of DLCs (summarized in Table 7). First, the high in-degree of
DLCs in LUSD’s social network for integrating technology with instruction, and their
corresponding lack of in-degree in LUSD’s social network for teaching the CCSS, explains why
DLCs are more active at brokering information on technology-enabled instruction than teaching
the CCSS (hypothesis 2). The distribution of actor in-degree in both networks further suggests
that information is mainly being shared among DLCs in the central office (Box 2A in Figure 2)
and communicated from DLCs in the central office to schools through top-down channels (Box
2B in Figure 2), which is consistent with earlier evidence presented on the brokering practices of
DLCs. The negative tie-sending effects of DLCs in LUSD’s technology reform network also
suggest that access to expertise is not a contributing factor to their influence as brokers. On the
contrary, the negative out-degree popularity effects in both reform networks suggests that the
most influential actors and brokers in LUSD are the least likely to seek advice from others
BUILDING NETWORKS FOR CHANGE
99
(hypothesis 3). This once again explains the prominent role of top-down brokering in both
reform networks and why DLCs are less engaged in brokering ties requiring them to seek advice
from school actors (e.g., itinerant coordination, liasing, and gatekeeping, Boxes 2C and 2D in
Figure 2).
In addition, I find low rates of network density in both reform networks could be
constraining the brokering practices of DLCs (hypothesis 4). While this constraint does not
appear to be as inhibiting for DLCs in LUSD’s technology reform network, the low density of
ties among schools in this network could explain why DLCs are less engaged in brokering
practices that involve initiating ties with school actors (e.g., itinerant coordination, liasing, and
gatekeeping) and conversely are more engaged in coordinating information with other DLCs
within the central office (where relations are denser) and representing information from the
central office into schools.
While I expected to find reciprocal ties to support DLCs in brokering information
(hypothesis 4), my social network results offer a more nuanced conclusion. Specifically, I find
that the high concentration of reciprocal ties among central office actors in LUSD’s technology
reform network creates an environment where DLCs can engage in rich dialogue with other
central office colleagues on instructional change and share these insights with school actors who
are less engaged in mutual dialogue and are therefore more open to receiving guidance from the
central office. In contrast, the high prevalence of reciprocal ties among school actors in LUSD’s
CCSS reform network seems to result in schools being more independent and less open to top-
down communication from the central office. Instead, school principals are positioned as a
central mediator between the central office and schools, translating and adapting information
from the central office to suit the specific goals and context of their schools. These results
BUILDING NETWORKS FOR CHANGE
100
together suggest that, while reciprocal exchange can facilitate district-wide brokering on
instructional change, understanding how reciprocal ties are distributed in social networks can
shed light on who has agency to broker information and the direction in which information is
being shared.
My analysis also reveals high rates of network clustering and significant homophily
effects in both reform networks (hypothesis 5). In the absence of power and influence over
instructional practice, these inward-looking ties seem to deter DLCs from having any capacity to
broker information on teaching the CCSS. In LUSD’s reform network for technology where
DLCs have more power and influence, these inward-looking ties could be inhibiting middle-up
and lateral brokering ties for coordinating information on technology-enabled instruction across
actor groups within the central office, preventing DLCs from accessing and sharing guidance for
positioning technology as a central resource for CCSS instruction. These inward-looking ties
could also be discouraging DLCs from mediating communication within and across schools, as
these lateral brokering ties are generally constrained in LUSD’s technology reform network.
BUILDING NETWORKS FOR CHANGE
101
CHAPTER SEVEN -- Multiple-Case Study on the Social Positioning of DLCs
Introduction
My social network results paint a detailed picture of how DLCs are brokering
information to support instructional change in LUSD. I show that DLCs have little to no capacity
for brokering information on teaching the CCSS. While DLCs are more active at brokering
information on technology-enabled instruction, they are mainly coordinating information
amongst themselves and sharing these insights with schools through top-down communication
channels. As discussed in chapter 6, these incomplete brokering practices have implications for
the theory of change of LUSD’s technology rollout, especially with regards to DLCs establishing
instructional coordination, coherence, and alignment to facilitate district-wide improvements in
teacher practice and student learning.
Because these incomplete brokering practices can detract from LUSD’s goals for teachers
to use technology to teach the CCSS, I delve deeper into their underlying causes. From the
organizational conditions that I have studied so far, power and influence (in-degree centrality)
seem to most consistently explain the brokering practices of DLCs. The qualitative component of
my dissertation therefore focuses on the conditions that position DLCs with power and influence
(or the lack thereof) for shifting instructional practice. Similar to other studies (e.g., Atteberry &
Bryk, 2011; Matsumura & Wang, 2014), I assume that DLCs are “socially positioned” to have
power and influence and that understanding the individual attributes of these coaches and social
conditions in their environment that bring about this positioning can explain more about their
brokering role and implications of their leadership for teacher learning and change. While my
social network analysis reveals some of conditions in LUSD’s social structure that could be
contributing to the social positioning of DLCs, I want to unpack these conditions further by
BUILDING NETWORKS FOR CHANGE
102
explicitly grounding them in the individual attributes of these coaches and the day-to-day social
context of educators in LUSD.
I use a multiple case-study research design to explore the social positioning of DLCs and
answer my second and third research questions. As a reminder, these research questions are: 2)
How does the organizational context of ed-tech coaches, as mediated by social structure,
influence their brokering practices; and 3) Are there similarities and/or differences in how ed-
tech coaches are brokering information on a) integrating technology with instruction and b)
teaching the CCSS, and if so, why? I answer research question 3 by examining how the social
positioning of DLCs allows them to be influential top-down brokers in LUSD’s social network
for integrating technology with instruction, but not so in LUSD’s social network for teaching the
CCSS. I focus on the top-down communication of DLCs because this is by far their greatest area
of activity in the district and is closely related to district leaders’ goals for coordinating LUSD’s
technology rollout with the CCSS and related signature practices. As shown in Figure 2, district
leaders expect the coordination of these instructional reforms within the central office (3A) to
inform how DLCs provide DFs with one-one-one support to align teacher practice with central
office goals for instructional change (3B). As such, exploring why DLCs are successful at
engaging teachers in top-down communication on integrating technology with instruction but
less so on teaching the CCSS provides a rich contrast for understanding the social positioning of
DLCs across these reform efforts and their ability to coordinate teacher practice accordingly.
I answer research question 2 by examining how, if at all, the social positioning of DLCs
leads them to support teacher collaboration on technology-enabled instruction. I focus on DLCs’
efforts to coordinate information sharing among teachers because this brokering practice is
arguably a critical step to spreading the benefits of their coaching to a broader pool of teachers
BUILDING NETWORKS FOR CHANGE
103
and building a district-wide social infrastructure for instructional change. As shown in Figure 2,
district leaders expect DLCs to build a school vision for improving student learning (3C),
facilitate teacher collaboration in learning (3D), and for these school-wide efforts to then lead to
inter-school networks for sharing best practices for instruction (3E) and build school expertise
for informing central office decision-making on instructional reform (3F).
While my results so far suggest that DLCs are not mediating information between
principals and teachers within schools to facilitate these intermediate outcomes (i.e., DLCs have
non-significant itinerant coordination scores), it is possible that DLCs are supporting school
efforts for instructional change in other ways. For instance, DLCs could be creating conditions
for school actors to directly communicate with each other and coordinate information for
instructional change, as evidenced by the positive coordination scores for school actors
highlighted in chapter 6. Alternatively, it is possible that DLCs are not invested in building
school networks for instructional change and that they are doing little to facilitate any kind of
communication between principals and teachers. I investigate these coaching practices, and their
connection to the social positioning of DLCs in LUSD’s technology reform network, where
DLCs have substantial power and influence.
Conceptual Framework
Figure 7 outlines the conceptual framework that guides my qualitative analysis. This
figure builds from Figure 2 by fleshing out the individual attributes of ed-tech coaches that
inform their brokering practices for shifting teacher practice, which lead to and are reinforced by
the social positioning of DLCs as figures of authority in schools. This relationship, in turn, is
shaped by social and other school conditions that give these coaches power and influence over
instruction. Below, I ground these aspects of my conceptual framework in findings from the
BUILDING NETWORKS FOR CHANGE
104
extant literature, providing more detail on how I conceptualize them and identifying knowledge
gaps that I address in my dissertation.
Individual Attributes of Ed-Tech Coaches
While power and influence are critical factors for enabling ed-tech coaches to affect
instructional change in schools, there is surprisingly little research on the individual attributes
that imbue ed-tech coaches with this authority. That said, there is growing evidence on the
qualities of other kinds of instructional coaches (e.g., literacy and math coaches) who are
successful at engaging teachers in instructional improvement efforts (e.g., Atteberry & Bryk,
2011; Bean et al., 2010; Marsh et al., 2010; Matsumura et al., 2010). This research suggests that
instructional expertise, strong interpersonal skills and prior experience in teaching and adult
education are critical qualities for coaches to build relationships with teachers focused on
instructional change. In addition, coach goals for instructional change, their perceptions of
teachers and school communities, and familiarity with school needs also influence the
authenticity of their work with teachers (Huguet et al., 2014; Marsh et al., 2010).
Brokering Practices of Ed-Tech Coaches
Because the individual attributes of coaches inform how they approach relationship
building with teachers, researchers have focused extensively on describing these relationships to
understand if and how coaches are supporting teacher learning and change. While I have
described these relationships in terms of structural definitions pertaining to brokering (i.e.,
representation for top-down brokering and itinerant coordination for within-school coordination),
these definitions do not provide a complete picture of how DLCs are communicating with
teachers. For instance, focusing on whether DLCs represent information from the central office
at schools does not shed light on the content and activities exchanged between DLCs and
BUILDING NETWORKS FOR CHANGE
105
teachers and how these transactions position DLCs as experts for technology-enabled instruction
but not for teaching the CCSS. As mentioned earlier, strictly focusing on whether DLCs operate
as middle-men who facilitate communication between principals and teachers is a narrow
perspective that might not capture how these coaches are supporting school-wide improvement
in other ways.
Given these limitations to defining brokering from a social network perspective, I draw
on other concepts from the extant literature to characterize brokering relationships between ed-
tech coaches and teachers in schools. As discussed in chapter 2, coaches can use bridging and/or
buffering strategies (Coburn & Woulfin, 2012) to motivate teachers to change their instructional
approach. These bridging and buffering strategies can be further characterized in terms of social
learning routines that coaches use to inform teacher learning and change, including exposing
teachers to specific kinds of dialogue and norms for instructional change (Coburn & Russell,
2008; Marsh et al., 2015), joint work that translates reform goals into meaningful tasks for
teachers and their colleagues (Marsh et al., 2015; Mayer et al., 2014), and tools and practices
that teachers continuously refer to and adapt to inform their professional growth (Bean et al.,
2010; Mudzimiri et al., 2014).
Research so far suggests that ed-tech coaches use gradual or phased-in bridging strategies
to support instructional change: (a) supporting teachers who are motivated to use technology in
their classrooms rather than working with teacher who are resistant to change, and (b) focusing
on teachers’ technological skills before addressing teachers’ pedagogical beliefs and other core
aspects of instruction (Kopcha, 2012; Sugar, 2005; Zhao et al., 2002). The reasoning behind this
approach is that technology adoption is less central to teachers’ belief systems which, in theory,
should make it easier for ed-tech coaches to shift teacher beliefs on technology adoption first (or
BUILDING NETWORKS FOR CHANGE
106
to engage teachers who have already bought into the idea of using technology for instruction)
and then use these peripheral beliefs to change higher-level goals such as teachers’ pedagogical
beliefs (Ertmer, 2005; Zhao & Cziko, 2001).
This strategy for teacher change of course means that ed-tech coaches might move slowly
in connecting technology with CCSS instruction, focusing first on building teacher self-efficacy
for using technology in their classrooms rather than pushing teachers to engage in personalized
and inquiry-based teaching practices to achieve CCSS learning objectives. Similarly, ed-tech
coaches might choose to work one-on-one with teachers who are motivated users of technology
rather than facilitating teacher collaboration and other school-wide efforts for instructional
change to broaden the footprint of technology adoption and instructional change. However, given
the rapid advancements to technology in the past decade, the close integration of technology with
the CCSS learning objectives, and the growing popularity of one-to-one computing programs in
districts, it is possible that ed-tech coaches are now engaging in more urgent bridging practices
than what has so far been described in the extant literature.
Social Positioning of Ed-Tech Coaches
The brokering practices of ed-tech coaches then inform, and are informed by, their social
positioning as figures of authority over teacher practice. As discussed earlier, this social
positioning can be observed in terms of the centrality of ed-tech coaches in their district’s social
structure (Daly et al., 2010; Hopkins et al., 2016). There are two kinds of social positioning of
DLCs in LUSD that I have detected in my social network data. The first is the low in-degree of
DLCs for advising central office and school actors on how to teach the CCSS. The second is the
high in-degree of DLCs for advising central office and school actors on technology-enabled
BUILDING NETWORKS FOR CHANGE
107
instruction, which may or may not preclude these coaches from coordinating instruction among
teachers to enable school-wide improvements in teaching and learning.
Contextual Factors: Social and Other School Conditions and District Context
Encompassing the relationship between the individual attributes of ed-tech coaches, their
brokering practices, and their social positioning is the organizational context of districts and
schools. Researchers have foregrounded social conditions in schools that enable instructional
coaches to socially engage teachers and influence teacher practice. As discussed in chapter 2, the
extent to which coaches exercise instructional leadership depends on district leaders embracing
an adaptive leadership approach that empowers central office and school actors who are
positioned lower in the organizational hierarchy of districts to lead instructional reform (Durand
et al., 2016; Ho & Ng, 2017). School leaders who legitimize the expertise and authority of
coaches are a critical part of this distributed leadership approach (Hopkins et al., 2016; Marsh et
al., 2010; Matsumura & Wang, 2014). This means that school principals need to share the same
goals for instructional change as coaches, the same understanding of the expected role of coaches
in supporting teacher practice, and establish school expectations, norms, and resources that are
supportive of this work (Atteberry & Bryk, 2011; Matsumura et al., 2010; Matsumura & Wang,
2014). While principals are central to motivating and supporting teachers to work with coaches
to improve instruction, their ability to foster this supportive environment also depends on the
existing culture and capacity of schools. As such, teacher perceptions of reform goals and
coaches (Coburn & Russell, 2008; Marsh et al., 2010) and school norms for teacher collaboration
and instructional change are also relevant social conditions that can influence the effectiveness
and outcomes of coaching programs (Anderson et al., 2014; Matsumura et al., 2010).
BUILDING NETWORKS FOR CHANGE
108
Apart from social conditions in schools, researchers have identified other contextual
factors that can influence coaching and instructional change, especially in the context of
education technology reform. Because most teachers and principals have had limited prior
exposure to technology in their teaching careers, their individual knowledge, beliefs, and
motivations are critical factors that can shape the work of ed-tech coaches in schools (Ertmer et
al., 2012). Research further suggest that teachers depend on external conditions outside of their
classroom to use technology effectively for instruction (Zhao et al., 2002). In addition to social
support from school leaders and teachers (discussed earlier), these external conditions include
access to adequate school infrastructure (e.g., technology resources, Internet bandwidth), human
capital in terms of qualified technical support and experts on technology-enabled instruction
(Bingham et al., 2016; Windschitl, 2002; Windschitl & Sahl, 2002), and scheduled time for
teachers to experiment with technology and improve on their instructional craft (Frank et al.,
2011). Other school conditions such as school size and demographics can also influence how
teachers engage ed-tech coaches for instructional support and ultimately use technology for
instruction (Warschauer & Matuchniak, 2010).
Finally, research on instructional coaching suggests that district context, especially in
terms of how coaching programs are designed to support teacher practice, can influence the
quality of teacher-coach interactions and the depth of instructional change realized in schools
(Coburn & Russell, 2008; Marsh et al., 2005, 2010). The recruitment and selection of coaches
and teacher participants in coaching programs, the articulation of coach responsibilities, the
assignment of coaches to schools, and district procedures for training and monitoring coaches are
all important design factors (Marsh et al., 2010). In the context of ed-tech coaching, these design
elements can dictate whether ed-tech coaches work with teachers across interrelated dimensions
BUILDING NETWORKS FOR CHANGE
109
of their instructional practice (i.e., using technology and teaching the CCSS) and whether ed-tech
coaches facilitate one-on-one versus more inclusive and collaborative learning opportunities for
teachers.
Contributions to the Extant Literature
In using this conceptual framework to guide my analysis, I make several contributions to
the extant literature. First, I build on existing evidence on individual attributes, brokering
practices, and social conditions that imbue coaches with power and influence by examining these
conditions at play in the context of ed-tech coaching. Ed-tech coaches are a unique case for
understanding this process of social positioning because of expectations from district leaders and
the public at large for these coaches to support multiple reforms in a range of instructional
settings (Herold, 2015; Superville, 2015). Moreover, because ed-tech coaches tend to be younger
teachers, they do not necessarily possess conventional attributes of expertise that would allow
them to claim power and influence in schools (e.g., ample years of teaching experience, prior
experience in adult education, in-depth pedagogical and content knowledge) (Flanigan, 2016).
This makes it important to understand how these coaches, who might otherwise be perceived as
lacking in instructional expertise and prior experience, are able to establish themselves as
influential leaders of instructional change in schools.
Aside from studying a policy-relevant case of instructional coaching, my research also
demonstrates how this social positioning can serve different purposes. Specifically, I
demonstrate how ed-tech coaches are socially positioned to engage teachers across multiple
reforms with interrelated goals for teacher learning and change (i.e., using technology for
instruction and teaching the CCSS) and to diffuse best practices for teaching and learning on a
school-wide basis. This analysis is different from merely studying whether ed-tech coaches
BUILDING NETWORKS FOR CHANGE
110
engage teachers in relationships focused on instructional change (which has been the focus of the
extant literature) because it demonstrates whether this engagement is coordinating teacher
practice across reforms and supporting communities of teachers and entire school staff to change
their instructional approach.
Case Study Approach
I use a case study methodology to investigate my research questions. The case study
methodology is appropriate for investigating a phenomenon in context, relying on detailed cases
to illicit information about a larger program or process of interest (Stake, 1995; Yin, 2014). The
case study methodology also places a strong emphasis on triangulating data from multiple
sources to strengthen claims about specific phenomena (Patton, 2002; Stake, 1995).
I use an instrumental, multiple-case study design for my qualitative study. Instrumental
case study is useful for learning about a specific phenomenon of interest (Stake, 1995) which, in
the case of my study, is understanding how the individual qualities and social environment of
DLCs inform their brokering practices and social positioning in schools. This makes it important
to examine DLCs as individual cases, tracking the conditions from which they draw social
influence over instructional practice. At the same time, because the social positioning of DLCs
depends on their social context (Matsumura & Wang, 2014; Woulfin, 2014), I consider the
schools in which these coaches work as sub-cases of interest. A multiple-case design – with
schools as distinct cases embedded within the DLCs as distinct cases – is an ideal approach. This
research design allows me to explore the social positioning of DLCs at multiple levels, as well as
observe points of convergence and divergence within and across DLCs and schools (Yin, 2014).
Case Site and Participant Selection
BUILDING NETWORKS FOR CHANGE
111
I follow two DLCs – Sarah and Cindy, each across two school sites: San Rafael and
Glendale Elementary for Sarah and Brand and York Elementary for Cindy.
47
I sample DLCs and
schools based on a maximum variation sampling strategy, allowing me to select diverse cases
that provide evidence of patterns of converge and divergence across a phenomenon of interest in
a sample (Cresswell, 2007). I achieve this maximum variation by selecting DLCs with a range of
in-degree centrality in LUSD’s social network for integrating technology with instruction (i.e.,
DLCs with high and low in-degree scores in LUSD’s technology reform network). I sample
DLCs based on in-degree centrality since this proxies their power and influence in LUSD’s
technology reform network and hence their social positioning in schools (Moolenaar, 2012;
Moolenaar et al., 2010). As shown in Table 13, while Sarah and Cindy both have low in-degree
centrality in LUSD’s CCSS reform network, their relative influence in this reform network still
parallels their ranking in LUSD’s technology reform network. As such, selecting these coaches
based on in-degree centrality provides variation for teasing out both unique and commonplace
conditions that inform the social positioning of these coaches.
In my analysis, Sarah represents the case of a DLC who is highly central in LUSD’s
technology reform network (i.e., high in-degree centrality) and who also has higher in-degree
centrality than other DLCs in LUSD’s CCSS reform network. She represents the coach with the
greatest potential for advising teachers on how to teach the CCSS. Moreover, her greater
connectivity to teachers in LUSD’s technology reform network provides her the most
opportunity to engage teachers in collaborative practices for experimenting with technology and
supporting school-wide instructional change. It is not by chance that Sarah finds herself in this
advantaged position. As Table 13 shows, Sarah has more experience in leading teacher PD and
47
To protect the anonymity of DLCs and their school sites, these are all pseudonyms.
BUILDING NETWORKS FOR CHANGE
112
working as a DLC, with 2014-15 being her second year of coaching at San Rafael and Glendale
EL. While Sarah has few years of teaching experience and only taught non-tested grade-levels
(kindergarten), these qualities do not seem to inhibit her from advising teachers on using
technology for instruction but could hold her back from advising teachers in tested grade-levels
on how best to teach the CCSS. As I discuss in greater detail below, Sarah previously worked
with the school principal at San Rafael from the central office during the initial stages of LUSD’s
technology rollout and taught Kindergarten at Glendale EL, allowing her to leverage existing
relationships at these school sites to further her work as a coach.
In contrast, Cindy represents the case of a DLC who is not as central in LUSD’s
technology reform network (i.e., low in-degree centrality relative to other DLCs) and has the
lowest in-degree centrality among DLCs in LUSD’s CCSS reform network. As such, she
represents the coach with the least potential for advising teachers on how to use technology for
teaching the CCSS. Because Cindy has limited connectivity to teachers at her school sites, she
also has fewer opportunities to mediate communication between principals and teachers to
facilitate knowledge sharing in schools and support school-wide instructional improvement.
There are several characteristics of Cindy that contribute to this social positioning, including her
lack of prior experience in coaching and leading teacher PD and lack of familiarity with her
assigned school sites. Like Sarah, Cindy also has limited prior teaching experience and has
mainly taught in non-tested grade-levels. Unlike Sarah, she has had no prior experience working
at any of her assigned schools.
I next sample Sarah’s and Cindy’s school sites based on their in-degree centrality in these
school communities, choosing one site where each DLC is centrally positioned as a source of
advice for technology-enabled instruction (i.e., high school in-degree) and another where each
BUILDING NETWORKS FOR CHANGE
113
DLC is less centrally positioned (i.e., low school in-degree). I limit my universe of schools to
those in Sample 2 where I collected survey responses from teaching staff to observe the in-
degree of DLCs with accuracy. I also sample schools serving different student populations within
each coach to account for how student demographics and learning needs might influence the
social positioning of DLCs (see Table 5). I prioritize diversity in student demographics since
prior research suggests this to be a relevant contextual factor for implementing technology-
enabled instruction in schools (Warschauer & Matuchniak, 2010). In addition, I bound my
analysis to DLCs working in elementary schools so that I can make comparisons across school
sites and minimize differences in contextual factors that are not the explicit focus of my study
(e.g., different grade-levels or subject area assignments of teachers, larger class sizes).
Amongst Sarah’s schools, San Rafael EL is a case of a smaller sized elementary school
that serves a more advantaged student population (i.e., lower proportion of low socioeconomic
status students, under-represented minority, and English language learner students) and a strong
track-record of performance on the CCSS assessments. In contrast, Glendale EL is a larger
elementary school with a more disadvantaged student population. While Glendale EL was one of
LUSD’s only high-performing Title 1 schools under the state’s standards-based assessments
prior to the CCSS, the school was not as successful as San Rafael EL under the CCSS
assessments for the 2014-15 school year. Perhaps because of the more advantaged school context
at San Rafael EL, Sarah is slightly more central in this school’s social network, although she
maintains relatively high-levels of connectivity to principals and teachers at both schools.
There is also interesting variation in Cindy’s school sites. Brand EL is a small Title 1
school that serves underprivileged students and is low-performing on the CCSS assessments. In
contrast, York EL is a larger and higher achieving school that serves a wealthier and less diverse
BUILDING NETWORKS FOR CHANGE
114
student population. These school conditions seem to translate into entirely different positions of
centrality for Cindy, who demonstrates some level of connectivity to the principal teachers at
Brand EL but almost no social ties to the principal and teachers at York EL. Unlike Sarah, Cindy
seems to find it more challenging to embed herself in the school community of her advantaged
school site and much easier to communicate with the principal and teachers at her lower-
performing school. Because these school settings map onto different levels of in-degree
centrality for each coach, suggesting that Sarah and Cindy are positioned differently as
instructional experts across their assigned schools, these selected school sites provide a rich
sample for observing social and other school conditions that inform the brokering practices and
social positioning of these DLCs.
In each school site, I focused my data collection on school principals (n=4) and a
purposive sample of teachers (n=15) who reported seeking instructional advice from their DLC
and not seeking instructional advice from their DLC as observed in my social network data. I
sample teachers in this way to observe a range of teacher perspectives on the instructional
expertise and role of DLCs. Where possible, I sampled teachers across grade-levels, who were
both novice and more experienced instructors, who were DFs and who were not DFs, and/or who
were reported by their colleagues as being a central source of advice for integrating technology
with instruction. Table 14 provides details on the teacher interview participants from each
school.
48
Data Sources
I rely on a combination of school network and interview data to inform my analysis.
First, I map the school networks for each of the school sites in my study. These network graphs,
48
While I was only able to speak with two DFs at Brand EL, I arranged for interviews with last year’s DLC and the
school-level CC and at this school site (and at York EL) to get more detail on Cindy’s relations with teachers at this
school.
BUILDING NETWORKS FOR CHANGE
115
shown in Figures 8 and 9, display the incoming and outgoing ties of DLCs, school-level CCs,
principals, and teachers in each school. I use these school networks to describe the social
positioning and brokering practices of Sarah and Cindy in each reform network. I then use this
information to triangulate my interview data on Sarah and Cindy’s brokering practices that
contribute to their social positioning in schools.
Interviews. I conducted interviews with DLCs along with principals and teachers in their
assigned school sites. I conducted these interviews in the Spring of 2015 following the collection
and initial analysis of my social network data. Interviews with DLCs lasted for one to two hours
and inquired about their prior experience in instruction and adult education; goals for improving
teacher practice; routines for meeting with teachers in schools; factors that enabled and/or
constrained their work with teachers; and interactions with district and school leaders and other
instructional support staff (e.g., CCs or other DLCs). When asking DLCs about their interactions
with others, I prompted DLCs to reflect on proximate time periods and, where relevant, share
artifacts from these interactions. For example, I asked DLCs to discuss the last time they met
with a teacher in their school, where they met, what they talked about, and who else was there.
My interviews with teachers lasted for 30 to 40 minutes and inquired about their
instructional background; goals for student learning; participation in PD related to LUSD’s
technology rollout and the CCSS, and their ego-networks for seeking advice on integrating
technology with instruction and teaching the CCSS. In discussing teachers’ social network
responses, I asked why they nominated certain colleagues and not others, and the routines that
informed their interactions with colleagues such as topics discussed, duration, setting, established
norms or procedures for interaction, and information and/or materials that were exchanged.
Across interview participants, I found that the individuals whom they mentioned as a source of
BUILDING NETWORKS FOR CHANGE
116
instructional advice matched the individuals nominated in my social network data collected in
the late-Fall of 2014, suggesting that teachers’ ego-networks for advice were relatively stable
across the school year.
As part of this discussion, I made sure to focus on teacher interactions with their DLC,
asking them to explain why they did (or did not) approach their DLC as a source of instructional
support and to describe their interactions. Once again, I prompted teachers to reflect on
proximate time periods. When interviewing DFs or teachers with high in-degree in their schools,
I also asked about their practices for sharing instructional advice with teachers and factors that
facilitated and/or hindered their efforts to collaborate with colleagues.
I also interviewed school principals at my case sites. Given the central role of school
principals in creating a school environment that is supportive of coaching, my interviews mainly
focused on understanding how school principals were leading efforts to integrate technology
with instruction of the CCSS. These interviews lasted for 30 to 40 minutes and inquired about
the instructional background of school principals, including their familiarity with technology and
standards-based instruction; their goals for student learning; their development of PD and
interactions with DLCs for improving teacher practice; their efforts to foster teacher
collaboration in learning and instructional change; and factors that facilitated and/or hindered
their efforts to support instructional change.
In addition to these school-level interviews, I interviewed central office leaders,
administrators, and support staff to provide further context for my findings. These central office
participants included LUSD’s Superintendent, Assistant Superintendent, Coordinator of
Education Technology, two elementary school CCs and another elementary school DLC. These
conversations focused on LUSD’s long-term priorities for teaching and student learning, the
BUILDING NETWORKS FOR CHANGE
117
theory of change behind LUSD’s technology rollout and intended outcomes of this reform,
district leaders’ efforts to communicate policy goals to school staff, and the design and
implementation of the DCP. Appendix C maps my interview questions onto categories and
constructs from my conceptual framework.
Data Analysis
I used rigorous analytical methods to identify themes and patterns in my interview data
that can shed light on the relationship between the social positioning of coaches and their
brokering practices for each individual case in my study (Stake, 1995). Because there is a
growing body of literature on the social positioning of coaches in schools and the conditions that
shape interactions between teachers and coaches on instructional change (e.g., Atteberry & Bryk,
2011; Marsh et al., 2010; Matsumura et al., 2010; Matsumura & Wang, 2014), I take a deductive
approach to analyzing my data, using a coding schema that pulls on key themes identified in this
literature as shown in my conceptual framework. To complement this approach and make sure
that I cover all the dimensions of each case in my data, I also draw on inductive coding
techniques to developing codes and identifying themes pertaining to my analytical questions.
Memoing and describing the case. Prior to coding my interview data, I read all
interview transcripts to re-familiarize myself with the data. As part of this process, I wrote
analytical memos to document initial reactions that further informed my coding process,
adjusting and adding new codes where relevant. For example, while my conceptual framework
suggests that DLCs can engage teachers in dialogue as a communication strategy for supporting
instructional change, I developed more specific constructs of what this dialogue could entail,
including dialogue on technical issues of using technology, student engagement, lesson
pedagogy, and lesson content standards and materials. Perhaps more importantly, I used this
BUILDING NETWORKS FOR CHANGE
118
initial memoing process to generate an understanding of each individual DLC and school-level
case. This involved looking across memos and notes for each interview to describe the
instructional background of each DLC and the social environment of their schools.
Coding procedures. Following my initial memoing, I classified and sorted my data into
categories and constructs derived from my conceptual framework. Specifically, I organized my
data into broader categories for the instructional background of coaches, their social
environment, other contextual factors in their surroundings, and their communication strategies
for engaging teachers on instructional change. Within these categories, I sorted my data into
specific constructs of interest. For instance, I organized interview data on the individual qualities
of coaches into dimensions of their instructional background, such as statements about their
technological, pedagogical and content knowledge, goals for teacher learning and change,
interpersonal skills, and so on. This sorting process allowed me to focus my analysis on relevant
areas for understanding the social positioning of DLCs, how this social positioning informs the
brokering practices of these DLCs, and how brokering practices reinforce their social
positioning. At this stage of my analysis, I remained open to inductive code that emerged as I
sorted my data into relevant categories and constructs. I coded each interview transcript several
times until I was confident that all data had been sorted into the appropriate category for
analysis.
After coding my data, I began my case analysis. Because I am using a multiple case
design, my case analysis evolved over several stages, beginning with drawing inferences from
my interview data for each school-level case and then aggregating my findings to each individual
DLC and then across all DLCs in my sample. This involved inductively coding my data across
the different categories and constructs of my conceptual framework to answer a series of
BUILDING NETWORKS FOR CHANGE
119
analytical questions (Neumann, 2006). For each school-level case, I asked the following
analytical questions:
1. What are the brokering practices of this coach at this school?
2. How do these individual qualities of this coach inform her brokering?
3. How does this school’s social context inform the brokering practices of this coach?
4. What other school and district conditions inform the brokering practices of this coach?
5. How do the brokering practices of this coach inform her social positioning at this school?
I then aggregated my responses to the above questions for to each DLC-level case by
answering the following analytical questions.
1. What are the similarities and differences in the social positioning of this DLC between her
two schools?
2. What factors (e.g., brokering practices, individual qualities, social conditions, other school
and district conditions) contribute to these similarities and differences?
Similarly, to identify points of convergence and divergence across the DLC cases in my sample,
I asked the same analytical questions outlined above except that I now compared my findings
across each DLC case:
1. What are the similarities/differences in the social positioning across Sarah and Cindy?
2. What factors (e.g., brokering practices, individual qualities, social conditions, other school
and district conditions,) contribute to these similarities and differences?
As implied by the above questions, I used this cross-case analysis to highlight similarities and
differences in the relationship between social positioning of DLCs across their assigned schools
and across the coaches in my sample, grounding these findings in specific pieces of evidence so
that I could contextualize divergent findings as needed.
Representation of Data
BUILDING NETWORKS FOR CHANGE
120
I use data tables, matrix displays, and figures to ensure the transparency of my findings,
present emerging themes, and highlight similarities and differences between DLC and school-
level cases. To ensure the transparency of my findings, I have prepared a data table that maps
findings for each individual DLC and their school to interview quotes supporting these
conclusions (see Appendix D). In this table, I also note the number of interviews with central
office administrators, DLCs, school principals, and teachers that provide evidence in support of
this claim, demonstrating that I triangulated findings across stakeholder groups before arriving at
final conclusions.
I also prepare matrix-displays and figures to visualize my main findings for each research
question. Tables 15 is a matrix-display that outlines district-wide conditions and Sarah and
Cindy’s brokering practices that contributed to these coaches not being central in LUSD’s CCSS
reform network (top-panel) and using their centrality in LUSD’s technology reform network to
varying degrees to foster teacher collaboration on instruction (bottom-panel). These tables are
color coded to reflect the degree to which DLCs are bridging the use technology in classrooms to
CCSS instruction and bridging teacher practice to central office goals for technology-enabled
instruction through collaborative coaching models (dark colors indicate a strong degree of
bridging and lighter shades indicate progressively less degrees of bridging). The text in these
tables provide a brief description of the district-wide conditions and brokering practices of DLCs
that contribute to their social positioning in schools.
To highlight the underlying factors that inform Sarah and Cindy’s brokering practices, I
then plot their brokering practices in Figures 10 and 11 according to the individual attributes
(vertical axis) and school conditions (horizontal axis) of these coaches. These plots demonstrate
if there is further variation in Sarah and Cindy’s brokering practices across school sites that
BUILDING NETWORKS FOR CHANGE
121
cannot be fully captured by way of matrix-displays. They also provide a spatial format for
comparing Sarah and Cindy directly to one another in terms of their brokering practices and
contextual factors, demonstrating which coaches are best positioned to guide teachers on CCSS
instruction and facilitate teacher collaboration in schools and the individual and school
conditions that drive these differences.
Trustworthiness
To build confidence in the value of my findings, I engaged various procedures to
minimize researcher bias and verify the repeatability of observations and interpretations
(Cresswell, 2007). Given that qualitative case study research draws upon experiential knowledge
to convey insights to the reader, I also contextualized my findings in sufficient detail and
accuracy to avoid misinterpretation and to support the transferability of claims (Cresswell, 2007;
Stake, 1995). I used several strategies to ensure the trustworthiness of my findings, including
prolonged engagement in the field, triangulation of data sources and methods, thick description,
and member checking.
When collecting my data, I used the entire Spring semester to speak with multiple
stakeholders at the central office and each school site. During this time, I developed a close
rapport with DLC participants so that I could fully grasp the context of their work and the school
sites. I also triangulated interview data from different stakeholders (i.e., central office leaders and
staff, DLCs, school principals, teachers) to reach my conclusions. As shown in Appendix C, I
used questions from my interview protocols with each group of study participants to collect data
on the different categories and constructs of my conceptual framework. This helped to ensure the
credibility and dependability of my findings, and to make sure that my biases as a researcher
were not influencing my interpretation of the data. My strategy to draw on ego-network data to
BUILDING NETWORKS FOR CHANGE
122
corroborate findings from my qualitative analysis further strengthened the confirmability of my
interpretations.
I also used thick-description and cross-case analyses to ensure that my findings were
contextualized in detail and can be transferrable to other contexts. The analytical questions that
informed my case analysis provided a clear structure for summarizing evidence and generating a
rich picture of each school and DLC-level case. The analytical questions that inform my cross-
case analysis strategy also help me focus on similar and different themes that emerge across
schools and DLCs, compelling me to provide explanations for when findings were transferable
and not transferable across contexts. As described above, I also used data representation
techniques (e.g., data tables, matrix-displays, and figures) to make my systematic process of data
analysis clear and situate my research conclusions in specific pieces of evidence. To the extent
possible, I also engaged in follow-up interviews and email exchanges with LUSD’s Coordinator
of Education Technology and DLC study participants to confirm my interpretation of the data.
49
Methodological Limitations
The qualitative component of my dissertation employs systematic data collection and
analysis procedures to generate trustworthy findings that explain the brokering practice of DLCs
in more detail beyond what I can observe through my social network data alone. That said, my
multiple-case study research design is subject to certain limitations. First, my sample of school
and DLC-level cases is small and not representative of the entire district. While I selected DLCs
and school sites to maximize variation in my data, they are not generalizable to all DLCs and
schools in LUSD. That said, because I intend to use this multiple-case study to provide in-depth
analysis of how the social positioning of DLCs rather than generalize my findings across DLCs
49
I plan to do more member checking once I have settled on the final interpretation of my qualitative data.
BUILDING NETWORKS FOR CHANGE
123
and schools (a research objective that I have come closer to accomplishing with my social
network data instead), this limitation is not so important.
A more substantial limitation is my reliance on self-reported data to account for the social
positioning and brokering practices of DLCs rather than observing these processes in action.
Capacity constraints and access to school sites limited my data collection efforts to interviews
only. This can be problematic to the extent that DLCs, principals, and teachers reflect and
discuss their professional practice in ways that are drastically different from real-life action. I
counteract this bias to the extent that I can by triangulating data across multiple interview
sources. Nevertheless, all discussions of my findings should still be considered within the
confines of this limitation. Finally, while I have engaged in various procedures to ensure the
trustworthiness of my findings, it is still possible that my findings are vulnerable to my own
biases as a researcher.
BUILDING NETWORKS FOR CHANGE
124
CHAPTER EIGHT -- Qualitative Findings on Social Positioning of DLCs
Introduction
This chapter presents my findings from my multiple-case study on the social positioning
of DLCs in schools. Figure 8 shows Sarah and Cindy to be isolates in their school networks for
teaching the CCSS, which can be seen by these coaches sharing zero social ties to other actors in
their school networks. In contrast, Figure 9 shows these DLCs to be central actors in most of
their school networks for integrating technology for instruction and to be surrounded by varying
degrees of teacher collaboration. This can be seen by the larger-sized nodes for DLCs in these
school networks and the greater frequency of inward-pointing social ties (i.e., there are more
arrows pointed toward DLCs in these networks than other school actors). In addition, these
school network graphs show varying levels of social ties exchanged among principals and
teachers who are in direct contact with these coaches, with there being higher levels of teacher
collaboration surrounding Sarah at San Rafael and Glendale EL, and lower levels of teacher
collaboration surrounding Cindy at Brand EL. York EL is an interesting case on its own as Cindy
is essentially an isolate in this network despite her formal status and expertise for integrating
technology with instruction.
As a reminder, I focus my analysis on investigating two sub-research questions related to
the social positioning of Sarah and Cindy. The first sub-question examines why these DLCs are
not engaging teachers on CCSS instruction. Specifically, I ask why DLCs are not at all engaged
in their school networks for teaching the CCSS? The second sub-question examines how these
DLCs might be using their power and influence in school networks for integrating technology
with instruction to coordinate information sharing among teachers. Specifically, I ask how, if at
all, are DLCs using their central positioning in school networks to support teacher collaboration
BUILDING NETWORKS FOR CHANGE
125
on technology-enabled instruction? To answer these sub-questions, I draw on qualitative data
collected from two case DLCs – Sarah and Cindy -- and four of their school sites (sub-cases). I
code these data according to my conceptual framework (Figure 7) that relates the social
positioning of these DLCs to their brokering practices, individual attributes, social and other
school conditions, and their broader district context (specifically LUSD’s design of the DCP). In
what follows, I present my findings for each of these sub-questions.
Weak Social Positioning of DLCs as CCSS Experts
Table 15 shows a combination of district conditions and brokering practices weakly
positioned Sarah and Cindy as experts for CCSS instruction in schools. Each of these factors are
color-coded to reflect whether they strongly (dark green), moderately (medium green), or hardly
(light green) position Sarah and Cindy as CCSS experts in schools. As this table demonstrates,
both coaches encountered district conditions and engaged in brokering practices that made them
less relevant for supporting CCSS instruction. First, the district central office articulated the
responsibilities of DLCs in ways that isolated their instructional support from signature practices
(i.e., Balanced Literacy) for improving student literacy as evaluated in CCSS assessments (light
green). Within this district context, both DLCs engaged in brokering practices that, to varying
degrees, buffered teachers’ use of technology from core instructional processes for improving
student achievement. Sarah was more process-oriented in her coaching approach, helping them
improve their instructional planning to develop students’ 21
st
century learning skills as evaluated
in the Common Core (dark green), while Cindy was more product-oriented and shielded teachers
from reflecting on instructional planning (light green). Both coaches, however, failed to
explicitly ground their support for teachers in Balanced Literacy instruction (light green),
resulting in teachers in perceiving Sarah and Cindy (and their instructional support) as
BUILDING NETWORKS FOR CHANGE
126
inconsequential to improving students’ reading proficiency and performance on CCSS
assessments. I find that the individual attributes of Sarah and Cindy as coaches, along with social
and other conditions in their school environment, contributed to similarities and differences in
their brokering practices. In this section, I discuss the district context of these DLCs, their
brokering practices, and the individual and school conditions that informed their brokering in
more detail.
Formal Responsibilities of DLCs Distanced from Balanced Literacy
As discussed in chapter 4, the LUSD’s superintendent intended for teachers to “teach
rigorous standards with engaging practices and support from technology.” While the expectation
was that these instructional practices would reinforce one another, the central office adopted
signature practices with different underlying philosophies to bring this vision to fruition. As
mentioned earlier, LUSD adopted the SAMR scale to guide teachers on how they should be
using technology in their classrooms, placing an emphasis on inquiry-led learning experiences
for students to develop their 21
st
century skills (the “4Cs”). In contrast, LUSD adopted Balanced
Literacy as a district-wide signature practice for supporting elementary instruction of the CCSS
in ELA, focusing on differentiated instruction to build student proficiency in reading and
writing.
50
These different signature practices – one focused on inquiry-led learning and another
focused on personalized instruction – distanced LUSD’s goals for teachers to “teach rigorous
standards” (via Balanced Literacy) from its simultaneous goal for teachers to teach with
“engaging practices and support from technology” (via the SAMR scale), with DLCs being
positioned to support teachers with the latter.
50
In this chapter, I focus on the juxtaposition of the SAMR scale and Balanced Literacy since my DLCs and their
case schools are at the elementary school level.
BUILDING NETWORKS FOR CHANGE
127
As noted earlier, LUSD created two frontline staff roles to support its signature practices;
one for CCs to develop curriculum maps and teaching guides to help teachers design lessons
under the Balanced Literacy framework and one for DLCs to support teachers to integrate
technology with instruction under the SAMR scale. DLCs and CCs were supposed to specialize
in their respective roles, while also coordinating resources and best practices among one another
to inform their work with teachers. This assumed coordination was crucial because, unlike
DLCs, CCs were not assigned to work at schools on a week-to-week basis, thus making it
imperative for DLCs to incorporate insights from CCs on Balanced Literacy instruction into their
coaching. However, while it was expected that DLCs and CCs would coordinate their
instructional support for teachers, principals and teachers did not perceive this to be the case,
instead seeing DLCs as providers of technical support and facilitators of 21
st
century learning
(i.e., the 4Cs) and CCs as content and pedagogical support providers for Balanced Literacy and
the CCSS. These perceptions were to a large extent shaped by how the central office defined the
work responsibilities of DLCs, only requiring DLCs to visit their school sites for one day in the
week and to dedicate most of their time toward supporting their DFs (two or three teachers per
site). As such, principals and teachers perceived DLCs as a luxury support staff, there to support
a select group of teachers (DFs) to add a technology component to instruction but not to improve
the core instructional work of schools.
Both Sarah and Cindy acknowledged that principals and teachers perceived them more as
technology experts rather than experts of content and pedagogy. As Sarah explained, teachers
“see me more as just a technology expert and then, as the year goes on, they are like, ‘Oh, we can
plan curriculum,’ and ‘Oh, I can become a better teacher.’ It is kind of an ‘ah-ha’ for them.” In
Cindy’s situation, both her schools (Brand and York EL) used their budgets to hire in-house CCs
BUILDING NETWORKS FOR CHANGE
128
for Balanced Literacy instruction (herein referred to as “Balanced Literacy coaches”) on the
assumption that Cindy would could not provide this support to teachers. In the case of York EL,
the principal relied almost exclusively on this Balanced Literacy coach, setting expectations and
making plans for this Balanced Literacy coach to meet with teachers across all grade-levels but
not coordinating with Cindy in the same way. While Sarah did not have Balanced Literacy
coaches at her school sites, both her schools (San Rafael and Glendale EL) hired external
consultants to provide PD on Balanced Literacy instruction, once again suggesting that principals
and teachers did not see her as a dedicated resource for this signature practice.
Process versus Product-Oriented Coaching
While LUSD’s configuration of teacher PD distanced DLCs and their instructional
support from CCSS instruction, Sarah and Cindy’s brokering practices further contributed to this
distancing to different degrees. Sarah was more process-oriented in that she supported teachers to
map out entire curriculum units of instruction and make deliberate choices in using technology to
facilitate student-led inquiry of content standards. This process-oriented approach focused on
instructional planning allowed Sarah to bridge her support for technology-enabled instruction to
teacher goals for instructing the CCSS. Sarah used project-based learning (PBLs) as a central
practice for pursuing this work with teachers. In facilitating these PBLs, she regularly engaged
DFs in reflective dialogue on how to push student autonomy and self-expression in learning,
bridging teachers toward a deeper understanding of how LUSD’s SAMR-scale related to their
instructional planning and anchor standards of the CCSS focused on 21
st
century learning skills.
This reflection is apparent in the following quote from one of her DFs (Teacher 6) at Glendale
EL:
BUILDING NETWORKS FOR CHANGE
129
It was always an open discussion of what do you want to work on and how can we make it
more engaging? How can we take it to the next level [on the SAMR scale]. It wasn't
substitution, but we would think of an activity, and then how could we make that activity
better with technology, so we always had these conversations of, ‘Okay, well I want my
kids to do this. What do you think? How can we do that in a way that is going to get them
to put their creativity into it and support their freedom of expression and things like that,
while collaborating with each other?’ It would just be this constant dialogue of, ‘What do
you think?’, ‘What apps have you heard of?’ that type of thing.
Unlike Sarah, Cindy was less concerned with developing teachers’ skills in curriculum
mapping and lesson planning and more focused on adding a technology component to teachers’
existing instructional routines. As such, she intended to engage teachers in more product-oriented
coaching strategies that buffered her support for technology-enabled instruction from teachers’
efforts to design lessons around content standards from the Common Core. As Cindy explained,
“my main responsibilities are to inspire teachers to really use and think about using technology
in their lessons to engage students…. ‘Show me what your lesson plan is. Show me what you're
doing, and let's see if we techify it.” Rather than engage DFs in prolonged phases of instructional
planning, Cindy exposed her fellows to lesson designs featuring new apps that she had already
created, assuming this exposure would then inspire their own creativity in instructional planning.
This product-oriented approach, however, restricted Cindy’s dialogue with DFs to focus mainly
on the use of technology in their classrooms and general conversations about student learning
(i.e., were students able to follow along in lessons or not), buffering teachers from more in-depth
conversations of pedagogy and content choices. The following quote from Teacher 11 at Brand
EL best captures Cindy’s coaching style in terms of providing DFs with ready-made solutions for
BUILDING NETWORKS FOR CHANGE
130
integrating technology with instruction at the avoidance of more complex, problem-solving
discussions on instructional planning.
If I was in a place where I wanted to do a lesson that incorporated technology and I
maybe just did not have an idea for it or to know where to go, she would create it. She
would say, “What if we did this? What do you think about this?” She would give me an
entire thing to use. I would go, “Oh my gosh, that is awesome,” and then we would use
it. I did not have to spend nights agonizing over what app am I going to use? What
project am I going to use for this? How am I going to do this? She would say, “Here are
some things that have already been created. Would any of these work in your
classroom?”
Although Sarah was more process-oriented than Cindy in terms of instructional planning,
both coaches were equally process-oriented in terms of helping teachers adopt classroom
practices that boost student engagement in learning. Both coaches recognized that teaching with
technology introduced a new element to classroom management that teachers were unprepared
for, such as scheduling time to teach students about technology, pacing lessons so that students
can focus on lesson content while also mastering technology and other 21
st
century learning
skills, and reconfiguring the physical layout of classrooms to support student sharing of iPads
(not all classrooms had one-to-one access to iPads) and collaborative exchanges between
students while using technology. Sarah and Cindy also felt strongly about facilitating student
knowledge exploration and self-expression to advance classroom instruction further along the
SAMR scale and promote inquiry-led learning. To create these shifts in teacher practice, Sarah
and Cindy regularly modeled student questioning strategies for teachers and provided feedback
to DFs on their classroom management, leaving teachers with concrete impressions of how to
BUILDING NETWORKS FOR CHANGE
131
engage students in higher-order thinking. The following quotes from teachers at their school sites
are particularly telling:
Teacher 11 (Brand EL): [In observing Cindy teach], I learned her mannerisms with
students, engaging students. I would try and pay attention to the questions she would ask
students when they got stuck. I would hear her say, “What would you do now?” or just
the way she would talk to them about their next steps. I was like, “That is a good way to
put it.”
Teacher 4 (San Rafael EL): Sarah planned a little script [for the students] … because
[she] reminded me, you only have 30 seconds on ChatterPix…It is kind of common
sense, but it was nice that she had [the student speaking points] on a piece of paper. I
thought, "Well, of course I am going to have the kids say, ‘Hi, my name is so and so,’
and then go straight into their content. [But] she formalized it a little bit more. I would
have had a more informal set up. I would have just had them on the carpets, saying,
"Hey, number one, you do this." I would have written it down on the board. "Number
two, you do this." Then I would have modeled it for them. She formalized it and I have
to say, that it worked better when implementing the lesson.
Minimal Focus of Instructional Support on Balanced Literacy
While Sarah and Cindy engaged in varying degrees of process-oriented strategies to
bridge their support for technology-enabled instruction to teacher goals for CCSS instruction,
both coaches failed to connect their instructional support explicitly to Balanced Literacy and
teachers’ goals for improving students’ reading proficiency. Because she was more process-
oriented in instructional planning, Sarah did make more of an effort to understand teachers’
learning objectives for Balanced Literacy and embed new digital apps in learning tasks for
BUILDING NETWORKS FOR CHANGE
132
accomplishing these goals. For instance, in helping a team of third-grade teachers at San Rafael
EL use technology for Reader’s Workshop, Sarah showcased an app (Padlet) that students could
use for shared and interactive reading, specific components of the Reader’s Workshop model that
the third-grade teachers happened to be working on at the time. In contrast, Cindy framed her
instructional support for teachers in Balanced Literacy as focusing on “how to” use technology
rather than helping teachers understand and implement specific teaching principles from
Balanced Literacy (e.g., conferring, shared reading). The following quote describing a joint PD
session between Cindy and the Balanced Literacy coach at York EL shows this to be the case.
We created a presentation. [The Balanced Literacy coach] talked about the reasons why
you would incorporate conferring. You need to talk to children, find out what their
interest levels are, if they are understanding what they are reading, things like that. I went
into how you can Confer using technology to make your life easier… My role was to
show, “You could write it on paper or you could do these things using technology.” It
was just the “how to.”
Despite Sarah being more intentional in helping teachers identify the connections
between technology and Balanced Literacy, the fact remains that both she and Cindy rarely
provided such instructional support to teachers. Moreover, in these rare instances, Sarah and
Cindy came up with lesson designs featuring simplistic uses of technology to support Balanced
Literacy and that did not advance their instructional practice further along the SAMR scale,
regardless of their different approaches to improving teacher practice. These lesson designs
buffered Sarah and Cindy’s instructional support from CCSS instruction, demonstrating that it
was not possible for teachers to excel in both the SAMR scale and Balanced Literacy at the same
time. Examples of these overly-simplified lesson designs include teachers using Google Drive or
BUILDING NETWORKS FOR CHANGE
133
Padlet for students to take notes while reading content (as developed in the third-grade classroom
discussed above), or students using QR codes to access comprehension questions about a text,
both of which would be at the level of substitution on the SAMR scale but might have been
helpful in facilitating specific teaching principles of Balanced Literacy. Sarah’s DFs also
expressed doubts as to whether inquiry-based approach of PBLs was conducive for supporting
student engagement in reading and could be merged with instructional routines from Reader’s
and Writer’s workshop, once again suggesting that Sarah did not engage teachers in explicit
conversations on how to reconcile principles of inquiry-based learning as found in the SAMR
scale with the differentiated teaching strategies of Balanced Literacy. In buffering technology-
enabled instruction from teachers’ ongoing work to teach Balanced Literacy, Sarah and Cindy
positioned themselves squarely as experts for helping teachers use technology for instruction and
not as experts for helping teachers teach the CCSS.
Individual and School Conditions Informed DLC Brokering Practices
Both the instructional background of Sarah and Cindy as coaches, along with conditions
in their school environment, informed their brokering practices and ultimately their weak
positioning as CCSS experts in schools. Sarah and Cindy had varying levels of instructional
expertise and prior experience in leading PD that contributed to their process versus product-
oriented coaching styles. In addition, both coaches were unfamiliar with Balanced Literacy as a
framework for teaching, which explains why they did not emphasize this signature practice as
part of their coaching. Sarah and Cindy’s brokering practices were also influenced by the extent
to which their schools demonstrated strong cultures for teacher learning and change, and had
prior exposure to technology and Balanced Literacy instruction so that teachers could identify
BUILDING NETWORKS FOR CHANGE
134
the connections between these reforms and approach Sarah and Cindy for more coordinated
support.
Figure 10 plots the influence of these individual and school conditions on Sarah and
Cindy’s brokering practices. The vertical axis in this figure represents Sarah and Cindy’s
instructional expertise and prior experience in leading teacher PD, while the horizontal axis
represents their school context in terms of the prevailing culture for teacher learning and change
and prior exposure to technology and Balanced Literacy. I plot Sarah and Cindy’s brokering
practices at each school on these axes, making for a total of four points (two for each coach at
their respective school sites). Points on the upper, far-right of this plot represent brokering
practices that closely bridge technology to CCSS instruction, whereas points on the lower, far-
left of this plot represent brokering practices that buffer technology from CCSS instruction. The
points on this figure demonstrate that Sarah benefited from individual and school conditions that
allowed her to engage in more bridging practices than Cindy. Sarah is positioned in the top-half
of this figure while Cindy is consistently positioned in the bottom-left square of the figure for
both schools. There does appear to be some variation in Sarah’s brokering practices across
schools, with San Rafael (upper-right square) proving to be a more conducive setting for
brokering practices that bridge technology to CCSS instruction than Glendale EL (upper-right
square but closer to the center). I discuss these individual and school conditions and their
influence on Sarah and Cindy’s brokering practices in more detail below.
Instructional expertise and PD experience of DLCs. Sarah and Cindy’s varied
instructional expertise and prior experience in leading teacher PD contributed to their different
coaching styles. While Cindy had more years of teaching experience than Sarah (see Table 13),
both coaches indicated that their recent exposure to using technology for instruction, which
BUILDING NETWORKS FOR CHANGE
135
occurred during the initial stages of LUSD’s technology rollout, had the most influence on their
coaching. This meant that, as a teacher who developed PD for the district on Haiku (LUSD’s
online learning content management system) and was now in her second year working as a DLC,
Sarah had more experience in guiding teachers on how to use technology for CCSS instruction.
As a first-year DLC who had not developed or led teacher PD in the past, Cindy had much less
of this valuable experience to inform her work in the classroom. These individual factors can
explain, in part, why Sarah was more adept at engaging teachers in process-oriented routines for
instructional planning, positioning Sarah on the top-half of Figure 10 and Cindy on the bottom-
half of Figure 10.
Additionally, while both coaches were motivated to make DFs better teachers,
recognizing that improving teachers’ technological skills alone would not improve teaching and
learning, they relied on different pools of expertise to inform this work. Sarah drew on her
expertise in supporting teachers across content areas and connecting technology to curriculum
content and learning standards, allowing her to engage teachers in conversations around
curriculum mapping and lesson planning. This is reflected in the following quotes from Sarah
and her DFs.
Sarah: This position has taught me a lot about technology and the different tools that are
out there. Knowing that and knowing the curriculum and what is expected of students, I
like to think that I'm good at putting those together so that it is more effective in the
classroom.
Teacher 8 (Glendale EL): After working together on Response to Intervention (RTI), we
went into math. I feel very passionate about math…Sarah started introducing flip lessons,
which completely opened me up to be able to work with my struggling students. We
BUILDING NETWORKS FOR CHANGE
136
would create flip lessons using a QR code, sometimes a Google Doc, sometimes they
would write on paper. They would watch videos and learn from the videos. It really
helped me with my math instruction because I have some kids that were really
struggling. I was able to work with them for an hour while the other kids were
independently learning.
In contrast, Cindy’s main strength as a coach involved helping teachers get comfortable
with using technology in their classrooms. Unlike Sarah, she expressed less self-efficacy in
supporting teachers across content areas, demonstrating a preference for language arts and social
studies instruction over mathematics. These factors likely contributed to Cindy focusing on
adding a technology component to teachers’ existing instructional plans, and guiding teachers on
classroom management and student questioning strategies that facilitated this technology use,
rather than supporting teachers in instructional planning. The following quotes from Cindy and
her DFs demonstrate that Cindy was mainly valued for her ability to build teacher self-efficacy
for integrating technology with instruction.
Cindy: I excel in getting to know the teachers and getting to know their strengths,
weaknesses, insecurities, things like that, and I start there. From there I feel like I can
introduce things. If there is a teacher that I realize, okay, they have no technology skills
at all, I am not going to go in and by like, “Oh, you can do this, this, this, this, and this. I
am going to change [all] this. This is what you should do."…A lot of times I have noticed
some of the younger teachers are more technologically advanced than the older teachers,
and more willing to just go for it and try new things, so just getting a sense of who they
are from the start I think helps a lot.
BUILDING NETWORKS FOR CHANGE
137
Teacher 10 (Brand EL): She was there to use the technology with the students… She
never pushed me to do something that I did not want to do. I do not know if this what
you mean but I was fearful in the beginning of the year [about using technology] but I
should not have been because it was an enhancement always.
Finally, both Sarah and Cindy were hindered by their lack of experience in using
Balanced Literacy as a framework for ELA instruction and in teaching students in tested grade-
levels. As discussed in chapter 4, LUSD did not recruit DLCs based on their familiarity with
Balanced Literacy because this signature practice had not yet been finalized at the time of DLC
selection. As such, neither Sarah nor Cindy mentioned using Balanced Literacy as a teaching
framework in their teaching careers. Sarah and Cindy also taught in non-tested grade-levels for
most of their teaching careers, which likely hindered their ability to help teachers prepare
students for mastery of content standards as evaluated in standards-based assessments. Sarah’s
unfamiliarity with Balanced Literacy and standards-based assessment took away from her
instructional expertise for guiding teachers on CCSS instruction, positioning her at the mid-way
point of the upper-half of the vertical axis in Figure 10. In contrast, Cindy’s unfamiliarity with
Balanced Literacy and standards-based assessments positioned her lower along the bottom-half
of the vertical axis.
School cultures for teacher learning and change. Sarah and Cindy’s coaching styles
were also shaped by their school cultures for teacher learning and change. At San Rafael, Sarah
benefited from having a principal who was proactive in building a culture of instructional
innovation through structured supports for teachers. These supports included problem-solving
sessions with teachers at the beginning of the school year on what technology-enabled
instruction should look like in the classroom and what it means to progress along the SAMR
BUILDING NETWORKS FOR CHANGE
138
scale, and purchasing additional iPads so that more teachers could have one-to-one access to
technology and hence more time to use technology with students. The principal also evaluated
teacher practice based on 21
st
century teaching skills and scheduled time for teacher
collaboration in PLCs. These conditions motivated teachers to learn about inquiry-based learning
and embrace a PBL approach to instruction, making it easier for Sarah to then work with DFs
and non-fellow teachers on improving their instructional planning and classroom practice. The
conducive school culture at San Rafael EL positioned Sarah further along the right-half of the
horizontal axis, providing the most ideal setting among all four school cases for bridging
coaching on technology-enabled instruction with instruction of the CCSS.
At Glendale EL, Sarah had worked with a similarly proactive principal during her first
year of coaching at the school. However, in her second year of coaching, the school was assigned
a new leader who was completely out of touch with the technology resources at her school and
did not have strong relations with teaching staff. This unfortunately meant that the principal was
very hands-off in her leadership style and provided minimal guidance or feedback to teachers on
their classroom practice. In this environment, Sarah largely depended on teachers’ own drive for
instructional innovation (sparked by their principal from the previous school year) and her
existing social capital with teachers (having previously taught at Glendale EL) to promote her
PBL approach. While Sarah was successful at promoting PBL as a joint enterprise with her DFs,
she was less successful at expanding PBL to non-fellow teachers in the absence of principal
support and coordination (which I discuss later in this chapter). Sarah also felt that the current
principal at Glendale EL did not value her joint enterprise with DFs to use PBL as a platform for
teaching the CCSS, pointing out that the principal did not renew the contract for one of her
trained DFs for the following school year, which Sarah perceived as a huge loss of human capital
BUILDING NETWORKS FOR CHANGE
139
for the school. This less innovative environment at Glendale EL positioned Sarah closer to the
center of the horizontal axis in Figure 10, making her less effective at bridging her support for
technology-enabled instruction to teachers’ goals for teaching the CCSS.
At Brand and York EL, Cindy encountered school cultures that were slow-moving in
embracing technology as a central resource for teaching and learning. As shown in Figure 10,
these slow-moving cultures positioned Cindy on the left-hand size of the horizontal axis, creating
environments that buffered Cindy from teacher instruction of the CCSS. This buffering
manifested itself in various ways, including principals framing technology as a non-essential
resource for classroom instruction and relying on Cindy for technical assistance instead of
pedagogical and content support.
First, both principals at both schools framed technology as a secondary resource for
instruction, recognizing that it was important for supporting CCSS instruction but that teachers
did not have to master using technology right away. This framing was in response to different
school conditions. At Brand EL, the principal was relatively new to being a school leader (this
was her second year as a school principal and working at Brand EL) and as such, did not have a
fully developed strategy for motivating teachers to use technology to transform student learning.
As she put it:
We are Reader’s and Writer’s school and next year, we have a bunch of teachers that are
ready to start working on CGI and they are already doing it themselves…As an English
language learner school, we need to really push for lots of discussions and high levels of
engagement, as much authentic learning as possible, bringing in project-based learning,
which is something we are learning right now. We are just dabbling in that piece, that is
sort of a new layer for us.
BUILDING NETWORKS FOR CHANGE
140
This slow-moving approach to instructional change, in turn, made it difficult for Cindy to
engage teachers in Brand EL on technology and instruction, much less to push teachers to
critically reflect on instructional planning and lesson design. For instance, unlike Sarah who
worked with DFs who were motivated to use technology for instruction, Cindy had to spend
more time persuading her DFs on the instructional benefits of technology and addressing their
fears about using technology in their classrooms.
At York EL, the principal was in charge of a school that had a long history of principal
turnover and strong personalities among the teaching staff who were very resistant to LUSD’s
vision for personalized and inquiry-led instruction. The following quote from the Balanced
Literacy coach at this school site capture the tensions at this school quite vividly:
It is like Lord of the Flies. What happens when you do not have a leader, leaders grow
within yourselves...There is a handful of teachers there that are extremely strong willed
and extremely strong voices on the campus to the point that some people feel like they
cannot say anything. But that also kept that school moving in a positive direction so I am
not going to negate that there were some benefits to it but in truth, that is also now
making it difficult for the shift.
Perhaps because of the school’s tumultuous history, the principal at York EL was sensitive to her
staff’s hesitations about adopting Balanced Literacy and the SAMR scale as signature practices
for instruction. Instead, the principal focused on encouraging teachers to make small shifts in
their teaching; a leadership strategy that Cindy disagreed with because it failed to create a sense
of urgency for change and consequently made it more challenging for her to position technology
as a central resource for teaching and learning. As she explained:
BUILDING NETWORKS FOR CHANGE
141
I have seen definite steps towards change, but they are small. I feel like some of that is
the principal because she said, "Okay, we are taking baby steps," which is great and I
think that that is a good way to start, but now it is time to move.
In addition to not making technology a priority for student learning, both the principals at
Brand and York EL did not fully integrate Cindy in school-wide PDs focused on CCSS
instruction. At Brand EL, the principal rarely asked Cindy to organize school-wide PD on
technology-enabled instruction, instead paying a stipend to formerly-trained DFs at her school to
organize these PDs. While the principal at York EL asked Cindy to showcase lesson designs at
occasional voluntary staff meetings, teachers often perceived these demonstrations as too generic
in nature, making Cindy appear disconnected from and disinterested in their specific classroom
needs. As Teacher 13 explained:
I already have to sit and come up with all this other stuff on my own that they are not
giving us this year so there is technology again. Cindy [organized a PD with] ideas of
things you could do with PicCollage. At that moment I was working on science and
soils…I walked away with something to do that day…[but] it is all so general like, “[You
can use it with this] or whatever it is that you are working on with your kids”.
Cindy also did not follow-up with teachers at York EL after these PDs to provide more targeted,
in-classroom support, nor was she encouraged to do so by the school principal who was cautious
of “pushing [teachers] too hard and too fast.” Eventually, Cindy stopped offering these PDs due
to low teacher attendance at these events, suggesting that teachers did not perceive her expertise
and guidance as essential to their own instructional improvement.
Finally, the principals at Brand and York EL treated Cindy as a technical support staff
member, asking her to help teachers download Apple IDs on iPads or demonstrate how to use
BUILDING NETWORKS FOR CHANGE
142
new digital app and webpages on Haiku. These principals would also engage their own teaching
staff in highly technical discussion about technology rather than embedding these conversations
in instructional matters (e.g., helping teachers understand the SAMR scale and brainstorm ideas
for using technology to achieve specific learning objectives). Given this over-emphasis on the
technical skills required to use technology in the classroom, principals and teachers at Brand and
York EL generally perceived technology (and Cindy’s instructional support) as somewhat of a
luxury item that was peripherally related to CCSS instruction. This is demonstrated in the
following quotes from principals and teachers at both schools.
Principal at York EL: I think because Balanced Literacy became our new reading
practice, again, without a tool, teachers knew, I'm going to die unless someone shows me
how to do this. So they were probably more open to support, “Yeah, show me, show me,
show me.” For technology again, they just have their head [that] this is a tool I can use
when I want, whereas I have to teach reading every single day for two hours.
Teacher 10 (Brand EL): Our principal approached me just a few weeks ago and asked me
to be on the Tech Task Force. She said, “You are becoming the tech guru and a leader on
campus with this. I would like you to [take on this role] so that people can go to you” …I
am so technology, technology, technology now that I think I have made it known on
campus how much I love it and how important I think it is that I have teachers who will
email me all the time or just come up to me and say, “Can you come to my room and
help me with this?” One time the teacher asked me to help her with an iPad issue and I
could not figure it out. It was driving me crazy…I have become this person where I have
to figure out [these technical issues] immediately. I cannot just say, “I do not know how
to do it.”
BUILDING NETWORKS FOR CHANGE
143
Limited school exposure to technology and Balanced Literacy. Across all school sites,
there was an absence of human capital to craft a coherent vision for the joint implementation of
LUSD’s technology rollout and adoption of Balanced Literacy. Most teachers had never used
Haiku, iPads, and apps for instruction until LUSD’s technology rollout. Teachers were also new
to the idea of developing their own curriculum content and lesson designs for CCSS instruction
under the Balanced Literacy framework (instead of using curriculum scripts and pacing guides).
In addition, the principals at my case schools were either familiar with technology and the
SAMR scale (San Rafael) or Balanced Literacy (Glendale, Brand and York EL) but not with
both signature practices, making it difficult for them to guide teachers on the connections
between these instructional programs.
Given the novelty of LUSD’s technology rollout and Balanced Literacy, teachers had a
poor understanding how to integrate these programs in practice and did not know how to draw on
Sarah and Cindy for more coordinated support. The following quotes demonstrate this to be the
case across Sarah and Cindy’s schools:
Teacher 9 (Glendale EL): Part of it last year was I didn't always have something, I would
be like, "I do not know what to do. I need ideas from you [Sarah] because you are the one
that has seen all this technology and all these new things." It is like, I cannot ask [about]
what I don't know.
Balanced Literacy Coach (Brand and York EL): For example, take me using Padlet for
an interactive read aloud. Until you really understand and can do interactive read aloud,
you are not going to see the benefit of the technology that is going to help increase
learning… [As teachers become] comfortable with technology [then] as they learn more
with Balanced Literacy, they will pull it in. Or as they move further along the road with
BUILDING NETWORKS FOR CHANGE
144
Balanced Literacy and we are bringing the technology to them, they will be like oh, that
would be good.
As noted earlier, because Sarah and Cindy were themselves unfamiliar with Balanced Literacy,
they failed to make the connections between their coaching and this signature practice more
explicit, further weakening their position in schools to guide teachers on CCSS instruction. In
Sarah’s case, this lack of human capital positioned her closer toward the center of the horizontal
axis in Figure 10. In Cindy’s case, this lack of human capital positioned her closer to the left-
extreme of the horizontal axis in Figure 10.
Different Collaborative Practices of DLCs for Technology-Enabled Instruction
Table 15 shows that a combination of district conditions and brokering practices resulted
in Sarah and Cindy supporting teacher collaboration on technology-enabled instruction to
varying degrees. Each of these factors are color-coded to reflect whether they strongly (dark
orange), moderately (medium orange), or hardly (light orange) enabled Sarah and Cindy to
facilitate teacher collaboration in schools. The range of colors in this table suggests that Sarah
facilitated teacher collaboration on technology-enabled instruction more effectively than did
Cindy. Once again, district-wide conditions informed how botch coaches worked in schools,
with the district central office articulating the responsibilities of DLCs in ways that privileged
their support for DFs over and above their work with non-fellow teachers. This messaging, in
turn, created mixed incentives for DLCs to engage teachers in collaborative routines where DFs
could share knowledge on technology-enabled instruction with non-fellow teachers (medium
orange). In this context, Sarah seemed to engage in brokering practices that were more
collaborative and inclusive of non-fellow teachers (dark orange) while Cindy engaged in
brokering practices that were more exclusionary toward less-tech-savvy DFs and non-fellow
BUILDING NETWORKS FOR CHANGE
145
teachers (light orange). I find that the individual attributes of Sarah and Cindy, as well as social
and other conditions in their schools, influenced their brokering practices. In what follows, I
discuss the district context of these DLCs, their brokering practices, and the underlying factors
that informed their brokering in more detail.
Formal Responsibilities of DLCs Prioritized DFs Over Other Teachers
In general, the articulation of responsibilities of DLCs from the central office sent mixed
signals on how these coaches should engage teachers in collaborative routines for supporting
technology-enabled instruction. On the one hand, the central office required DLCs to visit their
school sites one day in the week and to prioritize supporting their DFs, inhibiting these coaches
from working with a broader pool of teachers. On the other hand, the central office still expected
DLCs to work with school principals to bring about school-wide improvements in teaching and
learning, but did not set clear expectations for these DLC-principal partnerships. And because
schools were caught up with implementing multiple reforms, including LUSD’s technology
rollout and adoption of Balanced Literacy among other initiatives (e.g., mainstreaming special
education students in regular classrooms, piloting CGI for math instruction), principals had
limited time to set aside for helping teachers integrate technology with instruction. These
scheduling constraints, alongside LUSD’s non-prescriptive approach for articulating how DLCs
should be promoting technology-enabled instruction on a school-wide basis, resulted in varying
degrees of teacher collaboration on technology-enabled instruction across schools.
In addition, the district established the DCP as a voluntary coaching program for
teachers. This design feature was intentional, as LUSD’s CET understood that technology-
enabled instruction was a time-intensive endeavor and wanted to assign DLCs to support
teachers who were most committed to improving their instructional practice. Recognizing that
BUILDING NETWORKS FOR CHANGE
146
this voluntary approach could restrict technology-enabled instruction to a limited pool of early-
adopting teachers, the CET selected DFs across grade-levels and subject areas so that all teachers
could receive some exposure to technology-enabled instruction, either through direct training
from a DLC or through interaction with DFs in their grade-level or department teams. Moreover,
DF participation in the DCP was contingent on these fellows agreeing to showcase their
instructional work at district or school-wide PD events, although this requirement for
participation was not monitored in any kind of meaningful way. The voluntary nature of DF
training resulted in school settings where certain teachers had far more expertise and skills for
technology-enabled instruction than others, creating tensions among staff which Sarah and Cindy
addressed in different ways.
Collaborative versus Non-Collaborative Brokering Practices
Within this district context, Sarah and Cindy adopted drastically different approaches for
supporting teacher collaboration on technology-enabled instruction, with Sarah being more
inclusive in her vision for instructional change than Cindy. Sarah’s collaborative coaching style
was apparent in her efforts to include non-fellow teachers in her work with DFs, regularly
arranging opportunities for DFs to collaborate with grade-level colleagues on PBLs. Sarah saw
this collaborative work as breaking down walls, helping teachers feel part of a community of
learners and share ideas for driving instructional innovation. She also viewed teacher
collaboration as a smarter coaching practice, minimizing her workload for helping DFs build
PBLs from the ground up while still exposing teachers to a range of ideas and resources. As she
put it:
As a coach, I try to work smarter not harder, and so my teachers that are on the same
grade-level, I usually try to do the same kind of thing with them. Since it was all their
BUILDING NETWORKS FOR CHANGE
147
first PBL, which can be very daunting, I thought, "Let's put you guys together and then
you can talk and collaborate," and so they collaborated through Google Drive.
Sarah’s inclusive approach to coaching was also evident in her efforts to spark teachers’
interest in using technology, going out of her way to provide in-classroom support to non-fellow
teachers so that they could experience first-hand the benefits of using technology to achieve
learning objectives that were meaningful to them. In working with these non-fellow teachers,
Sarah deliberately chose apps that teachers could readily access and master to minimize the
perceived time burden of using technology and maximize the perceived net gain from her
instructional support. For instance, in describing her approach to training third-grade teachers on
technology-enabled instruction at San Rafael (a grade-level that had not volunteered any teachers
for DF training), Sarah explained that that:
[The principal and I] wanted to get something where kids could collaborate and I wanted
to show a tool that they could use instantly, that did not need a lot of teacher prep, a lot
of time, because I knew that if it did, it would scare them away, since they are not techy,
or they do not think they are. I also wanted it to be something that was easily
accessible…not something that we would have to collect all their iPads [for uploading a
new app]. It was something we could do quickly.
These deliberate actions increased Sarah’s outreach to teachers and her ability to foster dialogue
among them, thereby bridging a broader pool of teachers to LUSD’s vision for technology-
enabled instruction. While the quality of instructional support that Sarah provided to non-fellow
teachers was certainly not at the same level as her support for her DFs, simply because Sarah did
not have as much time to support non-fellow teachers and often chose instructional applications
of technology that were more simple and straightforward to implement, she nevertheless created
BUILDING NETWORKS FOR CHANGE
148
an inclusive atmosphere at schools that encouraged teachers to learn about technology and seek
support from her and their colleagues.
Cindy differed from Sarah in that she did not prioritize teacher collaboration as a
coaching practice. Instead, Cindy saw herself as a dedicated resource for her DFs, “inspiring”
them to transform their instructional approach through the digital tools, lesson designs, and
classroom practices that she shared with them. Through this intensive support, Cindy assumed
that DFs would eventually learn to innovate with instruction and share best practices with their
colleagues but that she did not need to directly facilitate these exchanges. Instead, Cindy
expected non-fellow teachers to approach her and her DF’s for guidance when needed. As she
explained, “Ultimately, my responsibility is toward my fellows, but I am working with the whole
staff, really, so if anybody else comes to me and asks questions or needs help, I can meet with
them.” Cindy’s non-collaborative coaching style effectively buffered non-fellow teachers from
LUSD’s goals for technology-enabled instruction and hindered collaboration between DFs and
these non-fellow teachers.
Unlike Sarah who took it upon herself to work with all kinds of teachers, Cindy invested
more of her time and energy in supporting teachers whom she perceived as having the greatest
potential for change, buffering other supposedly less-talented teachers from instructional support.
This was apparent in terms of her priority for supporting DFs who self-identified as motivated
users of technology as noted earlier. Even among her DFs, Cindy differentiated her training and
support by spending more time with younger and more technologically-savvy DFs whom she
perceived as being more willing to learn and flexible in their instructional approach. This was
especially apparent in Cindy interactions with her DFs at Brand EL, where she provided more in-
classroom support and feedback to her DF who was an early-stage teacher (Teacher 11) and less
BUILDING NETWORKS FOR CHANGE
149
of the same to her DF who was a far more experienced teacher (Teacher 10). Cindy also
expressed more pessimistic views of her DFs at York EL where most teachers at the school were
veteran teachers, acknowledging that she rarely communicated with these teachers in between
their weekly meetings and that her DFs would only use technology in their classrooms when she
was there to do the work for them.
I feel like my two fellows at York EL are still in the stages where we have to come up
with ideas together. They need me to come up with directions for them or they will not
do it. Or it is only the day that I am there is the technology day which was not my goal.
As these examples suggest, Cindy’s non-inclusive coaching style prevented her from making
technology-enabled instruction accessible to a broader group of teachers who did not entirely
embrace her vision and goals for instructional change.
Individual and School Conditions Informing DLC Brokering Practices
The instructional background of Sarah and Cindy as coaches, along with conditions in
their school environment, contributed to the above brokering practices and Sarah and Cindy’s
varied support for teacher collaboration. Specifically, Sarah was more open-minded in her
perceptions of teachers and their ability to change, was more experienced in leading teacher PD,
and was more familiar about teacher needs at her school sites, allowing her to be more
collaborative with teachers. In contrast, Cindy was narrower in her perceptions of teachers,
judging teachers and their potential to change based on their existing knowledge of technology,
and had not led teacher PD in the past nor worked at any of her assigned school sites. These
individual conditions made it more challenging for Cindy to engage teachers in collaborative
routine for instructional change. In addition, both coaches faced school cultures for teacher
learning and change (as described earlier) that supported varying degrees of teacher
BUILDING NETWORKS FOR CHANGE
150
collaboration. The culture for instructional innovation at San Rafael EL provided the most
conducive environment for Sarah to bridge gaps between teachers in their knowledge of
technology-enabled instruction through collaborative routines, while Glendale EL and Brand EL
had school cultures that were less supportive of teacher collaboration. York EL is a unique
school in that teachers at this school were closely-connected to one another in terms of
discussing and coordinating classroom instruction, but teachers were generally resistant to the
central office’s goal for instructional change and hence distrustful of Cindy and her instructional
support.
Figure 11 plots the influence of these individual and school conditions on Sarah and
Cindy’s brokering practices. The vertical axis in this figure represents Sarah and Cindy’s
perceptions of teachers and familiarity with their school sites, while the horizontal axis
represents their school context in terms of culture for teacher learning and change. I plot Sarah
and Cindy’s brokering practices at each school on these axes, making for a total of four points
(two for each coach at their school sites). Points on the upper, far-right represent brokering
practices that bridge teacher learning on technology to collaborative exchanges with colleagues,
whereas plots on the lower, far-left of this figure represent brokering practices that buffer teacher
learning on technology from collaborative exchanges with colleagues. This figure demonstrates
that Sarah benefited from individual and school conditions that allowed her to engage in more
collaborative brokering practices than Cindy. Sarah is positioned in the top-half of this figure
while Cindy is consistently positioned in the bottom-left square of the figure for both schools.
Once again, there appears to be some variation in Sarah’s brokering practices across school, with
San Rafael EL (upper-right square) proving to be more conducive for teacher collaboration on
technology-enabled instruction than Glendale EL (upper-left square). Similarly, there appears to
BUILDING NETWORKS FOR CHANGE
151
be some variation in Cindy’s brokering practices across schools, with York EL (far left of the
figure) proving to be more difficult school setting for Cindy to navigate than Brand EL (closer to
the center of the figure). I discuss these individual and school conditions and their influence on
Sarah and Cindy’s brokering practices in more detail below.
Coach perceptions of teachers. A distinguishing factor between Sarah and Cindy
concerned their perceptions of teachers and their ability to change. Sarah perceived teachers in
terms of their strengths and weaknesses across all domains of practice, including their
technological, pedagogical, and content knowledge and skills. This perspective motivated Sarah
to tailor her support for teachers so that they could see how technology could build on their areas
of strength. In other words, Sarah made it her responsibility to spark teachers’ interest in using
technology so that they would be motivated to invest more time and effort to learn more about
technology and shift their instructional approach. More importantly, Sarah did not think of
herself as a technology expert, explaining that she “knew enough about technology to know that
there is no way to be an expert it, because there is simply so much [out there].” At the same time,
Sarah did not perceive any of her teachers as being incapable of learning about technology,
recognizing that teachers may not think of themselves as being tech-savvy (or not tech-savvy)
but that they could all come to master using technology through different channels of support
(i.e., more hands-on support at first to learn about apps and then gradually building up to more
complex lesson designs and PBLs). This perspective once again made Sarah more approachable
to teachers, allowing her to connect her DFs to non-fellow teachers to support collaborative
learning experiences in schools. Sarah’s open-minded perception of teachers therefore positioned
her on the top-half of the vertical axis in Figure 11 for both her schools.
BUILDING NETWORKS FOR CHANGE
152
In contrast, Cindy generally perceived teachers as being resistant and unmotivated to
change. Moreover, she focused first and foremost on teachers’ familiarity with technology to
assess their capacity for instructional improvement. As someone who self-started her learning
about technology and instruction, bringing in old iPhones from home to her classroom for
students to use and experimenting with these resources with her grade-level partner, Cindy felt
that most teachers were unlike her and would not be willing to dedicate as much time and effort
toward improving their practice. Accordingly, Cindy gravitated toward supporting younger and
technologically-savvy teachers like herself. She was also less tolerant of teachers using
technology in ways that contrasted with her beliefs for teaching and learning. As she put it:
I do not want them to just do technology for the sake of doing technology. It should not
be just doing a project just because you have to use technology so many times during the
year. It should be because it is more engaging to the kids. The kids are learning more
from it ... They are gaining something from using technology, not just so that they can
have something to hang up on the wall that says, "Check, I did technology."
Given these views, Cindy was less willing than Sarah to deviate from her vision for instructional
change to spur interest in technology among a diverse group of teachers. This reluctance on
Cindy’s part not only created tensions in Cindy’s interactions with DFs who had different
instructional background and goals from her (such as her senior DF at Brand EL), but also
distanced Cindy from the teaching community at York EL, where most of the school’s teaching
staff were veteran teachers who had reservations about student-centric instruction, effectively
positioning Cindy as an outsider to the school community. Cindy’s rigid perceptions of teachers
therefore positioned her on the bottom-half of the vertical axis in Figure 11.
BUILDING NETWORKS FOR CHANGE
153
Coach prior experience in leading PD and familiarity with schools. Sarah’s prior
experience in leading teacher PD and familiarity with her school sites aided her collaborative
work with teachers. As a second-year DLC, Sarah had relevant experience supporting teacher
practice across a range of classroom settings and grade-levels. Having worked at both school
sites for two consecutive years, she had time to develop trust-based relationships with the
teaching staff and learn about their classroom needs. Because of these close relationships,
teachers at Sarah’s schools felt she was available to them, that they could openly discuss their
instruction with her, and that she could provide them with support tailored to their goals and
needs. As noted earlier, Sarah had previously taught at Glendale EL and had worked with the
principal at San Rafael from the central office in the initial stages of LUSD’s technology rollout,
once again affording her more access and buy-in to the school communities where she was
assigned. This social capital made a big difference at Glendale EL, where Sarah could still be an
integral part of the school community despite minimal involvement from the new school
principal. Once again, Sarah’s familiarity with schools positioned her at the top-half of the
vertical axis in Figure 11.
Unlike Sarah, Cindy had no experience in developing or leading teacher PD prior to
becoming a DLC. She was trained as a DF during the first year of LUSD’s DCP and then
transitioned immediately into working as a DLC. Cindy’s training as a DF, during which she
received one-on-one support from a DLC, likely motivated the exclusivity with which she then
approached her work as a coach, investing most of her energy toward supporting other self-
starting, ambitious teachers like herself. Cindy also did not have the advantage of working at any
of her assigned school sites in the past. In fact, she explained that she had “big shoes to fill”
because the teachers at her school sites were still very close to their DLC from the previous
BUILDING NETWORKS FOR CHANGE
154
school year. This absence of teacher motivation was especially acute at York EL, where teachers
still maintained ties with their DLC from last year and only started opening up to Cindy toward
the end of the school year. Altogether, these conditions explain why Cindy was less successful at
supporting teacher collaboration at Brand EL and why she had no social ties to teachers at York
EL, once again positioning Cindy in the bottom-half of the vertical axis in Figure 11.
School cultures for teacher learning and change. As mentioned earlier, because
teachers had limited prior exposure to technology and participation in the DCP was voluntary
(privileging in-depth training and support for a select number of DFs), there were large gaps in
teacher knowledge of technology at all four school sites. This uneven playing field left teachers
feeling uncomfortable in collaborating with their colleagues on technology-enabled instruction,
as they were more familiar with engaging their colleagues on equal terms where they could all
contribute expertise to shape classroom instruction.
Sarah encountered these tensions as a DLC, recounting an instance at one of her school
sites where her DF was “ostracized” by her grade-level team for sharing her classroom work at
school staff meetings because it made her colleagues look bad. She also mentioned that her
relationships with teachers at Glendale EL were strained by the fact that she used to work
alongside many of these teachers as an RTI aid but was now working above them as an
instructional coach. Cindy encountered similar barriers that prevented her DFs from sharing their
instructional work with colleagues, noting that the small size of school faculty and grade-level
teams made it especially difficult for teachers to push their colleagues to change.
[Brand and York EL] are smaller schools, [with] only two people per grade-level, and so
I think that sometimes when you have a smaller group, you have less ideas. It is easier to
BUILDING NETWORKS FOR CHANGE
155
do your own thing if it is a smaller [team], because there is only two people. It is like,
"Okay, I do not like your idea, I am going to do my idea."
From the perspective of DFs and other tech-savvy teachers, they too were aware that their
colleagues did not have the same levels of self-efficacy with using technology for instruction and
as such, did not think it was fair to expect their colleagues to keep pace with them. The following
quotes from Sarah and Cindy’s school sites reveal teachers’ reservations about pushing their
colleagues to adopt technology in their classrooms.
Teacher 1 (San Rafael EL): We are not doing technology lock and step together. Like I
said, time is the challenge for any teacher right now so I do not expect her [my grade-
level partner] to be lock and step with me in whatever I am doing.
Teacher 12 (York EL): We were thinking of doing a project in terms of the rainforest
animals we have been learning about where they could describe more about it. We just
talked about different options we could do for making books for different things and
having the students create them on their own. The problem was they did not push out the
app until just now and so we were thinking about it, it was probably maybe two weeks
ago when I spoke with her. We were planning on using it before but now we are not so
sure if we will have time…It looks like she is going to start implementing it, but I still do
not know if I will.
While these staff tensions were present at all school sites, they were mitigated to large
extent by school cultures that were supportive for teacher learning and change. As mentioned
earlier, the principal at San Rafael created a school culture for instructional innovation that was
inclusive of all teachers. For instance, the principal made a concerted effort to recognize the
successes of all teachers at her school, however big or small, and to showcase all teachers’
BUILDING NETWORKS FOR CHANGE
156
classrooms in district leader walkthroughs of her school (e.g., when central office administrators
and visitors from other districts would tour her school to observe teacher practices with
technology). More recently, the school had been nominated to receive a prestigious award for its
use of technology for instruction. Because of these developments, teachers who are at first
reluctant users of technology started showing more interest because they identified with their
school’s progress and did not want their students to be disadvantaged as they moved to higher
grade-levels where they would be assigned to more tech-savvy teachers. As Teacher 4 explained:
I know the apps that the kindergarten students were using were very simple. Almost like
a white board and they would write things in or to transfer pictures, [but] one of the
teachers recently commented, “Wow, if we are going to [receive an award], I should
prepare my students for all of the great things that Teacher 4 does in second-grade."
As this quote demonstrates, it is not that teachers at San Rafael did not experience any resistance
in persuading their colleagues to use technology in their classrooms, but that this resistance was
less systemic and better managed due to the school’s supportive culture for instructional change.
This supportive school environment, in turn, positioned Sarah in the far-right of the horizontal
axis in Figure 11, allowing her to be the most collaborative with teachers in this school.
At Glendale and Brand EL, the tensions arising from teachers’ differential knowledge of
technology was not as well managed due to the absent leadership of school principals. Teachers
at both schools admitted to having grade-level colleagues who were resistant to using technology
in their classrooms and that they did not know how to, nor were they interested in, addressing
this resistance to change. Teachers at these schools were also critical of their principals for not
doing more to encourage teacher collaboration on technology and instruction. At Brand EL, this
was because the principal had shifted the focus on school-wide PDs away from technology to
BUILDING NETWORKS FOR CHANGE
157
other instructional topics (e.g., Balanced Literacy) whereas at Glendale EL, the principal failed
to coordinate school-wide PD events with teachers. The following quotes from teachers convey
these challenges at each school:
Teacher 8 (Glendale EL): It is just a lot different because my principal last year was very
dynamic and very approachable. If there was something I would be like, "Hey, could I
have 10 minutes, 20 minutes in our meeting to share this?" And he would be like, "Of
course, please share it." I am sure she would do the same but things were a little more
organized. This year we get an email on Monday saying what we are doing on
Wednesday. It is harder to plan, so I have not really had the opportunities.
Teacher 11 (Brand EL): This year I have shared a few times [at school-wide PDs] …Last
year these [meetings] were all technology. This year they switched it. One time a month
it is about technology and another time it is book club. Another time it is ... I forget what
they all are. They change it up. The [meetings] have not been as well attended this year. I
personally have only gone to the technology ones to present to other teachers or one or
two other ones. I think the reason they are not so well-attended this year is because there
has been so much new stuff thrown at us…almost every day you have a meeting or this
or a that.
Because of these challenges, Sarah and Cindy had limited success in facilitating teacher
sharing on technology-enabled instruction. For instance, Sarah tried to pursue this collaborative
work at Glendale EL by implementing a PBL with her DFs and their grade-level colleagues, but
did not have much of an impact. In fact, one of her DFs expressed that the PBL did little to foster
a collaborative team culture and, if anything, resulted in the DF sacrificing her own training and
professional growth for Sarah to provide technical assistance to her grade-level colleagues who
BUILDING NETWORKS FOR CHANGE
158
were less familiar with technology. Similarly, other teachers expressed that their collaboration
with colleagues on technology-enabled instruction came from the previous school year with
support from Sarah and their old principal, and that they had not started any new collaborative
endeavors in the current school year. The change in the school culture at Glendale EL from year-
one to year-two of LUSD’s technology rollout meant that Sarah was positioned closer to the
center of horizontal axis in Figure 11.
As noted earlier, Cindy did little to support her DFs in sharing their instructional work
with colleagues. In fact, she only reached out to the grade-level colleagues of her DFs to resolve
tensions that she encountered in her training of DFs. For instance, Cindy started inviting the
grade-level partner of her senior DF (Teacher 10) at Brand EL to participate in their weekly
debriefs because she found this to improve the overall quality of their conversations. In contrast,
Cindy did little to support her more advanced DF (Teacher 11) to share technology-enabled
teaching practices with her grade-level partner, with this teacher instead resorting to teaching the
students of her grade-level partner herself so that they could have some exposure to learning with
technology. As these examples demonstrate, in the absence of a strong school culture for
instructional innovation, Cindy’s DFs were not able to engage their grade-level colleagues in
meaningful conversations about instructional improvement. This school environment therefore
positioned Cindy and her brokering practices on the left-hand side of the horizontal axis in
Figure 11.
As mentioned earlier, York EL is a unique case in that teachers at this school were part of
a cohesive community that was resistant to adopting student-centric approaches to instruction.
This seemed to be due, in large part, to York EL’s strong track record of student performance
under their state’s standards-based assessments preceding the CCSS, which made these teachers
BUILDING NETWORKS FOR CHANGE
159
reluctant to abandon teacher-led instructional practices that contributed to their past success. As
the Balanced Literacy coach for this school explained:
These teachers take their student achievement and they wear it like a badge of honor
which we have got to move [away] from. That is part of my shift in getting them to see.
Let the kids wear that badge of honor and let’s step back…building that relationship
there, I will tell you, has taken, I would say, until March, April. It has been very hard for
them to be willing to trust me enough to make a shift in their practice and be willing to
change. Failing is not an option for them and when you are shifting your practices, it is
going to get messy.
The disconnect between LUSD’s vision for instructional change and teacher expectations
for student performance at York EL can explain, in part, why teachers were distrustful of Cindy,
discounting the benefits of her coaching and perceiving her as an outsider to the school
community. As shown in the school’s social network in Figure 9, rather than reach out to Cindy
for instructional support on technology-enabled instruction, teachers at York EL were more
comfortable reaching out to a former DF at their school who was knowledgeable about
technology and embraced the school’s preferences for teacher-led instruction. And unlike the
Balanced Literacy coach at York EL who received clear direction and support from the school
principal to network with teachers and begin to shift their preferences for teacher-led instruction,
Cindy did not benefit from this principal involvement, requiring more time for her to build these
relationships with teachers. As she put it:
I feel like I'm a little bit more accepted [at Brand EL] honestly and more respected. I'm
starting to feel that way at York EL a little bit more…I think teachers now realize that
I'm not there to evaluate them, and they're a little bit more comfortable coming up to me
BUILDING NETWORKS FOR CHANGE
160
and asking me, "Can you help me with this? Can you do this?" Because they see the
benefit.
As the case of York EL demonstrates, it was not enough for Cindy to have a school culture that
supported teacher collaboration on instruction. In addition to these collaborative norms, Cindy
needed the school culture at York EL to be aligned with central office goals for instructional
change and for the principal at this school to champion these goals and help connect her to
teachers. In the absence of these school conditions, Cindy was positioned at the left-extreme of
the horizontal axis in Figure 11, presenting her with little to no opportunity to facilitate teacher
collaboration on technology-enabled instruction.
Cross-Case Conclusions
Overall, my analysis highlights several factors behind why Sarah and Cindy were not
central to guiding teacher practice for teaching the CCSS, and why these coaches supported
varying degrees of teacher collaboration in schools on technology-enabled instruction. First, the
ability of DLCs to bridge technology with CCSS instruction and foster teacher collaboration in
schools on technology-enabled instruction were inhibited from the very beginning by how LUSD
designed the DCP and articulated the responsibilities of these coaches. By selecting two different
signature practices for guiding instruction on literacy and with technology (i.e., Balanced
Literacy and the SAMR scale), creating two frontline staff roles to support these shifts in teacher
practice (i.e., DLCs for helping teachers to move along the SAMR scale and CCs for helping
teachers to teach in Balanced Literacy), and only assigning DLCs to visit their school sites and
meet with DFs once a week, the central office made it unclear whether DLCs should be
supporting teachers in core instructional work such as improving student literacy and
performance on ELA assessments for the Common Core. In addition, by requiring DLCs to
BUILDING NETWORKS FOR CHANGE
161
support their DFs first and foremost and requiring teachers to volunteer for coaching in the DCP,
the district central office positioned DLCs and their instructional support as an exclusive
resource, making them less accessible to guide other teachers, and created an uneven playing
field in schools that impeded teachers from cooperating with one another to promote technology-
enabled instruction.
Second, within this district context, I find that Sarah and Cindy engaged in brokering
practices that either bridged or distanced their coaching from CCSS instruction and from
building teacher collaboration in schools. Sarah came closest to bridging technology with CCSS
instruction, showing that a process-oriented approach to coaching teachers on using technology
also enables teachers to improve on core processes for teaching content standards (e.g.,
instructional planning, classroom management, student engagement). Sarah was also the most
central in her school networks for integrating technology with instruction and the most
collaborative in her work with DFs and non-fellow teachers. These patterns further suggest that a
process-oriented approach to coaching is valuable to teachers and lends itself to collaborative
routines for improving teacher practice, allowing Sarah to main a high-degree of power and
influence in her schools while supporting teachers to share best practices for instruction with
colleagues. In contrast, I find that Cindy created more distance between her instructional support
for DFs and CCSS instruction through engaging in more product-oriented coaching practices.
This coaching style also coincided with Cindy being less central in her school networks for
integrating technology with instruction and less collaborative in terms of supporting non-fellow
teachers. These patterns suggest that a product-oriented approach to coaching is not as valuable
to teachers and less conducive to teachers exchanging ideas for instructional improvement.
BUILDING NETWORKS FOR CHANGE
162
Third, I find that the above brokering practices were informed by the instructional
qualifications and backgrounds of Sarah and Cindy as coaches. Both DLCs needed guidance on
how to relate their coaching to teachers’ goals for improving student literacy, specifically with
regards to helping teachers implement practices from the SAMR scale and Balanced Literacy in
tandem. As a less experienced coach, Cindy also needed more training and support on navigating
challenging school conditions, working with teachers across subject areas and grade-levels,
engaging teachers in process-oriented and collaborative routines for instructional change.
Longevity at school sites also appeared to be an important factor in terms of Sarah and Cindy
being able to rally teachers to innovate with instruction and collaborate with one another. My
comparison of Sarah and Cindy also suggests that these coaches drew on different sources of
motivation to inform their work with teachers, resulting in these coaches having different
perceptions of teachers and providing teachers with varied levels of quality of instructional
support. This suggests that the individuals who are staffed as coaches can make or break the
efficacy of coaching programs such as the DCP, making it imperative for district central offices
to implement robust procedures for recruiting, selecting, training and evaluating coaches.
Fourth, while my results point to a range of school conditions that enabled and/or
constrained the brokering practices of DLCs, the common link across these school-level factors
appeared to be school leadership. To guide teachers on how to use technology to teach the CCSS
and build teacher communities dedicated to technology-enabled instruction, Sarah and Cindy
needed principals to provide them (and their teachers) appropriate instructional guidance to,
strategize with them on how to best utilize school time to support all teachers, and dedicate
institutional resources toward supporting teacher learning and change (e.g., purchasing additional
iPads and providing collaboration time for teachers to meet in PLCs). Principal decisions related
BUILDING NETWORKS FOR CHANGE
163
to workforce talent management (e.g., evaluating teacher practice, instructional goal setting)
were also critical for motivating teachers to work with Sarah and Cindy on core instructional
processes related to student learning and to collaborate with colleagues on instruction. This
suggests that, aside from investing in the distributed leadership practices of DLCs, districts such
as LUSD should also invest in principal capacity for managing knowledge-intensive reform.
Fifth, my findings demonstrate that LUSD’s technology rollout and implementation of
the CCSS placed large demands on human capital resources in schools that made it difficult for
principals and teachers to act quickly to improve teacher practice. Because most principals and
teachers had never used digital devices and apps for instruction prior to LUSD’s technology
rollout, and because principals and teachers had different levels of familiarity with Balanced
Literacy as a signature practice for literacy instruction, these school actors could not identify
connections between these instructional programs and hence failed to coordinate teacher practice
accordingly and draw on Sarah and Cindy for more comprehensive support. Similarly, the
unevenness in teacher knowledge and skills for technology-enabled instruction created a
challenging environment for Sarah, Cindy, and their DFs to promote best practices for
technology-enabled instruction among non-fellow teachers. This is because these non-fellow
teachers felt completely unprepared to experiment with technology in their classrooms and were
somewhat resentful of the privileged status and support afforded to DFs in their grade-level
teams. These findings suggest that district leaders need to provide time for schools to implement
knowledge-intensive reforms and to assess the existing capacity of schools when designing
instructional supports for these reforms.
BUILDING NETWORKS FOR CHANGE
164
CHAPTER NINE -- Implications for Policy, Practice, and Research
Review of Dissertation and Findings
As districts continue to invest in technology as a school resource for teaching and
learning, pouring millions if not billions of dollars into new machinery and related structures,
procedures, and resources for improving teacher practice, it behooves education leaders to
understand the impacts of these costly investments. Because research so far suggests that these
education technology reforms have inconclusive effects on student achievement (Means et al.,
2013; Zheng et al., 2016), there is a strong need for descriptive studies that take a closer look at
how these reforms are being implemented on the ground and the challenges facing educators as
they work to improve student learning. My research contributes to this area of inquiry by
exploring the instructional leadership of ed-tech coaches as a critical dimension of education
technology reform implementation. I focus on ed-tech coach leadership because research on the
implementation of education technology reform has revealed complex issues such as inadequate
access to school infrastructure, incoherent teacher PD, and misaligned expectations for school
success that require proactive, creative, and responsive leadership to be resolved (Bingham et al.,
2016; Tanenbaum et al., 2013). Since district leaders and technology advocates expect ed-tech
coaches to take on a larger share of these leadership responsibilities (Flanigan, 2016;
International Society for Technology in Education, 2011), I focus on the instructional leadership
of these coaches as a critical leverage point for instructional change.
Drawing on insights from prior research on instructional leadership (e.g., Daly &
Finnigan, 2016; Honig & Hatch, 2004), I investigate three leadership practices – instructional
coordination, coherence, and alignment – that are central to how ed-tech coaches should be
supporting district-wide improvements in teaching and learning. Recent research suggests that
BUILDING NETWORKS FOR CHANGE
165
districts are embracing a distributed approach to leadership to build instructional coordination,
coherence, and alignment in their school systems (Durand et al., 2016; Ho & Ng, 2017),
privileging frontline staff such as ed-tech coaches who are at the forefront of reform
implementation to engage expertise throughout districts to inform central office, school, and
educator practice. My research examines the leadership practices of ed-tech coaches from this
perspective, examining how ed-tech coaches build relationships with central office and school
actors throughout their district to improve instructional coordination, coherence, and alignment.
I use concepts from social network theory, along with social network and qualitative data
collected from LUSD’s Digital Coaching Program (DCP), to explore the brokering practices of
Digital Learning Coaches (DLCs) as evidence of their distributed leadership. I explore these
brokering practices for two interrelated reforms that LUSD implemented at the time, including a
one-to-one technology rollout of iPads and laptops to teachers and students and the district’s
simultaneous implementation of the Common Core State Standards (CCSS) to improve student
content knowledge and 21
st
century skills. In this setting, I test several hypotheses on how DLCs
are brokering information on these joint reform efforts among central office and school actors,
and the role of LUSD’s organizational context in shaping these brokering practices.
I find that DLCs are peripheral actors in LUSD’s reform network for teaching the CCSS
and as such, do not broker information on how to teach these content standards (Boxes 2A-2D in
Figure 2). In contrast, DLCs are central actors and active brokers in LUSD’s reform network for
integrating technology with instruction. Upon taking a closer look at the brokering practices of
DLCs in LUSD’s technology reform network, I find that these coaches are mainly involved with
coordinating information on technology-enabled instruction with one another (as opposed to
coordinating information on instruction with higher-ranked central office administrators and
BUILDING NETWORKS FOR CHANGE
166
other frontline staff such as Curriculum Coordinators) (Box 2A in Figure 2), and sharing this
information in a top-down manner from the central office into schools (Box 2B in Figure 2). I
find less evidence of DLCs brokering information within or between schools on technology-
enabled instruction (Box 2C in Figure 2), or of DLCs forging bottom-up communication
channels to guide central office decision-making based on school insights and needs for reform
implementation (Box 2D in Figure 2).
These brokering patterns suggest that DLCs are only fulfilling parts of the instructional
leadership processes expected to drive LUSD’s joint technology and CCSS rollout. Because
DLCs are not active brokers for sharing information on teaching the CCSS, these coaches are
failing to bridge the use of technology for instruction in general with instruction of these content
standards specifically (Box 3A in Figure 2). I find further evidence of this lack of coordination in
LUSD’s reform network for technology, where the siloed communication among DLCs in the
central office, and corresponding lack of communication between DLCs and central office
administrators and CCs, seems to have further isolated technology from core instructional
processes for teaching the CCSS. The fact that DLCs are communicating their insights on
technology-enabled instruction with teachers suggests that these coaches are aligning teacher
practice with some of the central office’s goals for instructional change (i.e., the use of
technology for instruction in general) and therefore helping teachers understand how these goals
are coherent with school efforts for instructional improvement (Box 3B and 3C in Figure 2).
However, because DLCs are less consistent at circulating information within or across schools,
or sharing insights from schools up to the central office, these coaches could still be doing more
to support instructional alignment and school coherence (i.e., Box 3D and 3E in Figure 2 by
BUILDING NETWORKS FOR CHANGE
167
sharing best practices for instruction across teachers and schools) and to build system coherence
for instructional change (Box 3F in Figure 2).
51
Through exploring a rich set of organizational conditions that influence information
brokering in LUSD (Box 4 in Figure 2), I find that the opinion leadership of DLCs –their power
and influence over instruction – to be strongly associated with their brokering patterns across
reform networks. I also find some evidence of access to expertise, density and reciprocity of ties,
and fragmented social structures contributing to the brokering practices of DLCs. For instance, I
find that DLCs broker information in LUSD’s technology reform network because they
demonstrate substantially more power and influence relative to other central office and school
actors. Moreover, DLCs broker information in a top-down manner from the central office into
schools because there is a negative association between power and influence and advice-seeking,
suggesting a top-down organizational structure where actors in positions of authority (i.e., DLCs
and other central office administrators and staff) are more likely to provide advice to, rather than
seek advice from, actors in lower-ranked positions of authority (i.e., school principals and
teachers). I also find evidence of denser and more reciprocated communication taking place
within the central office than among schools, suggesting that schools have few internal resources
to guide teachers on integrating technology with instruction and are therefore more open to
receiving direction and input from the central office. Finally, I find that fragmented social
structures with isolated social cliques can impede DLCs from brokering information in social
networks where they have limited power and influence to begin with, as well as constrain lateral
brokering ties that require DLCs to bring together different actor groups (e.g., coordination ties
51
As a reminder, I define school coherence as the process of translating central office goals for instructional change
into goals and strategies that specific to school needs, and system coherence as adjusting central office procedures
and resources to foster district conditions that are responsive to school needs (Honig & Hatch, 2004).
BUILDING NETWORKS FOR CHANGE
168
across different actor groups within the central office and coordination ties within and across
schools).
Based on these results, I use my qualitative data to dig deeper into the opinion leadership
of DLCs in schools, exploring the individual attributes of these coaches, the context and
activities of their brokering exchanges with teachers, and school and district conditions that
imbue them with power and influence for directing teacher practice. I use this descriptive detail
to unpack: (a) why DLCs have little power and influence over teacher practice for teaching the
CCSS, and (b) how DLCs might be using their power and influence for integrating technology
with instruction to facilitate direct communication and collaboration among teachers on
instructional change. For the latter question (b), I account for the possibility that, while DLCs
might not be middle-men who mediate communication between principals and teachers in the
same school (as seen in my social network data), these coaches could be coordinating
information within schools in ways that involve direct teacher-to-teacher engagement. My
qualitative results highlight several interrelated conditions that explain the opinion leadership of
DLCs across reform networks, which I use to discuss implications for teacher learning and
change and make recommendations for improving LUSD’s DCP.
Contributions to Research
My dissertation findings make several contributions to the extant literature. I provide
some of the first empirical evidence on the leadership practices of ed-tech coaches with regards
to building instructional coordination, coherence, and alignment in education reform. This
evidence represents a novel contribution to the extant literature not just because ed-tech coaches
are a policy-relevant case for investigating the leadership of frontline staff in knowledge-
intensive reform, but also because I describe their leadership in their entirety rather than focusing
BUILDING NETWORKS FOR CHANGE
169
on select dimensions of this leadership (i.e., just focusing on how ed-tech coaches support
instructional alignment which has so far been the conventional approach of the extent literature).
To this end, the conceptual framework that I have pieced together from several bodies of
literature to guide my analysis (Figure 1) is also a noteworthy contribution, offering a holistic
framework for scholars to make connections between instructional leadership, brokering, and
systemic change in future research.
I also offer rigorous evidence on the role of organizational context in shaping the
brokering practices of ed-tech coaches and the prospect for systemic instructional change.
Specifically, I observe how the organizational context of districts influence the system-wide
brokering practices of DLCs rather than just focusing on organizational conditions of schools
that influence how DLCs share information with teachers to support instructional alignment. My
use of social network theory to translate the brokering practices of ed-tech coaches and their
organizational context into properties that can be readily observed in the social structure of
districts is also informative, providing a common language to describe system-wide brokering
and complex organizational conditions that are difficult to observe in school systems
encompassing multiple schools with varied sub-cultures and work environments.
While the qualitative component of my study follows more closely in the footsteps of
prior studies by describing how ed-tech coaches negotiate power and influence within schools to
engage teachers on instructional change, I approach this topic from a different perspective.
Specifically, I shed light on how ed-tech coaches are positioned with power and influence to
guide teacher practice across interrelated reforms and dimensions of instruction. This research
focus is important given that knowledge-intensive reform increasingly requires teachers to
coordinate different instructional programs in their daily practice (Coburn et al., 2009; Durand et
BUILDING NETWORKS FOR CHANGE
170
al., 2016). My qualitative findings also offer insight on how ed-tech coaches use their power and
influence to engage communities of teachers on instructional change. This descriptive evidence
is important because knowledge-intensive reform – and education technology reform in
particular – depends on teachers collaboration to sustain and expand the footprint of instructional
innovation (Coburn et al., 2012; Frank et al., 2011; Sun et al., 2013).
Implications for Policy and Practice
My research findings have several implications for policy and practice concerning
education technology reform specifically and other kinds of knowledge-intensive reform more
broadly. Perhaps the most important takeaway for district leaders is that, even with all the bells
and whistles that come with education technology reform (i.e., new iPads, laptops, apps), these
reforms have to be implemented within the organizations of districts and, as such, depend on the
existing capacity of districts to execute these reforms effectively (Cuban, 2012; Tyack & Cuban,
1995). My results suggest that, with two years of implementing its technology rollout and
preparing teachers and students for CCSS assessments, LUSD is still in the early-stages of these
reform efforts and is not fully engaging in processes of instructional coordination, coherence,
and alignment needed to produce adaptive and systemic instructional change. This incomplete
and inconsistent leadership, in turn, could detract from critical intermediate and late-stage
outcomes expected from LUSD’s technology rollout.
I touch on two of these implementation shortfalls in my case-study findings. For instance,
I show that the failure of DLCs to engage teachers on CCSS instruction, or to coordinate insights
on technology-enabled instruction with central office administrators and CCs, resulted in
teachers perceiving efforts to personalize instruction for student reading under Balanced Literacy
to be at odds with the use of technology to facilitate inquiry-based learning experiences that
BUILDING NETWORKS FOR CHANGE
171
build student 21
st
century skills for learning. As shown in Figure 2, this uneven or inconsistent
implementation of personalized and inquiry-based instruction could have mixed effects on
student engagement and content knowledge, and ultimately student performance on CCSS
assessments and academic achievement across student sub-groups. I also show that DLCs varied
efforts to foster teacher collaboration in schools weakened the breadth of technology adoption
and instructional change in schools. This means that DLCs were not able to establish the use of
technology for inquiry-based instruction as a systemic practice which, as shown in Figure 2,
could further detract from LUSD’s efforts to improve student engagement, content knowledge,
and performance on CCSS assessments.
Therefore, while knowledge-intensive reform is driven by a theory of action where
combined efforts in instructional coordination, coherence, and alignment are supposed to
generate systemic improvements in teaching and learning, my dissertation show that this theory
of change is slow moving in practice and can have mixed effects on teacher practice and student
learning in the short-term. This is perhaps not surprising given that standards-based instructional
reforms have historically been top-down and centrally-driven by design (Cuban, 2012; Smith &
O’Day, 1991; Tyack & Cuban, 1995), making it difficult for education practitioners to embrace
autonomous and decentralized leadership practices for facilitating instructional experimentation
and disseminating non-traditional teaching practices at scale. While the novelty of technology
and the non-prescriptive nature of CCSS instruction should provide an impetus for doing away
with these top-down implementation structures, districts need to be more deliberate in designing
these instructional reforms and reconfiguring their leadership approach to challenge the status
quo (Durand et al., 2016; Ho & Ng, 2017; Hopkins et al., 2016).
BUILDING NETWORKS FOR CHANGE
172
For instance, my results suggest that the opinion leadership of ed-tech coaches (Box 4 in
Figure 2) is an important aspect of their leadership and brokering, suggesting that district leaders
should focus on how these coaches are positioned in the social structure of districts to affect
instructional change (Coburn & Russell, 2008; Hopkins et al., 2016; Spillane et al., 2015). In
particular, since ed-tech coaches must negotiate for power and influence from other central office
and school actors, district leaders could take proactive steps to imbue their school systems with
distributed leadership practices that enhance the influence of ed-tech coaches (Ho & Ng, 2017).
My qualitative results suggest that district articulation of coach responsibilities is also crucial in
this regard, informing whether ed-tech coaches can claim authority over multiple domains of
teacher practice and how these coaches balance their work between providing one-on-one
support to teachers and supporting school-wide improvements in teaching and learning.
I also find that school principals are critical gatekeepers with the potential to legitimize
(or discredit) the presence of ed-tech coaches in their schools. This suggests that districts should
focus on supporting principals to understand the goals of knowledge-intensive reform and how
they can work with frontline staff such as ed-tech coaches to improve teacher practice (Hopkins
et al., 2016; Matsumura & Wang, 2014). Finally, research suggests that districts should establish
routines for joint decision-making among administrators and educators to grant influence to
frontline staff who may not occupy conventional roles of authority in their district (Durand et al.,
2016; Liou, 2016). While LUSD started to implement these collaborative routines by organizing
weekly meetings for DLCs and monthly meetings between DLCs and CCs, my results suggest
that district leaders might need to provide more guidance on the purpose of these meetings and
how they should inform the coaching of DLCs. There is probably more about these meetings that
BUILDING NETWORKS FOR CHANGE
173
I could explore from my interviews with central office administrators, DLCs, and CCs that I
have not yet unpacked.
In addition to creating an ideal social setting for ed-tech coaches to exercise instructional
leadership, the individual attributes of ed-tech coaches and their approach to brokering can also
inform their ability to affect instructional change. This is consistent with what other studies have
revealed about the implementation of coaching programs (Coburn & Russell, 2008; Huguet et
al., 2014; Marsh et al., 2010), suggesting that district procedures for recruiting, selecting,
training, and evaluating coaches to be important design features of coaching programs. Central
office procedures for assigning coaches to schools, privileging locations where coaches have
established relationships with principals and teachers and allowing coaches to work at assigned
schools for extended time, are also design elements that can enhance the leadership and efficacy
of coaches (Marsh et al., 2010).
Finally, my results show that districts are complex organizations with various factors that
influence information sharing and the brokering practices of ed-tech coaches (Bridwell-Mitchell
& Cooc, 2016; Daly, 2010; Moolenaar, 2012). While opinion leadership of ed-tech coaches is
arguably one of the most important organizational conditions highlighted in my results, I find
other normative conditions to be important as well, including the advice-seeking behaviors of
central office and school actors, the density and reciprocity of ties in district networks, and the
formation of social cliques (Box 4 in Figure 2). This implies that district leaders should attend to
the social structures of their school systems prior to designing interventions to build capacity for
instructional change. For example, administering a socio-metric survey to central office and
school actors to map out social networks for discussing education technology and other
knowledge-intensive reform topics can provide valuable information on who are the opinion
BUILDING NETWORKS FOR CHANGE
174
leaders in these reforms, the levels of social engagement and depth of communication among
central office and school actors on instructional change, and the extent to which cliques and
homophilous ties predict tie formation and information sharing. This information, in turn, could
allow districts to develop instructional coaching programs and other instructional procedures that
leverage existing capacity in their school systems while targeting specific areas of need.
In the case of LUSD, collecting this social network data prior to the DCP would most
likely have revealed a vacuum of school expertise on guiding teachers to use technology for
instruction, making the district’s decision to recruit expert teachers (i.e., existing opinion leaders
in their district) to coach teacher practice in schools a well-informed decision. In contrast, this
pre-reform social network data would most likely have shown district leaders that schools are
engaging in more developed conversations on how to teach the CCSS and that school principals
are critical opinion leaders for guiding teacher practice in this area. Based on this initial data,
district leaders could have implemented more targeted strategies for allowing DLCs to contribute
to existing school efforts in CCSS instruction. These targeted strategies include building
collaborative relationships between DLCs and school principals and articulating the
responsibilities of DLCs to include facets of CCSS instruction that are meaningful to school
implementation of these content standards (e.g., focus on improving student literacy). In
addition, this initial social network assessment could have shown that communication on
instructional reform, especially within the central office and within and across schools, is
fragmented according to formal role and that more deliberate structures and routines. This
evidence, in turn, could have promoted district leaders to understand the underlying causes for
this fragmented communication and develop more explicit structures, resources, and routines to
overcome these barriers to communication.
BUILDING NETWORKS FOR CHANGE
175
Future Research
My dissertation opens doors to new questions that should be explored in future research.
For instance, while I have described in detail the brokering practices of ed-tech coaches in the
early-stages of LUSD’s technology rollout and implementation of the CCSS, I am unable to
speak to how these brokering practices might change over time as LUSD continues with
implementing these reforms. From conversations with district leaders, I learned that the district
has already included many of the recommendations highlighted in my research, such as
expanding the formal role of DLCs to include advising teachers on both technology and
Balanced Literacy instruction at elementary grade-levels, and recruiting principals to participate
in coaching on technology and instructional management. These kinds of procedural changes,
along with other efforts to improve school capacity for experimenting with instruction, could
result in DLCs engaging in different brokering practices that should be explored. More recent
studies on instructional coaches and brokering suggest that coaches do in fact broker information
in different ways over time, starting with more top-down communication and then forming
lateral ties to build intra and inter-school networks for instructional change (Hopkins et al.,
2016).
While I have described the leadership and brokering practices of ed-tech coaches as a
specific case of frontline staff supporting knowledge-intensive reform, future research is also
needed to describe instructional leadership and brokering in other kinds of knowledge-intensive
reform settings (e.g., multiple-measure teacher evaluation systems, data-driven decision-
making). It could be that there is something unique about education technology reforms or the
role of ed-tech coaches in supporting instructional change that are influencing my findings and
that might not hold in other settings, which is important to tease out when guiding districts on
BUILDING NETWORKS FOR CHANGE
176
how to generate systemic instructional change. Similarly, while I chose to investigate LUSD’s
DCP because of the generalizability of the district’s organizational structure and student
population to other districts in the country, the fact remains that my findings are not translatable
to large urban districts serving large concentrations of under-privileged students and operating
with large administrative bureaucracies. How feasible might it be to expect ed-tech coaches and
other frontline staff to demonstrate instructional leadership and broker information in these more
complex organizational settings? This can only be studied by extending my line of research to
urban districts that are also grappling with leadership and implementation challenges related to
knowledge-intensive reform.
Summary
This study provides a first look at the instructional leadership and brokering practices of
ed-tech coaches. I have established that ed-tech coaches are not engaging in all the leadership
practices expected from them by district leaders, technology advocates and the media, and have
identified various organizational factors that enable and/or constrain their instructional
leadership. This study contributes to our understanding of ed-tech coach leadership and
education technology reform specifically, as well as frontline staff leadership and knowledge-
intensive reform implementation more broadly. Beyond the insights from this study, there are
other questions yet to be investigated regarding frontline staff leadership and systemic
instructional change.
If district leaders take the insights from this work and apply it to their own school
systems, I believe we will be taking a step toward supporting coordinated, coherent, and aligned
instructional practice in education technology reform and better learning outcomes for students.
Given that the number of districts rolling out personalized learning initiatives and one-to-one
BUILDING NETWORKS FOR CHANGE
177
computing continues to grow, even making slight improvements to the leadership overseeing
these reforms could have a significant impact on public education.
BUILDING NETWORKS FOR CHANGE
178
REFERENCES
Adler, P. S., & Kwon, S. (2002). Social capital: Prospects for a new concept. Academy of
Management Review, 27(1), 17–40.
Albrecht, T. L., & Bach, B. W. (1997). Communication in complex organizations: A relational
approach. New York, NY: Harcourt Brace.
Anderson, R., Feldman, S., & Minstrell, J. (2014). Understanding relationship: Maximizing the
effects of science coaching. Education Policy Analysis Archives, 22(0), 54.
https://doi.org/10.14507/epaa.v22n54.2014
Atteberry, A., & Bryk, A. S. (2011). Analyzing teacher participation in literacy coaching
activities. The Elementary School Journal, 112(2), 356–382.
Aubusson, P., Burke, P., Schuck, S., Kearney, M., & Frischknecht, B. (2014). Choosing rich
tasks: The moderating impact of technology on student learning, enjoyment and
preparation. Educational Researcher, 43(5), 219–229.
Barrow, L., Markman, L., & Rouse, C. E. (2009). Technology’s edge: The educational benefits
of computer-aided instruction. American Economic Journal: Economic Policy, 1(1), 52–
74.
Bean, R. M., Draper, J. A., Hall, V., Vandermolen, J., & Zigmond, N. (2010). Coaches and
coaching in Reading First schools: A reality check. The Elementary School Journal,
111(1), 87–114.
Bingham, A. J. (2016). Drowning digitally? How disequilibrium shapes practice in a blended
learning charter school. Teachers College Record, 118(1), 1–45.
BUILDING NETWORKS FOR CHANGE
179
Bingham, A. J., Pane, J. F., Steiner, E. D., & Hamilton, L. S. (2016). Ahead of the curve:
Implementation challenges in personalized learning school models. Educational Policy,
1–36. https://doi.org/10.1177/0895904816637688
Blanchard, M., LePrevost, C., Tolin, A. D., & Gutierrez, K. S. (2016). Investigating technology-
enhanced teacher professional development in rural, high-poverty middle schools.
Educational Researcher, 45(3), 207–220. https://doi.org/10.3102/0013189X16644602
Borgatti, S. P., & Foster, P. C. (2003). The network paradigm in organizational research: A
review and typology. Journal of Management, 29(6), 991–1013.
Borgatti, S. P., & Ofem, B. (2010). Overview: Social network theory and analysis. In A. J. Daly
(Ed.), Social network theory and educational change (pp. 17–30). Cambridge, MA:
Harvard Education Press.
Bourdieu, P. (1986). The forms of social capital. In J. G. Richardson (Ed.), Handbook of theory
and research for the sociology of education (pp. 241–258). New York: Greenwood.
Bridwell-Mitchell, E. N. (forthcoming). Them that’s got: How tie formation in partnership
networks gives high schools differential access to social capital. American Educational
Research Journal.
Bridwell-Mitchell, E. N., & Cooc, N. (2016). The ties that bind: How social capital is forged and
forfeited in teacher communities. Educational Researcher, 45(1), 7–17.
https://doi.org/10.3102/0013189X16632191
Bryk, A. S., & Schneider, B. (2002). Trust in schools: A core resource for improvement. New
York, NY: Russell Sage Foundation.
BUILDING NETWORKS FOR CHANGE
180
Bryk, A. S., Sebring, P. B., Allensworth, E., Luppesco, S., & Easton, J. Q. (2010). Organizing
schools for improvement: Lessons from Chicago. Chicago, IL: University of Chicago
Press.
Burt, R. S. (1992). Structural holes: The structure of competition. Cambridge, MA: Harvard
University Press.
Burt, R. S. (2000). The network structure of social capital. Research in Organizational Behavior,
22, 345–423.
Cavanagh, S. (2014a). State digital education programs grow, but disparities persist. Education
Week. Retrieved from
http://blogs.edweek.org/edweek/DigitalEducation/2014/11/state_digital_education_proga
ms_grow_and_disparities_persist.html?print=1
Cavanagh, S. (2014b, June 11). K-12 district leaders evolving into smarter ed-tech consumers.
Education Week. Retrieved from http://www.edweek.org/ew/articles/2014/06/11/35ii-
sophisticated.h33.html?tkn=TZZFDfaq3GmTmzdN0hW1tndl3Z%2Fav4hMgczD&print=
1
Cavanagh, S. (2015). Blunt advice for harried ed-tech “coaches” offered at ISTE. EdWeek.
Retrieved from
http://blogs.edweek.org/edweek/DigitalEducation/2015/07/blunt_advice_for_ed-
tech_coaches_offered_at_ISTE.html?cmp=ENL-DD-NEWS2-RM
CCSS. (2014a). Common Core State Standards for English language arts & literacy in
history/social studies, science, and technical subjects. Washington, D.C.: Author.
Retrieved from http://www.corestandards.org/
CCSS. (2014b). Common Core State Standards for mathematics. Washington, D.C.: Author.
BUILDING NETWORKS FOR CHANGE
181
Coburn, C. E., & Russell, J. L. (2008). District policy and teachers’ social networks. Educational
Evaluation and Policy Analysis, 30(3), 203–235.
Coburn, C. E., Russell, J. L., Kaufman, J. H., & Stein, M. K. (2012). Supporting sustainability:
Teachers’ advice networks and ambitious instructional reform. American Journal of
Education, 119(1), 137–182.
Coburn, C. E., & Stein, M. K. (2006). Communities of practice theory and the role of teacher
professional community in policy implementation. In M. I. Honig (Ed.), New directions
in education policy implementation: Confronting complexity (pp. 25–46). Albany, NY:
SUNY.
Coburn, C. E., & Talbert, J. E. (2006). Conceptions of evidence-based practice in school
districts: Mapping the terrain. American Journal of Education, 112, 469–495.
Coburn, C. E., Toure, J., & Yamashita, M. (2009). Evidence, interpretation, and persuasion:
Instructional decision making at the district central office. Teachers College Record,
111(4), 1115–1161.
Coburn, C. E., & Woulfin, S. L. (2012). Reading coaches and the relationship between policy
and practice. Reading Research Quarterly, 47(1), 5–30.
Coleman, J. S. (1988). Social capital in the creation of human capital. The American Journal of
Sociology, 94, S95–S120.
Cresswell, J. W. (2007). Qualitative inquiry and research design: Choosing among five
approaches. Thousand Oaks, CA: SAGE Publications.
Cuban, L. (2012). Inside the black box of classroom practice: Change without reform in
American education. Cambridge, MA: Harvard Education Press.
BUILDING NETWORKS FOR CHANGE
182
Daly, A. J. (2010). Social network theory and educational change. Cambridge, MA: Harvard
Education Press.
Daly, A. J., & Finnigan, K. S. (2011). The ebb and flow of social network ties between district
leaders under high stakes accountability. American Educational Research Journal, 48(1),
39–79.
Daly, A. J., & Finnigan, K. S. (2012). Exploring the space between: Social networks, trust, and
urban school district leaders. Journal of School Leadership, 22, 493–530.
Daly, A. J., & Finnigan, K. S. (2016). Thinking and acting systemically: Improving school
districts under pressure. Washington, D.C.: American Educational Research Association.
Daly, A. J., Finnigan, K. S., Moolenaar, N. M., & Che, J. (2014). The critical role of brokers in
the access and use of evidence at the school and district level. In A. J. Daly & K. S.
Finnigan (Eds.), Using research evidence in education: From the schoolhouse door to
capitol hill (Vol. 2, pp. 13–32). Switzerland: Springer International Publishing.
Daly, A. J., Moolenaar, N. M., Bolivar, J. M., & Burke, P. (2010). Relationships in reform: The
role of teachers’ social networks. Journal of Education Administration, 48(3), 359–391.
Daly, A. J., Moolenaar, N. M., Liou, N. M., Tuytens, M., & Del Fresno, M. (2015). Why so
difficult? Exploring negative relationships between educational leaders: the role of trust,
climate, and efficacy. American Journal of Education, 122(1), 1–38.
Darling-Hammond, L., Wei, R. C., Andree, A., Richardson, N., & Orphanos, S. (2009).
Professional learning in the learning profession: A status report on teacher development
in the United States and abroad. Palo Alto, CA: National Staff Development Council.
DeArmond, M., & Gross, B. (2016). Personalized learning can’t ignore school leaders.
Washington, Seattle: Center for Reinventing Public Education. Retrieved from
BUILDING NETWORKS FOR CHANGE
183
http://www.crpe.org/thelens/PL-school-
leaders?utm_source=newsletter&utm_medium=email&utm_content=Read%20more&ut
m_campaign=Notes%20From%20the
Desimone, L. M. (2009). Improving impact studies of teachers’ professional development:
Toward better conceptualizations and measures. Educational Researcher, 38(3), 181–
199.
DiPaola, M. F., & Tschannen-Moran, M. (2005). Bridging or buffering? The impact of schools’
adaptive strategies on student achievement. Journal of Educational Administration,
43(1), 60–71.
Durand, F. T., Lawson, H. A., Wilcox, K. C., & Schiller, K. S. (2016). The role of district office
leaders in the adoption and implementation of the Common Core State Standards in
elementary schools. Educational Administration Quarterly, 52(1), 45–74.
https://doi.org/10.1177/0013161X15615391
Ertmer, P. A. (2005). Teacher pedagogical beliefs: The final frontier in our quest for technology
integration? Educational Technology Research and Development, 53(4), 25–39.
Ertmer, P. A., Ottenbreit-Leftwich, A. T., Sadik, O., Sendurur, E., & Sendurur, P. (2012).
Teacher beliefs and technology integration practices: A critical relationship. Computers
& Education, 59(2), 423–435.
Finnigan, K. S., & Daly, A. J. (2010). Learning at a system level: Ties between principals and
low-performing schools and central office leaders. In A. J. Daly (Ed.), Social network
theory and educational change. Cambridge, MA: Harvard Education Press.
Firestone, W. A., & Martinez, M. C. (2007). Districts, teacher leaders, and distributed leadership:
Changing instructional practice. Leadership and Policy in Schools, 6(1), 3–35.
BUILDING NETWORKS FOR CHANGE
184
Flanigan, R. L. (2016, June 9). Ed-tech coaches becoming steadier fixture in classrooms.
Education Week. Retrieved from http://www.edweek.org/ew/articles/2016/06/09/ed-tech-
coaches-becoming-steadier-fixture-in-classrooms.html?cmp=eml-enl-eu-news1
Frank, K. A., Zhao, Y., Penuel, W. R., Ellefson, N., & Porter, S. (2011). Focus, fiddle, and
friends: Experiences that transform knowledge for the implementation of innovations.
Sociology of Education, 84(2), 137–156.
Freeman, L. C. (1978). Centrality in social networks: Conceptual clarification. Social Networks,
1, 215–239.
Garet, M. S., Porter, A. C., Desimone, L. M., Birman, B. F., & Yoon, K. S. (2001). What makes
professional development effective? Results from a national sample of teachers.
American Educational Research Journal, 38(4), 915–945.
Glazer, E. M., & Hannafin, M. J. (2008). Factors that influence mentor and teacher interactions
during technology integration collaborative apprenticeships. Journal of Technology and
Teacher Education, 16(1), 35–61.
Glazer, E. M., Hannafin, M. J., Polly, D., & Rich, P. (2009). Factors and interactions influencing
technology integration during situated professional development in an elementary school.
Computers in Schools: Interdisciplinary Journal of Practice, Theory, and Applied
Research, 26(1), 21–39.
Goodreau, S. M. (2007). Advances in exponential random graph (p*) models applied to a large
social network. Social Networks, 29, 231–248.
Goodreau, S. M., Kitts, J. A., & Morris, M. (2009). Birds of a feather, or a friend of a friend?
Using exponential random graph models to investigate adolescent social networks.
Demography, 46(1), 103–125.
BUILDING NETWORKS FOR CHANGE
185
Gould, R. V., & Fernandez, R. M. (1989). Structures of mediation: A formal approach to
brokerage in transaction networks. Sociological Methodology, 19, 89–126.
Granovetter, M. S. (1973). The strength of weak ties. American Journal of Sociology, 1360–
1380.
Gray, L., Thomas, N., & Lewis, L. (2010). Teachers’ use of educational technology in U.S.
public schools: 2009 (NCES 2010-040). Washington, D.C.: National Center for
Education Statistics, Institute of Education Sciences, U.S. Department of Education.
Gronn, P. (2000). Distributed properties: A new architecture for leadership. Educational
Management Administration and Leadership, 28(3), 317–338.
Grossman, P., Smagorinsky, P., & Valencia, S. (1999). Appropriating tools for teaching English:
A theoretical framework for research on learning to teach. American Journal of
Education, 108(1), 1–29.
Hansen, M. T. (1999). The search-transfer problem: The role of weak ties in sharing knowledge
across organization subunits. Administrative Science Quarterly, 44(1), 82–111.
Hargreaves, A., & Fink, D. (2005). Sustainable leadership. San Francisco, CA: Jossey-Bass
Publishers.
Harris, A. (2004). Distributed leadership and school improvement: Leading or misleading?
Educational Management Administration and Leadership, 32(1), 11–24.
Harris, A. (2011). System improvement through collective capacity building. Journal of
Educational Administration, 49, 624–636.
Heifetz, R., Grashow, A., & Linsky, M. (2009). The practice of adaptive leadership: Tools and
tactics for changing your organization and the world. Boston, MA: Harvard Business
Press.
BUILDING NETWORKS FOR CHANGE
186
Herold, B. (2015, June 10). Why ed-tech is not transforming how teachers teach: Student-
centered, technology-driven instruction remains elusive for most. Education Week.
Retrieved from http://www.edweek.org/ew/articles/2015/06/11/why-ed-tech-is-not-
transforming-
how.html?tkn=OPRFoHJGLvKAMhV6awlIV3D5sAGGQT29NQRy&print=1
Ho, J., & Ng, D. (2017). Tensions in distributed leadership. Educational Administration
Quarterly, 53(2), 223–254. https://doi.org/10.1177/0013161X16681630
Honig, M. I. (2003). Building policy from practice: District central office administrators’ roles
and capacity for implementing collaborative education policy. Educational
Administration Quarterly, 39(3), 292–338.
Honig, M. I. (2004). Where’s the “up” in bottom-up reform? Education Policy, 18(4), 527–561.
Honig, M. I. (2012). District central office leadership as teaching: How central office
administrators support principals’ development as instructional leaders. Educational
Administration Quarterly, 48(4), 733–774.
Honig, M. I., & Hatch, T. (2004). Crafting coherence: How schools strategically manage
multiple, external demands. Educational Researcher, 33(8), 16–30.
https://doi.org/10.3102/0013189X033008016
Honig, M. I., & Ikemoto, G. S. (2008). Adaptive assistance for learning improvement: The case
of the Institute for Learning. Peabody Journal of Education, 83(3), 328–363.
https://doi.org/10.1080/01619560802222327
Hopkins, M., Ozimek, D., & Sweet, T. M. (2016). Mathematics coaching and instructional
reform: Individual and collective change. The Journal of Mathematical Behavior.
https://doi.org/http://dx.doi.org/10.1016/j.jmathb.2016.11.003
BUILDING NETWORKS FOR CHANGE
187
Huguet, A., Marsh, J. A., & Farrell, C. C. (2014). Building teachers’ data-use capacity: Insights
from strong and developing coaches. Education Policy Analysis Archives, 22(52).
Retrieved from http://dx.doi.org/10.14507/epaa.v22n52.2014
Huisman, M. (2014). Imputation of missing network data: Some simple procedures.
Encyclopedia of Social Network Analysis and Mining, 707–715.
Hunter, D. R., Goodreau, S. M., & Handcock, M. S. (2008). Goodness of fit of social network
models. Journal of the American Statistical Association, 103(481), 248–258.
Ibarra, H. (1993). Personal networks of women and minorities in management: A conceptual
framework. Academy of Management Review, 18(1), 56–87.
Inan, F. A., & Lowther, D. L. (2009). Factors affecting technology integration in K-12
classrooms: a path model. Educational Technology Research and Development, 58(2),
137–154. https://doi.org/10.1007/s11423-009-9132-y
International Society for Technology in Education. (2011). ISTE Standards Coaches. Author.
Retrieved from http://www.iste.org/docs/pdfs/20-14_ISTE_Standards-C_PDF.pdf
Jabbar, H. (2015). Competitive networks and school leaders’ perceptions: The formation of an
education marketplace in Post-Katrina New Orleans. American Educational Research
Journal, 52(6), 1093–1131.
Kadushin, C. (2012). Understanding social networks: Theories, concepts, and findings. New
York, NY: Oxford University Press.
Kopcha, T. J. (2012). Teachers’ perceptions of the barriers to technology integration and
practices with technology under situated professional development. Computers &
Education, 59, 1109–1121.
BUILDING NETWORKS FOR CHANGE
188
Lake, R., Hill, P. T., & Maas, T. (2015). Next Generation school districts: What capacities do
districts need to create and sustain tools that are ready to deliver on Common Core.
Seattle, WA: Center for Reinventing Public Education. Retrieved from
http://crpe.org/research/innovation-and-technology
Larson, A. (1992). Network dyads in entrepreneurial settings: A study of governance of
exchange relations. Administrative Science Quarterly, 51(2), 123–140.
Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. New
York, NY: Cambridge University Press.
Lawless, K. A., & Pellegrino, J. W. (2007). Professional development in integrating technology
into teaching and learning: Knowns, unknowns, and ways to pursue better questions and
answers. Review of Educational Research, 77(4), 575–614.
Lazonder, A. W., & Harmsen, R. (2016). Meta-analysis of inquiry-based learning: effects of
guidance. Review of Educational Research, 86(3), 681–718.
Lin, N. (2001). Social capital: A theory of social structure and action. New York, NY:
Cambridge University Press.
Liou, Y. (2016). Tied to the Common Core: Exploring the characteristics of reform advice
relationships of educational leaders. Educational Administration Quarterly, 1–48.
https://doi.org/10.1177/0013161X16664116
Little, J. W. (2002). Locating learning in teachers’ communities of practice: Opening up
problems of analysis in records of everyday work. Teachers and Teaching: Theory and
Practice, 8(3), 345–354.
Louis, K. S., Marks, H. M., & Kruse, S. (1996). Teachers’ professional community in
restructuring schools. American Educational Research Journal, 33(4), 757–798.
BUILDING NETWORKS FOR CHANGE
189
Lusher, D., Koskinen, J., & Robins, G. (2012). Exponential random graph models for social
networks: Theory, methods, and applications. Cambridge, UK: Cambridge University
Press.
Marsden, P. V. (1990). Network data and measurement. Annual Review of Sociology, 16, 435–
463.
Marsden, P. V. (2011). Survey methods for network data. In J. Scott & P. Carrington (Eds.), The
SAGE handbook of social network analysis. Thousand Oaks, CA: Sage.
Marsh, J. A., Bertrand, M., & Huguet, A. (2015). Using data to alter instructional practice: The
mediating role of coaches and professional learning communities. Teachers College
Record, 117(4), 1–40.
Marsh, J. A., Bush-Mecenas, S., & Hough, H. (2016). Local control in action: Learning from the
CORE Districts’ focus on measurement, capacity building, and shared accountability
(No. 16-4) (pp. 1–14). Stanford, CA: Policy Analysis for California Education, Stanford
Graduate School of Education. Retrieved from https://www.edpolicyinca.org/publications
Marsh, J. A., Kerr, K., Ikemoto, G. S., Darilek, H., Suttrop, M., Zimmer, R. W., & Barney, H.
(2005). The role of districts in fostering instructional improvement. Santa Monica, CA:
RAND Corporation.
Marsh, J. A., McCombs, J. S., & Martorell, F. (2010). How instructional coaches support data-
driven decision making policy implementation and effects in Florida Middle Schools.
Educational Policy, 24(6), 872–907.
Matsumura, L. C., Garnier, H. E., & Resnick, L. B. (2010). Implementing literacy coaching the
role of school social resources. Educational Evaluation and Policy Analysis, 32(2), 249–
272.
BUILDING NETWORKS FOR CHANGE
190
Matsumura, L. C., & Wang, E. (2014). Principals’ sensemaking of coaching for ambitious
reading instruction in a high-stakes accountability policy environment. Education Policy
Analysis Archives, 22(51). https://doi.org/http://dx.doi.org/10.14507/epaa.v22n51.2014
Mayer, A., Woulfin, S., & Warhol, L. (2014). Moving the center of expertise: Applying a
communities of practice framework to understand coaching in urban school reform.
Journal of Educational Change, 16(1), 101–123.
McLaughlin, M. W., & Talbert, J. (2003). Reforming districts: How districts support school
reform. Seattle, WA: University of Washington Press.
Means, B., Toyama, Y., Murphy, R., & Baki, M. (2013). The effectiveness of online and blended
learning: A meta-analysis of the empirical literature. Teachers College Record, 115(3),
1–47.
Miller, L., Gross, B., Maas, T., Hernandez, J., Lu, A., & Lake, R. (2016). Financing
personalized learning: What can we learn from first-generation adopters? Seattle, WA:
Center for Reinventing Public Education. Retrieved from
http://crpe.org/publications/financing-personalized-learning-what-can-we-learn-first-
generation-adopters
Mishra, P., & Koehler, M. (2006). Technological pedagogical content knowledge: A framework
for teacher knowledge. The Teachers College Record, 108(6), 1017–1054.
Molnar, M. (2017, January 26). Schools can evaluate their own ed-tech with “Rapid Cycle”
coaching tool. Education Week. Retrieved from
https://marketbrief.edweek.org/marketplace-k-12/schools-can-evaluate-ed-tech-rapid-
cycle-coaching-tool/?cmp=eml-enl-ii-news2
BUILDING NETWORKS FOR CHANGE
191
Moolenaar, N. M. (2012). A social network perspective on teacher collaboration in schools:
Theory, methodology, and applications. American Journal of Education, 119(1), 7–39.
Moolenaar, N. M., Daly, A. J., & Sleegers, P. J. C. (2010). Occupying the principal position:
Examining relationships between transformational leadership, social network position,
and schools’ innovative climate. Educational Administration Quarterly, 46(5), 623–670.
Moolenaar, N. M., Daly, A. J., Sleegers, P. J., & Karsten, S. (2014). Social forces in school
teams: How demographic composition affects social relationships. In D. Zandvliet, P. den
Brok, T. Mainhard, & J. van Tartwijk (Eds.), Interpersonal relationships in education
(pp. 159–182). Rotterdam, The Netherlands: Sense Publishers.
Moolenaar, N. M., Sleegers, P. J., & Daly, A. J. (2011). Teaming up: Linking collaboration
networks, collective efficacy and student achievement. Teaching and Teacher Education,
28(2), 251–262.
Mudzimiri, R., Burroughs, E. A., Luebeck, J., Sutton, J., & Yopp, D. (2014). A look inside
mathematics coaching: Roles, content, and dynamics. Education Policy Analysis
Archives, 22(53). https://doi.org/10.14507/epaa.v22n53.2014
Nahapiet, J., & Ghoshal, S. (1998). Social capital, intellectual capital, and the organizational
advantages. Academy of Management Review, 23(2), 377–401.
National Center on Standards & Assessment Implementation. (2016). State of the states.
Retrieved March 23, 2016, from http://www.csai-online.org/sos
Neumann, A. (2006). Professing passion: Emotion in the scholarship of professors at research
universities. American Educational Research Journal, 43(3), 381–424.
Obstfeld, D. (2005). Social networks, the tertius lungens orientation, and involvement in
innovation. Administrative Science Quarterly, 50(1), 100–130.
BUILDING NETWORKS FOR CHANGE
192
Pane, J. F., Griffin, B. A., McCaffrey, D. F., & Karam, R. (2013). Effectiveness of Cognitive
Tutor Algebra 1 at scale (WR-984-DEIES). Santa Monica, CA: RAND.
Papay, J. P., & Kraft, M. A. (2015). Productivity returns to experience in the teacher labor
market: Methodological challenges and new evidence on long-term career improvement.
Journal of Public Economics, 130, 105–119.
Patton, M. Q. (2002). Designing qualitative studies (3rd ed.). Thousand Oaks, CA: Sage
Publications, Inc.
Penuel, W. R., Riel, M., Krause, A., & Frank, K. A. (2009). Analyzing teachers’ professional
interactions in a school as social capital: A social network approach. Teachers College
Record, 111(1), 124–163.
Penuel, W. R., Sun, M., Frank, K. A., & Gallagher, H. A. (2012). Using social network analysis
to study how collegial interactions can augment teacher learning from external
professional development. American Journal of Education, 119(1), 103–136.
Portes, A. (2000). The two meanings of social capital. Sociological Form, 15(1), 1–12.
https://doi.org/10.1111/ 1467-8551.00220
Putnam, R. D. (1993). Making democracy work. Princeton, NJ: Princeton University Press.
Reagans, R., & McEvily, B. (2003). Network structure and knowledge transfer: The effects of
cohesion and range. Administrative Science Quarterly, 48(2), 240–267.
Reagans, R., & Zuckerman, E. W. (2001). Network diversity and productivity: The social capital
of R&D teams. Organization Science, 12, 502–517.
Robins, G., Pattison, P., Kalish, Y., & Lusher, D. (2007). An introduction to exponential random
graph (p*) models for social networks. Social Networks, 29(2), 173–191.
Rogers, E. M. (2003). The diffusion of innovations. New York, NY: Free Press.
BUILDING NETWORKS FOR CHANGE
193
Rouse, C. E., & Krueger, A. B. (2004). Putting computerized instruction to the test: A
randomized evaluation of a “scientifically based” reading program. Economics of
Education Review, 23, 323–338.
Scott, J. (2000). Social network analysis: A handbook (2nd ed.). Newbury Park, CA: Sage.
Smith, M. S., & O’Day, J. (1991). Systemic school reform. In S. H. Furhman & B. Malen (Eds.),
The politics of curriculum and testing (pp. 233–267). Washington, D.C.: Falmer Press.
Snijders, T. A. B., van de Bunt, G. G., & Steglich, C. E. G. (2010). Introduction to stochastic
actor-based models for network dynamics. Social Networks, 32, 44–60.
Snoeyink, R., & Ertmer, P. A. (2001). Thrust into technology: How veteran teachers respond.
Journal of Educational Technology Systems, 30(1), 85–111.
Spillane, J. P. (2013). The practice of leading and managing teaching in educational
organizations. In Leadership for 21st century learning (pp. 59–82). Paris, France:
Organization for Economic and Cooperative Development.
Spillane, J. P., & Burch, P. (2004). Leading from the middle: Mid-level district staff and
instructional improvement. Chicago, IL: Cross City Campaign for Urban School Reform.
Spillane, J. P., & Diamond, J. B. (2007). Distributed leadership in practice. New York, NY:
Teachers College Press.
Spillane, J. P., Halverson, R., & Diamond, J. B. (2004). Theory of leadership practice: A
distributed perspective. Journal of Curriculum Studies, 36(1), 3–34.
Spillane, J. P., Hopkins, M., & Sweet, T. (2015). Intra- and interschool interactions about
instruction: Exploring the conditions for social capital development. American Journal of
Education, 122(1), 71–110. https://doi.org/10.1086/683292
BUILDING NETWORKS FOR CHANGE
194
Spillane, J. P., Kim, C. M., & Frank, K. A. (2012). Instructional advice and information
providing and receiving behaviour in elementary schools: Exploring tie formation as a
building block in social capital development. American Educational Research Journal,
49(6), 1112–1145.
Stake, R. (1995). The art of case study. Thousand Oaks, CA: Sage Publications, Inc.
Stuart, T. E. (1998). Network positions and propensities to collaborate: An investigation of
strategic alliance formation in a high-technology industry. Strategic Management
Journal, 43, 668–698.
Sugar, W. (2005). Instructional technologist as a coach: Impact of a situated professional
development program on teachers’ technology use. Journal of Technology and Teacher
Education, 13(4), 547–571.
Sun, M., Penuel, W. R., Frank, K. A., Gallagher, H. A., & Youngs, P. (2013). Shaping
professional development to promote the diffusion of instructional expertise among
teachers. Educational Evaluation and Policy Analysis, 35(3), 344–369.
Superville, D. R. (2015, January 21). School districts turn to teachers to lead. Education Week.
Retrieved from http://www.edweek.org/ew/articles/2015/01/21/school-districts-turn-to-
teachers-to-lead.html
Swan, B., & Dixon, J. (2006). The effects of mentor-supported technology professional
development on middle school mathematics teachers’ attitudes and practice.
Contemporary Issues in Technology and Teacher Education, 6(1), 67–86.
Swinnerton, J. (2007). Brokers and boundary crossers in an urban school district: Understanding
central-office coaches as instructional leaders. Journal of School Leadership, 17, 195–
221.
BUILDING NETWORKS FOR CHANGE
195
Tanenbaum, C., Le Floch, K., & Boyle, A. (2013). Are personalized learning environments the
next wave of K-12 education reform? Washington, D.C.: American Institutes for
Research. Retrieved from
http://www.air.org/sites/default/files/AIR_Personalized_Learning_Issue_Paper_2013.pdf
Tsai, W. (2000). Social capital strategic relatedness and the formation of intra-organizational
linkages. Strategic Management Journal, 21, 925–939.
Tsai, W., & Ghoshal, S. (1998). Social capital and value creation: The role of intrafirm networks.
Academy of Management Journal, 41(4), 464–476.
Tyack, D., & Cuban, L. (1995). Tinkering toward utopia: A century of public school reform.
Cambridge, MA: Harvard University Press.
Uzzi, B. (1997). Social structure and competition in interfirm networks: The paradox of
embeddedness. Administrative Science Quarterly, 42(1), 35–67.
Valente, T. W. (2010). Social networks and health: Models, methods, and applications. New
York, NY: Oxford University Press.
Wang, L., Ertmer, P. A., & Newby, T. J. (2004). Increasing preservice teachers’ self-efficacy
beliefs for technology integration. Journal of Research on Technology in Education,
36(3), 231–250.
Warschauer, M., & Matuchniak, T. (2010). New technology and digital worlds: Analyzing
evidence of equity in access, use, and outcomes. Review of Research in Education, 34(1),
179–225.
Wilhelm, A. G., Chen, I., Smith, T. M., & Frank, K. A. (2016). Selecting expertise in context:
Middle school mathematics teachers’ selection of new sources of instructional advice.
American Educational Research Journal, 53(3), 456–491.
BUILDING NETWORKS FOR CHANGE
196
Wimmer, A., & Lewis, K. (2010). Beyond and below racial homophily: ERG models of a
friendship network documented on Facebook. American Journal of Sociology, 116(2),
583–642.
Windschitl, M. (2002). Framing constructivism in practice as the negotiation of dilemmas: An
analysis of the conceptual, pedagogical, cultural, and political challenges facing teachers.
Review of Educational Research, 72(131), 131–175.
Windschitl, M., & Sahl, K. (2002). Tracing teachers’ use of technology in a laptop computer
school: The interplay of teacher beliefs, social dynamics, and institutional culture.
American Educational Research Journal, 39(1), 165–205.
Woulfin, S. L. (2014). Charting the research on policies and politics of coaching. Education
Policy Analysis Archives, 22(50), http://dx.doi.org/10.14507/epaa.v22n50.2014.
Yin, R. K. (2014). Case study research: Design and methods (5th ed.). Thousand Oaks, CA:
Sage Publications, Inc.
Zhang, X., Anderson, R. C., Morris, J., Miller, B., Nguyen-Jahiel, K. T., Lin, T., … Hsu, J. Y.
(2016). Improving children’s competence as decision makers: Contrasting effects of
collaborative interaction and direct instruction. American Educational Research Journal,
53(1), 194–223.
Zhao, Y., & Cziko, G. A. (2001). Teacher adoption of technology: A perceptual control theory
perspective. Journal of Technology and Teacher Education, 9(1), 5–30.
Zhao, Y., Pugh, K., Sheldon, S., & Byers, J. (2002). Conditions for classroom technology
innovations. The Teachers College Record, 104(3), 482–515.
BUILDING NETWORKS FOR CHANGE
197
Zheng, B., Warschauer, M., Lin, C., & Chang, C. (2016). Learning in one-to-one laptop
environments: A meta-analysis and research synthesis. Review of Educational Research,
1–33. https://doi.org/10.3102/0034654316628645
BUILDING NETWORKS FOR CHANGE
198
TABLES AND FIGURES
Table 1
Brokering Properties from Social Network Theory
Brokering Property Brokering Practice Social Network Definition
Betweenness
All brokering
practices
The total number of two-path ties where an actor is situated
between two other actors. This measure represents the total
efforts of ed-tech coaches to broker information in their school
systems, including coordinating information within the central
office, within and across schools, and facilitating top-down and
bottom-up communication between the central office and schools
Coordination
Central office
coordination
Brokering information between actors within the same group to
support the overall goals of this group. This property represents
ed-tech coaches’ efforts to coordinate information with other
central office administrators and frontline staff.
Itinerant coordination
Coordination within
schools
Mediating between two insiders in a group to support their goals.
The broker does not share in the group’s identity. This property
represents ed-tech coaches’ efforts to coordinate information
among school leaders and teachers in school.
Liasing
Coordination across
schools
A scenario where the broker, as well as the two actors that he/she
is mediating between, are affiliated with different groups and are
building connections between groups that do not have prior
allegiance to one another. This property represents ed-tech
coaches’ efforts to coordinate information among school leaders
and teachers in different schools
Representation
Top-down
communication
A scenario in which the broker and the actor who is sharing
information are from the same group, while the actor who is
receiving this information is an outsider from another group.
Representation involves negotiating how insights from the
broker’s group are shared with outsiders. This property represents
ed-tech coaches’ efforts to broker information from the central
office into schools and advise teachers on how to change their
instructional practice to align with the goals of the central office.
Gate-keeping
Bottom-up
communication
A scenario where the broker and the actor who is receiving
information are from the same group while the actor who is
sharing information is an outsider from another group. The
broker is a gatekeeper who decides whether to grant an outsider
access to his/her group, as well as how to present this outside
information to his/her group. This property represents ed-tech
coaches’ efforts to facilitate and/or buffer communication
channels from schools to the central office to inform central
office procedures and resources for supporting instructional
change.
Note. This table defines brokering properties from social network theory (Gould & Fernandez, 1989) and relates
them to ed-tech coach brokering practices. The yellow circle in these visuals represents ed-tech coaches in their
brokering role. The blue circles in these visuals represent the actors between whom ed-tech coaches broker
information. The clear circle in these visuals indicates if ed-tech coaches share the same group affiliation with these
other actors (i.e., actors within the clear circle are in the same group).
BUILDING NETWORKS FOR CHANGE
199
Table 2
Social Network Properties for Describing the Social Structure of Districts and Schools
Property Definition Influence on Brokering
In-degree centrality
Number of ties that an actor
receives in his/her network
Indicates the extent to which actors control the flow of resources in their network and operate as
opinion leaders regardless of formal role and status. In-degree centrality implies that an actor has the
necessary power and influence to broker information across groups.
Out-degree centrality
Number of ties that an actor
sends in his/her network
Indicates the extent to which actors engage in advice-seeking. Out-degree is central to how brokers
build expertise and credibility to share with others (via building personal networks for advice).
Network density
Proportion of ties that exist
among all possible connections
between actors in a network
Creates an efficient social infrastructure for disseminating and coordinating information, thus
facilitating brokering ties and the elimination of structural holes in networks.
Network reciprocity
The proportion of all ties that are
reciprocated out of all possible
connections between actors in a
network
Suggestive of interdependency, coordination, flexibility, and ultimately trust at the relational
(reciprocal ties) or organizational level (network reciprocity). This relational dependency can
facilitate complex instructional tasks that require a shared sense of purpose and coordinated action,
which is consistent with the idea of brokering.
Reciprocal ties
Whether ties are reciprocated
between two actors
Homophily
The tendency for actors to send
ties to actors who share similar
attributes (e.g., physically
proximate, work in the same
role, share the same gender)
Homophilous ties, while engendering reliability, trust, and common understanding that facilitates
information sharing, can also restrict communication to cliques of similar actors that can be
challenging for brokers to gain access to.
Network Closure
(Clustering)
The proportion of all triads in a
network that are closed through
transitive ties between all three
actors
The formation of actor cliques, while providing a dense set of relations for disseminating
information among clique members, can reinforce group-think and norms, circulate redundant
information, and shied individuals from external sources of influence such as brokers.
BUILDING NETWORKS FOR CHANGE
200
Table 3
Descriptive Comparison of DLCs to Teaching Staff in LUSD
DLCs Teachers
Prior/current experience in
elementary grade levels 57.14% 54.44%
Prior/current experience in middle
school grade levels 42.86% 23.70%
Prior/current experience in high
school grade levels 12.50% 26.23%
Prior/current experience in ELA 21.43% 20.87%
Prior/current experience in math 35.71% 15.51%
Prior/current experience in science 14.29% 12.41%
Prior/current experience in social
science 14.29% 19.46%
Female 80.00% 78.28%
Racial/ethnic minority 14.00% 2.95%
Master's degree or higher 73.30% 58.96%
Average years district experience 8.13 10.66
Sample of DLCs/teachers 15 709
BUILDING NETWORKS FOR CHANGE
201
Table 4
Comparison of Survey Respondents to Non-Respondents in Sample 1
All District Employees
Respond
(n=229)
Non-Respond
(n=65) Diff
Female 0.76 0.78 -0.03
Minority 0.03 0.07 -0.03
Years district experience 9.99 12.02 -2.03*
Master’s Degree 0.72 0.59 0.13*
Central office 0.06 0.06 0.00
School principal 0.17 0.00 0.17***
DLC 0.06 0.02 0.05
DF 0.62 0.78 -0.16*
TTC representative 0.13 0.12 0.00
Elementary School 0.49 0.46 0.03
Middle School 0.32 0.25 0.07
High School 0.20 0.23 -0.03
Minority students 0.53 0.55 -0.03
Low-SES students 0.44 0.46 -0.02
ELL students 0.23 0.25 -0.03
SPED students 0.08 0.08 0.00
Female students 0.49 0.48 0.01+
Met/exceed ELA 0.57 0.53 0.04
Met/exceed math 0.48 0.44 0.05
Note. +=p<0.10, *=p<0.05, **p<0.01, ***=p<0.001. This table compares respondents to non-respondents in Sample
1 based on demographics, job position, school-level, school demographics and school performance on Common
Core-aligned standards-based assessments for the 2014-15 school year. For school demographics and performance,
employees working from the district’s central office are assigned district-wide values for these data points.
BUILDING NETWORKS FOR CHANGE
202
Table 5
Descriptive Statistics on Schools in Sample 2
Student Demographics Teacher Demographics Resp Rate
Enroll SD Min ELL SPED m/e ELA m/e math MA Min Yrs Exp #DFs
Central
Office
-- -- -- -- -- -- -- -- -- -- --
0.84
E1*^ 353 0.48 0.64 0.26 0.14 0.49 0.51 0.71 0.00 12.07 5 0.88
E2^ 678 0.89 0.89 0.66 0.07 0.34 0.23 0.22 0.05 9.96 6 0.60
E3*^ 384 0.89 0.93 0.58 0.11 0.26 0.17 0.47 0.00 9.40 4 0.69
E4*^ 452 0.43 0.58 0.25 0.08 0.60 0.57 0.88 0.00 13.06 3 0.79
E5*^ 376 0.08 0.25 0.09 0.08 0.72 0.79 0.54 0.08 11.85 4 0.67
E6*^ 523 0.96 0.98 0.79 0.10 0.19 0.18 0.48 0.00 9.48 3 0.70
E7* 540 0.52 0.69 0.33 0.12 0.44 0.41 0.45 0.00 10.60 4 0.71
E8* 582 0.10 0.26 0.05 0.09 0.74 0.69 0.61 0.07 9.79 4 0.67
E9* 679 0.25 0.44 0.16 0.07 0.52 0.55 0.64 0.00 13.20 6 0.80
M1* 938 0.71 0.77 0.31 0.12 0.47 0.39 0.58 0.00 10.03 10 0.70
M2 763 0.91 0.95 0.40 0.13 0.32 0.21 0.50 0.04 6.25 9 0.55
M3* 958 0.21 0.37 0.06 0.08 0.60 0.49 0.64 0.00 10.00 12 0.69
M4* 1171 0.13 0.19 0.05 0.04 0.77 0.75 0.46 0.00 6.74 16 0.89
M5 1248 0.12 0.20 0.04 0.05 0.64 0.64 0.53 0.05 11.38 13 0.61
M6
945 0.63 0.71 0.21 0.11 0.46 0.30 0.61 0.10
9.83 11
0.56
Elementary
max 916 0.96 0.98 0.79 0.14 0.89 0.81 0.88 0.09 15.71 16
min 353 0.04 0.16 0.02 0.04 0.19 0.17 0.22 0.00 8.65 3
median 577 0.45 0.61 0.25 0.08 0.51 0.53 0.58 0.00 11.82 5
Middle
max 1248 0.91 0.95 0.40 0.13 0.77 0.75 0.64 0.10 11.38 16
min 763 0.12 0.19 0.04 0.04 0.32 0.21 0.46 0.00 6.25 9
median 952 0.42 0.54 0.13 0.09 0.54 0.44 0.56 0.02 9.92 12
Note. * indicates schools with staff response rates of 67% or higher. ^ indicates case schools for qualitative data analysis. All numbers based on data
for 2014-15 school year. The first panel in this table provides the survey response rate for all central office employees. The second panel in this table
provides data on school demographics, school staff, and school survey response rates. The bottom panel provides summary statistics for all
elementary and middle schools in the district. SD=percent students with low socioeconomic status; min=percent students who are underrepresented
minorities; ELL=percent students who are English Language Learners; SPED=percent students who are special education; met/exc ELA and
math=percent students who met or exceeded performance standards in ELA and math; MA (tchr)=percent teachers with Master’s degree; Min (tchr)=
percent teachers who are underrepresented minorities; Yrs Exp = average years of experience of teachers; #DFs=number of DFs in 2013-14 and
2014-15; and #DLCs=number of DLCs in 2014-15.
BUILDING NETWORKS FOR CHANGE
203
Table 6
Comparison of Survey Respondents to Non-Respondents in Sample 2
All District Employees
Respond
(n=265)
Non-Respond
(n=123) Diff
Female 0.83 0.80 0.03
Minority 0.03 0.04 -0.01
Years district experience 9.95 10.25 -0.30
Master’s Degree 0.65 0.46 0.19***
Central office 0.05 0.03 0.02
School principal 0.08 0.00 0.08***
DLC 0.05 0.01 0.04*
DF 0.32 0.19 0.13*
TTC representative 0.06 0.03 0.03
Other teacher 0.46 0.73 -0.27***
Teachers
Respond
(n=210)
Non-Respond
(n=117) Diff
Female 0.86 0.07 0.79
Minority 0.03 0.00 0.03
Years district experience 10.05 10.41 -0.36
Master’s Degree 0.60 0.16 0.44*
DF 0.38 0.18 0.20***
TTC representative 0.08 0.04 0.03
Other teacher 0.59 0.78 -0.19***
Elem multi subject 0.55 0.12 0.44*
Ela subject 0.25 -0.06 0.31
% Math subject 0.12 -0.10 0.22*
% Sci subject 0.12 -0.03 0.15
% Socsci subject 0.25 -0.05 0.30
% Kindergarten 0.08 -0.01 0.09
% Grade 1 0.10 0.03 0.08
% Grade 2 0.10 -0.02 0.11
% Grade 3 0.12 0.05 0.08
% Grade 4 0.12 0.08 0.04*
% Grade 5 0.14 0.03 0.11
% Grade 6 0.24 -0.14 0.38*
% Grade 7 0.32 0.01 0.31
% Grade 8 0.36 -0.09 0.45
Note. +=p<0.10, *=p<0.05, **p<0.01, ***=p<0.001. This table compares respondents to non-respondents in Sample
2 based on demographics, job position, and courses and grade-levels of instruction for all district employees (panel
1) and for all teachers (panel 2).
BUILDING NETWORKS FOR CHANGE
204
Table 7
Research Hypotheses and Supporting Evidence
Research Hypothesis Evidence from Social Network Analysis Supported?
Hypothesis 1: Ed-tech coaches are prominent
brokers (i.e., have high-levels of betweenness
and brokering relations for coordination,
itinerant coordination, liasing, representation
and gatekeeping) for (a) integrating
technology with instruction and (b) teaching
the CCSS.
Global brokerage score: DLCs are more likely to be
prominent brokers in social networks that have a positive
and significant global betweenness score. DLCs are more
likely to coordinate information within the central office
and within and among schools in social networks that have
positive and significant coordination, itinerant coordination
and liasing scores respectively. DLCs are more likely to
represent information from the central office into schools,
and filter information upward from schools into the central
office, in social networks that have positive and significant
global representation and gatekeeping scores respectively.
Individual brokerage score: DLCs are prominent brokers
relative to other central office and school actors in social
networks where they have high standardized brokerage
scores that exceed 1.96 in value (p<0.05). This can be seen
by comparing the mean standardized brokerage scores for
DLCs relative to other actor groups and by comparing the
distribution of standardized brokerage scores for DLCs to
that of other actor groups.
Tech: Partially supported. This reform network
has positive and significant global betweenness
coordination, and representation scores. In this
network, DLCs have large, positive
standardized betweenness, coordination, and
representation scores. However, DLCs mainly
seem to be coordinating information with other
DLCs in the central office rather than
communicating with central office
administrators and CCs. Moreover, I do not
find significant evidence of itinerant
coordination, liasing, or gatekeeping at the
global/network-level nor at the individual level
for DLCs.
CCSS: Not supported. This reform network has
positive and significant global betweenness,
coordination, and gatekeeping scores, but
DLCs do not have any significant standardized
brokerage scores in the network.
Hypothesis 2: Ed-tech coaches with power
and influence (i.e., high-levels of in-degree
centrality) are more capable at brokering
information on (a) integrating technology with
instruction, and (b) teaching the CCSS.
Whole-network analysis: Ed-tech coaches demonstrate
higher brokerage scores in social networks where they have
high in-degree centrality
ERG model: Ed-tech coaches demonstrate higher brokerage
scores in social network where they have positive and
significant tie-receiver effects.
Tech: Supported. In this reform network, DLCs
have significantly higher in-degree centrality
and positive and significant tie-receiver effects.
CCSS: Supported. In this reform network,
DLCs have significantly lower in-degree
centrality and positive but insignificant tie-
receiver effects.
Hypothesis 3: Ed-tech coaches with access to
expertise (i.e., high-levels of out-degree
centrality) are more capable of brokering
information on (a) integrating technology with
instruction, and (b) teaching the CCSS.
Whole-network analysis: Ed-tech coaches demonstrate
higher brokerage scores in social networks where they have
high out-degree centrality
ERG model: Ed-tech coaches demonstrate higher brokerage
Tech: Not supported. In this reform network,
DLCs have significantly higher out-degree
centrality, but negative and significant tie-
sender effects. I also find a negative and
significant out-degree popularity effect,
BUILDING NETWORKS FOR CHANGE
205
Research Hypothesis Evidence from Social Network Analysis Supported?
scores in social network where they have positive and
significant tie-sender effects relative to other central office
and school actors. They should also demonstrate higher
brokerage scores in social networks with a more positive
and significant out-degree popularity effect.
indicating that the most influential actors in this
reform network are the least likely to send ties
to others.
CCSS: Not supported. In this reform network,
DLCs have significantly higher out-degree
centrality, but this difference no longer holds in
my ERG models where I control for other
endogenous network conditions. I also find a
negative and significant out-degree popularity
effect, indicating that the most influential actors
in this reform network are the least likely to
send ties to others.
Hypothesis 4: Ed-tech coaches are more
capable of brokering information on (a)
integrating technology with instruction and (b)
teaching the CCSS in dense and reciprocated
networks.
Whole-network analysis: Ed-tech coaches demonstrate
higher brokerage scores in social networks with higher rates
of (a) network density and (b) network reciprocity.
ERG model: Ed-tech coaches demonstrate higher brokerage
scores in social networks with a more positive and
significant effect for: (a) edges and (b) reciprocal ties on tie
formation.
Tech: Partially supported. I find low rates of
network density and reciprocity in this network
even though DLCs are active brokers. That
said, I do find more dense and reciprocated ties
among central office actors, which are
conducive to DLCs coordinating information
within the central office and representing this
information in schools. As noted above, these
are the most prominent brokering practices of
DLCs in this network.
CCSS: Supported. I find low rates of network
density and reciprocity in this network and that
DLCs are not active brokers. I also find that
reciprocal ties are concentrated among school
actors and as such, could be shielding schools
from seeking instructional advice from the
central office.
Hypothesis 5: Ed-tech coaches are less
capable of brokering information on (a)
integrating technology with instruction and (b)
teaching the CCSS in fragmented networks
with actor cliques (i.e., homophilous ties and
network clustering).
Whole-network analysis: Ed-tech coaches demonstrate
lower brokerage scores in social networks with higher rates
of network clustering relative to network density.
ERG model: Ed-tech coaches demonstrate lower brokerage
scores in social networks with more positive and significant
Tech: Partially supported. I find higher rates of
network clustering (relative to network density)
in this network. I also find significant and
positive effects for homophily and cyclic
closure in my ERG models. Nevertheless,
DLCs are prominent brokers in this network,
BUILDING NETWORKS FOR CHANGE
206
Research Hypothesis Evidence from Social Network Analysis Supported?
effects for homophily and cyclic closure on tie formation. suggesting that inward-looking ties are not
inhibiting the capacity of these coaches to
broker information in the district. That said, it
is possible that LUSD’s fragmented social
structure is inhibiting DLCs from engaging in
lateral brokering ties, such as coordinating
information across actor groups within the
central office or within and across schools,
which are more constrained in LUSD’s
technology reform network.
CCSS: Supported. I find higher rates of
network clustering in this network (relative to
network density). I also find significant and
positive effects for homophily and cyclic
closure in my ERG models. Alongside these
trends, I find that DLCs are not prominent
brokers in this network.
BUILDING NETWORKS FOR CHANGE
207
Table 8
ERGM Covariates and Research Hypotheses
Model Parameters Research Hypothesis / Control
Tie-Receiver/Sender Covariates
Formal position: Binary indicators identifying tie-
receivers and senders who are CCs, DLCs, Principals,
DFs, TTC representatives, and other teachers (reference
group is central office administrators)
Hypothesis 2: Ed-tech coaches with power and
influence (i.e., high-levels of in-degree centrality)
are more capable at brokering information on (a)
integrating technology with instruction, and (b)
teaching the CCSS.
Hypothesis 3: Ed-tech coaches with access to
expertise (i.e., high-levels of out-degree
centrality) are more capable of brokering
information on (a) integrating technology with
instruction, and (b) teaching the CCSS.
Subject area expertise: Binary indicators identifying
tie-receivers and senders with prior experience (current
in the case of teachers) in secondary ELA/social science,
secondary math, and secondary science instruction
(reference group is elementary multiple subjects). I also
include tie-receiver and sender indicators for expertise
in English language learner (ELL) and special
education (SPED) instruction.
Control
Grade-level expertise (DFs, TTC representatives,
and other teachers only): Binary indicators for tie-
receivers and senders who are teachers assigned to
tested grade-levels for CCSS-aligned assessments.
Control
Years of teaching experience: Binary indicators
identifying tie-receivers and senders with two or fewer
years of teaching experience, three to five years of
teaching experience, and six to nine years of teaching
experience (reference group is actors with 10 or more
years of teaching experience)
Control
Percent of students who meet proficiency standards
in CCSS-aligned ELA assessments: Tie-receiver and
sender covariates for the percent of students who meet
proficiency standards in 2014-15 CCSS-aligned ELA
assessments (% met/exceed ELA). This variable is mean-
centered around the district-wide proficiency rate in
ELA.
Control
Gender: Binary indicator for tie-receivers and senders
who are female.
Control
Educational attainment: Binary indicator for tie-
receivers and senders who have a master’s degree of
higher (Master’s degree plus).
Control
Dyadic Controls
Same location: Binary indicator for whether tie-
receivers and senders both work from the central office
or in the same school (same location).
Hypothesis 5: Ed-tech coaches are less capable of
brokering information on (a) integrating
technology with instruction and (b) teaching the
CCSS in fragmented networks with actor cliques
(i.e., homophilous ties and network clustering).
Same formal role: Binary indicator for whether tie-
receivers and senders work in the same formal role
(same formal role).
Same subject area expertise: Binary indicator for
whether tie-receivers and senders share the same subject
BUILDING NETWORKS FOR CHANGE
208
Model Parameters Research Hypothesis / Control
area expertise (same subject area expertise)
Same grade-level expertise (DFs, TTC representatives,
and teachers only): Binary indicator for whether tie-
receivers and senders are assigned to teach at the same
grade-level (same grade-level)
Same years of teaching experience: Binary indicator
for whether tie-receivers and senders fall into the same
categories for years of teaching experience (e.g., novice,
three to five years, six to nine years, and 10 or more
years (same years teaching experience)
Absolute difference in percent of students who meet
proficiency standards in CCSS-aligned ELA
assessments: Absolute difference between tie receivers
and senders in the percent of students who meet
proficiency standards in CCSS-aligned ELA
assessments in their work location (absolute difference
in ELA perf)
Same gender: Binary indicator for whether tie-receivers
and senders both have the same gender (both female).
Same education: Binary indicator for whether tie-
receivers and senders both have a Master’s degree or
higher (both Master’s degree plus).
Network Parameters
Out-degree popularity: The tendency for actors who
nominate more colleagues as a source of support to be
increasingly popular in their network (out-degree
popularity)
Hypothesis 3: Ed-tech coaches with access to
expertise (i.e., high-levels of out-degree
centrality) are more capable of brokering
information on (a) integrating technology with
instruction, and (b) teaching the CCSS.
Network density: The proportion of all possible ties
that are realized. In ERG models, network density is
accounted for through edges, the baseline propensity for
tie formation, which equates to network density in an
unconditional model (edges).
Network reciprocity: The proportion of all social ties
that are reciprocated. In ERG models, network
reciprocity is accounted for through reciprocal or mutual
ties, which measure the likelihood of observing
reciprocal ties in the network.
Hypothesis 4: Ed-tech coaches are more capable
of brokering information on (a) integrating
technology with instruction and (b) teaching the
CCSS in dense and reciprocated networks.
Network clustering: The proportion of all triads in a
network that are closed. When network clustering is
greater than network density, there is a “high” level of
clustering present which suggests that communication is
occurring in cliques with sparser regions in between. In
ERG models, network clustering is accounted for
through cyclic closure, which represents the likelihood
of observing cyclic triads.
Hypothesis 5: Ed-tech coaches are less capable of
brokering information on (a) integrating
technology with instruction and (b) teaching the
CCSS in fragmented networks with actor cliques
(i.e., homophilous ties and network clustering).
Note. Italic terms refer to how I name each covariate in my model output.
BUILDING NETWORKS FOR CHANGE
209
Table 9
Standardized Scores of Brokering ties in LUSD’s Reform Networks for Technology and CCSS
Sample 1 Sample 2 Central Office
Technology
Betweenness (B) 3.99 *** 3.37 *** -0.84
Coordination (C) 112.72 *** 133.67 *** 1.70 .
Itinerant coordination (IC) -6.14 *** -7.16 *** -1.56
Liasing (L) -11.15 *** -12.13 *** 0.79
Representation (R) 75.84 *** 58.00 *** -0.66
Gatekeeping (G) 6.69 *** 11.34 *** -2.35 *
CCSS
Betweenness (B) 1.50 -1.26 -1.77 .
Coordination (C) 65.92 *** 85.26 *** 0.26
Itinerant coordination (IC) -1.29
-3.74 *** -2.38 *
Liasing (L) -7.34 *** -10.11 *** -0.79
Representation (R) 24.49 *** 12.56 *** -0.23
Gatekeeping (G) 18.44 *** 19.41 *** -2.59 **
Note. +=p<0.10, *=p<0.05, **=p<0.01, ***=p<0.001. This table reports the global brokerage scores for
betweenness, coordination, itinerant coordination, liasing, representation, and gatekeeping for each of LUSD’s
social networks (Samples 1 and 2 and among all central office actors in the district). These standardized scores are
derived from the total number of ties observed across all network actors for each brokering role and the expected
value and standard deviation of these brokering ties given each network’s density and the total number of actor
groups in each network (n=30 in Sample 1, n=16 in Sample 2, and n=3 for central office). A positive standardized
score indicates that the total count of brokering ties exceeds their expected value and that these brokering ties are a
prominent feature of the social network.
BUILDING NETWORKS FOR CHANGE
210
Table 10
Brokerage, In-degree and Out-degree Scores for Actors in LUSD’s Reform Networks
(1) Technology
Sample 1 c.o. admin CCs DLCs Principals DFs & TTC Other teachers
B 5.76 2.01 11.26 -0.83 -0.93 --
C 23.31 12.86 26.06 2.81 4.90 --
IC -0.46 -0.46 0.03 -0.47 -0.47 --
L -0.25 -0.73 -0.71 -1.04 -1.09 --
R 13.63 5.38 34.72 -0.03 -0.14 --
G 0.81 0.05 -0.29 1.12 0.63 --
in-degree 0.80* 0.20 2.40*** -0.24 -0.22*** --
out-degree 0.52 0.86** 1.45*** 0.21 -0.23*** --
Sample 2 c.o. admin CCs DLCs Principals DFs & TTC Other teachers
B 3.65 2.50 11.49 -0.17 -0.57 -0.89
C 28.37 15.70 31.70 4.64 6.94 2.84
IC -0.57 -0.06 0.43 -0.45 -0.58 -0.56
L -0.99 -0.75 -0.66 -0.89 -1.05 -1.07
R 9.55 7.41 37.34 0.03 -1.07 -0.26
G -0.09 0.14 -0.23 3.22 1.32 0.34
in-degree 0.65* 0.31 2.61*** -0.04 -0.08 -0.28***
out-degree 0.22 1.00** 1.42*** 0.55* 0.08 -0.31***
(2) CCSS
Sample 1 c.o. admin CCs DLCs Principals DFs & TTC Other teachers
B 2.97 4.91 1.02 0.86 -0.71 --
C 10.09 18.61 4.18 2.45 4.45 --
IC -0.32 0.23 -0.12 1.04 -0.39 --
L -0.04 0.18 -0.61 -0.27 -0.85 --
R 6.10 9.29 3.46+ 0.51 -0.01 --
G 1.29 1.39 0.40 8.01 0.27 --
in-degree 1.50*** 1.53*** 0.55* 0.18 -0.29*** --
out-degree 0.21 0.71* 0.92*** 0.44** -0.23*** --
Sample 2 c.o. admin CCs DLCs Principals DFs & TTC Other teachers
B 1.10 3.49 0.14 2.52 -0.46 -0.72
C 10.34 19.08 4.30 4.67 7.72 4.03
IC -0.44 2.19 -0.32 1.15 -0.49 -0.47
L -0.85 -0.58 -0.77 -0.35 -0.89 -0.90
R 4.22 7.68 1.52 0.57 0.00 -0.110
G -0.37 0.82 0.15 14.20 0.74 0.15
in-degree 1.37*** 1.69*** 0.31 0.77*** -0.15+ -0.28***
out-degree -0.10*** 0.68 0.68*** 0.80*** 0.02*** -0.23***
Note. +=p<0.10, *=p<0.05, **=p<0.01, ***=p<0.001. This table reports the mean standardized brokerage, in-
degree, and out-degree scores for each group of central office and school actors. The stars in this table indicate if
standardized in-degree and out-degree scores are significantly different from those of other actors in the network. I
do not t-test the mean brokerage scores for actor groups since these brokerage scores are already standardized.
BUILDING NETWORKS FOR CHANGE
211
Table 11
ERG Model Results Predicting Tie Formation in LUSD’s Reform Networks
Technology CCSS
Sample 1 Sample 2 Sample 1 Sample 2
(1) Formal role
CC (receiver) 0.81
1.15
1.87 *** 3.09
**
*
CC (sender) 0.81
1.14
0.74
1.06
DLC (receiver) 3.80 *** 4.16 *** 1.40 + 1.30
DLC (sender) 0.40 ** 0.45 ** 1.12
1.39
Principal (receiver) 0.17 *** 0.38 *** 1.04
1.62 *
Principal (sender) 5.06 *** 5.55 *** 6.75 *** 6.81
**
*
DF (receiver) 0.21 *** 0.25 *** 0.28 *** 0.22
**
*
DF (sender) 1.62 * 2.53 *** 1.64 ** 1.66 **
TTC rep (receiver) 1.25
1.20
0.36 *** 0.40 **
TTC rep (sender) 1.53 * 1.34
1.60 * 1.80 **
Other teacher (receiver) --
0.16 *** --
0.15
**
*
Other teacher (sender) --
2.13 *** --
1.57 *
(2) Homophily effects
Same location (central office or school) 9.42 *** 9.71 *** 8.55 *** 8.92
**
*
Same formal role 1.35 * 1.24 * 1.12 1.39
**
*
Same subject area expertise 3.41 *** 3.69 *** 3.38 *** 3.68
**
*
Same grade-level (teachers only) 1.52 ** 2.02 *** 3.58 *** 6.97
**
*
Same years teaching experience 1.32 ** 1.21 * 1.11 1.18
Absolute difference in ELA perf 0.97 *** 0.97 *** 0.96 *** 0.94
**
*
Both female 1.31 ** 1.21 + 1.39 * 1.23
Both Master's degree plus 1.18 1.03 1.26 1.19
(3) Network
Out-degree popularity 0.45 *** 0.55 *** 0.80 ** 0.86 *
Edges 0.03 *** 0.01 *** 0.00 *** 0.00
**
*
Mutual/reciprocity 2.22 ** 1.33
1.30
1.96 **
Cyclic closure 3.01 *** 3.28 *** 3.02 *** 2.33
**
*
Control variables Y Y Y Y
Alkaline Information Criterion 4,267
5,125
3,205
3,860
Bayesian Information Criterion 4,673
5,563
3,611
4,298
Log Likelihood -2,087
-2,514
-1,556
-1,882
# individuals 229
265
229
265
# dyads 52,212
69,960
52,212
69,960
Note. +=p<0.10, *=p<0.05, **=p<0.01, ***=p<0.001. Results reported in odds ratios. Control variables in this
model influence tie-receiver, -sender, and dyadic effects for subject area experience, grade-level assignment
(teachers only), years of teaching experience, district mean-centered school demographics, gender, and educational
attainment. I also include dyadic effects for whether tie-receivers and –senders share the same formal role and work
in the same location (i.e., both in the central office or in the same school). Complete model results available in
Appendix B.
BUILDING NETWORKS FOR CHANGE
212
Table 12
Network Statistics for LUSD’s Reform networks for Technology and the CCSS
Technology Size Density Reciprocity Cluster
Sample 1: central office + schools 229 0.01 0.16 0.15
- Central office 37 0.12 0.33 0.40
- Sample 1: schools 192 0.00 0.25 0.20
Sample 2: central office + schools 265 0.01 0.15 0.17
- Central office 37 0.12 0.33 0.40
- Sample 2: schools 228 0.01 0.19 0.22
CCSS Size Density Reciprocity Cluster
Sample 1: central office + schools 229 0.01 0.17 0.22
- Central office 37 0.07 0.21 0.35
- Sample 1: schools 192 0.01 0.18 0.30
Sample 2: central office + schools 265 0.01 0.23 0.22
- Central office 37 0.07 0.21 0.35
- Sample 2: schools 228 0.01 0.27 0.29
Note. In this table, I report the network size, density, reciprocity, and clustering of LUSD’s social networks for
integrating technology with instruction and teaching the CCSS. I report these statistics in several ways. First, I report
these statistics for the entire network for each sample. I then report these statistics for interactions among central
office actors and then among school actors in each sample.
Table 13
In-Degree of DLCs in LUSD’s Reform Networks for Technology and the CCSS
In-degree Demographics Prior Experience
Tech CCSS Gender Race Education Teach Coach Grade
School
(max)
Sarah* 4.74 0.23
F Minority Masters 6 2 KG 6
Chloe 3.59 0.52
F White Masters 16 2 KG-1, 3-5 2
Marsha 3.02 0.52
F White Masters 8 2 4-5 0
Danielle 3.16 -0.06
F White Masters 13 2 KG-2, 5 12
Cindy* 1.16 -0.34
F White Masters 14 1 KG, 2 0
max 4.74 0.52
-- -- -- 6 2 12
median 3.16 0.23
-- -- -- 13 2 2
min 1.16 -0.34
-- -- -- 16 1 0
Note. * indicates that DLC selected for multiple case study analysis. For prior experience, I report the number of
years of teaching experience (Teach), coaching experience (Coach), maximum years working at assigned school
sites (School (max)), and prior grade-level teaching experience of coaches.
BUILDING NETWORKS FOR CHANGE
213
Table 14
Characteristics of Teacher Interview Participants
Teacher Grade Gender Years Teaching DF DLC
Talk to
DLC?
E1 = San Rafael EL
Teacher 1 3 M 16 N Sarah N
Teacher 2 1 F 2 Y (14-15) Sarah Y
Teacher 3 5 F 3 Y (13-14) Sarah Y
Teacher 4 2 F 16 Y (14-15) Sarah Y
Teacher 5 4 M 10 Y (13-14) Sarah N
E2 = Glendale EL
Teacher 6 5 F 3 Y (13-14) Sarah Y
Teacher 7 3 F 3 Y (14-15) Sarah Y
Teacher 8 2 F 2 Y (13-14) Sarah Y
Teacher 9 KG F 3 Y (13-14) Sarah Y
E3 = Brand EL
Teacher 10 1 F 13 Y (14-15) Cindy Y
Teacher 11 5 F 4 Y (14-15) Cindy Y
E4 = York EL
Teacher 12 1 F 2 N Cindy N
Teacher 13 2 F 7 N Cindy N
Teacher 14 4 F 1 N Cindy N
Teacher 15 RSP M 6 N Cindy Y
Note. This table reflects the range of teachers whom I could interview at my case schools. To protect the identity of
teacher participants, I have masked their grade-level assignments by linking each teacher to the grade-levels of other
interview participants. The last column (Talk to DLC?) indicates whether teachers nominated their DLC as a source
of advice for integrating technology with instruction in my social network data.
BUILDING NETWORKS FOR CHANGE
214
Table 15
Matrix-Display of District Conditions and Brokering Practices that Influence the Social
Positioning of DLCs
Social Positioning
District Articulation of
DLC Responsibilities
Sarah’s Brokering
Practices
Cindy’s Brokering
Practices
DLCs and their
support for
technology-enabled
instruction are (or
are not) central to
teaching the CCSS
DLCs isolated from
instructional support for
Balanced Literacy
(assigned to CCs
instead).
Process-oriented
coaching
Minimal focus on
Balanced Literacy
Product-oriented
coaching
Minimal focus on
Balanced Literacy
DLCs facilitate
teacher
collaboration in
supporting teachers
to use technology
for instruction
DLCs mainly responsible
for supporting DFs.
Supporting DF
collaboration with non-
fellow teachers is a
secondary priority.
Collaborative coaching
Non-collaborative
coaching
Note. This table summarizes the district conditions and brokering practices of Sarah and Cindy with regards to
bridging technology to instruction of the CCSS (top panel) and using teacher collaboration to bridge teacher practice
with central office goals for technology-enabled instruction (bottom panel). The cells in the top-panel are color-
coded to reflect whether they strongly (dark green), moderately (medium green), hardly (light green) position Sarah
and Cindy (and their instructional support) as bridging technology to CCSS instruction. The cells in the bottom-
panel are color-coded to reflect whether they strongly (dark orange), moderately (medium orange), hardly (light
orange) position Sarah and Cindy (and their instructional support) as enabling teacher collaboration on technology-
enabled instruction.
BUILDING NETWORKS FOR CHANGE
215
Figure 1. Conceptual framework on the instructional leadership and brokering practices of ed-tech coaches, and the mediating role of
social structure and organizational context. This figure illustrates the instructional leadership practices of ed-tech coaches, how these
leadership practices can be observed through brokering, and aspects of organizational context and social structure that shape their
leadership. Text in red are aspects of this conceptual framework that I flesh out in greater detail when reviewing my qualitative data
and methods (Figure 7).
(1) Instructional Leadership
(2) Brokering Practices
Coordination
(1A.1) Develop central office instructional
vision that is coordinated across programs
and services
Alignment
(1C) Aligningteacher practice with central
office and school goals
Coherence
(1B.1) Translate central office vision into
school goals and strategies for instructional
improvement
(2A) Coordinating information within
central office
-middle-up
-lateral communication
(2B) Top-down communication from
central office to schools
-bridging and/or buffering
-social learning routines (dialogue and
norms, joint work, shared repertoire of
tools)
(4) Social structure
(2C) Coordinate information within and
across schools
-bridging and/or buffering
-social learning routines (dialogue and
norms, joint work, shared repertoire of
tools)
(2D) Bottom-up communication channels
from schools to central office
(1A.2) Develop procedures and resources to
support this vision
(1B.2) Adjusting central office procedures,
structures, and resources in response to
school implementation of reform
• Brokering properties
• Opinion leadership / in-degree
• Out-degree
• Density, reciprocity, homophily and
clustering
(3) Organizational Context
• Power and influence
• Access to expertise
• Informal communication patterns such as dense ties, reciprocal exchange, and closure
BUILDING NETWORKS FOR CHANGE
216
Figure 2. Theory of action for LUSD’s technology rollout and the Digital Coaching Program (DCP). This figure outlines the theory of
change for LUSD’s technology rollout and the role of the DCP in building a social infrastructure for systemic instructional change.
Teachers use technology to
deliver personalized and inquiry-
based instruction grounded in the
CCSS
Improved student
engagement
Improved student content
knowledge and 21
st
century skills
Improved student
performance on CCSS
assessments and narrowing
achievement gaps
Improve school access to
technology
• New digital tools and apps
• Online learning content
management system
• Internet bandwidth
• Technical support
Adoption of signature practices
for instruction (e.g., Balanced
Literacy, SAMR scale)
Digital Coaching Program
(1) Instructional Leadership
(1A) Coordination
(1B) Coherence
(1C) Alignment
(2) Brokering
(2A) Coordinate information
within central office
(2B) Top-down
communication from central
office to schools
(2C) Coordinate information
within and across schools
(2D) Bottom-up
communication from
schools to central office
(3A) Position
technology as a central
resource for supporting
CCSS instruction
(3B) Build DF and
teacher knowledge,
beliefs, and motivation
for using technology to
teach the CCSS
(3C) Develop school
vision for using
technology to improve
student learning
(3D) Facilitate collaboration
between DFs and teachers on
instruction
(3G) Build social
infrastructure for
disseminating and
sustaining instructional
improvement
(3E) Build interschool networks
for instructional change
(3F) Schools advise DLCs and
central office on how to support
school improvement efforts
(4) Social structure
• Opinion leadership / in-degree
• Out-degree
• Density, reciprocity, homophily,
and clustering
Theory of Action of LUSD’s Technology Rollout
BUILDING NETWORKS FOR CHANGE
217
Figure 3. Box plot of standardized brokerage scores for central office administrators (CAs), CCs,
DLCs, principals and DFs and TTC representatives in LUSD’s social network for integrating
technology with instruction (Sample 1). This graph shows the distribution of standardized
individual brokerage scores for each actor group in LUSD’s social network for integrating
technology with instruction. The top panel shows the distribution of these scores for overall
betweenness, coordination, and itinerant coordination. The bottom panel shows the distribution
of these scores for liasing, representation, and gatekeeping. The red lines in these graphs indicate
threshold values of 1.96 and -1.96 to demonstrate when individualized brokerage scores are
statistically significant.
0 50 100 150 200
betweeness zscore
CA CC DLC Prin TTC/DF
0 50 100 150 200
coordinator zscore
CA CC DLC Prin TTC/DF
0 50 100 150 200
itinerant coordinator zscore
CA CC DLC Prin TTC/DF
0 50 100 150 200
liason zscore
CA CC DLC Prin TTC/DF
0 50 100 150 200
representative zscore
CA CC DLC Prin TTC/DF
0 50 100 150 200
gateskeeper zscore
CA CC DLC Prin TTC/DF
BUILDING NETWORKS FOR CHANGE
218
Figure 4. Box plot of standardized brokerage scores for central office administrators (CAs), CCs,
DLCs, principals and DFs and TTC representatives in LUSD’s social network for teaching the
CCSS (Sample 1). This graph shows the distribution of standardized individual brokerage scores
for each actor group in LUSD’s social network for teaching the CCSS. The top panel shows the
distribution of these scores for overall betweenness, coordination, and itinerant coordination. The
bottom panel shows the distribution of these scores for liasing, representation, and gatekeeping.
The red lines in these graphs indicate threshold values of 1.96 and -1.96 to demonstrate when
individualized brokerage scores are statistically significant.
0 50 100 150 200
betweeness zscore
CA CC DLC Prin TTC/DF
0 50 100 150 200
coordinator zscore
CA CC DLC Prin TTC/DF
0 50 100 150 200
itinerant coordinator zscore
CA CC DLC Prin TTC/DF
0 50 100 150 200
liason zscore
CA CC DLC Prin TTC/DF
0 50 100 150 200
representative zscore
CA CC DLC Prin TTC/DF
0 50 100 150 200
gateskeeper zscore
CA CC DLC Prin TTC/DF
BUILDING NETWORKS FOR CHANGE
219
Figure 5a. Network graph of LUSD’s social network for integrating technology with instruction (Sample 1). This network graph is
derived from Sample 1. Network actors are color coded based on their formal role in the district as follows: central office leaders and
administrators = light blue, CCs = light green, DLCs = dark blue, principals = red, DFs = orange, and TTC representatives = yellow.
BUILDING NETWORKS FOR CHANGE
220
Figure 5b. Network graph of LUSD’s social network for integrating technology with instruction (Sample 2). This network graph is
derived from Sample 2. Network actors are color coded based on their formal role in the district as follows: central office leaders and
administrators = light blue, CCs = light green, DLCs = dark blue, principals = red, DFs = orange, TTC representatives = yellow,
elementary school classroom teachers = pink, and middle school classroom teachers = purple.
BUILDING NETWORKS FOR CHANGE
221
Figure 6a. Network graphs of LUSD’s social networks for teaching the CCSS (Sample 1). This network graph is derived from Sample
1. Network actors are color coded based on their formal role in the district as follows: central office leaders and administrators = light
blue, CCs = light green, DLCs = dark blue, principals = red, DFs = orange, and TTC representatives = yellow.
BUILDING NETWORKS FOR CHANGE
222
Figure 6b. Network graphs of LUSD’s social networks for teaching the CCSS (Sample 2). This network graph is derived from Sample
2. Network actors are color coded based on their formal role in the district as follows: central office leaders and administrators = light
blue, CCs = light green, DLCs = dark blue, principals = red, DFs = orange, TTC representatives = yellow, elementary school
classroom teachers = pink, and middle school classroom teachers = purple.
BUILDING NETWORKS FOR CHANGE
223
Figure 7. Conceptual framework for the relationship between the social positioning of DLCs (i.e., their power and influence or in-
degree), their brokering practices, and individual, school, and district conditions that contribute to this social positioning.
Individual attributes of coach
• Instructional expertise (technological,
pedagogical and content knowledge)
• Interpersonal skills
• Prior experience in teaching and adult
education
• Goals for instructional change
• Perception of teachers and school
communities
• Familiarity with schools culture and
needs
Brokering Practices
• Bridging and/or buffering strategies
• Social learning routines
• Dialogue and norms
• Joint work
• Shared tools and practices
Social Positioning of DLCs (in-degree)
• Not central to advising teachers on
teaching the CCSS
• Central to advising teachers on how
to integrate technology with
instruction
Other school conditions: Knowledge, beliefs, and goals of teachers and principals, school infrastructure, human capital, other institutional
resources, and school size and demographics
Social conditions
• School principal support
• Principal and teacher perceptions of reform goals
• Principal and teacher perceptions of role of coach
• Norms for teacher collaboration and instructional change
• Institutional resources
District conditions:Design of coaching programs (selection and recruitment of coaches and teachers, articulation of coach responsibilities, assignment
of coaches to schools, training and monitoring of coaches)
BUILDING NETWORKS FOR CHANGE
224
San Rafael EL Glendale EL
Brand EL York EL
Figure 8. Social networks for teaching the CCSS at San Rafael EL, Glendale EL, Brand EL, and
York EL. Nodes are sized according to in-degree centrality of actors. Network actors are color
coded based on their formal role in the district as follows: CCs = light green, DLCs = dark blue,
principals = red, DFs = orange, TTC representatives = yellow, and elementary school classroom
teachers = pink.
BUILDING NETWORKS FOR CHANGE
225
San Rafael EL Glendale EL
Brand EL York EL
Figure 9. Social networks for integrating technology with instruction at San Rafael EL, Glendale
EL, Brand EL, and York EL. Nodes are sized according to in-degree centrality of actors.
Network actors are color coded based on their formal role in the district as follows: CCs = light
green, DLCs = dark blue, principals = red, DFs = orange, TTC representatives = yellow, and
elementary school classroom teachers = pink.
BUILDING NETWORKS FOR CHANGE
226
Figure 10. Plotting DLC brokering practices for bridging/buffering technology to/from CCSS
instruction according to individual attributes of coaches and school conditions. The vertical axis
in this figure spans from individual attributes that allow these coaches to bridge technology with
CCSS instruction (top-half) to individual attributes that lead these coaches to buffer technology
from CCSS instruction (bottom-half). The horizontal axis spans from school conditions that
buffer technology from CCSS instruction (left-half) to school conditions that bridge technology
to CCSS instruction (right-half). Each circle represents the brokering practices of DLCs at their
respective schools.
Instructional expertise and PD
experiences that bridgetechnology
with CCSS instruction
Instructional expertise and PD
experiences that buffertechnology
with CCSS instruction
School culture and prior exposure to
technology and Balanced Literacy that
bridgetechnology with CCSS
instruction
School culture and prior exposure to
technology and Balanced Literacy that
buffertechnology with CCSS
instruction
Sarah
San Rafael
Sarah
Glendale
Cindy
Brand & York
BUILDING NETWORKS FOR CHANGE
227
Figure 11. Plotting DLC brokering practices that enable/disable teacher collaboration on
technology-enabled instruction according to individual attributes of coaches and school
conditions. The vertical axis in this figure spans from individual attributes that enable these
coaches to engage teachers in collaborative routines for integrating technology with instruction
(top-half) to individual attributes that hinder these coaches from engaging teachers in
collaborative routines (bottom-half). The horizontal axis spans from school conditions that
hinder these coaches from engaging teachers in collaborative routines for integrating technology
with instruction (left-half) to school conditions that support these coaches to engage teachers in
collaborative routines (right-half). Each circle represents the brokering practices of DLCs at their
respective schools.
PD experience and familiarity with
schools that enableteacher
collaboration
PD inexperience and unfamiliarity
with schools that disableteacher
collaboration
School culture that enablesteacher
collaboration
School culture that disablesteacher
collaboration
Sarah
San Rafael
Sarah
Glendale
Cindy
Brand
Cindy
York
BUILDING NETWORKS FOR CHANGE
228
APPENDIX A: Goodness of Fit Tests for ERG Models
Figure A.1. Model fit for LUSD’s social network for integrating technology with instruction (Sample 1). This figure plots the
goodness-of-fit diagnostics for minimum geodesic distance, edge-wise shared partners, in-degree, and out-degree, comparing these
observed network features (thick black line) to box plots indicating the distribution of these features from 100 network simulations.
BUILDING NETWORKS FOR CHANGE
229
Figure A.2. Model fit for LUSD’s social network for teaching the CCSS (Sample 1). This figure plots the goodness-of-fit diagnostics
for minimum geodesic distance, edge-wise shared partners, in-degree, and out-degree, comparing these observed network features
(thick black line) to box plots indicating the distribution of these features from 100 network simulations.
BUILDING NETWORKS FOR CHANGE
230
APPENDIX B: Complete ERG Model Results
Table B.1.
Complete ERG model results predicting tie formation in LUSD’s social networks for integrating
technology with instruction and teaching the CCSS.
Technology CCSS
Sample 1 Sample 2 Sample 1 Sample 2
Formal role
CC (receiver) 0.81
1.15
1.87 *** 3.09 ***
CC (sender) 0.81
1.14
0.74
1.06
DLC (receiver) 3.80 *** 4.16 *** 1.40 + 1.30
DLC (sender) 0.40 ** 0.45 ** 1.12
1.39
Principal (receiver) 0.17 *** 0.38 *** 1.04
1.62 *
Principal (sender) 5.06 *** 5.55 *** 6.75 *** 6.81 ***
DF (receiver) 0.21 *** 0.25 *** 0.28 *** 0.22 ***
DF (sender) 1.62 * 2.53 *** 1.64 ** 1.66 **
TTC rep (receiver) 1.25
1.20
0.36 *** 0.40 **
TTC rep (sender) 1.53 * 1.34
1.60 * 1.80 **
Other teacher (receiver) --
0.16 *** --
0.15 ***
Other teacher (sender) --
2.13 *** --
1.57 *
Same location 9.42 *** 9.71 *** 8.55 *** 8.92 ***
Same formal role 1.35 * 1.24 * 1.12
1.39 ***
Control variables
Subject area experience
Secondary ELA/socsci (receiver) 0.82 + 0.76 ** 0.88
0.61 ***
Secondary ELA/socsci (sender) 0.82
1.10
0.70 * 0.79 +
Secondary math (receiver) 0.78 + 0.71 ** 0.82
1.21
Secondary math (sender) 0.95
1.32 + 0.50 ** 0.82
Secondary science (receiver) 0.87
0.91
0.52 *** 0.40 ***
Secondary science (sender) 1.11
0.95
1.38 * 1.00
ELL (receiver) 0.67 ** 0.65 *** 1.35 * 1.45 **
ELL (sender) 1.00
1.05
1.61 ** 1.25
SPED (receiver) 0.21 ** 0.33 ** 0.95
0.35 ***
SPED (sender) 0.53 * 0.48 ** 0.86
0.59 *
Same subject area expertise 3.41 *** 3.69 *** 3.38 *** 3.68 ***
Grade-level assignment
Tested grade-level (receiver) 0.56 *** 0.91
0.65 ** 0.79
Tested grade-level (sender) 1.40 * 1.06
1.40 * 1.10
Same grade-level (teachers) 1.52 ** 2.02 *** 3.58 *** 6.97 ***
Years teaching experience
Two or fewer (receiver) 1.46 ** 1.06
1.19
0.96
Two or fewer (sender) 0.90
1.20
0.83
1.36 *
Three to five (receiver) 1.23
1.22
1.54 ** 1.13
BUILDING NETWORKS FOR CHANGE
231
Technology CCSS
Sample 1 Sample 2 Sample 1 Sample 2
Three to five (sender) 0.78
0.98
0.77
0.89
Six to nine (receiver) 1.76 *** 1.55 *** 1.10
0.98
Six to nine (sender) 0.84
1.05
1.01
1.09
Same years teaching experience 1.32 ** 1.21 * 1.11
1.18
Student demographics
% met/exceed ELA (receiver) 1.00
1.01 * 1.00
1.01 **
% met/exceed ELA (sender) 0.99 * 1.00
0.99 * 0.98 **
Absolute difference in ELA perf 0.97 *** 0.97 *** 0.96 *** 0.94 ***
Personal characteristics
Female (receiver) 0.51 *** 0.65 *** 1.23
1.38 *
Female (sender) 1.11
0.87
0.99
0.81
Both female 1.31 ** 1.21 + 1.39 * 1.23
Master's degree plus (receiver) 1.33 * 1.16
0.98
1.21
Master's degree plus (sender) 0.59 *** 0.69 ** 0.54 *** 0.61 ***
Both Master's degree plus 1.18
1.03
1.26
1.19
Network
Out-degree popularity 0.45 *** 0.55 *** 0.80 ** 0.86 *
Edges 0.03 *** 0.01 *** 0.00 *** 0.00 ***
Mutual 2.22 ** 1.33
1.30
1.96 **
Cyclic closure 3.01 *** 3.28 *** 3.02 *** 2.33 ***
Alkaline Information Criterion 4,267
5,125
3,205
3,860
Bayesian Information Criterion 4,673
5,563
3,611
4,298
Loglikelihood -2,087
-2,514
-1,556
-1,882
# individuals 229
265
229
265
# dyads 52,212
69,960
52,212
69,960
Note. +=p<0.10, *=p<0.05, **=p<0.01, ***=p<0.001. Results reported in odds ratios.
BUILDING NETWORKS FOR CHANGE
232
APPENDIX C: Interview Questions linked to Conceptual Framework
Table C.1.
Crosswalk linking interview questions for DLCs, school principals, teachers, and central office
leaders, administrators, and instructional support staff to categories and constructs from my
conceptual framework.
Participant Interview questions liked to conceptual framework
Individual attributes of coach
DLC Do you have any previous experience with using technology for
instruction? (Technological, pedagogical, and content knowledge)
How would you describe your expertise as a digital coach?
(Technological, pedagogical, and content knowledge, interpersonal
skills)
What positions did you hold before becoming a coach (e.g., a teacher,
curriculum specialist, etc.)? How and why did you decide to become a
coach?
(Prior experience in teaching and instructional leadership,
interpersonal skills)
Did you have any previous experience working with adult learners?
Leading professional development?
(Prior experience in teaching and instructional leadership,
interpersonal skills)
How long you have been a coach at LUSD? A coach at schools [A] and
[B] in particular?
(Familiarity with school culture, goals, and needs)
What are your main responsibilities as a coach?
a) Are your responsibilities different between schools [A] and [B]? If
so, how? And why do you think that is?
b) Do teachers generally understand your role? How does this
understanding differ between schools [A] and [B]?
(Goals for teacher learning and change, coach perceptions of
teachers and school communities)
Principal I understand that [Name of Coach] is the digital learning coach assigned
to your school. How long have you known or been working with her in a
professional capacity?
(Familiarity with school culture, goals, and needs)
Brokering practices
DLC Please describe a typical day at school [A]? At school [B]?
**Note: The following questions ask about DLCs’ routines with their
BUILDING NETWORKS FOR CHANGE
233
Participant Interview questions liked to conceptual framework
DFs. I asked similar questions about their routines when hosting teacher
workshops, working with teachers in professional learning communities,
and supporting school/department/grade-level meetings.
Can you describe the teachers whom you are working with as digital
fellows?
a) Who decided that you should work with these teachers?
b) Why were these teachers selected?
c) Do you generally approach these teachers or wait for them to
approach you?
d) When do you typically work with teachers and for how long?
e) What is the general focus of this work? Or does it vary by individual?
(Social learning routines, design of coaching program – selection and
recruitment of teachers for training)
I would now like you to think of a teacher whom you have been working
with for the school year. Who is this teacher?
a) At the start of the year, how did you get a sense of or measure this
teacher’s prior knowledge/skills for integrating technology with
instruction?
b) What has been the focus of your work with this teacher?
c) Is there a strategy, model or framework that guides your approach to
coaching this teacher?
d) What learning materials, resources and/or tools (e.g., lesson plans,
samples of student work, etc.) have you shared with this teacher?
Could you provide me with a sample of these items?
e) What are your plans for working with this teacher for the remainder
of the school year?
f) How do you plan to follow up with this teacher in the following
school year?
(Bridging and/or buffering strategies, social learning routines)
What are the benefits of working one-on-one with a teacher to integrate
technology with instruction? What do you think are the drawbacks?
(Bridging and/or buffering strategies, social learning routines,
individual attributes of coach and teachers, school and district
conditions)
What factors facilitate your relationship with each of your teachers?
What factors get in the way or make it difficult?
(Bridging and/or buffering strategies, social learning routines,
individual attributes of coach and teachers, school and district
conditions)
Can you think of a time where you felt like you were successful or
BUILDING NETWORKS FOR CHANGE
234
Participant Interview questions liked to conceptual framework
effective in helping a teacher integrate technology with instruction?
a) What characteristics of the teacher made it easy or contributed to
his/her success?
b) How did your approach to coaching this teacher contribute to your
success in this situation?
c) Were there any other factors that made it easy?
(Buffering and/or bridging strategies, social learning routines,
individual attributes of coach and teachers, school and district
conditions)
And vice versa – can you think of a time when it was particularly
difficult to help a teacher implement technology-enabled instruction?
a) What characteristics of the teacher made it difficult?
b) How did your approach to coaching this teacher make the situation
more difficult?
c) Were there any other factors that made it difficult?
(Buffering and/or bridging strategies, social learning routines,
individual attributes of coach and teachers, school and district
conditions)
Principal What professional development programs have you organized at your
school to support teachers in integrating technology with instruction?
a) What has been the focus of these trainings?
b) What is your role in these trainings?
c) What, if any, is the role of DLCs in these trainings?
d) How are teachers recruited to attend these trainings?
e) Is there a strategy, model or framework that informs how these
workshops are conducted?
f) What learning materials, resources and/or tools do you share with
teachers at these trainings? Could you provide me with a sample of
these items?
g) Do you follow up with workshop participants with regards to the
content/tools/activities learned in this workshop? If so, how?
(School principal support for instructional role of coach, institutional
resources)
Can you recall an instance when a digital fellow was particularly
effective in sharing a new instructional resource or approach with
colleagues? If so, please describe this instance.
a) In what setting was the digital fellow able to share what she had
learned with her colleagues (e.g., school, district-wide, or grade-level
meeting, one-on-one conversation, etc.)
b) What was your role in facilitating this instance of knowledge sharing?
What about the digital coach?
c) What materials, resources, and/or tools did the digital fellow use to
convey her knowledge to colleagues?
BUILDING NETWORKS FOR CHANGE
235
Participant Interview questions liked to conceptual framework
d) What factors facilitated the digital fellow in reaching out to her
colleagues?
(Buffering and/or bridging strategies, social learning routines, school
principal support for instructional role of coach, other school
conditions)
Teacher **NOTE: I asked this question both for teachers’ ego-networks for
seeking advice on technology-enabled instruction and teaching the
CCSS.
In the fall semester, I administered a survey where I asked you to
nominate teachers and other staff members whom you approach for
advice for integrating technology with instruction. I have printed out a
list of the individuals whom you nominated in this survey. Can you
describe these people for me?
a) Are there other people whom you approach for advice on integrating
technology with instruction, not on this list? If so, whom?
b) If you approach certain people more or less frequently since
responding to this survey, how has this changed? Why?
c) How long have you known these staff members?
d) Why do you approach these staff members, in particular, for advice?
(Buffering and/or bridging strategies, social learning routines)
I would now like you to think back to the last time you approached these
people for advice on technology-enabled instruction?
a) Whom did you meet with?
b) What was the configuration of this meeting?
c) For how long did you meet with these staffs?
d) What aspects of your work did you focus on?
e) What routines did you engage in to go about this work?
f) What materials, resources, and/or tools were exchanged as part of
this meeting? Would you mind providing a sample of these items?
(Buffering and/or bridging strategies, social learning routines)
I noticed that in this list of people, you have (or have not) listed the
digital coach – [Name of Coach] – assigned to your school as a person of
contact. Can you explain why you rely (or do not rely) on her as a source
of advice?
(Buffering and/or bridging strategies, social learning routines,
teacher perceptions of coach, institutional resources)
**Note: I only asked this question of teachers with high in-degree in their
school networks for technology-enabled instruction
Is your school principal aware of your role in advising teachers to
integrate technology with instruction? What about the digital coach
assigned to your school?
BUILDING NETWORKS FOR CHANGE
236
Participant Interview questions liked to conceptual framework
a) Do you collaborate with the digital coach at your school in advising
teachers on how to integrate technology with instruction? If so, how?
b) Does your school principal support your efforts to advise other
teachers If so, how?
(Buffering and/or bridging strategies, social learning routines, school
principal support for instructional role of DLCs)
What other factors facilitate or hinder your work in advising teachers on
how to integrate technology with instruction?
(Individual attributes of teachers, institutional resources, and other
school and district conditions)
Social conditions in school
DLC What are your main responsibilities as a coach?
c) Who determines your responsibilities as a coach?
d) Are your responsibilities different between schools [A] and [B]? If
so, how? And why do you think that is?
e) Do teachers generally understand your role? How does this
understanding differ between schools [A] and [B]?
(Design of coaching programs – articulation of responsibilities,
teacher perceptions of coach, other school conditions)
What are your priorities or goals for your work this year at school [A]?
At school [B]?
a) Is there a certain aspect of teachers’ instructional practice that you are
targeting?
b) Who decided on those priorities (e.g. you, the principal, the district)?
c) Why were these priorities selected?
d) How did your priorities change over the course of the school year? If
so, how?
(Design of coaching programs – articulation of responsibilities)
I am also trying to get a sense of the school climate.
a) How would you characterize the nature of teachers’ interactions with
each other?
b) How open are teachers to experimenting with new teaching practices?
(Norms of risk-taking and teacher collaboration, coach perceptions
of teachers and school communities)
What support does your district or school provide to you as a coach?
a) What is the content and frequency of that support?
b) To what extent has this support been helpful to you?
c) Is there any additional support you would like?
d) Are you expected or required to participate in a certain amount of
professional development each year? If so, how much/how often?
(Design of coaching programs – training and support)
BUILDING NETWORKS FOR CHANGE
237
Participant Interview questions liked to conceptual framework
To what extent does your principal support your work with teachers
around technology-enabled instruction? Is there anything she/he does in
particular that supports or hinders your work?
(School principal support for instructional role of coach)
Principal What are your goals for student learning?
a) What intermediate goals have you set for teaching staff to reach these
goals?
b) How does LUSD’s technology rollout relate to these priorities?
c) How have your goals for student learning changed over the course of
the school year? What about your goals for teaching staff?
(School principal perception of reform goals)
I understand that [Name of Coach] is the digital learning coach assigned
to your school. How was this coach assigned to your school? Was there a
specific reason for this assignment?
(Design of coaching programs – assignment to schools)
What are the responsibilities of this digital learning coach at your school?
a) Who decided these responsibilities?
b) How do these responsibilities further your goals for student learning?
Goals for teaching staff?
(School principal perception of coach, design of coaching programs –
articulation of responsibilities)
How do you support the digital learning coach in fulfilling these
responsibilities?
(School principal support for instructional role of DLCs in schools)
How often do you meet with the digital learning coach to check in on her
work?
a) Do you wait for the digital coach to approach you or do you approach
her?
b) When do you typically meet with the coach and for how long?
c) What is the main focus of your conversations?
(School principal support for instructional role of DLCs in schools)
What are some of the challenges that you experience in supporting the
work of digital coaches at your school?
(School principal support for instructional role of DLCs in schools,
other school conditions)
I understand that another objective of the DCP is for digital fellows to
share best practices for instruction with colleagues. How do you facilitate
these opportunities for sharing at your school?
BUILDING NETWORKS FOR CHANGE
238
Participant Interview questions liked to conceptual framework
(School norms for teacher collaboration and instructional change,
school principal support for instructional role of DLCs in schools)
Have you encountered any challenges in working with digital fellows to
share best practices with colleagues? If so, please describe these
challenges.
(School norms for teacher collaboration and instructional change,
school principal support for instructional role of DLCs in schools,
other school conditions)
I understand that there might be some other teachers (i.e., fellows from
previous years, tech-savvy teachers) who can offer valuable expertise to
their colleagues.
a) Who are these teachers at your school?
b) How do you support these teachers in advising their colleagues who
might have less experience using technology for instruction?
c) Have you encountered any challenges in working with these tech-
savvy teachers to share best practices with colleagues? If so, please
describe these challenges.
(School norms for teacher collaboration and instructional change,
school principal support for instructional role of DLCs in schools,
other school conditions)
What support does your district provide to you to lead instructional
change at your school?
a) What is the content of this support?
b) How often is this support provided?
c) Can you please describe the general setting in which you receive this
support?
d) To what extent has this support been helpful to you?
e) Is there any additional support you would like?
f) Are you expected or required to participate in a certain amount of
professional development each year? If so, how much/how often?
(Institutional resources, other school and district conditions)
Teacher As you know, LUSD has been implementing a technology-based
instructional reform for the past two years. Could you please describe the
goals of this reform in your own words?
(Teacher perceptions of instructional reform goals)
How do the goals of this reform relate to your own priorities for
classroom instruction?
(Teacher perceptions of instructional reform goals)
What have been your priorities or goals for student learning this year?
a) How do these goals relate to the needs of the students in your
classroom?
BUILDING NETWORKS FOR CHANGE
239
Participant Interview questions liked to conceptual framework
b) How do these goals relate to the district’s goals for student learning?
Your school’s goals for student learning?
c) How did your priorities change over the course of the school year?
(Teacher perceptions of instructional reform goals)
I noticed that you have (or have not) listed the digital coach – [Name of
Coach] – assigned to your school as a person who provides you with
instructional support. Can you explain why you rely (or do not rely) on
her as a source of advice?
(Teacher perceptions of coach)
**Note: I only asked this question of teachers with high in-degree in their
school networks for technology-enabled instruction
I noticed that a lot of teachers at your school have listed you as a regular
source of advice for integrating technology with instruction.
a) How would you describe your role in this regard?
b) How did you come to assume this advisory role?
(School norms for teacher collaboration and instructional change)
I would now like you to think back to the last time you met with a
teacher(s) to provide guidance on in integrating technology with
instruction.
a) What is the configuration of this meeting?
b) For how long did you meet?
c) What aspects of teachers’ work did you focus on?
d) What routines did you engage in to go about this work?
e) What materials, resources, and/or tools did you exchange as part of
this meeting? Would you mind providing a sample of these items?
(School norms for teacher collaboration and instructional change)
Other school conditions
DLC I noticed that the student demographics in schools [A] and [B] are quite
different from one another. How does the difference in these student
populations influence your work with teachers?
(Other school conditions)
Principal How long you have been a school principal at LUSD? At this school in
particular?
(Individual attributes of school principal)
What positions did you hold in LUSD before becoming a school
principal (e.g., a teacher, curriculum specialist, etc.)? How and why did
you decide to become a school principal?
(Individual attributes of school principal)
Do you have any previous experience working with adult learners?
Leading professional development?
BUILDING NETWORKS FOR CHANGE
240
Participant Interview questions liked to conceptual framework
(Individual attributes of school principal)
Have you had prior experiences in using technology for instruction? If so,
please describe these experiences.
(Individual attributes of school principal)
Beyond what has already been discussed in our interview, are there other
factors at the school or district level that constrain or enable your efforts
to lead teachers in integrating technology with instruction?
(Other school or district conditions)
Teacher What is your current position at [Name of School]?
a) What are your main responsibilities in this position?
b) How long have you been working in this position?
(Individual attributes of teacher)
How long have you worked in an elementary school setting?
a) Have you worked in other roles/positions in an elementary school
setting? If so, what roles?
(Individual attributes of teacher)
What prior experiences have you had with using technology for
instruction?
(Individual attributes of teacher)
District conditions
DLC What are your main responsibilities as a coach?
f) Who determines your responsibilities as a coach?
g) Are your responsibilities different between schools [A] and [B]? If
so, how? And why do you think that is?
h) Do teachers generally understand your role? How does this
understanding differ between schools [A] and [B]?
(Design of coaching programs – articulation of responsibilities,
teacher perceptions of coach, other school conditions)
What are your priorities or goals for your work this year at school [A]?
At school [B]?
e) Is there a certain aspect of teachers’ instructional practice that you are
targeting?
f) Who decided on those priorities (e.g. you, the principal, the district)?
g) Why were these priorities selected?
h) How did your priorities change over the course of the school year? If
so, how?
(Design of coaching programs – articulation of responsibilities)
To what extent do central office administrators support your work with
teachers on technology-enabled instruction? Is there anything that they
BUILDING NETWORKS FOR CHANGE
241
Participant Interview questions liked to conceptual framework
do in particular that supports or hinders your work?
(District conditions)
Principal I understand that [Name of Coach] is the digital learning coach assigned
to your school. How was this coach assigned to your school? Was there a
specific reason for this assignment?
(Design of coaching programs – assignment to schools)
What are the responsibilities of this digital learning coach at your school?
c) Who decided these responsibilities?
d) How do these responsibilities further your goals for student learning?
Goals for teaching staff?
(School principal perception of coach, design of coaching programs –
articulation of responsibilities)
Beyond what has already been discussed in our interview, are there other
factors at the school or district level that constrain or enable your efforts
to lead teachers in integrating technology with instruction?
(Other school or district conditions)
Central office
leaders and
staff
How long you have been in [current role] at LUSD?
(Individual attributes of central office leaders and staff)
What positions did you hold in LUSD before [name of role] (e.g., a
teacher, curriculum specialist, etc.)?
(Individual attributes of central office leaders and staff)
Do you have any previous experience working with adult learners?
Leading professional development?
(Individual attributes of central office leaders and staff)
Have you had prior experiences in using technology for instruction? If so,
please describe these experiences.
(Individual attributes of central office leaders and staff)
What are the long-term priorities of the school district?
a) How does LUSD’s technology rollout fit into these priorities?
b) How has the leadership in LUSD communicated the goals of its
technology rollout to school principals? Teachers?
(Communication of central office goals for instructional change)
I understand that LUSD had supported multiple capacity building
measures to support teachers in integrating technology with classroom
instruction – in particular the DCP. Please describe these instructional
support programs.
a) Why did LUSD decide to pursue multiple capacity building
strategies? What is the purpose of each of these strategies?
BUILDING NETWORKS FOR CHANGE
242
Participant Interview questions liked to conceptual framework
b) In your opinion, how are these different capacity building strategies
supposed to work together (if at all)?
c) Do you think that teachers will find certain forms of capacity building
more helpful than others? If so, why?
d) What challenges have teachers experienced so far in participating in
these different capacity building programs?
e) What changes would you like to make to these capacity building
strategies for teachers?
(Design of coaching program, other district conditions)
I understand that you are specifically responsible for overseeing the
digital coach program at LUSD. Can you please describe how digital
coaches are hired to their position in LUSD?
(Design of coaching program – selection and recruitment of coaches)
What are the main responsibilities of digital coaches?
a) Who determines these responsibilities?
b) How are these responsibilities communicated to digital coaches? To
the principals and teachers at their school sites?
(Design of coaching program – articulation of responsibilities)
I understand that digital coaches are assigned to work across multiple
school sites. How are these school site assignments determined?
a) What is the typical caseload for digital coaches in terms of school
sites? Digital fellows?
(Design of coaching program – assignment to schools)
How does the central office support digital coaches in fulfilling these
responsibilities?
(Design of coaching program – training and support)
What measures, if any, are in place to evaluate the work of digital
coaches in schools?
(Design of coaching program – evaluation)
I understand that one of the main responsibilities of digital coaches is to
provide one-to-one guidance to digital fellows. I would like to learn more
about how teachers in LUSD are recruited to participate as digital
fellows. Can you describe the application process?
(Design of coaching program – selection and recruitment of teachers
for training)
What are the school district’s expectations of digital fellows?
(Design of coaching program – expected outcomes for trained
teachers)
BUILDING NETWORKS FOR CHANGE
243
APPENDIX D: Data Matrix Supporting Qualitative Findings
Table D.1.
Supporting evidence for why Sarah and Cindy are not central in school networks for teaching the CCSS.
Supporting Evidence From…
Finding & Sample Quote
Central
Admin
DLC Principal Teachers
Formal responsibilities of DLCs isolated from Balanced Literacy
Administrator: So, we are driving down this road, right, and we have our technology people doing
amazing work, they are experts in their subjects, they are terrific in teaching their grade level, and then
we started new initiatives for Common Core. That is Balanced Literacy – Reader’s and Writer’s
Workshop -- and CGI. And so, we are driving down this lane on two paths. I really wanted to make it
sort of join. And so, next year, our DLCs will become Connect coaches, and so they will be expected
to work on both sides of the house. We hired super strong instructional people who want to teach with
technology. And we hired our DLCs, who are very solid in their content and amazing at technology,
and we are going to merge them.
2/3 3/4 4/4 7/15
Sarah engaged in more process-oriented coaching strategies to bridge technology with CCSS instruction
Teacher 6: It was always an open discussion of what do you want to work on and how can we make it
more engaging? How can we take it to the next level [on the SAMR scale]. It was not substitution, but
we would think of an activity, and then how could we make that activity better with technology, so we
always had these conversations of, ‘Okay, well I want my kids to do this. What do you think? How can
we do that in a way that is going to get them to put their creativity into it and support their freedom of
expression and things like that, while collaborating with each other?’ It would just be this constant
dialogue of, ‘What do you think?’, ‘What apps have you heard of?’ that type of thing.
n/a 1/1 2/2 7/9
Sarah rarely focused on Balanced Literacy as part of her instructional support to teachers
Teacher 2: With Balanced Literacy, it is really a focus on them. Working up that stamina and being
able to read for an extended period of time independently and so if I have them doing a lot, like I might
have them do something maybe after their reading on the iPads but I have not incorporated a ton of the
technology with the Balanced Literacy just because I want them to be engaged in their books and
reading. I mean they have done a little bit of reading with the apps that we have on the iPad but mostly
that is probably the extent of it.
n/a 1/1 2/2 4/9
BUILDING NETWORKS FOR CHANGE
244
Supporting Evidence From…
Finding & Sample Quote
Central
Admin
DLC Principal Teachers
Sarah had relevant instructional expertise and PD experience for bridging technology with CCSS
instruction, but was not familiar with Balanced Literacy
Sarah: This position has taught me a lot about technology and the different tools that are out there. Knowing
that and knowing the curriculum and what is expected of students, I like to think that I'm good at putting
those together so that it is more effective in the classroom.
n/a 1/1 2/2 7/9
Sarah benefited from an innovative school culture at San Rafael that facilitated her process-oriented
approach to coaching.
Principal: Basically, it is not technology for technology's sake but it is more of technology and how are
we teaching those 21st century skills. That is when we were talking about creativity, collaboration,
communication and critical thinking. That is what we always focus on. We went through a lot of PD to
talk about, "This is not just about opening an e-book or it is not just sitting there, practicing math facts
skills"…We did lots of PD and talking about how we want to move forward.
n/a 1/1 1/1 4/5
But there was limited human capital at San Rafael to bridge technology with Balanced Literacy
Teacher 5: That is what I am talking about, at the district level because these are people that have
educational backgrounds. You have to bring in experts on technology but then they do not have
backgrounds on education necessarily so it is really bridging two entirely different fields of work and
then trying to get that to blend. That has been a challenge, I think.
n/a 1/1 1/1 5/5
Sarah benefited from a less innovative school culture at Glendale EL, limiting her process-oriented approach
to coaching to those teachers who were individually driven to be innovative and with whom she had strong
relations
Sarah: [The principal from San Rafael] and I sat down ahead of time and planned this all out … …
because she had certain teachers that she wanted me to support … I guess in that case she did say,
"Can you focus on this, can you help me out with this team and whatever?" At Glendale EL, that has
not happened.
n/a 1/1 1/1 3/4
There was limited human capital at Glendale EL to bridge technology with Balanced Literacy
Teacher 9: Part of it last year was I didn't always have something, I would be like, "I do not know what
to do. I need ideas from you [Sarah] because you are the one that has seen all this technology and all
these new things." It is like, I cannot ask [about] what I don't know.
n/a 1/1 1/1 4/4
BUILDING NETWORKS FOR CHANGE
245
Supporting Evidence From…
Finding & Sample Quote
Central
Admin
DLC Principal Teachers
Cindy engaged in more product-oriented coaching strategies that buffered technology from CCSS
instruction
Teacher 11: If I was in a place where I wanted to do a lesson that incorporated technology and I maybe
just did not have an idea for it or to know where to go, she would create it. She would say, “What if we
did this? What do you think about this?” She would give me an entire thing to use. I would go, “Oh my
gosh, that is awesome,” and then we would use it. I did not have to spend nights agonizing over what
app am I going to use? What project am I going to use for this? How am I going to do this? She would
say, “Here are some things that have already been created. Would any of these work in your
classroom?”
n/a 1/1 1/2 4/6
Cindy rarely focused on Balanced Literacy as part of her instructional support to teachers
We created a presentation. [The Balanced Literacy coach] talked about the reasons why you would
incorporate conferring. You need to talk to children, find out what their interest levels are, if they
are understanding what they are reading, things like that. I went into how you can Confer using
technology to make your life easier… My role was to show, “You could write it on paper or you
could do these things using technology.” It was just the “how to.”
n/a 1/1 0/2 3/6
Cindy demonstrated less instructional expertise (more knowledge about technology than instructional
planning) and less PD experience for bridging technology to CCSS instruction. She was also not familiar
with Balanced Literacy
Cindy: I excel in getting to know the teachers and getting to know their strengths, weaknesses,
insecurities, things like that, and I start there. From there I feel like I can introduce things. If there is
a teacher that I realize, okay, they have no technology skills at all, I am not going to go in and by
like, “Oh, you can do this, this, this, this, and this. I'm going to change [all] this. This is what you
should do."…A lot of times I have noticed some of the younger teachers are more technologically
advanced than the older teachers, and more willing to just go for it and try new things, so just
getting a sense of who they are from the start I think helps a lot.
n/a 1/1 2/2 3/6
BUILDING NETWORKS FOR CHANGE
246
Supporting Evidence From…
Finding & Sample Quote
Central
Admin
DLC Principal Teachers
Cindy encountered a slow-moving school culture at Brand EL that made it difficult for her to engage
teachers in process-oriented coaching strategies
Principal: We are Reader’s and Writer’s school and next year, we have a bunch of teachers that are
ready to start working on CGI and they are already doing it themselves…As an English language
learner school, we need to really push for lots of discussions and high levels of engagement, as
much authentic learning as possible, bringing in project-based learning, which is something we are
learning right now. We are just dabbling in that piece, that is sort of a new layer for us.
n/a 1/3* 1/1 2/2
There was limited human capital at Brand EL to bridge technology with Balanced Literacy
Balanced Literacy Coach: For example, take me using Padlet for an interactive read aloud. Until
you really understand and can do interactive read aloud, you are not going to see the benefit of the
technology that is going to help increase learning… [As teachers become] comfortable with
technology [then] as they learn more with Balanced Literacy, they will pull it in. Or as they move
further along the road with Balanced Literacy and we are bringing the technology to them, they will
be like oh, that would be good.
n/a 3/3* 1/1 2/2
Cindy encountered a slow-moving school culture at York EL that made it difficult for her to engage teachers
in process-oriented coaching strategies
Cindy: I have seen definite steps towards change, but they are small. I feel like some of that is the
principal because she said, "Okay, we are taking baby steps," which is great and I think that that is a
good way to start, but now it is time to move.
n/a 3/3* 1/1 3/4
There was limited human capital at York EL to bridge technology with Balanced Literacy
Teacher 13: I already have to sit and come up with all this other stuff on my own that they are not
giving us this year so there is technology again. Cindy [organized a PD with] ideas of things you
could do with PicCollage. At that moment I was working on science and soils…I walked away with
something to do that day…[but] it is all so general like, “[You can use it with this] or whatever it is
that you are working on with your kids”.
n/a 3/3* 1/1 3/4
Note: This table provides a sample quote for each finding and indicates the number of central office administrator, DLC, principal, and teacher interviews that
provide evidence in support of this finding. For findings pertaining to district context, I count all interviews that I conducted with central office administrators,
CCs, DLCs, and in schools. For findings pertaining to each coach, I count all interviews that I conducted these coaches and the principals and teachers at both of
their assigned schools. For findings pertaining to each school, I count my interviews each DLC and the principal and teachers at that specific school. *For Brand
and York EL, I count a total of three DLCs at these schools, including Cindy, the DLC at this school from the previous school year, and the current Balanced
Literacy coach at this school.
BUILDING NETWORKS FOR CHANGE
247
Table D.2. Supporting evidence for why Sarah is more collaborative than Cindy in coaching teachers on technology-enabled
instruction.
Supporting Evidence From…
Finding
Central
Admin
DLC Principal Teachers
Formal responsibilities of DLCs focused on supporting DFs first and foremost, followed next by other
teachers
Administrator: Our digital learning coaches provide not only one-on-one training to their fellows
but are also available to do trainings after school.
2/3 4/4 4/4 8/15
Sarah engaged in collaborative coaching strategies that allowed her to reach out to non-fellow teachers
Sarah: As a coach, I try to work smarter not harder, and so my teachers that are on the same grade-
level, I usually try to do the same kind of thing with them. Since it was all their first PBL, which
can be very daunting, I thought, "Let's put you guys together and then you can talk and
collaborate," and so they collaborated through Google Drive.
n/a 1/1 2/2 4/9
Sarah had open-minded perceptions of teachers that took into account all of their strengths and weaknesses,
motivating her to tailor instructional support according to teachers needs
Sarah: She is a very strong teacher, and so even though she wasn’t very tech savvy, she was very
content savvy and really good with classroom management and everything like that.
n/a 1/1 n/a n/a
Sarah’s prior experience in leading teacher PD and familiarity with her school sites allowed her to be more
accessible and collaborative in her coaching
Principal (San Rafael EL): Sarah has actually only taught for three years before we pulled her out
but she was incredible. I was working with her on other PD projects. I knew she was incredible,
period.
n/a 1/1 2/2 9/9
Sarah benefited from an inclusive school culture at San Rafael that encouraged teacher collaboration on
instruction
Sarah: [The principal and I] wanted to get something where kids could collaborate and I wanted to
show a tool that they could use instantly, that did not need a lot of teacher prep, a lot of time,
because I knew that if it did, it would scare them away, since they are not techy, or they do not
think they are. I also wanted it to be something that was easily accessible…not something that we
would have to collect all their iPads [for uploading a new app]. It was something we could do
quickly.
n/a 1/1 1/1 5/5
BUILDING NETWORKS FOR CHANGE
248
Supporting Evidence From…
Finding
Central
Admin
DLC Principal Teachers
Sarah benefited from a less inclusive (more isolating) school culture at Glendale EL, limiting her outreach
to non-fellow teachers at the school
Teacher 8: It is just a lot different because my principal last year was very dynamic and very
approachable. If there was something I would be like, "Hey, could I have 10 minutes, 20 minutes
in our meeting to share this?" And he would be like, "Of course, please share it." I am sure she
would do the same but things were a little more organized. This year we get an email on Monday
saying what we are doing on Wednesday. It is harder to plan, so I have not really had the
opportunities.
n/a 1/1 0/1 4/4
Cindy engaged in non-collaborative coaching strategies that restricted her accessibility to non-fellow
teachers
Cindy: Ultimately, my responsibility is toward my fellows, but I am working with the whole staff,
really, so if anybody else comes to me and asks questions or needs help, I can meet with them
n/a 1/1 2/2 4/6
Cindy had a narrow-minded view of teachers and their willingness to change, motivating her to support
other DFs who were like her and shared her vision for instructional change
Cindy: I do not want them to just do technology for the sake of doing technology. It should not be
just doing a project just because you have to use technology so many times during the year. It
should be because it is more engaging to the kids. The kids are learning more from it ... They are
gaining something from using technology, not just so that they can have something to hang up on
the wall that says, "Check, I did technology."
n/a 1/1 n/a n/a
Cindy had no experience leading teacher PD and not worked at any of her assigned school sites in the prior
year, challenging her to be more accessible to and collaborative with teachers
At Brand EL, I feel like I'm a little bit more accepted, honestly, and more respected. I'm starting to
feel that way at York EL a little bit more, but I feel like I came in, I had big shoes to fill because
[their DLC from last year] was awesome and so I feel like they really had a connection with her,
and when I came in, they were like, "What does she know? She was a teacher last year."
n/a 1/1 2/2 2/6
Cindy encountered a less inclusive (more isolating) school culture at Brand EL that made it difficult for her
and her DFs to reach out to non-fellow teachers
Cindy: [Brand and York EL] are smaller schools, [with] only two people per grade-level, and so I
think that sometimes when you have a smaller group, you have less ideas. It is easier to do your
own thing if it is a smaller [team], because there is only two people. It is like, "Okay, I do not like
your idea, I am going to do my idea."
n/a 2/3* 1/1 1/2
BUILDING NETWORKS FOR CHANGE
249
Supporting Evidence From…
Finding
Central
Admin
DLC Principal Teachers
Cindy encountered a more cohesive school culture at York EL that was strongly opposed to the central
office’s goals for student-centric instruction, positioning her as an outsider to the school community who
was unable to engage teachers much less facilitate collaborative routines for teacher learning
Balanced Literacy Coach: These teachers take their student achievement and they wear it like a
badge of honor which we have got to move [away] from. That is part of my shift in getting them to
see. Let the kids wear that badge of honor and let’s step back…building that relationship there, I
will tell you, has taken, I would say, until March, April. It has been very hard for them to be
willing to trust me enough to make a shift in their practice and be willing to change. Failing is not
an option for them and when you are shifting your practices, it is going to get messy.
n/a 3/3* 1/1 4/4
Note. This table provides a sample quote for each finding and indicates the number of central office administrator, DLC, principal, and teacher interviews that
provide evidence in support of this finding. For findings pertaining to district context, I count all interviews that I conducted with central office administrators,
CCs, DLCs, and in schools. For findings pertaining to each coach, I count all interviews that I conducted these coaches and the principals and teachers at both of
their assigned schools. For findings pertaining to each school, I count my interviews each DLC and the principal and teachers at that specific school. *For Brand
and York EL, I count a total of three DLCs at these schools, including Cindy, the DLC at this school from the previous school year, and the current Balanced
Literacy coach at this school.
Abstract (if available)
Abstract
In the past decade, school districts have faced increasing pressure from federal and state policies to use technology to innovate teaching and learning. In response to these pressures, districts are spending billions of dollars to outfit their schools with new technologies (e.g., laptops, iPads, broadband internet, new software), and to hire expert teachers to work as education technology coaches or “ed-tech coaches” to guide teachers in using technology creatively and effectively in the classroom. While ed-tech coaching programs are gaining prominence among districts, these programs make strong assumptions about the capacity of ed-tech coaches to shift teacher practice that are grounded in little evidence. In this dissertation, I focus on the instructional leadership responsibilities of ed-tech coaches, especially with regards to understanding how these coaches circulate information on instructional change among central office staff, school leaders, and teachers to facilitate district-wide improvements in instruction. I explore the instructional leadership of these coaches in a mid-sized urban district that recently completed a one-to-one rollout of laptops and iPads to support instruction of the Common Core State Standards. In this reform setting, I use social network theory and methods, and along with complementary qualitative data collected from four case schools, to describe how ed-tech coaches broker information on instructional change between and among central office administrators and schools and the organizational conditions that shape their brokering practices.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Teacher self-efficacy and instructional coaching in California public K-12 schools: effective instructional coaching programs across elementary, middle, and high schools and the impact on teacher...
PDF
Instructional coaching, educational technology, and teacher self-efficacy: a case study of instructional coaching programs in a California public K-12 school district
PDF
Instructional coaching and educational technology in California public K-12 school districts: instructional coaching programs across elementary, middle, and high schools with educational technolo...
PDF
Instructional coaching in California public K-12 school districts: instructional coaching programs in elementary, middle, and high schools and the impact on teacher self-efficacy with educational...
PDF
From meaning‐making to expansive learning: how contradictions shape teachers' implementation of technology‐based personalized learning
PDF
A case study of technology-embedded instruction: a student-centered approach to enhance teaching and learning in a K-12 school
PDF
The role of leadership in using data to inform instruction: a case study
PDF
Preparing teachers for social emotional learning driven instruction and practice
PDF
The effects of coaching on building and sustaining effective leadership practice of an urban school administrator
PDF
Student mobility in policy and poverty context: two essays from Washington
PDF
High school reform to improve mathematics achievement
PDF
Tough conversations and missed opportunities: implementing district policies for racial equity
PDF
The relationship between ethnicity, self-efficacy, and beliefs about diversity to instructional and transformational leadership practices of urban school principals
PDF
Instructional technology integration in a parochial school district: an evaluation study
PDF
Instructional coaching: disrupting traditional math practices through inquiry and action research
PDF
Leadership and implementation of 1:1 technology: considering teacher self-efficacy in the implementation process
PDF
Overcoming the cultural teaching gap: an evaluative study of urban teachers’ implementation of culturally relevant instruction
PDF
Secondary math reform and the role of policy, practice, and instructional leadership on math achievement
PDF
Elements of a 1:1 computer laptop program in a Los Angeles County high school and implications for education leaders
PDF
Transformational technology practices in K-12 schools: a case study
Asset Metadata
Creator
Hashim, Ayesha Khalid
(author)
Core Title
Building networks for change: how ed-tech coaches broker information to lead instructional reform
School
Rossier School of Education
Degree
Doctor of Philosophy
Degree Program
Urban Education Policy
Publication Date
06/01/2017
Defense Date
04/12/2017
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
education reform,education technology,instructional coaching,instructional leadership,OAI-PMH Harvest
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Strunk, Katharine O. (
committee chair
), Marsh, Julie A. (
committee member
), Valente, Thomas (
committee member
)
Creator Email
ahashim.usc@gmail.com,ahashim@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c40-375050
Unique identifier
UC11255862
Identifier
etd-HashimAyes-5351.pdf (filename),usctheses-c40-375050 (legacy record id)
Legacy Identifier
etd-HashimAyes-5351.pdf
Dmrecord
375050
Document Type
Dissertation
Rights
Hashim, Ayesha Khalid
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
education reform
education technology
instructional coaching
instructional leadership