University of Southern California Dissertations and Theses

A comparative study of self-efficacy, outcome expectancy, and retention of beginning urban science teachers

(USC Thesis Other)

A comparative study of self-efficacy, outcome expectancy, and retention of beginning urban science teachers

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

A COMPARATIVE STUDY OF SELF-EFFICACY, OUTCOME EXPECTANCY,
AND RETENTION OF BEGINNING URBAN SCIENCE TEACHERS

by

Nina Klein

A Dissertation Presented to the
FACULTY OF THE ROSSIER SCHOOL OF EDUCATION
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF EDUCATION

May 2009

Copyright 2009 Nina Klein

ii

Dedication

This dissertation is dedicated to my family. Thank you to my dad Werner, my
mom Doris and my sister Julia: Without your unwavering support this journey
would not have been possible.

iii

Acknowledgments
I want to thank my friends who believed in me and my ability to persist through
this arduous and often seemingly impossible process, first and foremost I want to
thank you, George: Thank you for all your support and advice, and your patience
throughout this time.

Finally, thanks to everyone at USC who I met on this journey: Dr. Margo
Pensavalle, for chairing my committee and for your advice, Dr. Gisele Ragusa for
keeping me on my toes, Dr. Fred Freking, who made the journey across town to
be a part of this, and finally thanks to my friend Kathryn, who I would not have
met otherwise.

iv

Table of Contents

Dedication ii

Acknowledgments iii

List of Tables vii

List of Figures viii

Abstract ix

Chapter I: Overview of the Study 1
Introduction 1
Statement of the Problem 10
Purpose of the Study 11
Importance of the Study 13
Limitations, Delimitations, and Assumptions 15
Theoretical Framework 16
Definition of Terms 17

Chapter II: Literature Review 22
Introduction 22
Teacher Shortage: Background 24
Attrition and Retention 26
Teacher Quality and Student Achievement 30
Recruitment of Well-Qualified Candidates 31
Recruitment of Alternatively Certified Teachers 33
Teacher Preparation: Traditional and Alternative
Credentialing 34
Background
Alternative Credentialing and Student 38
Achievement
The Achievement Gap: Background 45
The Achievement Gap in California 52
Self-Efficacy and Outcome Expectancy of Teachers: Background 56
Teacher Efficacy and Student 61
Achievement
Literature-Based Rationale for Methodology 63
Surveys 63
Interviews 64
Retention Data 65

v
Summary and Conclusion 66

Chapter III: Methodology 69
Introduction 69
Sample and Population 72
Instrumentation 74
Survey: Modified Version of the Science Teachers’ Efficacy 75
Beliefs Instrument (STEBI)
Interviews 80
Retention Data 82
Data Collection and Timeframe 83
Data Analysis 83
Reliability and Validity: Confidence of Truthfulness of the Study 85
Summary 86

Chapter IV: Data Analysis and Interpretation of Findings 88
Introduction 88
Data Overview 90
Means (Averages) of Survey 91
Kurtosis and Skewness 93
Data Collected for RQ 1a 96
Findings from the Surveys 96
Findings from the Interviews 106
Summary and Interpretation of Findings RQ 1 a 117
Data Collected for RQ 1b 122
Findings from the Surveys 122
Findings from the Interviews 132
Summary and Interpretation of Findings RQ 1 b 139
Data Collected for RQ 2 143
Findings from the Self-Reported Retention (Survey) 143
Findings from the Reported Retention for Each Program 146
Findings from National and California State Retention Data 149
Summary and Interpretation of Findings RQ 2 152

Chapter V: Summary, Conclusions, and Implications 155
Summary 155
Recommendations and Implications 159
Suggested Areas for Further Research 163
Conclusion 165

References 167

vi

Appendices 181
Appendix A: Teacher Efficacy Survey Data 182
Appendix B: Outcome Expectancy Survey Data 184
Appendix C: Retention Data 186
Appendix D: Instrumentation of Study 187
Appendix E: IRB Exemption Traditional (U) Program 191

vii
List of Tables

Table 1: Two Perspectives on the Causes and Consequences 27
of School Staffing Problems

Table 2: Methodology Chart 71

Table 3: Group Mean Descriptives from Surveys 90

Table 4: Teacher Efficacy Survey Statistics Overview 105

Table 5: Outcome Expectancy Survey Statistics Overview 132

Table 6: Percentage of Teachers Choosing Each Self-Reported Option 146

Table 7: Program Reported Retention of Science Teachers 148

Table 8: Retention Data of First-Year Teachers in Comparison 151

viii

List of Figures

Figure 1: Relationship of Problem, Research Questions (RQ), and..............74
Methods of Data Collection (Data Sources)

Figure 2: Rankings by scale group means.....................................................92

Figure 3: Kurtosis and Skewness for Efficacy (TE) and Outcome..................95
Expectancy (OE)

Figure 4: Percentage of Teachers Choosing Each Self-Reported Option.....146

Figure 5: Program Reported Retention of Science Teachers Entering..........149
their Second Year in 2008/2009, in Percent

Figure 6: Retention Data of First-Year Teachers in Comparison, in..............151
Percent

ix
Abstract

The purpose of the multi-tiered study presented is to compare the effect of
credentialing route on the self-efficacy, outcome expectancy, and retention of
beginning urban science teachers serving students in a large urban school
district in Southern California. Candidates from one traditional, university-based
teacher education program and from two alternative programs, the Teach for
America and District Intern Programs, were surveyed and interviewed during the
second semester of their first year of teaching. To determine the potential of a
difference in self-efficacy and outcome expectancy, the study gave teachers a
modified version of the Science Teachers’ Efficacy Belief Instrument (STEBI),
developed and validated by Riggs and Enochs (1989). Two representative
candidates from each program were then interviewed in order to probe for
deeper understanding of possible sources of their efficacy and outcome
expectancy. The final part of the study is an evaluation of retention data from the
three programs, each to triangulate this information with data collected from the
surveys, and comparing these retention rates with published data. The study
provides data on unresearched questions about traditionally and alternatively
credentialed science teachers in urban settings in California.

1
Chapter I. Overview of the Study

Although for over a century our nation has advanced the ideal that a high-
quality and excellent public education is the birthright of all children, our
schools cannot fulfill this ambitious and noble purpose unless all of us [...]
commit ourselves to sustaining education as a public trust and a promise
to future generations. (Nieto, 2005)

Introduction
Public education was established with the idea of leveling the playing field
for all children so the brightest can emerge as leaders in society. Unfortunately,
the public education system has fallen short of its promise to offer equal
education to all children (Noguera, 2003). Historically, a lack of equal educational
access was mainly based on race and ethnicity; but through later events,
socioeconomic status became an increasingly significant factor (Shaw, 1944).
Today, California’s public schools are not instruments of equalization. Even
schools within the boundaries of one city’s school district vary enough to be
unable to equally fulfill this promise (Oakes, 2002). Having been through the
cycle of segregation, desegregation, and resegregation, minority children are still
experiencing an inferior educational setting at a higher rate than other groups
(Stephan, 1986; Valencia, 1991). Access to qualified teachers is one factor
contributing to the quality of education for students according to this research.
Therefore, studies of what makes teachers qualified are a necessity.

2
Teaching in K-12 public schools is still a profession largely filled by young
female White adults from the middle and upper class (Amobi, 2007; Garcia,
1991): As the students in K-12 public schools are increasingly from outside of
this ethnic group, the resulting ethnic discrepancy can cause tension (Katz, 1999;
Thompson, 2007). The incongruence between the backgrounds of teachers and
their students creates an array of problems, starting with a shortage of qualified
teachers in areas that are the least like White middle-class neighborhoods.
Minority students from urban areas with a high poverty rate have limited access
to highly qualified teachers (Darling-Hammond, 2004). An increased focus on
multicultural education (Bennett, 2001) and critical race theory in teacher
education (Landon Beyer, 2001) has been the result of this demographical trend,
as many teachers report they do not feel prepared to teach minority populations.
Therefore, studies of programs preparing teachers to do well in diverse urban
classrooms are important to further an understanding of this educational context.
Teachers seeking employment in urban settings often face conditions that
set them up to fail, such as overcrowded schools, large class sizes, a lack of
resources, and a high number of newly hired and inexperienced teachers paired
with a defeated veteran staff (Ingersoll, 1991). As a result, in the last decade,
there has been an observable increase in teacher migration rates out of these
challenging school settings and into other schools, or attrition out of the
profession altogether (Marvel, Lyter, Peltola, Strizek, Morton, & Rowland, 2007;

3
Murnane, Singer, Willett, Kemple, & Olsen, 1991; Schlechty & Vance, 1981;
Shen 1997). That the most academically proficient teachers are those leaving
(Schlechty & Vance, 1983; Smith, 1993) is a very alarming trend for public
education in urban areas.
The high teacher migration and attrition rate is most pronounced and
persistent in low-income schools, which are traditionally hard to staff (Darling-
Hammond, 2004). This problem is caused not only by a higher retirement rate of
teachers, but also by teachers leaving after a short time in the profession.This
phenomenon has been labeled as the revolving door (Ingersoll, 2001) and
warrants further study for several reasons.
High teacher attrition endangers the continuity of educational programs. If
a school experiences high teacher turnover for several consecutive years, it is
difficult to establish professional learning communities or a cohesive professional
development program (Ingersoll, 2001). Studies of new teachers document that
teacher learning and effectiveness increases during the first four years of
teaching (Rivkin, Hanushek, & Kain, 2005), so if new teachers continue to leave
during that time, school sites will not benefit from this learning.
The trend of leaving the teaching profession is partially due to beginning
teachers struggling to navigate the demands of the profession without adequate
support in settings alien to their own educational experiences (Oakes, 2002).
Further, lack of training on the procedures and mechanisms of the first days of

4
school, heavy paperwork demand, and practical questions overwhelm new
teachers at the very beginning getting them off to a bad start (Wong & Wong,
1998). Even if they experience student teaching in a traditional program, few
beginning teachers have seen the beginning of a school year from an adult’s
perspective until they are the teachers themselves. Only after a period of
adjustment do these new teachers develop the ability to utilize strategies learned
in their credentialing programs, and strategies employed by successful and
experienced teachers that work with the same student population, referred to as
best practices (Strong, 2006).
Through research, a fairly consistent portrait of the teachers leaving the
profession early has been created: Teachers of mathematics, science, and
special education leave at higher rates, and so do men and younger teachers
(Ingersoll, 1991). The largest part of the teacher turnover is primarily due to job
dissatisfaction and the desire to find better career opportunities (Quartz, 2003).
Efforts to recruit and hire large numbers of teachers increases district
expenditures, especially in urban districts that traditionally have less access to
additional resources. In consideration of these reasons according to Darling-
Hammond and Sykes (2003), teacher retention is a greater concern than teacher
recruitment in addressing the shortage of teachers, especially in math and
science and in hard-to-staff schools. Studies are needed to further understand

5
the needs of these groups of teachers and of factors that could lead to higher
retention.
Teachers who feel unsuccessful in their assignments are more likely to
leave the profession. According to Wilson, Floden, and Ferrini-Mundy (2001), a
major gap exists in knowledge related to the preparation of teachers for urban
schools. One effort to increase support is through the mentoring of new teachers
during the first years in the profession. The high teacher turnover may be
attributed to the complex array of challenges and difficulties beginning teachers
face when they first enter the classroom, as they are expected to fulfill basically
the same duties as a veteran teacher (Lipton & Wellman, 2003). This challenge
often leads them to doubt their own efficacy and initial reasons for entering the
teaching profession, taking them into a form of survival mode (Bartell, 2005).
Studies of beginning teachers (Boccia, 1991; Corley, 1998; Grant & Zeichner,
1981; Odell, Loughlin, & Ferraro, 1986; Veenman, 1984) reveal several different
problems they encounter, such as classroom management, obtaining resources,
and communicating with colleagues. Mentoring of beginning teachers has been
shown to reduce the effect of the revolving door, keeping new teachers in the
classroom, and helping them through some of their first-year challenges and
difficulties described above (Bey & Holmes 1992; Manley, Siudzinski, & Varah,
1989; Smith 1993). Mentoring is not only associated with higher retention, but
also with higher teacher efficacy (Anthony & Kritsonis, 2007). In turn, higher

6
teacher self-efficacy is linked to higher student achievement (Goddard, Hoy, &
Woolfolk-Hoy, 2000; Henson, Kogan, & Vacha-Haase, 2001).
If schools are not successful in retaining their teachers, they need to
recruit new teachers, which is challenging and costly. Thus, a recent
accommodation has been the issuance of emergency and alternative credentials
in attempts to fill these hard to fill positions.
Until the early 1990s, most people who wanted to teach in the public
schools needed to complete an undergraduate program of teacher
preparation. By 1993, 40 states had created postbaccalaureate alternate
routes into teaching, as a way of reducing shortages in critical areas such
as mathematics and science […] (Wilson et al., 2001)

According to existing research literature, these alternatively credentialed
teachers are often outperformed by their traditionally prepared counterparts.
Alternatively credentialed teachers contribute to a lower quality teaching staff,
comprised of individuals with less content knowledge or less pedagogical
knowledge, and they have been linked to lower student achievement (Darling-
Hammond, 2000; Ferguson & Womack, 1993; Goldhaber & Brewer, 2000). As
this phenomenon continues, studies of the effects of underprepared teachers are
important to educational research in order to facilitate a full understanding of the
effects of alternative credentialing. This is true especially in teacher shortage
areas such as science.
These concerns are more pressing in a shortage area of qualified
teachers, such as science. Focusing studies on science teachers in particular,

7
three areas of consideration are critical if current trends are to be reversed. First,
the recruitment of science teacher candidates who participated in preparation
programs to teach in an urban school is a focal point. These teachers are able to
work successfully with traditionally low-performing students, and this is the first
step toward ensuring more equitable access for all students. Second, an
improvement in the retention rate of these successful teachers in urban and low-
income educational settings is needed. To achieve this, the prevention of the
current high attrition rates is necessary, which can be effected through studies of
known contributing factors to attrition and the elimination of these factors
wherever possible. Finally, professional development training to assist the
teaching staff at urban sites to develop successful teaching practices that can
predict high student outcomes is required. According to current research, such as
Darling-Hammond (2000), Haycock (1998), and Rivkin et al. (2005), the greatest
factor influencing student success is the teacher and teacher quality. Therefore,
the focus of educational research related to teacher preparation needs to be on
the development of quality teachers, on their retention, and on ensuring equitable
access to these teachers for all students.
While research has linked qualified teachers to a higher achievement rate
(Goldhaber, 2003; Darling-Hammond, 2006), questions remain about how to
make sure that teachers with the potential to increase achievement end up with
the students that need them the most. This concern arises out of the necessity to

8
satisfy the declared purpose of public education to serve as the great equalizer
for all children and to close the achievement gap that separates and segregates
American students along socioeconomic and racial lines.
Several mandates (e.g., parts of the No Child Left Behind Act, 2001) have
tried to address the issue of underachievement on standardized measures
through legislation necessitating certain teacher qualifications in order to address
the gap. In an attempt to decrease the achievement gap by controlling for
teacher quality, legislation under the NCLB act now limits the influx of less
qualified teachers into the classroom. However, NCLB legislation defines “highly
qualified teachers” as based on subject matter competence, and not on
pedagogical preparation. Thus, legislation still allows teachers to enter the
profession who, though qualified in their subject matter are merely enrolled in a
teacher preparation program and who do not possess the expert knowledge of
pedagogical strategies needed to yield high learning outcomes in urban settings.
Schools with difficulties filling their vacancies with traditionally prepared
candidates have a higher percentage of alternatively credentialed teachers,
especially in mathematics, science, and special education (Darling-Hammond,
2004). Apart from other problems these schools are already facing, such as
multi-track schedules and overcrowding (Oakes, 2002), additional hardships are
placed upon them by hiring alternatively credentialed teachers. Alternatively
credentialed teachers are described as underprepared because they have not

9
been required to prove competency in pedagogy and methodology, and have not
received instruction on how to successfully teach their subject matter to English
language learners (Darling-Hammond, 2002).
Literature has found certain teacher characteristics are related to student
achievement. Teacher self-efficacy and outcome expectations are such
characteristics, as self-efficacy has been connected to student achievement as
early as 1976 in a study conducted by the RAND Institute (Armor, Conry-
Osegura, Cox, King, McDonnell, Pascal, et al.). In this study, high teacher
efficacy was found related to higher student reading outcomes. According to
Bandura (1986, 1997), teacher self-efficacy denotes a teachers’ judgment about
his or her abilities to impact student learning. In the past 20 years, studies have
found links among student achievement, student efficacy, and teacher efficacy,
as well as the collaborative efficacy of a school (Goddard et al. 2000; Pajares,
1996; Ross, 1992, 1994, 1998; Tschannen-Moran & Woofolk-Hoy, 1998). In light
of the findings linking high teacher efficacy to student achievement, teacher
efficacy can be used as a validated contributing factor with a positive effect on
student achievement. Outcome expectancy has been examined in some of these
studies (Ross, Cousins and Gadalla, 1996). While teacher self-efficacy focuses
on the beliefs if the teacher personally can bring about student learning, outcome
expectancy is referred to as the belief that certain teacher actions will bring about
student achievement (without linking making statements about the teacher’s

10
personal ability to perform these actions). It is sometimes referred to as general
teaching efficacy (Woolfolk, Rosoff and Hoy, 1990) or as teacher locus of control
(Smylie, 1990).

Statement of the Problem
The problem this study will examine is a possible effect of alternative and
traditional credentialing on factors contributing to student achievement: teacher
self-efficacy, outcome expectancy, and retention. Current literature establishes
these three factors as bearing an influence on student achievement. The
achievement gap of urban student populations has to be addressed and
effectively resolved. Even as the gap currently exists, a greater number of
alternatively credentialed teachers are serving urban student populations.
Therefore, students in urban settings are potentially being underserved by less
qualified teachers. This study investigates the potential connections among these
variables: differences in levels of teacher self-efficacy, outcome expectancy, and
retention to the credentialing route of teachers. The reason for the focus on
beginning science teachers is that recruitment and retention of fully certified
science teachers, along with mathematics and special education teachers,
seems particularly elusive.

11
Purpose of the Study
This study draws a data-driven conclusion comparing different teacher
credentialing routes in the areas of teacher self-efficacy, outcome expectancy,
and retention. The study examines middle and high school teachers of science
according to credentialing route in the urban context of a large metropolitan
school district. The purpose of this study is to explore whether there are
differences in teacher self-efficacy, outcome expectancy, and retention among
traditionally and alternatively certified beginning science teachers. To examine
these differences, the following research questions have been developed:
(1) (a) How do differently prepared (traditional vs. alternative)
beginning, urban, secondary science teachers compare in their
self-efficacy?
(1) (b) How do differently prepared (traditional vs. alternative)
beginning, urban, secondary science teachers compare in their
outcome expectancy?
(2) After the first year of teaching, how do the retention rates of
traditional and alternative teacher populations compare?
These research questions frame data collection and analysis, as well as
the discussion of findings. In order to answer these research questions, the
following data was collected. Likert-scale self-efficacy and outcome expectancy
surveys, asking beginning teachers from two alternative and one traditional

12
program about their self-efficacy and outcome expectancy beliefs. The survey is
a modified version of the Science Teacher Efficacy Belief Instrument (STEBI),
developed and validated by Riggs and Enochs (1989). These surveys address
the first research question (RQ 1), the difference in self-efficacy and outcome
beliefs of differently prepared science teachers, by providing quantitative data
that was analyzed using descriptive statistics. In-depth interviews were held with
two new teachers from each program for a total of six beginning teachers. These
interviews further explore the teachers’ beliefs, generating qualitative data, and
thus allowing for triangulation. Participants in the interviews were chosen based
on their survey averages being representative of their program for self-efficacy
and outcome expectancy. Retention rates from the alternative and traditional
programs after the first year of teaching were used to determine teacher
retention. This data was triangulated with a self-statement on the survey cover
sheet regarding retention, and this data can be used to answer the second
research question (RQ 2) through comparison with the other programs, data from
the state of California, and national data.
The hypotheses for this study are:
(1) A measurable difference will be observed in teacher self-efficacy and
outcome expectancy among beginning alternatively and traditionally
prepared teachers, with traditionally prepared teachers showing
greater self-efficacy and higher outcome expectancy. This difference

13
will be evident in the quantitative surveys as well as in the qualitative
interviews.
(2) The retention of traditionally prepared teachers will be greater in
comparison with alternatively prepared teachers. More traditionally
prepared teachers will express that they are going to stay in urban
teaching, and through the program’s self-reporting after the first year
the traditional program will show higher retention.

By looking at quantitative and qualitative data generated, this study
analyzes differences between alternatively and traditionally credentialed
teachers. A triangulation of the quantitative and qualitative data increases the
validity and reliability of the findings of the study (Campbell & Fiske, 1959; Jick,
1979; McGrath, 1982). The mixed-methods approach (Johnson, 2007), which
collects quantitative and qualitative data, provides greater validity to the study as
well (Johnson, 1997, on validity and mixed-method approaches).

Importance of the Study
This study is of importance to professionals in the field of teacher
education and teacher credentialing. The analyzed results contribute important
data findings related to self-efficacy and the retention of beginning teachers.
Teacher educators in traditional and alternative programs are able to look at

14
these findings to increase or decrease the length of their programs and to focus
on enhancing efficacy beliefs of their candidates and on increasing retention
rates. A sample group of new science teachers showing a much greater sense of
efficacy, outcome expectancy, or retention can be studied further to provide
insights into the reasons for the observed success. The triangulation of
quantitative and qualitative data gives teacher educators a picture of their
program candidates as they develop as professionals.
The data generated can inform legislators to make important research-
based decisions about alternative credentialing, and its impact on teacher quality
and retention, and thus on student outcomes. In essence, the study provides
findings that enable policy makers to address current issues in public education
through a variable that they can control: teacher preparation and credentialing.
For accountability and evaluative purposes, the study includes a
quantifiable difference of efficacy and retention data from one traditional program
and two alternative programs. District administrators are able to determine the
best recruitment strategies and school-site administrators are able to use the
findings to suggest faculty hiring and professional development strategies. The
findings can lead to new interventions to help certain beginning teachers develop
their efficacy and outcome expectancy, and increase their retention.

15
Limitations, Delimitations, and Assumptions
The study’s main limitation is that the scope of the study does not allow for
a control of the quality of any program’s implementation. Also, due to the study’s
voluntary structure, only one traditional program was included, as mailed surveys
to a second university program were not returned in sufficient numbers.
The study has several delimitations in its design. It is looking specifically at
the urban context, as the achievement gap is most pronounced here. Therefore,
the transferability of any findings to suburban or rural settings might be limited.
With the study including only selected programs and only their science teacher
candidates, the results may not be transferable to elementary teachers or to
teachers of other subject areas.
The last point leads to the assumptions made in this study, which are that
the two chosen alternative programs allow a broad scope and a good overview of
the pool of alternatively credentialed teachers, and that the chosen university
program provides the same general picture for traditionally prepared teachers.
Finally, due to the lack of a quality control in the programs’ respective
implementations, it is assumed that each program is consistently implemented
and internal variations are minimal.

16
Theoretical Framework
The proposed study is based on the theoretical framework of the social
cognitive theory of learning, as efficacy is a notion firmly grounded in this theory.
Social cognitive theory states that people can learn through observation, that
learning is an internal process that may or may not be reflected in behavior, that
behavior is goal-directed and self-regulated, and that learning and behavior are
influenced indirectly through reinforcement (Ormrod, 2006). In teacher education,
student teaching is one of the most powerful factors influencing the learning of
new teachers, and student teaching is largely based on the concept of modeling.
Modeling is based on the social cognitive notion of learning through observation.
Self-efficacy is another concept out of the social cognitive theory of
learning. It is the judgment a person makes about his or her ability to perform a
task or to reach a goal. As such, self-efficacy is influenced through previous
successes and failures (Bandura, 1986), messages received from others (Zeldin
& Pajares, 2000), successes and failures of others (Zeldin & Pajares, 2000), or
successes and failures of a group of people (collective self-efficacy, Bandura,
1997). When looking at teacher preparation, these sources of self-efficacy are
important, as high efficacy has been linked to positive teaching behaviors and
positive student outcomes (Tschannen-Moran & Woolfolk-Hoy, 2001). Mentoring
by a master teacher or upon first entering the classroom can provide the second
source of development of efficacy: messages from others. If new teachers do not

17
receive these positive influences to develop a high sense of self-efficacy,
socialization might occur, which in urban, high-poverty and low-performing
schools is often a negative experience, as the veteran teachers may share their
sense of low self-efficacy.

Definitions of Terms
Achievement Gap: “The achievement gap is measured by national
average test-score differences between racial and ethnic groups based on the
National Assessment of Educational Progress (NAEP) and SAT results” (Lee,
2002). For the purpose of this study, the term achievement gap will be used to
describe the discrepancy of performance on these measures between groups of
students.
Alternatively Prepared Teachers, also “undercertified,”
“underqualified”: “Teachers on emergency, temporary and provisional
certificates” (Laczko-Kerr & Berliner, 2002). As California stopped the issuing of
emergency permits, except for CLAD waiver purposes, the term will be used for
the purpose of this study to describe teachers entering the profession through
intern credentials. The two alternative programs examined are Teach for America
(TFA) and the District Intern Program (DI), see below.

18
Attrition: “Teachers exiting the teaching profession” (Ingersoll,
2001). For the purpose of this study, attrition due to reasons other than
retirement will be examined.
Certified Teachers: “Teachers from accredited university programs
that have met all state requirements for receiving the regular initial certificate to
teach” (Laczko-Kerr & Berliner, 2002). For the purpose of this study, “certified”
and “traditionally prepared” will be used synonymously, although in some
literature the term “traditionally prepared” is reserved for teachers who have
completed a 4-year undergraduate program in education. Both terms are
different from “highly-qualified,” see below.
District Intern Program: “[California] has adopted an internship
model as the type of alternative pathway that the state funds and accredits.
Internship programs may be run exclusively by districts (District Intern programs)
or as partnerships between universities and districts” (Chin & Young, 2007). For
the purpose of this study, one district intern program’s teachers will be
participating as a population sample. The studied school district is a large
metropolitan district in California.
Highly-qualified Teacher: [NCLB] set the important goal that all
students be taught by a “highly qualified teacher” (HQT) who holds at least a
bachelor’s degree, has obtained full State certification, and has demonstrated
knowledge in the core academic subjects he or she teaches” (Letter from

19
Margaret Spellings, US Department of Education, October 21, 2005). For the
purpose of this study, it is important to distinguish “highly qualified” and
“traditionally prepared” teachers, as alternative candidates have fulfilled the
highly qualified challenge by obtaining an intern credential, but have not
undergone a teacher preparation program that includes student teaching prior to
entering the classroom.
Leavers: Former teachers that have left the teaching profession
(US Department of Education, Marvel et al., 2007).
Movers: Teachers that changed school sites but stayed in
classroom teaching (US Department of Education, Marvel et al., 2007). Also
included for the purpose of this study in the group of “movers” are teachers that
stayed at a school site but left the classroom to become coaches, coordinators,
or administrators.
NCLB: The No Child Left Behind Act of 2001, requiring states,
districts, and schools to increase performance so that all students achieve
proficiency in state reading/ language arts and mathematics assessments by
2014 (Public Law 107-110).
Outcome Expectancy: people enact certain behaviors to produce
desirable outcomes; this is called outcome expectancy (Bandura, 1977).
Behavior is based on both outcome expectancy and self-efficacy, according to
Bandura (1977).

20
Retention: Opposite of “Attrition,” see above. Teachers staying in
classroom teaching, leading to a more experienced teaching force.
Self-efficacy: “Expectations of personal mastery, affecting initiation
and persistence of coping behavior.” (Bandura, 1977). For the purpose of this
study, self-efficacy is used to describe teachers’ expectations of personal
mastery of teaching behaviors and confidence in their teaching ability. Behavior
is based on both outcome expectancy and self-efficacy, according to Bandura
(1977).
Stayers: Teachers that stay at the same school site in a teaching
position (US Department of Education, Marvel et al., 2007).
Teach for America Program: A non-profit organization, operating
a corps of members that teach in low-income public schools, which are most
affected by the achievement gap. They operate under the following “theory of
change”: “Teach for America’s mission is to enlist our most promising future
leaders in the movement to eliminate educational inequality. We accomplish this
by building a diverse, highly selective national corps of outstanding college
graduates - of all academic majors and career interests - who commit two years
to teach in urban and rural public schools in our nation’s lowest-income
communities and become lifelong leaders for expanding educational opportunity”
(Teach for America website at
http://www.teachforamerica.org/mission/theory_of_change.htm).

21
Traditionally Prepared Teachers: For the purpose of this study,
“certified” and “traditionally prepared” will be used synonymously, although in
some literature the term “traditionally prepared” is reserved for teachers having
completed a 4-year undergraduate program in education. Both terms are
different from “highly-qualified,” see below.
Undercertified or Underqualified Teachers: see above under
“Alternatively Prepared Teachers.”
Urban Schools: For the purpose of this study, urban denotes
public schools in a large metropolitan area in Southern California, serving a high
percentage of language, racial and ethnic minority students, and low-income
students.

22

Chapter II: Literature Review

When citizens and policy makers are unhappy with schools, they often
look for a place to assign blame. Labaree (1996) suggests that it is always
open season on teacher education, but Lucas (1997) observes that in the
late 1950s and early 1960s, attacks on educator preparation were
especially forceful. Although the policy context of 40 years ago and today
differ, criticism of teacher education then and now and the policy
community’s response are similar in many ways. (Earley, 2000)

Introduction
Throughout the history of K-12 education, teacher education and teacher
quality have been a focal point of discussions related to the academic
performance of students. Opinions on how to improve students’ academic
performance seem to have shifted toward an emphasis on teachers’ content
knowledge and away from pedagogical knowledge, as expressed in the NCLB
legislation. “Highly qualified” has been used to describe an aspiring teacher with
regard to subject (content) knowledge alone, and not pedagogical knowledge,
thus narrowing the definition of the knowledge teachers need (Berry, Hoke, &
Hirsch, 2004). Teachers of certain subject areas are in demand in urban, low-
SES school districts, and teacher retention seems to be elusive, especially in
these settings (Darling-Hammond and Sykes, 2003). Public urban school
settings, in particular, have seen a teacher shortage in mathematics and science,

23
Open teaching positions in these schools, which could not be filled with fully
credentialed candidates, are often filled by uncredentialed or undercredentialed
teachers from alternative programs (Laczko-Kerr & Berliner, 2002). Although the
state of California has stopped issuing “Emergency Credentials” (CL 533 P),
districts still issue them for substitute teachers, CLAD or BCLAD certification
(now required in districts serving high percentages of English Language
Learners, or ELL students). Additionally, the California Commission on Teacher
Credentialing (CCTC) still issues “Intern Credentials,” which allow non-
credentialed beginning teachers to be the teachers of record as long as they are
enrolled in a credentialing program. “Intern Teachers” very often end up in the
most challenging, urban, low-SES school sites (Lankford, Loeb, & Wycoff, 2002).
These schools are also marked by underperformance of their students on
mandated state and national tests, a phenomenon often referred to as the
achievement gap (Ladson-Billings, 2006). As this study explores the differences
in self-efficacy, outcome expectancy, and retention between science teachers
from alternative programs and traditional programs, the underlying question it
addresses is whether the lack of traditionally prepared teachers contributes to the
low student achievement found in the same schools.
For this literature review, the following key areas have been identified:
1. Teacher Shortage
2. Preparation of Teachers: Traditional and Alternative Credentialing

24
3. The Achievement Gap
4. Self-Efficacy of Teachers
Before offering a summary, a literature-based rationale for the
methodologies used in the study to collect data is presented.

Teacher Shortage
Background
Contemporary educational theory holds that one of the pivotal causes of
inadequate school performance is the inability of schools to adequately
staff classrooms with qualified teachers. This theory also holds that these
school staffing problems are primarily due to recent increases in teacher
retirements and student enrollments. [...] The results of the analysis
indicate that school staffing problems are not primarily due to teacher
shortages, in the technical sense of an insufficient supply of qualified
teachers. (Ingersoll, 2001)

In his analysis of factors behind the staffing problems of some schools,
Ingersoll (2001) found that certain schools are generating an “excess demand”
for new teachers due to their high turnover of staff. Using data from three cycles
of the Schools and Staffing Survey (SASS) and its supplement, the Teacher
Follow-up survey (TFS), the author analyzed the findings from an organizational
perspective, which meant that characteristics of the schools were examined,
unlike previous studies that focused solely on teacher characteristics. Data
showed, among other findings, a higher turnover rate of teachers in high-poverty

25
schools and in urban schools. Reasons found to drive this turnover are large
class sizes, intrusions on classroom time, lack of planning time, lack of
community support, and interference with teaching. Also, young teachers were
found more likely to leave teaching. This information is important, as Ingersoll
states that the dominant response to school staffing issues are recruitment
initiatives, for example the Teach for America program. New teachers recruited
through these programs are usually younger than teachers that went through a
traditional credentialing program, and are thus more likely to leave. The
comparison found in the Ingersoll study of a young teacher departing compared
to a middle-aged teacher is 171%, with the young teacher being much more
likely to leave. Although this study helps educational decision makers understand
that recruitment is only one piece in solving the complexities of staffing hard-to-
staff schools, it is much easier to devise recruitment incentives than to
fundamentally change the organizational nature of urban high-poverty schools
that leads to the phenomenon of the “revolving door.”
As several mandates (e.g., the NCLB legislation from 2001) try to address
the issue of underachievement on standardized measures, and research has
linked qualified teachers to a higher achievement rate (Goldhaber, 2003; Darling-
Hammond, 2006), questions remain about how to make sure that teachers with
the potential to increase achievement end up with the students that need them
the most. Although recruitment of candidates who underwent a preservice

26
training preparing them to teach in an urban, high-poverty school and to work
successfully with this group of students is not the only part that needs to be
addressed, it also should not be overlooked. Apart from incentives and special
programs that offer alternative credentialing, suggestions in the literature on
successful teacher preparation programs include insights into positive factors
found in some teacher education programs (Darling-Hammond, 2006). Another
aspect crucial to solving the staffing problem at some schools is the prevention of
the high attrition, which may be addressed through increased support and
elimination of the factors identified as driving this attrition, such as large class
sizes (Ingersoll, 2001).

Teacher Shortage: Attrition and Retention
The high teacher attrition rate in urban, low-income schools, which are
traditionally hard to staff, is a persistant problem. This problem has been labeled
the issue of the “revolving door” (Ingersoll, 2001) and warrants further study in
research for several reasons. High employee turnover is found to be both the
cause and the effect of ineffectiveness and low performance, according to
studies cited in Ingersoll (2001). In schools, high teacher attrition endangers the
continuity of an educational program (Darling-Hammond, 2006). High teacher
turnover for several consecutive years makes it difficult to establish professional
learning communities in a professional development program. Further, keeping

27
all teachers trained through in-services at comparable levels will prove a complex
challenge (Ingersoll, 2001). High teacher turnover, especially at the last minute—
as is the case in many urban schools— makes it very challenging for an
administrative team to plan and map out the school year. The recruitment of
large percentages of teaching staff also increases district expenditures,
according to Ingersoll’s (2001) findings.
The problem of high teacher attrition is shown as having effects seen from
two different perspectives in the table below.
Table 1: Two Perspectives on the Causes and Consequences of School
Staffing Problems (Ingersoll, 2001, p. 506)

28

The table shows that, as seen from the perspective of contemporary educational
theory, teacher shortages lead to staffing problems, and from an organizational
perspective teacher turnover will also lead to school staffing problems. From both
perspectives, the end result is a decrease in the quality of a school’s
performance.
Another problem is the supply of new and qualified teachers in certain
fields, for example mathematics, science, and special education (Darling-
Hammond and Sykes, 2003). Open positions at hard-to-staff school-sites
frequently get filled with teachers that are not fully credentialed. Placement of a
lower quality teaching staff comprised of individuals with less content knowledge
or less pedagogical knowledge into classrooms has been linked to lower student
achievement, as shown in several studies (e.g., Ferguson & Womack, 1993).
According to this body of literature, many hard-to-staff schools are already low-
performing, and thus the practice of hiring less qualified teachers does not serve
student populations
The first step in ensuring retention is to hire teachers that are prepared to
teach and be successful in the urban setting (Darling-Hammond, 2006; Oakes,
2002). Once these teachers are recruited, retention becomes easier, as
published data from an urban teacher education program demonstrates (Quartz,
2006). Through preparation and support, teachers are less likely to leave:

29

Studying the graduates of urban teacher education programs is one way
of gauging the success or social benefits of what Marilyn Cochran-Smith
calls the new multicultural teacher education- a culturally relevant,
community sensitive, social justice-based approach to preparing teachers
for work in urban schools. (Quartz, 2006)

Also worth considering is why so many teachers leave the classroom
while still having a genuine interest in education. Olsen and Anderson (2004)
introduce the distinction between “leavers” and “shifters,” defined as teacher
education graduates who stay in education but shift out of classroom teaching.
One deciding factor for this shift, their study states, is working conditions that
often encourage a lack of professionalism, a factor also found by Ingersoll
(2001). While well-prepared teachers are able to establish an environment
conducive to learning, they are often overwhelmed by the lack of support.
Factors leading to attrition, according to Ingersoll (2001), are large class sizes,
lack of support and cohesion among teachers, administrators, and community,
and interferences or intrusions into the classroom. In a follow-up publication,
Ingersoll and Smith (2003) found that 39% of teachers leave to pursue another
line of work, and 29% leave due to job dissatisfaction. When probed further,
these teachers stated poor salaries (79%), student discipline problems (35%),
and poor administrative support (26%) as the top three reasons for leaving.
Unless at least some of these factors are addressed, it will not be possible to
achieve equitable access to well-qualified teachers for all students, as it will be

30
impossible to recruit and retain them (Darling-Hammond & Sykes, 2003;
Ingersoll, 2001; Olsen & Anderson, 2004).

Teacher Shortage: Teacher Quality and Student Achievement
According to Ingersoll and Smith (2003), between 40 and 50% of
beginning teachers leave the profession within the first five years. After the first
five years, attrition rates decrease until teachers approach retirement age. More
experienced teachers in general have been shown to be more effective than new
teachers (Murnane & Philips, 1981; Klitgaard & Hall, 1974), although experience
alone is not a linear indicator of high effectiveness. According to Rosenholtz
(1986), the effect of experience levels off after about five years. Therefore, other
teacher quality indicators besides experience need to be considered.
The strongest factors influencing student achievement, according to
Darling-Hammond (2000), are that the teacher has a major in the field he or she
teaches and that he or she is fully credentialed. This finding supports Shulman’s
(1986) development of different types of knowledge in teaching (propositional,
case, and strategic knowledge) and his distinction of content knowledge,
pedagogical content knowledge, and curricular knowledge. If teacher quality is
only defined by content knowledge— as the NCLB definition of highly qualified
teachers suggests— according to Shulman and Darling-Hammond, important
pieces of teacher education are missing. Linda Darling-Hammond (2000)

31
presents a study analyzing data from state surveys of policies, case study
analyses, the Schools and Staffing Survey (SASS), and the National Assessment
of Educational Progress (NAEP) to find out how teacher qualifications and other
school input are related to student achievement. While she acknowledges the
strong relationship of student demographic characteristics on student outcomes,
Darling-Hammond concludes that findings related to teacher quality variables are
even more influential. This importance of teacher preparation and credentialing
leads into the following review of differently credentialed teachers.

Teacher Shortage: Recruitment of Well-Qualified Candidates
Certain aspects of teacher education programs prepare its graduates to
be successful in urban, high-poverty and low-performing settings, according to
Darling-Hammond (2006). According to the findings, preparing teachers to be
effective in achieving high student outcomes, to teach a diversity of students, and
to persist through the challenges of teaching as a professional career while
continuing to reflect on and improve their practice are some of the desired
outcomes of teacher preparation. Certain beliefs are at the core of these teacher
education programs: the necessity to integrate knowledge, or to “enable teachers
to learn about practice in practice” (p. 287), to connect the university experience
with fieldwork over an extended amount of time, and to retain highly-qualified
teachers through support systems. Further, any teacher preparation program

32
needs to ensure that its graduates gain not only the content knowledge, but also
the content pedagogical knowledge and the general pedagogical knowledge they
need to teach their students and increase learning outcomes (Shulman, 1986).
Therefore, the first step in solving the problem of equity in access to well-
qualified teachers for an urban, high-poverty school is to form a strong
partnership with certain university programs that prepare their candidates to
teach in such settings, and to recruit teachers from such programs (Darling-
Hammond, 2006).
Recruitment processes also need to address the reality that while there is
a growing diversity of students, the teaching staff remains mainly White and
female (Bennett, 2001). As such, some programs may explicitly emphasize the
goal of attracting more diverse candidates into the teaching profession and do so
by showing an “instructional initiative designed to increase the number of
students from underrepresented minorities […] who enter a teacher education
program” (Bennett, 2001, p. 21). This commitment will help the school succeed,
or at least minimize the incongruence of ethnicity and thus resolve the culture
gap between prospective teachers and their students. Schools should be
sensitive to the culture of their student body and surrounding communities, and
try to recruit a more diverse teaching staff. According to Ingersoll (2001), minority
teachers are also less likely to leave the profession prematurely, which would
help to remedy the high attrition out of urban high-poverty schools.

33

Teacher Shortage: Recruitment of Alternatively Certified Teachers
Many hard-to-staff schools are having trouble recruiting well-qualified
candidates for their open positions (Oakes, 2002). Therefore, alternative
credentialing provides a pool of candidates willing to teach in settings for which
traditionally prepared teachers may be hard to find. The literature on alternative
credentialing is polarized. Some research strongly advocates against the harmful
consequences of hiring alternatively qualified teachers (for example, studies
presented by Darling-Hammond and Berliner). Other research finds positive
outcomes from graduates of programs such as Teach for America (TFA) and
their generally strong sense of mission to reform and equalize education, while
achieving high student outcomes (for example, a study by Glazerman, 2005, on
TFA). In a study of the Los Angeles Unified School District Intern program, Trish
Stoddart (1990) summarizes her findings on comparing interns and traditionally
prepared teachers succinctly:

Developing alternative route programs which primarily serve to socialize
teacher candidates into prevailing school practice, while providing
teachers, will not help improve instruction for at-risk students. Universities
need to work with school districts to develop programs which recruit
teachers willing to teach in multicultural inner-city schools and provide
them with state-of-the-art professional preparation.

34
In order to address underperformance in urban public school systems, as
Stoddart clearly evinces, simply filling rooms with teachers will not end unequal
access to qualified teachers, nor will it improve instructional practices or learning
outcomes at these sites but rather may perpetuate the low achievement already
found at these schools.

Teacher Preparation
Traditional and Alternative Credentialing Background
Becoming a credentialed teacher includes several important steps, which
can vary widely according to the state in which a credential is sought. This
differential has led to an “erosion” of the value of degrees in teacher education
and a loss in status by schools granting these degrees (Murray, 2000). In
California, aspiring teachers need to demonstrate basic skills, as well as subject
matter competency for secondary candidates, or general competency for
elementary candidates and competencies on several exams, often before
entering a teacher education program. Upon completing a traditional university-
based teacher education program, candidates may then enroll in a graduate
school, which may (University of California, Berkeley) or may not (University of
California, Los Angeles) require the GRE, or enroll in a school of education (e.g.,
the California State Universities). All of these university-based programs offer the
required coursework as well as the clinical experience (student teaching), ending

35
in a state-mandated (SB 2042) performance assessment such as the
Performance Assessment for California Teachers (PACT).
Besides fulfilling coursework to earn a traditional credential before
entering the classroom, one of the main distinctions of traditional programs is
teaching experience, in which a student teacher is placed in the classroom of an
experienced teacher to observe and eventually to teach several classes. If done
well, student teaching is the point of “theory to practice” (Carlson, 1999), whereby
students apply their theoretical knowledge to a classroom experience without
many of the issues faced by beginning teachers (usually student teachers are
only responsible for a few periods, and they always have the cooperating teacher
in the room).
However, many teacher education programs lack relevance, according to
their students and K-12 educators, and are often perceived as too removed from
the reality of K-12 education, leading to a gap between theory and practice
(Brouwer & Korthagen, 2005). Student teaching is one part of the professional
education of a university-based teacher education program, requiring a close
relationship between the school of education and K-12 schools (Morey, Bezuk, &
Chiero, 1997) and, as such, bridging this gap. In an attempt to solve this
dilemma, professional development schools (PDS) were introduced, as a place
to change K-12 schools, to initiate preservice teachers, and to offer continuing
professional development (Lieberman & Miller, 1990). If student teaching

36
happens in isolation from the university, its impact is limited when compared to
PDS settings (Abdal-Haqq, 1998).
Besides the theoretical training of university coursework and everyday
routines and challenges of clinical experience, a major goal of preservice teacher
education is increasingly to develop reflective practice as a habit of mind in their
students (Webb, 1999). Especially in student teaching, a reflective practice
model, as introduced by Dewey (1933), enables preservice teachers to learn new
ideas, as well as to grow professionally after leaving the teacher education
program (Lee, 2005).
Apart from these traditional routes into the teaching profession, there are
alternative ways to enter after providing evidence of basic subject matter
competency, such as the Teach for America (TFA) or district intern programs
(DI). In order to understand the possibilities alternative credentialing may offer, it
is helpful to examine the executive summary of teacher preparation research
compiled by Wilson et al. (2001) for the US Department of Education. This meta-
analysis of published research includes 57 out of 300 studies published in peer-
review journals. The summary highlights the following findings, without
dismissing or whitewashing alternative credentialing: There is a positive
connection between teacher preparation in the subject matter they teach and
student outcomes, although it is not conclusive how much preparation is
sufficient. Further, pedagogical aspects of teacher preparation matter, as they

37
affect teaching practice as well as student achievement. Clinical experiences
(student teaching) are a powerful element of teacher education. There is a
difference in the success of alternative programs in their recruitment of a more
diverse teaching force, in attracting “the best and the brightest,“ and in their
ability to prepare teachers. However, it seems that longer alternative programs
that more closely resemble traditional programs are more successful in preparing
qualified teachers. Allen (2003) prepared a document for the Education
Commission of the States (ECS), in which he succinctly summarizes the debate
over alternative credentialing routes:

No issue related to teacher preparation has generated more debate than
the issue of the effectiveness of alternative route preparation programs.
Proponents insist alternative routes play a critically important role in
expanding the pool of teachers, and in particular provide a pathway for
unusually capable candidates who otherwise would be lost to the
profession. Critics argue alternative route programs shortchange both
teacher candidates and the students they teach because their preparation,
particularly in pedagogy, is inadequate.

As mentioned above, the research literature is polarized, with few articles that
attempt to examine the effects in an unbiased manner. This bias may exist
because articles are either published by teacher educators or their affiliations,
which usually dismiss anything but traditional credentialing and are backed by
reports such as the National Commission on Teaching and America’s Future
(NCATF) from 1996: What Matters Most: Teaching for America’s Future, or are

38
prepared for policymakers trying to find a more cost-effective and swift solution to
the teacher shortage in hard-to-staff schools and are thus often found to support
alternative routes to credentialing.

Traditional and Alternative Credentialing and the Preparation of Teachers:
Alternative Credentialing and Student Achievement
Literature mainly supporting alternative routes of credentialing, for
example research articles by Chin and Young (2007), Glazerman et al. (2005),
and Goldhaber and Brewer (2000), emphasize overcoming barriers to entering
the teaching profession, and depict alternative credentialing routes as a way to
diversify the teaching force and address the current needs of the labor market.
The study done by Goldhaber and Brewer (2000), has in fact been cited by both
sides of the alternative credentialing debate, as it reports mixed findings on the
effect of certification status on student achievement. This study empirically tested
12th grade students’ math and science achievement desegregated by teacher
certification status. It included students of teachers with standard (traditional)
credential, probationary, emergency, private, and no certification. Drawing on the
National Educational Longitudinal Study of 1988, Goldhaber and Brewer
accessed data of 24,000 nationally representative 8th grade students that were
re-surveyed as 10th graders in 1990, and as 12th graders in 1992. As students

39
provided comprehensive information about their background (race, income,
family structure, etc.), the researchers were able to control for these variables.
In short, it was found that in mathematics students of teachers with a
standard, probationary, or emergency credential did better than students of
teachers with a private or no credential. Surprisingly, students of teachers with
emergency credentials did similar to students of teachers with standard
credentials, a finding contrary to previous findings by NCTAF and others. In
science, the results were less statistically significant but showed the same trends
as in mathematics. One caveat about this study is that only 12th grade students’
scores were used, which may reduce its overall generalizability, as it is
documented that especially minority students in urban school settings drop out of
school before that time. It might have been more useful to compare elementary
school reading and mathematics scores, as other studies have done, in order to
provide a fuller picture.
Glazerman, Mayer, and Decker (2005) examined the impact of Teach for
America on student outcomes in several areas across the nation where TFA is
operating in an effort to generate a nationally representative sample. The total
number of students is close to 1800, with 785 taught by TFA teachers, 279 by
other (control) novice teachers, and 701 by veteran control teachers. Notably, of
the control novice teachers, only 38% held a credential, whereas 51% of the TFA
teachers held a credential when the survey was held. Therefore, it is clear that

40
this study is not a comparison of beginning teachers with traditional and
alternative credentials and its impact on student outcomes, but rather a study of
teachers that already existed at the school sites, as long as they were recruited
through TFA between 2000 and 2002. It might therefore be a comparison of a
TFA teacher that had been teaching for two or more years, and was retained in
the profession to the control teachers, which varied greatly in their experience
and education, according to the researchers.
In this comparison, the TFA teachers were found to have a positive impact
on the math achievement of the students and were not found to have an impact
on reading achievement, with achievement being measured by the Iowa Test of
Basic Skills. In conclusion, the researchers find that Teach for America is a cost-
effective solution to offer schools in poor communities an appealing pool of
teacher candidates.
In the literature opposing alternative credentialing on the basis of student
achievement, one of the most published and outspoken authors in the field is
Darling-Hammond. Reviewing state policy evidence (2000), her conclusion of
teacher quality and student achievement states the following:

Among variables assessing teacher “quality,“ the percentage of teachers
with full certification and a major in the field is a more powerful predictor of
student achievement than teachers’ education levels. [...] The effects of
well-prepared teachers on student achievement can be stronger than the
influences of student background factors, such as poverty, language
background, and minority status.

41

These conclusions are based on several previous studies, such as Ferguson
(1990), who analyzed data from 900 school districts, controlling for variables of
student background and district characteristics. Ferguson found that teachers’
expertise (licensure scores, degree, and experience) accounted for more
variation in grade 1-11 test scores than did students’ socioeconomic status.
When evaluating evidence on the impact of teacher certification, Darling-
Hammond (2001) addresses findings by Goldhaber and Brewer (2000) indicating
that students of emergency certified teachers did as well as students of fully
certified teachers. Due to the very small sample of only 34 emergency
credentialed mathematics teachers— in comparison to the large overall sample
of over 3,400 mathematics teachers in the study— she cautions against drawing
conclusions from the findings, as a sampling error may have caused this effect.
She then summarizes the literature supporting the conclusion that fully certified
teachers have indeed a positive effect on student outcomes.
For mathematics, Darling-Hammond cites the work by Hawk, Coble, and
Swanson (1985), which was also cited by Goldhaber and Brewer (2000), and
found a positive effect of certification on teacher performance in general
mathematic and in algebra, as measured by student outcomes. Monk (1994)
found in the data of the Longitudinal Study of American Youth (LSAY) that
subject matter courses, and even more so coursework in education, have a

42
positive effect on student achievement. Another study done previously by Begle
(1979) and based on the findings of the National Longitudinal Study of
Mathematical Abilities noted that the number of credits a teacher had taken in
mathematics methods courses was a stronger predictor of student performance
than the number of credits taken in mathematics courses.
In science education, 65 studies found a positive impact of teachers’
backgrounds in both education and science courses on student achievement
(Druva & Anderson, 1983). These findings seemed to stem from the observation
that teachers with training in science methods were more likely to use hands-on
science laboratory experiments, whereas the teachers without the educational
coursework emphasized memorization.
According to Darling-Hammond (2001), of critical importance when using
the general terms of traditional (or full) credentialing and alternative credentialing
is the “substantial variability among both “traditional” and “alternative” programs”
(p.69). She continues:

As one review of nontraditional program models noted, there are major
differences between the design and outcomes of alternate routes to
certification that maintain standards but concentrate coursework and
clinical training and those of alternative certification approaches that
reduce requirements and leave both training and licensing responsibilities
to local districts […]

43
A California study cited in “How Teacher Education Matters” (Darling-
Hammond, 2000) found a strong negative relationship between high school
students’ mathematics scores and the number of emergency certified teachers,
after controlling for student poverty rates (Fetler, 1999). The Los Angeles County
Office of Education (1999) similarly found that elementary students’ reading
achievement relates strongly to the proportion of fully certified teachers.
When examining the possible cause of these results in a study published
in 2002, Darling-Hammond, Chung, and Frelow examined the sense of
preparedness of new teachers, which surveyed 2,956 New York City teachers.
Teachers were asked to rate their preparedness and their personal views about
teaching, including their sense of efficacy and their plans to stay in teaching. The
study found highly significant differences, with traditionally certified teachers
feeling better prepared than non-certified teachers except in the use of
technology. When further disaggregating the data according to programs, only
Teach for America had enough participants to be included in the alternate route
analysis, as well as 18 traditional programs. The study found that TFA recruited
teachers rated their preparedness lower than the average teacher education
graduate on 39 out of 40 items.
In her study from 1997 on “schools that work,” Darling-Hammond
researched evaluations of Teach for America teachers and found that, according
to several participants, TFA does not prepare candidates to be successful with

44
students. Former TFA teachers also documented this inadequacy, such as
Jonathan Schorr (1993), who describes the insufficient training and preparation,
and the negative impact on his students.
A study examining the performance of students in the classes of under-
certified (emergency, temporary, and provisionally credentialed) and certified
(credentialed) teachers in selected districts in Arizona was published by Ildiko
Laczko-Kerr and David Berliner in 2002. They also included TFA teacher recruits
in their study pool of under-certified teachers. Using SAT 9 data from 1997-2000,
the researchers obtained information about the teachers’ credentialing status
from the five participating districts, which all served inner-city student populations
made up mainly of minority populations. All districts participate in the Teach for
America program, and all schools in the district state they experience difficulty in
filling open positions. The study included teachers hired in 1998-2000, a total
sample size of N=293. Of these, 159 were credentialed, 89 emergency-
credentialed, 19 temporary credentialed, and 26 provisionally credentialed.
An ANOVA analysis of the three groups of under-certified teachers
established that the student achievement between these three groups was not
statistically different, and thus the three groups were analyzed as one group
(under-certified teachers). This information means that students of TFA recruited
teachers did not perform better than students of other under-certified teachers in
this study, which contradicts findings published by Teach for America. For all

45
subjects on the SAT-9 test in the Laczko-Kerr and Berliner study, students of
credentialed teachers significantly outperformed students of under-certified
teachers (1998-1999, and 1999-2000, in reading, mathematics, and language).
This finding led the researchers to refer to under-certificated programs as a
“harmful public policy” that should be abandoned, as they found that under-
certified teachers from TFA and other programs have a negative effect on
student achievement. As school achievement is measured on standardized tests,
just like the SAT 9 used in this study, a continued evaluation of the effects of
alternative credentialing seems appropriate.

The Achievement Gap
Background
The term achievement gap describes discrepancies in student academic
achievement as based on standardized tests (Ladson-Billings, 2006). While
student achievement can be measured in various ways, in order to compare
schools across a district, state, or nationwide, standardized test measures are
applied. According to their test results, schools find themselves ranked in
comparison to other schools. Although one can argue about the appropriateness
of determining a school’s achievement based solely or mainly on this measure, in
the high-stakes accountability environment found in public education since the
1990s, standardized tests and their implications are the primary, if not the only,

46
tool of evaluation. Therefore, a school’s students’ high performance on these
tests is crucial, as budgetary decisions as well as sanctions are tied to the
outcomes, as dictated by NCLB. Unfortunately, schools labeled low-performing
are usually the schools serving a high percentage of minority students and
students living in poverty (Oakes, 2002). The term achievement gap describes
this difference of student achievement between minority or low-SES students and
White middle-class or more affluent students.
Jeakyung Lee (2002) provides a history of the achievement gap. Following
the measures taken after the civil-rights movement in the 1960s, the gap
narrowed during the 1970s and the 1980s. The achievement gap can provide a
broad picture of racial and ethnic equity in educational achievement and is of
importance as an indicator of racial and ethnic equity in society. When trying to
examine which factors precisely led to a narrowing of the gap in the 1970s and
80’s, scholars identified a complexity and interrelation of possible factors, thus
making a definitive analysis difficult (Grissmer, Flanagan, & Willamson, 1998;
Hedges & Nowell, 1998; Smith & O’Day, 1991). Among the factors they cite are
socioeconomic and family conditions, youth culture and behavior, and school
conditions and tracking practices. A contributing factor in the achievement gap
between African American and White students is school segregation. As schools
became less segregated in the 1960s, traditionally low performing students had
better educational opportunities, which brought about a narrowing of the

47
achievement gap after a few years (in the 70s and 80s). Later, as desegregation
stopped and resegregation began, the achievement gap widened again (Orfield
& Yun, 1999). The achievement gap between Latino and White students,
however, did not follow this pattern. As the achievement gap has widened again
since the 1990s and sanctions against schools have been tied to standardized
test performance, closing the gap has become a central topic on the political
stage, leading, for example, to the national No Child Left Behind Legislation in
2001. However, the gap persists, especially in certain K-12 settings:
Despite countless school reform efforts during the last two decades of the
20th century, we begin the 21st century with continuing gaps in academic
achievement among different groups of students. These gaps in
achievement appear by income and by race and ethnicity. (Johnson,
2002, p. 4)

In her book about the achievement gap, Ruth Johnson (2002) summarizes
the evidence of the persistent gaps in student academic achievement between
African American, Latino and Native American students on one hand, and White
and Asian students on the other hand, emphasizing the importance of these
outcomes in the current high-stakes standardized testing and accountability
climate of public education, a climate that has led to the above-mentioned labels
“low-performing” and “program improvement” schools. Unable to meet set
improvement targets, these schools face sanctions by the state. Johnson
suggests a data-based approach to solve the issue of educational inequity for
which the continuing achievement gap provides evidence.

48
Similarly, in her summary of student achievement in America for The
Educational Trust, Ruiz (2000) provides data from the US Department of
Education and the National Assessment of Educational Progress (NAEP), and
unpublished data from the Educational Testing Service (ETS), showing details of
the achievement gap in reading and mathematics between racial and socio-
economic groups of students. Ruiz documents a difference in access to a
rigorous curriculum, as well as a difference in graduation rates. Ruiz also
acknowledges a discrepancy in access to both resources and qualified teachers
for racially and ethnically diverse groups of students. The data on student
academic achievement shows that Latino and African American students, as well
as poor students, perform lower in mathematics and in reading. This same group
of students traditionally has had less access to college-preparatory curricula at
the high school level. Further, it has had less access to resources and teachers
with a degree in their subject area in comparison to its White, Asian, and more
affluent counterparts. This group also graduates high school and college at lower
percentage rates in comparison to entering freshmen numbers. In conclusion, the
achievement gap is an indicator of a much deeper social inequity based on race
and socioeconomic status.
Berliner and Biddle (1995) note that there is a wide discrepancy in the
education of American students when compared to their international
counterparts. While US secondary students as a group often score below the

49
median in mathematics and science, students from some states actually score as
high as the top-ranked international students, while others score as low as the
bottom-ranked ones. This data evidences that the achievement gap is an issue of
social inequity and not an overall failure of the American public school system.
Hammond (2000) links the underperformance of students in mathematics
and physical science to differences in teacher qualification, stating that teacher
qualification is directly linked to student achievement. Similar findings are
reported by Jordan, Mendro, & Weerashinghe (1997), Sanders & Rivers (1996),
and Wright, Horn, & Sanders (1997). Qualified teachers are not distributed
equally in public schools, and in fact studies show evidence of deep inequality in
the assignment of the least-qualified teachers to minority, especially African
American, students (Jordan et al., 1997; Sanders & Rivers, 1996). When taken
together, these findings are of great concern to educational researchers,
practitioners, and legislators, as they point toward maintenance of the
achievement gap. The question remains how much of the achievement gap can
be attributed to teacher qualification. Time and retention rates need to be taken
into consideration, as a multitude of complex factors play a confounding role in
the persistence of the gap, as discussed above.
In her AERA presidential address from 2006, Gloria Ladson-Billings
suggests that the focus on the achievement gap is misplaced; instead,
consideration of the “education debt” is needed. She argues that in education, as

50
in economics, deficits accumulate over time into debt, and this debt cannot be
solved by simply eliminating the current deficit. She suggests that looking at the
education debt may explain more about the widening or narrowing of the
achievement gap, which is nothing more than a short-range picture of
performance of students on a very limited set of measures. According to Ladson-
Billings, this debt is accumulating and needs to be addressed by full and
sustained school desegregation and funding equity in schools.
While Ladson-Billings follows the line of thought of culturally responsive
teaching to increase student learning outcomes (e.g., Ladson-Billings, 1994 and
1995) and of Haberman’s notion of a “pedagogy of poverty” (1991), she also
points to teachers without adequate preparation (Ladson-Billings, 1997) to
explain the lag in achievement of groups of students in mathematics and science.
She cites the Third International Mathematics and Science Study (TIMSS) from
the US Office of Education as a source establishing this lag. While she
emphasizes that tracking practices in low-ability classes as well as other
circumstances are contributing factors to low achievement, one of her
conclusions about raising achievement pertains to teachers and teacher
knowledge: “Effective pedagogical practice involves in-depth knowledge of
students as well as of subject matter” (p. 704). This notion starkly contrasts to a
definition of “highly qualified” teachers as those merely knowledgeable in subject
matter, and is in line with Shulman’s (1987) belief that beyond content

51
knowledge, teachers also need the pedagogical knowledge. In order to raise
student achievement, Ladson-Billings (1997) proposes culturally relevant
pedagogy, which, she observes, results in high student outcomes, Similar to
Gloria Ladson-Billings, Lisa Delpit (2006) focuses on culturally responsive
teaching in order to overcome low achievement for poor, urban students. She
postulates teaching more content, rigorous content and critical thinking in urban
classrooms with knowledge of the situations and circumstances of students. In
order to assess students’ strengths and use diverse strategies for teaching and
learning, teachers need to evaluate the appropriateness of different
methodologies.
In conclusion, to qualify teachers to respond to the achievement gap and
to increase learning outcomes for urban, low-SES, and minority students,
teachers need to be well-prepared— although neither Ladson-Billings nor Delpit
explicitly states teacher preparation as a focal point. It is clear, however, that
teachers unprepared to develop curriculum, diverse teaching strategies, and
alternative assessments, will be less likely to develop into culturally responsive
teachers, and more likely to maintain the existing “pedagogy of poverty” in urban
schools.

52
The Achievement Gap in California
In California, an increasing number of minority students has been
identified as achieving lower on standardized tests in comparison to its
counterparts. This data concerns educators, as the number of these groups has
been steadily increasing in public K-12 school enrollment. The percentage of
recent immigrant children with “Limited English Proficiency” (LEP students), has
increased, with concentrations in certain areas of the state and in identified
schools (California Department of Education, Language Census Report 1997).
California enrolls approximately 40% of all LEP students in the country, and LEP
students make up 25% of all students served in California public schools.
However, California does not have a plan for bilingual education for these
students. Indeed, Proposition 227 (passed on June 2nd, 1998) limits the use of
languages other than English in public schools (Garcia, 2002), which contradicts
the findings on quality bilingual education and its positive impact on student
achievement and success (August & Hakuta, 1997; Cummins, 2000; Greene,
1998; Krashen, 1999; Krashen & Biber, 1998; Ramirez, Yuen, & Ramey, 1991;
Stanford Working Group 1993; Willig, 1985). In Colorado such results led to the
stop on Amendment 31 (Escamilla, Shannon, Carlos, & Garcia, 2003).
Increasing enrollment of LEP students in California, paired with a legislative
context prohibiting the most effective instruction of these students, perpetuates

53
the cycle of low academic achievement of minorities on California’s high-stakes
standardized tests.
Besides English proficiency, another contributing factor to the low
achievement of minority and high poverty students is tracking practices at
schools. The most obvious form of such tracking is the Concept 6 model,
otherwise known as multi-track calendar. This calendar format is mainly found in
overcrowded, urban schools, which started implementing it as early as the
1980s, supposedly as a temporary fix for overcrowding (Oakes, 2002). Concept 6
allows schools to accept 33% more students than can be accommodated, as one
of three tracks is on a two-month vacation, while the other two are in session.
thus, allowing them to maximize enrollment. The three tracks rotate their
vacations, with one track off, while the other two are in school. There are,
however, many disadvantages for students in Concept 6 schools, as (a) the
schools remain large, (b) instructional time is lost, as there are fewer but longer
school days, so instructional minutes are comparable to traditional calendar
schools), (c) there are fewer nights for homework, and (d) there is lost time due
to the need to review material when a track returns after two months. “B-Track”
students in particular experience their vacation at a most inopportune time: the
middle of their academic semester.
Another concern about Concept 6 and other multi-track schools is
adequate access to enrichment, remediation, and college preparatory programs,

54
such as Advanced Placement. Often schools do not offer these programs on all
three tracks due to limited resources and qualified teachers, and students on
other tracks cannot access them. Although Concept 6 is known as a last resort to
many in education (e.g., State Superintendent of Public Instruction Delaine
Eastin in a letter to Assembly member Goldberg, 2002, cited in Oakes, 2002),
approximately 17% of California public school K-12 students still are on multi-
track calendars. These students are mostly urban, low SES, and minority
students. Thus Concept 6 may be another important contributing factor to the
achievement gap.
Within the same school, students in California may still have very different
educational experiences, even if the school is operating on a single calendar or
students are attending the same track. Students are divided into ability-grouped
classes, either into Gifted and Talented (GATE or honors) or English as a
Second Language (ESL) classes. Such placement affords students very different
access to rigorous and college-preparatory curricula. Often classes for students
who are not on a GATE, college-preparatory, or honors track focus less on
critical thinking or problem-solving and provide less hands-on experiences,
instead focusing on reading texts and filling out worksheets (Oakes & Lipton,
1999). Often placement in low-level classes becomes a powerful self-fulfilling
prophecy for students, especially when LEP students belong to otherwise
stigmatized social groups (Ferguson, 2003; Harber, 2005; Weinstein, Gregory, &

55
Strambler, 2004). Because they travel as an isolated group, students may find
themselves in “low-ability classes” throughout their school day and in all subjects,
even though their primary need for support is in English class. Tracking by ability
groups can therefore be considered another major contributing factor to the
achievement gap phenomenon.
In her article “Educators Respond to Accountability,” Woody (2004)
examines the effect of California’s accountability system on schools, stating that
policymakers in California assumed three main points when implementing the
state’s accountability system. First, they acknowledged a sense of urgency to
implement an accountability system, even if it was not completely developed.
Second, they were hoping “that educators would be motivated to reach and
sustain high levels of student achievement by rewards (monetary, public
recognition) and sanctions (low-performance labeling, threat of school takeover)”
(p. 15). Finally, policymakers felt that disaggregated achievement data would
lead to more equity. In her findings, Woody reports teachers recognized certain
positive aspects to the accountability system, such as the attention given to
student achievement and the introduction of standards; however, they criticized
the impacts of the system on curriculum and instruction, believing that their
professionalism had been compromised. The desegregation of achievement data
showed certain achievement gaps, which Woody found were often excused by
teachers with an underlying belief that certain subgroups are not as apt to

56
perform high on these tests. Teacher expectations are therefore another factor
leading to the achievement gap and its persistence.
In conclusion, the achievement gap is well-documented in the literature on
a national scale, as well as in California. As the public educational system
increasingly uses sanctions as a measure of accountability for schools
underperforming on standardized tests, patterns arising out of students’ racial
and ethnical background, as well as their socioeconomic status are concerning. If
this trend continues, groups of children and of society will be left behind in a
world increasingly emphasizing the importance of education, literacy, and
problem-solving skills. Factors that decrease the achievement gap and can be
controlled by policy standards, such as teacher certification and teacher qualities
like efficacy, need to be studied closely in order to help all students succeed.

Self-Efficacy and Outcome Expectancy of Teachers
Background
The term teacher efficacy describes a teacher’s judgment about his/her
ability to positively affect student outcomes, even and especially with students
considered difficult to teach. Literature relates teacher efficacy to students’ own
sense of efficacy (Anderson, Greene, & Loewen, 1988), motivation (Midgley,
Feldlaufer, & Eccles, 1989), and, perhaps most importantly, student achievement
(as early as Armor et al., 1976). A high sense of efficacy positively impacts a

57
teacher’s effort, commitment, persistence, and resiliency (studies cited in
Tschannen-Moran & Woolfolk-Hoy, 2001), all important factors when considering
the high attrition in low-income public schools, Self-efficacy measurement and
studies of self-efficacy are grounded in social-cognitive learning theory.
Considering these findings about the links between teacher and student
efficacy, studies need to examine the originary sources of efficacy. To this end,
Bandura’s (1977) seminal work on self-efficacy offers some insights. He
describes four main sources of efficacy information for individuals: performance
accomplishments, vicarious experience, verbal persuasion, and emotional
arousal. Traditional teacher credentialing program enable most of these sources
through student teaching experiences, while alternative approaches do not. For
example, performance accomplishments are highly influenced by participant
modeling, which guided student teaching experiences offer. But modeling alone
produces much lower efficacy and less successful performances than
participation in the process, according to Bandura (1977). According to his
findings, “self-efficacy is derived from partial enactive mastery ... on stressful
tasks that the individuals had never done before” (p. 211).
Thus, even if student teaching does not expose a preservice teacher to all
potential classroom teaching situations, according to the research findings
exposure to and mastery of stressful situations in the student teaching context
translates into performance in novel situations. Besides participant modeling,

58
other modes of induction lead to efficacy, such as performance exposure,
suggestion, and desensitization to a difficult situation or task, all part of the
observational phase and student teaching in a traditional credentialing program.
Based on Bandura (1977), Mulholland and Wallace (2001) confirmed that
mastery experiences are the most important source of efficacy beliefs for novice
teachers.
When first entering the classroom and over the first year of teaching, a
teacher’s sense of efficacy and his/her attitudes often change greatly and, most
often, in a negative way, according to several studies (Bullough, 1989; Corcoran,
1981; Day 1959; Gaede, 1978; Hogben & Petty, 1979; Kuhlman & Hoy, 1974;
Ligana, 1970; Wright & Tuska, 1968). Chester and Beaudin (1996) examined
factors influencing these changes in newly hired, urban teachers in Connecticut.
They examined variations in the efficacy changes according to teacher
characteristics, according to elementary, secondary assignments, and special
education assignments, and according to school practices. This study found that
age in consideration of experience played a role in the decline of efficacy. Older
novice teachers experienced a smaller negative change in their efficacy beliefs,
whereas the efficacy beliefs of experienced teachers first entering a public, low-
income school setting did not decrease as much as did those of novice teachers,
regardless of their age. Therefore, the researchers found that teaching

59
experience or a higher age upon first entering teaching can positively impact
efficacy in new hires.
Novice teachers further showed that opportunities for collaboration had a
significant positive effect on efficacy, as did supervisor observation. Availability of
resources alone did not positively affect efficacy, but in combination with
collaboration and supervision it showed a positive impact. The study established
the importance of early and positive feedback and collaboration, important
factors when considering traditional and alternative credentialing. In traditional
programs, student teachers receive feedback and guidance from a master
(cooperating) teacher, as well as from university supervisors. This resource may
contribute positively to efficacy beliefs later in their professional career, as during
their first teaching experiences they were exposed to two factors found to have a
positive impact by Chester and Beaudin (1996).
In a later study Woolfolk-Hoy and Spero (2005) examined the changes in
teacher efficacy over the first years of teaching, from entry into a (traditional)
teacher credentialing program to the end of the induction year. This longitudinal
study extends the work of Chester and Beaudin (1996), as it includes an
assessment and monitoring of preservice efficacy developments. As their study
included novice teachers that had undergone a one-year internship and had
received support throughout their first teaching experiences, it did not show the
decline in efficacy when first entering the classroom, as other studies have. The

60
study confirmed that efficacy directly relates to perceptions of support, and added
to the knowledge that teachers placed in low-SES settings feel less supported.
Therefore, schools serving low-SES student populations should offer new
teachers more support in order to counteract their decline in efficacy.
Measures of teacher efficacy consist of two dimensions: general teaching
efficacy and personal teaching efficacy (Hoy & Woolfolk, 1990). While general
teaching efficacy is based upon a teacher’s attitude toward education and the
possibility of educating children labeled difficult, personal teaching efficacy is the
teacher’s personal sense of ability to teach difficult children. The study presented
here is based upon the construct of personal teaching efficacy of teachers, which
has been shown by Woolfolk-Hoy and Spero (2005) to be independent of general
efficacy. The findings of the 1990 study conclude that while general teaching
efficacy declined in preservice student teaching, personal teaching efficacy
increased, and the organizational socialization of the student teaching
experience led to a more custodial and less humanistic stance of preservice
teachers. These findings reveal disillusionment with the educational
establishments and schools in general, but an increased belief in the personal
ability to make a difference and be effective in a classroom. If a new teacher has
gone through an alternative program, this “reality check” and adjustment of
teaching efficacy does not occur until entering the classroom. This study confirms
student teaching as valuable for establishing a sense of personal resiliency in

61
preservice teachers and for effecting a higher personal teaching efficacy in the
face of known adversary conditions, as presented by the real classroom
situations and a decline in general efficacy. Resiliency is an important factor in
leading to a higher retention of teachers in the classroom (Oakes, 2002).

Teacher Efficacy and Student Achievement
Teacher efficacy relates to student achievement most likely through a
causal chain effect, as evidenced in work by McLaughlin and Marsh (1978). The
researchers found that teacher efficacy influences teacher behavior, which in turn
influences student efficacy and behavior, and thus impacts student outcomes.
Ashton and Webb (1986) built further on this chain by providing evidence that
general teaching efficacy relates to math scores, while personal teaching efficacy
impacts language scores through teachers’ instructional practice.
Ross (1992) conducted a study of the effects of efficacy and coaching on
student outcomes in Canada. His sample included 18 history teachers in Ontario.
Apart from finding a positive impact from intensive coaching, he established that
student achievement was higher in classrooms with teachers who had high
efficacy. He used the Gibson and Dembo (1984) efficacy survey to measure
teacher efficacy, and the student outcomes were measured in September and
May using multiple choice items from the Ontario Assessment Instrument Pool.
Further, cognitive skills of students were assessed through open-ended

62
instruments developed and tested in previous studies. Ross found personal
teaching efficacy more salient in determining student achievement than general
teaching efficacy, a point of distinction in the next study, which relates collective
teacher efficacy to student achievement.
Collective teacher efficacy, according to Goddard et al. (2000), is a
“group-level attribute” (p. 482), expressing the group’s shared belief in its
capability to be successful. Collective efficacy is therefore an attribute of schools,
and not of individuals, but insofar as it is formed by the members of a schools’
staff, individual efficacy belief will influence collective efficacy. The researchers
found that when comparing 47 elementary schools, collective efficacy had a
great impact on achievement: a one unit increase of collective efficacy resulted in
an increase of over 40% of a standard deviation in student achievement.
One of the earliest studies linking student achievement to teacher efficacy
was conducted by Armor et al. in 1976. The RAND Corporation study showed the
reading outcomes of the School Preferred Reading Program of students in Los
Angeles schools serving primarily minority populations. The researchers found
that the greater the teachers’ sense of efficacy, as measured on the original
RAND efficacy measure, the more the students advanced in reading
achievement.
Although the literature has linked teacher efficacy to student achievement,
more studies of this link are needed.

63

Literature-Based Rationale for Methodology
Surveys
The first step in the data collection process of the current study was to
administer surveys to new science teachers from three programs (one university
program and the two alternative programs: Teach for America and District
Intern). The survey includes questions regarding the teachers’ efficacy and
outcome expectations, with 24 questions to be answered on a 5-point Likert-
scale. The survey is a modified version of the “Science Teachers’ Efficacy Belief
Instrument” (STEBI) developed and validated by Riggs and Enochs (1989), and
is included in the appendix. The survey was modified by deleting one item in
order to make it applicable for secondary science teachers. This decision was
made through consultation with a panel of experts. After this deletion, the item
loading on the factors according to Riggs and Enochs was completed, with 12
items loading on efficacy and 12 items loading on outcome expectancy of the
new teachers. The surveys were analyzed for differences between the three
sample groups and the two subgroups using descriptive statistics. The survey
measures two subgroups of teacher beliefs, namely efficacy and outcome
expectancy, and its cover sheet identifies a participant as belonging to either the
traditionally or alternatively credentialed group as well as to one of the three
programs. Differences between the groups were analyzed using descriptive

64
statistics. The cover sheet also includes a self-identification question regarding
retention, as discussed below.

Interviews
The interviews were conducted as a follow-up to the surveys, with a
purposefully selected small sample to provide a deeper understanding of the self-
efficacy and outcome expectancy beliefs of the teachers. The interview
participants were selected according to their survey results as representative of
their respective programs. According to Patton (2002), interviews enable the
interviewer to enter another person’s perspective and ask questions about things
that cannot easily be observed. Efficacy is a belief that can only be observed to a
certain degree, usually manifesting itself in specific classroom behaviors; for
example, a high sense of efficacy has been observed to lead to less in-seat time
instruction and a more positive classroom climate while teachers with a low
sense of efficacy have a less humanistic stance towards teaching (Pupil Control
Ideology, PCI, Willower, Eidell & Hoy, 1967). Therefore, establishing a teacher’s
self-efficacy on the basis of observation can be difficult.
Although interviews rely upon self-reported data (unlike observation),
Patton (2002) affirms that they are not less meaningful, desirable, or valid. He
cautions researchers, however, that the quality of the obtained data largely
depends on the interviewer. The interviews collected data using a standardized,

65
open-ended interview approach with fully formulated questions. The interview
protocol included standard probes to ensure that the interviewed teachers
received the same stimuli. According to Patton, advantages of this approach to
interviews include availability for inspection, a decrease of variation in the
presentation of interview questions, focus and time-efficiency, and facilitated data
analysis, as responses are easy to find and compare.
The interview consisted of eight questions that were revised through
consultation with a specialist in the field of qualitative research, and reviewed by
an expert panel. Questions specifically address the two subgroups of teacher
beliefs from the survey (efficacy and outcome expectancy), as well as two
questions about the respective preparation programs. Further information on the
interview methodology will be addressed in the following chapter on
methodology.

Retention Data
The quantitative retention data is represented by the following samples:
the retention of the three analyzed programs after the first year of teaching, the
self-declared expectation of retention of the surveyed population, and the
retention data published for the State of California and the nation. If teachers are
not retained, it is not possible to build teacher experience and expertise of the
specific teaching context at a school site, both of which contribute positively to

66
student achievement, according to the literature discussed above. As long as
teachers continue to leave the profession after only a few years, the issue of the
revolving door (Ingersoll, 2001) will continue.

Summary and Conclusion
The debate around alternative credentialing routes into teaching is valid.
The numerous studies conducted often provide polarized findings, either
condemning alternative routes into teaching as harmful policy, or dismissing
traditional teacher education as a relic without justification in the modern labor
market. The question underlying this issue, however, has not been studied in
much detail: specifically, what the differences are between teachers coming out
of alternative versus traditional programs. While the debate around the impact of
their teaching abilities is marked by entrenched positions (for example Darling-
Hammond, 2000, and Laczko-Kerr & Berliner, 2002, vs. Smith, 2006, and
Glazerman et al., 2005), alternative routes may be here to stay. Legislation
defines highly-qualified teachers mostly through subject-matter competency, as
in California where intern credentials are still available for new teachers that have
not completed coursework in education.
Teacher efficacy and outcome expectancy may very well be one of the
most important characteristics to study when comparing teachers from different
credentialing routes, as they have been linked to teacher behavior, as well as to

67
student efficacy and achievement (Tschannen-Moran & Woolfolk Hoy, 2001).
Teacher educators, program directors for alternative programs, as well as policy
makers should therefore be concerned if one credentialing route channels
teachers that have a markedly low sense of efficacy and outcome expectancy
into urban classrooms.
Advocates for alternative programs claim that there is indeed a greater
need for teachers than can be met through traditional credentialing programs and
that alternative certification is helping teachers to overcome the barriers to
entering the profession. This debate is highly problematic, as it shows the
general value attached to public education and to the status of teaching (Darling-
Hammond, 2000). According to this rationale, qualifying limitations are not
necessary to enter the teaching profession. This logic—only one step from the
conclusion that most people that are educated can teach— endangers the
professionalism of teachers. Ingersoll (2001), in his study on the teacher
shortage, found that recruitment is only a temporary fix. As long as retention
rates of teachers do not stabilize to levels comparable to other professions,
recruitment alone will not effectively address the teacher shortage. In the long
run, underprepared teachers leave the profession earlier, thus preventing the
schools they serve from developing a stable instructional program. Departures
also cost districts more money due to the constant need to hire new teachers
(Darling-Hammond, 2000). The proposed focus of the study to include only

68
secondary science teachers is due to the findings that science, along with
mathematics and special education, is the subject in which recruitment and
retention of teachers is especially difficult (Ingersoll, 2001).

69
Chapter III: Methodology
Introduction
This study examines the relationship between factors contributing to
student achievement in urban schools and credentialing route. The factors
studied through data collection are teacher self-efficacy, outcome expectancy,
and teacher retention. An evaluation of the teachers’ self-efficacy, outcome
expectancy, and their retention rates was analyzed according to certification
route. High teacher self-efficacy and low teacher turnover are contributing factors
to high student achievement, as established in existing literature.
In order to examine these factors, the following research questions were
developed to be examined through qualitative and quantitative data collection
and analysis:

(1) (a) How do differently prepared (traditional vs. alternative)
beginning, urban, secondary science teachers compare in their
self-efficacy?
(1) (b) How do differently prepared (traditional vs. alternative)
beginning, urban, secondary science teachers compare in their
outcome expectancy?

70
(2) After the first year of teaching, how do the retention rates of this
teacher population from traditional and alternative programs
compare?
In order to answer these questions, a mixed-methods approach was used
in the collection of data. Quantitative data was collected through Likert-Scale
surveys administered to the sample teacher population. Further, interviews
generated qualitative data for the analysis on teacher efficacy and outcome
expectancy. Retention data was collected through a document review of the
respective programs after the first year of teaching. Each program’s data was
compared, as well as self-reported data from the surveyed teachers. This
collected data was then compared to the retention data published for the state of
California, as well as to published national data for similar groups of teachers.

71
Table 2: Methodology Chart
Data
Source
(Type of
Method)
Rationale and Function
of Method
Result of Method Population Method of
Analysis
Survey:
Science
Teacher
Efficacy
Belief
Instrument
(STEBI),
Riggs &
Enoch,
1989,
modified
Teacher efficacy is
linked to student
achievement
Survey will expose
differences in teacher
efficacy and outcome
expectancy according
to preparation format
(traditional or
alternative)
Quantitative data
through 5-point
Likert-scale on 24
survey questions
N not
predetermined
all beginning
traditionally
prepared science
teachers from
one university
program
all beginning
alternatively
prepared
teachers from
two programs
Pattern
sampling to
include all
science
teachers from
the programs
Descriptive
statistical
analysis of
Likert- scale
data
Interviews Teacher efficacy is
linked to student
achievement
Interviews will expose
differences in teacher
efficacy and outcome
expectancy according
to preparation format
(traditional or
alternative)
Triangulation with
Survey data
Qualitative data
through interview
data transcription
and coding on 6
questions relating
to teacher efficacy
and outcome
expectancy
N= 6
2 teachers from
each program
are interviewed
Purposeful
Sampling
(Multi-tiered
data
collection)
according to
survey
Interviewees
were selected
to be
representativ
e of their
sampling
population:
their survey
average
scores were
representativ
e of the
average
score of the
program
Retention
Data
High teacher turnover
is identified as
contributing to low
student achievement
Differences in retention
data of traditionally
and alternatively
prepared teachers are
exposed
Quantitative data
of retention from
participating
programs after the
first year, self-
reporting, and
published
retention data
N not
predetermined
Traditional and
alternative
program data of
retention after
year 1
Self-reported
data
Analysis of
retention
rates in
comparison
to each
other, state-
wide and
national data

72
Sample and Population
As described, the study’s purpose is to compare the effects of teacher
preparation formats and outcomes on contributing factors to student
achievement, namely teacher efficacy, outcome expectancy, and high teacher
retention. Thus, the primary unit of analysis for this study are the teachers
completing either format of preparation. One population consists of new urban
science teachers who have completed a traditional, university-based program.
The other population is urban science teachers who are new, first-year teachers
without a preliminary credential but who are currently enrolled in either the district
intern program of a large urban school district or the Teach for America Program
and hold a university intern credential. A “secondary science teacher” describes
a full-time teacher in grades 6-12 (middle or high school) in a large urban district
in Southern California, all teachers have proven subject-matter competency as
established by NCLB.
Two beginning science teachers from each program were selected for
interviews; therefore, two interviews were held with traditionally prepared
candidates, and four with alternatively prepared candidates. Participants were
selected based upon survey answers representative of their sample groups. The
surveys were statistically analyzed and then the prospective interview
participants whose surveys showed average (mean) values for their group were
contacted. In order to keep the data comparable, only teachers holding full-time

73
classroom positions for less than 12 months in a large, urban school district were
studied. Data collection took place from April to July at the end of the first year of
teaching. This designation allowed data to be reflective of a comparable
experience in the classroom for all teachers, because data collection occurred at
the same time. Further, data collection occurred after a time when the beginning
teachers had the opportunity to experience an array of classroom situations.
The group of traditionally prepared teachers consisted of recent science
teacher candidates from a large local university. The candidates studied were
finishing a master’s degree at the end of the first year of teaching, and had
finished their teacher preparation coursework prior to beginning the school year;
thus, they already held preliminary credentials. The university has a strong focus
on preparing teachers for urban, low-income schools in the city in which it is
located. This focus makes it comparable to the alternative program Teach for
America, as the mission of Teach for America states:
Teach for America's mission is to enlist our nation's most promising future
leaders in the movement to eliminate educational inequality. We
accomplish this by building a diverse, highly selective national corps of
outstanding recent college graduates—of all academic majors and career
interests—who commit two years to teach in urban and rural public
schools in our nation's lowest-income communities and become lifelong
leaders for expanding educational opportunity. (Retrieved from
http://www.teachforamerica.org/mission/theory_of_change.htm)
For the second alternatively certified science teachers group, science teachers
from the district’s intern program were sampled. Both alternative programs, DI

74
and TFA, prepare their candidates for the kind of teaching situations encountered
in urban school districts.
Instrumentation

Figure 1: Relationship of Problem, Research Questions (RQ), and Methods
of Data Collection (Data Sources)

Low
Teacher
Self-
Efficacy
and
Outcome
Expectancy
High
Teacher
Attrition
Leading to
Low
Teacher
Experience
Contributing
Factor
Contributing
Factor
RQ 2:
Does
teacher
preparatio
n format
impact
science
teacher
retention?
Surveys
Interviews
Program
Retention
Data
State &
National
Retention
Data
RQ 1:
Does the
format of
teacher
preparation
impact
science
teacher self-
efficacy/
outcome
expectancy?

D
A
T
A

S
O
U
R
C
E
S
Measure of
Teacher
Preparation
Format’s
Impact on
Achievement
Achievement
Gap in Urban
Schools
Self-
Reported
Intention

75
The data collected for this study consists of a survey, interviews, and
retention data from three sources, discussed in detail below.
Survey: Modified Version of the Science Teachers’ Efficacy Beliefs Instrument
(STEBI)
The STEBI combines two instruments, the Personal Science Teaching
Efficacy Belief scale and the Science Teaching Outcome Expectancy Scale. It
was modified for use with secondary teachers. The survey includes questions
regarding the teachers’ efficacy and outcome expectancy, with 24 questions to
be answered on a five-point Likert-scale. The survey is based on the Science
Teachers’ Efficacy Belief Instrument, developed and validated in a study by
Riggs & Enochs (1989) to accurately reflect elementary teachers’ science
efficacy beliefs and science outcome expectancies.
This survey scale answers the first research question, which compares the
influence of teacher credentialing route on science teacher efficacy and outcome
expectancy. If, after analyzing the data using descriptive statistics, a difference in
efficacy and outcome expectancy of traditionally and alternatively prepared
teachers is found, the effect can be traced to credentialing route. All other
factors are alike: all teachers are teaching in comparable urban settings, are at
the end of their first year of teaching, and are teaching science in secondary
public schools in large metropolitan districts in Southern California. The survey
was administered to all new science teachers from the traditional university

76
program and all currently enrolled new science teachers in the alternative
programs. On a background survey attachment, the teacher’s gender,
preparation program, and teaching experience calculated in months was
recorded. Further, the background sheet asks teachers to rate their retention
expectation in urban teaching.
A study published by Riggs and Enochs in 1989 established validity for the
survey scale. After reviewing the existing literature on science teaching and
factors impacting science teaching at the elementary level, the researchers found
that efficacy beliefs of teachers need to be examined due to their influence on
teaching behavior.
The scale was developed by the researchers to measure elementary
teachers’ beliefs, as the literature suggests that they do not teach science as a
high priority (Stake & Easley, 1978; Schoeneberger & Russell, 1986), although
exposure to science and scientific principles was found to be critical by the
National Science Board Commission on Precollege Education in Mathematics,
Science and Technology (1983). The scale was specifically designed to assess
both teacher efficacy beliefs and outcome expectancies for students.
Although the scale was developed to measure elementary teachers’
science efficacy beliefs, a consultation with content experts suggests it is also
useful for determining secondary teachers’ beliefs. Beginning teachers often feel
overwhelmed (Bartell, 2005), experiencing a survival mode and low self-efficacy

77
during their first year. Literature has shown efficacy to be content-specific; thus,
an examination of efficacy of teachers in a specific subject matter will result in a
higher generalizability. Bandura (1981) defined efficacy as a situation-specific
construct, rather than a global one. Secondary teachers of science have an
undergraduate degree in science; however teachers at the middle-school level or
in the integrated science classes at high school will often be placed outside their
areas of expertise. Further, the student outcome expectancy questions are not
based on efficacy beliefs but are rather questions about the perception of the
locus of control, asking teachers if they believe that they are able to impact
student learning.
The scale keeps the constructs of teacher efficacy and outcome
expectancy distinct in order to facilitate an evaluation of both aspects separately.
It was originally modeled after Gibson and Dembo’s Teacher Efficacy Scale
(1984), which measured both aspects in general. The researchers modified the
items to be setting-specific to science classrooms and to load either on efficacy
beliefs or expectancy outcome. Two scales resulted, one measuring efficacy
beliefs and the other measuring outcome expectancies, which were combined
into the STEBI. Several items were added to create a larger item pool.
Two evaluation rounds by experts were conducted. A measurement expert
edited all items for clarity, resulting in a 50-item pool. A panel of five experts then
evaluated the items, classified the dimension of each item, and rated the scale,

78
items, and their representativeness. These steps contributed to the content
validity of the instruments. Items classified inconsistently by the experts were
eliminated from the pool.
The first trial study of the instrument included 71 practicing elementary
teachers enrolled in graduate classes and aimed at refining the item pool further.
The researchers conducted a factor analysis on both subscales (efficacy and
outcome expectancy) after discovering inherent flaws in the outcome expectancy
scale. Items to be included in the next study were selected based on factor
loading. Items that showed a crossloading on both efficacy and outcome
expectancy were excluded. Several new, negatively phrased items were added
to the scale. Further, validity data was collected based on past correlations to
teaching efficacy beliefs. After analysis of the reliability and item correlations, six
items with the lowest corrected correlation from the efficacy scale and four items
from the outcome expectancy scale were omitted. The resulting items provided
two clearly distinct and homogenous scales.
The researchers then conducted a follow-up study with a population of
331 rural and urban elementary teachers, with a one-tailed t-test establishing that
no statistically significant differences existed between urban and rural teachers.
Due to this finding, the results of the proposed study may be transferable to other
groups of science teachers, although only urban teachers will be included. Again,
reliability was established though internal consistency comparison and items

79
without high discrimination were dropped from the scale. Factor analysis was
conducted as described above in the tryout study. After item analysis and
omission of several items, the resulting scale had 25 items, 13 loading on
teacher efficacy and 12 loading on outcome expectancy. The resulting alpha
from the efficacy belief scale was 0.92 for the outcome expectancy 0.77. The
researchers conducted a specific test (Cattell, 1966), in order to establish that
only the two factors (efficacy and outcome expectancy) should be considered in
following analyses. A second factor analysis further supported two discrete
factors, enhancing construct validity.
With such preparation, the instrument seems promising to accurately
assess efficacy beliefs of science teachers, and their outcome expectancies. The
lower reliability of the outcome expectancy scale is in line with earlier findings
that consistently document difficulty in measuring this construct (Gibson &
Dembo, 1984). Nevertheless, due to the thorough reliability and validity testing
performed by Riggs and Enochs (1989), the STEBI will be able to provide reliable
data to analyze differences in science teachers’ beliefs, according to their
preparation route. The STEBI was thus chosen for this study over other
measures, as it seems more specific in its approach toward capturing teachers’
beliefs. As the instrument collected data from secondary teachers, one item was
deleted. Item 3 asks respondents to rate the following statement: “Even when I
try very hard, I do not teach science as well as I do most subjects.” As secondary

80
teachers do not teach any other subjects, the item was omitted for the purpose of
this study.
Therefore, the instrument used in the study includes 12 (not 13) items
loading on teacher efficacy beliefs, and 12 items loading on outcome expectancy.
In the sentence “I understand science concepts well enough to be effective in
teaching [elementary] science,” the word “elementary” was deleted from item 12
for the same reason as cited above.

Interviews
The interviews generating qualitative data were conducted with six
purposefully selected teachers from the surveyed population. Two beginning
science teachers from each program were selected; therefore, two interviews
were held with traditionally prepared candidates, and four with alternatively
prepared candidates. Participants were selected based upon survey answers
representative of their sample groups. The surveys were statistically analyzed
and then the prospective interview participants were contacted. Their surveys
showed average (mean) values on self-efficacy and outcome expectancy for their
group.
Interviewees were compared across different programs and, in order to
emphasize differences in values, opinions, and emotions, a structured interview
type was used. This procedure limits the variance of answers due to factors such

81
as different flow of question order, etc. All interviewees were presented with the
same questions in the same order. However, in order to allow for a more
thorough understanding, several probing questions were asked at the
researcher’s discretion in order to get a deeper understanding of the
interviewee’s perspective on certain questions. To ensure comparability of
experiences, each interviewee was interviewed once at the end of his or her first
year of teaching.
Each interview consisted of eight open-ended questions (Appendix C).
The interview collects data to answer the first research question comparing
teacher self-efficacy and outcome expectancy according to credentialing route.
The interview was first generated by the researcher in collaboration with a panel
of qualitative research experts through consultation of efficacy literature and to
align to the survey items. The original 12-question interview went through a
review process by the panel and was condensed to 8 questions, which are
structured in three subgroup sections: personal teaching efficacy (opinion and
emotion questions), outcome expectancy (opinion questions), and questions
about their preparation program (opinion and experience questions).
After the selection of the survey instrument, interview questions were
reviewed in order to reflect a stronger alignment with the survey and to allow the
triangulation of data. The first section of the interview focuses on teacher efficacy
and includes three questions; the second section focuses on the outcome

82
expectancy of the teacher and also includes three questions. The last section
asks the interviewee to evaluate his or her preparation program in two questions.
Probing questions for the first two sections are included in the interview plan to
ensure a depth and complexity of answers while maintaining continuity and
comparability of question stimuli.

Retention Data
The retention data used is (1) a self-reporting of expectations to stay in
urban teaching, (2) an overview of each program’s retention between the first
and second year of teaching, and a comparative figure of retention data nation-
and statewide. The question used for self-reporting is included on the survey
cover (Appendix D). The question asks participants to rank their responses
through one of four answer choices. For analysis reasons, the first choice (will
continue teaching in an urban setting as long as I am able), was assigned the
value 4, the second choice was assigned the value 3, the third choice was
assigned the value 2, and the last choice was assigned a value of 1. All sets of
retention data are compared: the self-reported intent, the retention of the
analyzed programs, the retention data published by the State of California (Reed
et al, 2006), and published national data (Marvel et al, 2007).

83
Data Collection and Timeframe
All data for the study was collected in the spring and summer of 2008 by
the researcher. Data collection began after IRB approval and the appropriate
contacts were made, with the surveys given no later than May 2008. Interviews
were audio-taped and transcribed by the researcher in May, June, and July 2008.
The researcher collected all data, which was handled confidentially. All
participants’ names were changed to protect their identities, as were the names
of their school sites or other identifying factors. Original data is stored in a locked
file cabinet to ensure confidentiality.

Data Analysis
The quantitative data generated through the survey was analyzed using
descriptive statistics after the item loading on each factor (efficacy and outcome
expectancy) had been calculated and reverse items had been changed in their
score. After the item loading on the factors according to Riggs and Enochs
(1989) was completed, the surveys were analyzed for differences between the
traditional and two alternative programs. As the survey includes a cover sheet
that identifies a participant as belonging to one group (traditional or alternative
program), differences between these groups were statistically analyzed for each
item and for self-efficacy and outcome expectancy as a whole. With participants
belonging to mutually exclusive groups, a comparison can be made to determine

84
a difference in the responses according to program. The responses were sorted
according to program and according to the subgroup (TE and OE), and analyzed
using the SPSS program. Averages (statistical means) were generated for each
item and each program, and these averages were then compared. Averages
were also calculated for all answers to items for each subgroup, and the
programs could be compared for their participants’ TE and OE. Due to the small
sample size, statistical significance cannot be established. But differences found
between the traditional and alternative certification routes may still be used to
consider the hiring practices of urban districts. Future studies with larger sample
sizes may be considered, if differences are found.
The interview data generated in the 6 interviews was transcribed, coded
and analyzed, using the two established subgroups of the survey (personal
teacher efficacy and outcome expectancy, TE and OE) as themes, as well as
general efficacy, and comments about the program. Another theme (focus on
standardized testing as indicator of teaching success) arose after transcription
and is thus included in the analysis.
The quantitative retention data was analyzed through regression analysis,
which is able to detect differences between the quantitative data samples. This
retention data is used to answer the second research question about
comparative teacher retention. If teachers are not retained, building teacher
experience and expertise of the specific teaching context at a school site, both of

85
which contribute positively to student achievement, is impossible. The study also
tested if TFA teachers indeed fulfill the two-year commitment to teaching that is
expected by the program. Again, the small sample size of studied teachers
prohibits establishing statistical significance for the retention data of the three
programs.

Reliability and Validity: Confidence of Truthfulness of the Study
While the study is based on several assumptions, such as whether the
district intern program and TFA in combination will provide an accurate scope of
alternatively credentialed teachers, certain aspects increase confidence in its
findings. For example, the triangulation of data to include surveys, interviews,
and retention data emphasizes potential patterns: If a trend arises in how
participants answer a survey question related to their efficacy, and the interviews
elaborates this trend, credibility increases. However, if the two data sources
provide conflicting responses, a discussion of these results is indicated for further
analysis. The scale to evaluate teacher efficacy and outcome expectancy has
been authorized for content validity and reliability by the researchers. The
interview has been validated through a peer-review process involving several
stages of redesign, and has been modified to mirror parts of the survey in order
to be able to triangulate results. Through this process, the interview can also rely
on the content validity of the questions from the survey. The retention data will be

86
drawn from published or unpublished data provided by the programs themselves,
and is therefore not generated by the researcher, but rather accumulated,
presented, and interpreted through meta-analysis. The self-reported data on the
survey cover sheet helps to establish a further outlook on the commitment to the
profession by all surveyed teachers should be subject to self-reporting critique,
but may also be used to project further retention in the profession.
Although the study does not represent a true experiment, several factors
strengthen its internal validity by counterbalancing possible interfering variables.
The content specific approach is in line with Bandura’s notion of efficacy as
situation-specific and not globally measurable, as discussed above. It also
improves the comparability of findings between the sample groups. Further,
purposeful sampling through a selection process of the interview participants as
representative according to their answers on the survey provides an impression
of each program’s teachers and not an exceptional case.

Summary
This chapter provided the research methodology for this study, including a
description of the research design, sample, frameworks, and instruments.
Further, an explanation of the data collection and analysis was provided. The
proposed study will be able to draw a data-based conclusion comparing teachers
from different credentialing routes in their self- efficacy, outcome expectancy, and

87
retention, which have been established as contributing factors to student
achievement. The study will rely on self-reporting through surveys and
interviews, and will not collect observational data. Therefore, the goal of the
study is to comparatively measure perceived self-efficacy, outcome expectancy
and the program reported retention from two alternative and one traditional
program.

88
Chapter IV: Data Analysis and Interpretation of Findings
Introduction

This chapter presents an analysis of the data collected by the researcher
in the current study for the purpose of investigating the influence of teacher
credentialing route on teacher efficacy, outcome expectancy, and retention of
urban science teachers after their first year of teaching. The focus of the study is
to determine and analyze any potential differences in beginning science teachers
from three different preparation programs (Teach for America, District Intern, and
a university program). The data for the study was collected using the following
instruments:
Likert scale surveys, given to all first-year science teachers from the three
programs
Interviews with two teachers from each program
Retention data as reported from each program
Self-reported expectation of retention from all surveyed teachers
Retention data published by the state of California and national data
The data obtained from the surveys was segregated according to the RQ it
addressed (RQ 1a or 1b), and scores were reversed for the negatively worded
items, according to Riggs and Enochs (1989). Survey data was then statistically
analyzed, using descriptive statistics with the following sample populations TFA:

89
n=25, DI: n=15, university program: n=11; total N=51. The two interviews from
each program (N= 6) were transcribed, coded, and analyzed. The retention data
reported from each program was compared with the other programs as a
retention percentage, and compared with self-reported data as well as published
state data. All data analysis was done in order to answer one of the following
research questions:
(1) (a) How do differently prepared (traditional vs. alternative)
beginning, urban, secondary science teachers compare in their
self-efficacy?
(1) (b) How do differently prepared (traditional vs. alternative)
beginning, urban, secondary science teachers compare in their
outcome expectancy?
(2) After the first year of teaching, how do the retention rates of teacher
populations of traditional and alternative programs compare?

The data is presented and analyzed, according to the research question it
addresses. Table 3 provides an overview of the analyzed survey data, followed
by the instrumentation within each research question.

90
Data Overview: Data Collected for RQ 1a and b
Table 3: Group Mean Descriptives from Surveys

Group Membership
N Mean
Standard
Deviation Skewness Kurtosis
Statistic Statistic Statistic Statistic Std. Error Statistic
Std.
Error
TFA TE Scale Mean
25 4.0433 .45651 .329 .464 -1.000 .902
OE Scale Mean 25 3.7533 .25851 -.632 .464 1.077 .902
Valid n 25

DI TE Scale Mean
15 3.9778 .45038 -.436 .580 -.977 1.121
OE Scale Mean 15 3.0556 .35588 .583 .580 .221 1.121
Valid n 15

U TE Scale Mean
11 3.9545 .40202 -1.367 .661 2.295 1.279
OE Scale Mean 11 3.3939 .40840 -1.201 .661 2.717 1.279
Valid n 11

Note. For data presentation and analysis, the following abbreviations are used:
TFA= Teach for America
DI= District Intern
U= University Program
TE= Teacher Efficacy Scale
OE= Outcome Expectancy Scale
Std= Standard
N= number of cases in a study
n= typically used as the number of cases in
each sub-group of sample.
Therefore, N= all cases in the study (51) and TFA n= 25, DI n= 15, U n= 11
OR Alternatively Prepared Teachers n= 40 and Traditional n=11.

92
Means (Averages) of Survey
For the purpose of describing the data collected in the study descriptively,
the means, or averages, of the survey items will be compared. In order to
compare and determine the resulting data for statistical significance, the standard
deviations of each mean are necessary; they are presented in the Table 3. Using
the means and standard deviations may show that, statistically speaking, there is
not a true difference between the means. However, due to the small sample size,
the rank order of the means for descriptive purposes will also be reported, which
is what is expected in descriptive statistics. The standard deviations are included
in Table 3 (p. 86) for reference, but will not be discussed, as this is not customary
unless trying to establish statistical significance.

Figure 2: Rankings by scale group means
TE Programs: U (3.95) < DI (3.98) < TFA (4.04)
Traditional and Alternative: U (3.9545) < TFA + DI (4.0155)
OE Programs: DI (3.06) < U (3.39) < TFA (3.75)
Traditional and Alternative: TFA + DI (3.40445) > U (3.3939)
Both groups round to 3.4, therefore are approximately equal

Based on this figure, the average (mean) for responses regarding efficacy
for U is 3.95, for DI is 3.98, and for TFA is 4.04, with a difference Δ of 0.09

93
between the programs. Also based on this table, the average (mean) for
responses regarding outcome expectancy for U is 3.39, for DI is 3.06, and for
TFA is 3.75, with a Δ of 0.36 between the programs, a much greater difference
than on the efficacy items. Based upon this data, teachers prepared in the
university (U) program show lower efficacy beliefs on average than both
alternative (TFA and DI) programs. Further, U teachers show lower outcome
expectancy beliefs than TFA teachers. If the two alternative programs are taken
together and compared to a traditional preparation program, the alternative
program teachers show higher efficacy beliefs than traditional teachers, but are
similar in their outcome expectancy.

Kurtosis and Skewness
Kurtosis and skewness of a distribution examine if the data collected in a
study follows a normal distribution curve, or bell curve. Both are important in
establishing the representativeness of a data sample. Kurtosis of a distribution
can be normal, leptokurtic, or platykurtic (Huck, 2004). A leptokurtic distribution is
shown in the data collected from the U program teachers, meaning that the
distribution curve is more peaked or pointier than a normal distribution. The
number for a normal distribution is zero (0), whereas negative numbers as in the
scores for efficacy (TE) in the alternative programs mean a flatter curve, and
positive numbers mean a pointier curve (Vogt, 1999). Therefore, the alternative

94
programs show a platykurtic distribution for teacher efficacy. The traditional
program data shows an unusually large number of scores in the center of the
curve, making it pointier and thus showing less variation in the responses. Most
teachers from this program answered similarly, resulting in the pronounced point
of the distribution curve. The alternative programs show a less pronounced
center of the distribution, which means that the scores varied more greatly than
in the traditional program and the teachers did not agree as much in their answer
choices. The TFA teachers show a much greater consensus and a leptokurtic
distribution for outcome expectancy, even if it is less pronounced than in the
U program group.
Skewness is used to describe an asymmetrical distribution of scores. The
value of skewness for a normal distribution is zero (0). The scores from the
U program show a negative skew in their distribution for efficacy and outcome
expectancy. This information means that the scores are drawn out to the lower
end of the efficacy and outcome expectancy spectrum, with more answers than
in a normal distribution bunched up in the higher scores. This means that most
teachers’ sense of efficacy and outcome expectancy is above the median value
for the answer scale (above 2). As only skewness values more than twice their
standard error are an indicator of asymmetrical distribution (Vogt, 1999), only the
U program shows a departure from symmetry: -1.3 (TE) and -1.2 (OE) for a
standard error of 0.6.

95
Figure 3: Kurtosis and Skewness for Efficacy (TE) and Outcome
Expectancy (OE)
TE Scale
Kurtosis-
TFA= -1.0 DI= -.977 U= 2.925
Skewness-
TFA= .329 DI= -.436 U= -1.367
OE Scale
Kurtosis-
TFA= 1.077 DI= .221 U= 2.717
Skewness-
TFA= -.632 DI= .583 U= -1.201

In conclusion, on both scales, TFA and DI means met the general criteria
for a normal distribution when considering kurtosis and skewness. The
U distribution did not. Based on the U kurtosis coefficients for both scales, the
distributions were leptokurtic. Additionally, U means were negatively skewed on
the TE and OE Scales. This result may be due to the much smaller population
size in the U population, as discussed below.

96
Data Collected for RQ 1a: How do differently prepared (traditional vs.
alternative) beginning, urban, secondary science teachers compare
in their self-efficacy?

This research question focused on establishing if there is a difference in
self-efficacy of the studied population according to its preparation format. In order
to answer this question, survey data was collected, and interviews were
administered with three different preparation programs: one university-based,
traditional program, whose beginning teachers experienced pedagogical
coursework and student teaching before entering as first-year teachers, and two
alternative programs, whose beginning teachers had no or minimal preparation
before entering the profession, but received pedagogical coursework while
already teaching. The surveys were administered in late spring (April to May
2008), and the interviews were conducted between May and July 2008.

Findings from the Surveys
The survey was a modified version of the STEBI instrument, developed
and validated by Riggs and Enochs (1989). The administered survey form is
attached in the appendix. From this survey, questions 2, 4, 5, 7, 11, 16, 17, 18,
20, 21, 22, and 23 are used to establish teacher self-efficacy; the remaining
questions will be discussed later in this chapter, as they pertain to outcome

97
expectancy beliefs (RQ 1b). The items are clustered to correspond with the three
interview questions regarding efficacy beliefs.

Cluster 1: Items 2, 4, 5, 7, 11, 22, and 23.
Items in cluster 1 of the survey address the level of perceived
effectiveness by a teacher of secondary science, contributing factors to this
effectiveness and supporting evidence for being effective. It is the largest cluster
of the three and some items in cluster 1 (item 4, 5, 22, and 23) influence cluster 2
as well. Item 2 addresses the notion of life-long learning while being an educator.
This question shows a high and above mean or average
1
efficacy belief in all
three groups (U, TFA, and DI); with the teachers from the alternative TFA
program scoring highest (4.48). The Δ between the three programs is 0.28 for
this question. Both, TFA and the university program show answer scores
2
of 4
and 5 only, while teachers from the DI program show scores ranging from 1 to 5,
suggesting that they are not certain about their ongoing learning and their
improvement in pedagogical skills. Of the three programs, the DI group shows

1
Note: As established above in Chapter 4, Introduction, the average (mean) for an answer
regarding efficacy for U is 3.95, for DI is 3.98, and for TFA is 4.04. Accordingly, efficacy scores
are ranked by average, as above or below average, for the purpose of this analysis.

2
The answer score range states what answers teachers from this group gave (1= Strongly Agree,
2= Agree, 3= Undecided, 4= Disagree, 5= Strongly Disagree).

98
the lowest mean for this question (4.20). However, all groups show that the
majority of their teachers are continually learning to improve their teaching by
scoring above 3.
Item 4 suggests that there is a step-by-step process to effective science
teaching and attempts to explore if teachers feel they have mastered this
process. This item and the next, item 5, show the lowest efficacy belief overall;
teacher efficacy decreased a full point in comparison to question 2 (3.48). Again,
TFA scored the highest value, with U second (3.36), and DI third (3.33). The Δ of
the programs is therefore 0.15 for this question. All programs showed a much
greater range of answers on this question: TFA from 2-4, DI from 2-4, and U from
1-4. This data suggests that some of the teachers surveyed have not yet
accessed a step-by-step methodological approach to teaching science, but the
majority has by scoring above 3.
The next item in the cluster, item 5, addresses classroom management in
the science lab and during hands-on investigations. This question shows again
low efficacy scores in comparison to the overall mean: TFA teachers have the
high score of 3.48, followed by DI (3.40), and U (3.27). The Δ of the programs is
0.21 for this question. TFA and DI show answer scores ranging from 2-5, U from
2-4. This suggests that both alternative programs teachers’ are comfortable with
hands-on science experiments in their classrooms. All programs scored above 3,

99
denoting that all teachers feel generally effective in their monitoring of
experiments.
Item 7 addresses perceived teaching ability, showing an average high
efficacy score of 4.07 from the DI teachers. This information makes the efficacy
for the item above the DI group average for all items (3.98). The U teachers
scored second on the item (3.91), and third TFA (3.80). Both TFA and U scored
below average on this item, but above the “Undecided” score of 3. The Δ of the
programs is 0.27 for this question. TFA and U teachers show the biggest range in
answers, from 2-5 for TFA and 1-5 for U. The scores for DI reach from 3-5. The
data suggests that overall the DI teachers are most convinced of their science
teaching ability, but all participants having scored above 3 all surveyed teachers
indicate the belief that they are effective.
Item 11 addresses teacher perceptions of subject mastery sufficient to
effective teaching. It contains both subject matter competency and teaching
ability components. This item shows a high average of 4.45 for U teachers. TFA
scored 4.40, and DI 4.13; thus, the Δ of the programs is 0.32 for this question,
which is bigger difference than all other items thus far. It is the second item to
score above average for all groups. TFA and DI range in answer scores from 2-5,
U from 4-5. With high scores well above 3 from all teachers in the three
programs, all of these programs seem to be recruiting candidates with a strong
academic background that are sure about their content knowledge, although DI

100
teachers show a slightly lower efficacy on this question when compared with U
and TFA.
The next item, 22, explored comfort level with the unexpected and with
students directing the flow of the lesson. This item addresses classroom
management and control issues in the classroom. All three groups scored well
above average: TFA the highest with 4.64, U second with 4.55, and DI third with
4.53. The Δ is 0.11, slightly above the 0.09 average. The answer range for TFA
and U on the item is from 4-5; DI shows a range of 1-5, with the 1 being a
possible misreading (only one participant answered 1, all others 4-5). Teachers
from all groups therefore seem comfortable with their ability to manage and share
control in their classroom by scoring not only above the “Undecided” score of 3,
but also above 4.
The last item in cluster 1, item 23, explores whether teachers feel they can
overcome motivational issues in their students. It addresses whether teachers
feel they have a variety of strategies to reach all learners and the background to
make subject matter relevant. This item shows below average scores for all
groups, with TFA scoring highest (3.88), followed by U scores (3.64), and DI
(3.60). The Δ for the item is therefore 0.28, the third largest difference of all
efficacy items. The answer score range for TFA and DI is 2-5, whereas U
teachers show a range of 2-4. The data suggests that although beginning
teachers sometimes feel at a loss about how to inspire unmotivated students— a

101
struggle often described as “disillusionment”— overall they are confident in their
abilities to interest students, as indicated by average scores above 3 for all
groups.
In summary, this cluster shows some items scoring higher than average
efficacy beliefs for all groups, and other items showing lower efficacy beliefs. The
programs reveal a wide range of answer choices for this cluster, but all scores
are above 3, suggesting strong efficacy beliefs. The teachers from alternative
programs scored highest on six out of the seven items, suggesting that they have
a higher perceived effectiveness of their teaching than their traditionally prepared
counterparts. The interview data for the first question will follow and elaborate on
reasons for this higher perceived effectiveness in alternatively credentialed
teachers.

Cluster 2: Items 16, 17, 18, and 21.
Cluster 2 items address the perceived challenges, difficulties, and
strengths of the beginning teachers, as well as the causes for them. Item 16
addresses the teachers’ abilities to explain hands-on learning and to make it
accessible to students. U teachers scored highest on this item (4.00), scoring
slightly above their average for all items. TFA scored 3.92, which constitutes
below average for the group, the same for the DI teachers, with a score of 3.87.
Therefore, the Δ of all programs is 0.13 for this question, a smaller range than on

102
most other items. The answer scores for TFA and U are in a range from 2-5, and
for DI from 3-5. This data either means that teachers stay within their safety zone
in their first year, conducting only experiments that they can explain very well, or
that they are all well-qualified and able to explain whatever they may encounter
in a science context. All groups showed a score above 3; thus the data reports
that the majority of teachers do not perceive difficulty in explaining experiments
to their students.
Item 17 tries to elicit teacher beliefs on how well prepared they feel in
subject matter competency. On this item, TFA and DI scored the high score of
4.40, whereas U teachers scored 4.18. The item shows an above average
efficacy belief for all groups, with a Δ of the programs of 0.22. The answer scores
for TFA and U are in the range of 4-5, for DI from 2-5. Therefore, a much greater
spectrum of efficacy is found in DI teachers’ efficacy beliefs concerning
answering student questions than is found in other programs, although the mean
is higher in DI than in U teachers. Again, this finding may speak to the perceived
subject matter competency of respective teachers. All teachers surveyed scored
above 4, and perceive the ability to answer student questions.
Item 18 asks the teachers to rate if they believe they have acquired the
skills for teaching science. The U teachers scored highest and above average on
this item (4.18), DI second and still above average (4.07); only TFA teachers
scored below average (3.88). The Δ for this item is therefore 0.20, an above

103
average difference, which is notable but maybe to be expected, as the alternative
programs do not provide methods instruction to their students prior to entering
the teaching profession, and this question specifically asked about the skills of
teaching science (methodology). The answer scores for U range from 3-5, for DI
and TFA from 2-5. This data suggests that the traditionally prepared teachers
perceive themselves as having acquired the skills to teach science at a higher
rate; however, all teachers believe they have these skills by scoring well above
the “Undecided” score of 3.
The last item, 21, tried to elicit whether the beginning teachers feel they
have a variety of strategies to explain science concepts. All groups scored above
average on this item. The U teachers scored the highest (4.09), TFA second
(4.08), and DI third (4.00). The Δ for the three groups is therefore 0.09, which
equals the average difference. The answer ranges for U and DI are 3-5, for TFA
2-5. This data suggests that U teachers feel they have slightly more instructional
strategies to meet students needs, but all groups feel efficacious on the number
of strategies, as shown by scoring above average and above the threshold of 3
(Undecided).
In summary, the U teachers scored strongest in this cluster of efficacy
questions. They scored highest in three out of four items, and are the only group
that has average scores of 4 or higher in their responses for each of these four
items. This data suggests that the traditionally prepared teachers may be more

104
perceptive and analytical about the reasons for their strengths and difficulties. In
their responses to these items it is suggested that they were able to develop a
deeper understanding of the pedagogy and methodology (items 16, 18 and 21) of
science teaching.

Cluster 3: Item 20.
Cluster 3 consists of only one item directly, item 20, as it asked beginning
teachers to rate their comfort level about being evaluated by an administrator.
This item shows the greatest Δ (0.68) of any items on the efficacy survey. DI
teachers scored highest (4.13) and above average, TFA teachers scored second
(4.08) and slightly above average, and U teachers scored third (3.45) and well
below average. The answer ranges for DI are from 3-5, for U 2-5, and for TFA 1-
5. Thus, the item not only shows the greatest intergroup difference, but also the
greatest intragroup difference in their answers. The U teachers in particular do
not perceive administrative visits to their classroom as beneficial as do the other
groups, and prefer to opt out at higher numbers if given the choice. The interview
data will be used to elaborate upon the possible sources of the low scores of the
traditional teachers on this item. Notably, however, all groups scored well above
3, showing that the majority does not explicitly feel uncomfortable about
administrative evaluation in their classroom.

105
Below is a summary table of answer data from the teacher efficacy part of
the survey and an accompanying graph. Detailed frequency tables and graphs
for each question can be found in the appendix. Notably, all groups on all items
scored above 3, showing that they all feel efficacious in their teaching for the
items included in this survey.
Table 4: Teacher Efficacy Survey Statistics Overview
Statistics
Group
Membership Q2 Q4 Q5 Q7 Q11 Q16 Q17 Q18 Q20 Q21 Q22 Q23
TFA N Valid 25 25 25 25 25 25 25 25 25 25 25 25
Missing 0 0 0 0 0 0 0 0 0 0 0 0
Mean 4.48 3.48 3.48 3.80 4.40 3.92 4.40 3.88 4.08 4.08 4.64 3.88
Std.
Deviation
.510 .586 .823 .816 .707 .862 .500 1.092 1.038 .759 .490 .781
DI N Valid 15 15 15 15 15 15 15 15 15 15 15 15
Missing 0 0 0 0 0 0 0 0 0 0 0 0
Mean 4.20 3.33 3.40 4.07 4.13 3.87 4.40 4.07 4.13 4.00 4.53 3.60
Std.
Deviation
.941 .724 .910 .799 .834 .640 .828 .961 .640 .535 1.060 .828
U N Valid 11 11 11 11 11 11 11 11 11 11 11 11
Missing 0 0 0 0 0 0 0 0 0 0 0 0
Mean 4.36 3.36 3.27 3.91 4.45 4.00 4.18 4.18 3.45 4.09 4.55 3.64
Std.
Deviation
.505 .924 .647 1.136 .522 .775 .405 .603 1.293 .539 .522 .674

106
Findings from the Interviews
The researcher conducted the interviews in May, June, and July 2008.
Two teachers from each program were selected, based on showing
representative survey scores for the respective program. The interviewed
teachers will be referred to by their program and number 1 or 2, e.g., TFA 1 and
DI 2. The interview had three parts: (1) three questions related to the teachers’
efficacy beliefs, which will be discussed in this part of the chapter, (2) three
questions related to their outcome expectancy beliefs, which will be addressed in
the discussion of RQ 1b, and (3) two questions describing their preparation
program, providing background information.
All six interviews were conducted and audiotaped by the researcher, then
transcribed and coded. The coding variables emerging from the efficacy
questions on the interview were (1) the level of perceived effectiveness as a
science teacher and supporting evidence for this effectiveness, (2) perceived
challenges and difficulties and reasons for these challenges and difficulties,
perceived strengths and reasons for these strengths, and (3) comfort level about
being evaluated by administrators.
Perceived Effectiveness and Supporting Evidence
This question was formulated to align with items 2, 4, 5, 7, 11, 22, and 23
on the survey to triangulate data and findings. Both candidates from the TFA
program state that they feel “quite effective” (TFA 1) or “relatively effective” (TFA

107
2). Interestingly, both teachers trace this belief about effectiveness to periodic
assessments (TFA 1) and to mastering standards on a checklist, as determined
by standardized tests (TFA 2).
Right now I think I am quite effective. In the beginning I was definitely not
very confident and I know that the first time I got any feedback was the
first periodic assessment through the district in ICS and Chemistry and I
didn’t know how effective I was until they took the test and then they
scored pretty well [...] and they are actually mastering the knowledge that
is required in order to do proficient or advanced on standardized exams.
(TFA 1, Interview, p. 1)

I’d say relatively effective because [...] I have this tracking system that I
use on Excel to sort of see which content standards each student is
mastering for each general unit and based on their percentage I can gage
how effective I am or if I have to go back and reteach. (TFA 2, Interview,
p. 1)

Both U teachers cite standardized tests as one part of evidence for their
performance, but the U teachers also include other indicators for their
performance. Both U teachers also feel effective and prepared.
I think surprisingly I am more effective and prepared than some of the
other first-year teachers in my department. When I come into my class, I
know what I am doing, I feel confident, I know my tools, and I feel at this
point that I know how to differentiate content [...]. I teach honors and
regular biology and I have been able to push my honors students so they
are not just passing the classes and getting ok grades, but I actually push
them to excel and they are now required to do more than just the basics.
And I think I have been really effective in pushing the higher end of the
students. In low achieving schools we focus so much on the kids that are
failing to figure out how we can prevent them from failing that we don’t do
anything for the kids that are doing ok. I feel like I really made a change to
push them to the next level also on CST, which would be Advanced. [...]
(U 2, Interview, p. 1-2)

108
[...]I know I am effective because they are very interested in the material
and accomplish the tasks I have assigned them, on which I spend a
considerable amount of time planning to ensure that it will be of interest to
them. I hear my students tell each other how much they enjoy my class,
and that enjoyment shows in their test scores. (U1, Interview, p.1)

This data suggests that the TFA teachers in the study have very high efficacy
beliefs, based on a single, easily measurable variable: standardized testing and
acquisition of standards. Notably, while the U teachers cite standardized test
performance as one indicator, they also use other variables to determine their
effectiveness, such as student interest, “pushing students to excel” (U 2), raising
student interest in the material, and making class enjoyable (U1), as well as
being confident in knowing how to differentiate material and knowing the “tools”
(U 2). Therefore, the qualitative data shows a greater depth of analysis and depth
of the efficacy beliefs in the U teachers.
Both DI teachers express doubts about their preparedness. One of them
also mentions standardized tests as an indicator of effectiveness.

Actually after I finished my first year I don’t know if I am that effective. I
have over 15 years teaching experience at a different level and I am very
effective at that. I never took a teaching class per se on how to teach
middle school and how to teach this population in particular. [...] I am
learning a lot on how to interact with this group of students and how to
better encourage them to do things that I ask. (DI 1, Interview, p. 1-2)

I don’t think I am terribly effective because this was my first year and they
tell us in the intern program the main thing you have to do in your first year
is survive and then you will never have another year like it. I have children

109
that are very challenging but I guess everybody says that. [...] I think for a
first year teacher though I have done pretty good. How do I know? Well, if
you have the discussion if benchmarks are at all an indicator on how well
they are doing, they did pretty well on their assessments. They did better
on the second semester than the first, and that is good, it shows
improvement. [...] students really like me, and I love the kids, and the fact
that I love the kids is also a strength. They are difficult, that doesn’t mean I
don’t love them. [...] I think you only get good work out of children who
love you or at least respect you with a certain rapport.
(DI 2, Interview, p. 1)

The data shows no connection to standardized assessments for DI 1, the only
teacher that did not use this measure as one indicator of effectiveness. The
second teacher uses benchmarks as an indicator, but emphasizes care and
concern (“love”) as being crucial to effective teaching as well.
Looking at the answers to interview question 1 may explain the generally
high efficacy score of TFA teachers on the surveys: If benchmarks are used as
the only measure to establish success and effectiveness in teaching, such
effectiveness becomes relatively easily measurable and maybe temporarily
easier to achieve. However, the less clear the definition of “effectiveness”
becomes, or the more variables that are used to observe such effectiveness, the
less convinced beginning teachers are of their own efficacy. Teacher U 2, for
example, uses 7 indicators for her effectiveness to answer question 1: her
preparedness, her confidence, knowing the tools, knowing how to differentiate,
pushing students to excel in the curriculum, pushing them to the next level on
CST, and classroom management. According to her survey responses, she has

110
one of the lower average efficacy scores in her program, and her score is lower
than any of the other interviewees (3.3 vs. 3.75 for U 1, 4.04 for TFA 1, 4.375 for
TFA 2, 3.7 for DI 1, and 3.54 for DI 2). In this case, a comparison of the interview
data and survey data suggest initially a discrepancy between the quantitative and
the qualitative data collected for the three programs. Possibly there is an
implication that the broader a beginning teacher sees his or her responsibility, the
more difficult it is for him or her to feel efficacious. The qualitative data generated
by the interviews with the U teachers show a greater depth and more variables in
their description of efficacy, and thus a greater sense of efficacy for the U
teachers. The surveys showed the alternative TFA program teachers as having
greater efficacy. This incongruence and conflict of quantitative and qualitative
data will be addressed in the discussion of this section.
Perceived Challenges, Difficulties, and Strengths, and Their Reasons
This question was formulated to find more reasons why teachers feel
effective or not, and more indicators that lead them to feel effective or not
effective. The question therefore aligns with survey items 2, 4, 5, 7, 11, 16, 17,
18, and 21. For this question, data responses divide into two categories. Most
interviewed beginning teachers see their difficulties and challenges as being
rooted in the school or the students (inherent in their situation and not giving
them much to do about it), the other two (TFA 2, and U 2) see them as personal
challenges (which they could address and overcome) and as more temporary in

111
nature. Below are some examples of the first category, in which beginning
teachers that see their difficulties and challenges as inherent in their situation,
and therefore without much possibility for change.
[Difficulties are] also the logistics of the school like kids being pulled out,
kids coming in, moving out and it is just really hard to keep everything
straight. [...] some difficulties are making that connection with the families
when it is so hard to contact home sometimes [...] (TFA 1, Interview, p. 2)

I think the challenges are multi-faceted but one is getting to do students to
do homework. My students come to school and they want to socialize, [...]
I think one of the other challenges is the textbook, which I use rarely, is
written way above the comprehension level [...] Most of them don’t use
higher levels of thinking. They are using low levels of thinking, kind of
reaction types of thinking, not why does this happen and how do you know
that this is happening, their logic is flawed.
My difficulties come with the parents. The parents have had such a difficult
life and they don’t want their kids to have a difficult life, so they make their
life so easy that they get whatever they want without having any
consequences for not doing things. I think that is really detrimental to
them. (DI 1, Interview, p. 3)

I was talking about the English learners already, that was a challenge,
plus I had interspersed in my classes with no rhyme or reason children
that should be in special ed., but they tell us at my school that most
classes are sheltered classes anyway, so I have made a survey through
looking at computer and cumulative files to see how many have IEPs and
how many are English learners, what their CELDT scores are, and things
like that, and I have data to back up my opinion. (DI 2, Interview, p. 2)

Their [students] basic skill levels in reading, writing, arithmetic, and
organization are very low and their attendance is problematic. (U 1,
Interview, p. 1)

This data reveals that the majority of the interviewed beginning science teachers
(four out of six) believed that (to some degree) their challenges lay outside their
circles of influence. For example, they cannot change the parents nor the

112
difficulty of contacting them, they cannot influence the English learner status of
their students, their lack of skills in English and or mathematics, or their poor
work, study, or homework habits. Only the interviewed teachers TFA 2 and U2
saw the challenges as completely within her area of influence, as described in
the next paragraph.
The interviewed teachers TFA 2 and U 2 saw the opportunity for change in
something they can pursue in the classroom.
I think my challenge has been so far trying to instill in them that work habit
that they haven’t got so far, and keep it up, even though the other
teachers have given up. In my class, for example, I give homework at
least four times a week, which is almost every day and I have a lot of 9
th

graders that just came from middle school and they are used to not
passing their classes and still being able to graduate or go on to high
school. (U 2, Interview, p. 3-4)

Challenges for me right now are behavior management more than
anything because I know if you have difficulty managing your classes it is
difficult to get real learning across. I’d say that my strength would be my
ability to scaffold and differentiate material and I think if I would be a little
better at behavior management and dealing with different kids and
learning modalities I would be a lot more effective with my scaffolding.
(TFA 2, Interview, p. 2)

Both teachers above state challenges they can influence, instilling work
habits in their students and behavior management in their classroom.
Additionally, five out of six interviewed teachers see some challenges as their
own, which they can overcome, and which are therefore influenced by them, like
the two examples above. Below are examples of such challenges and difficulties.

113
Definitely a challenge right now is the time commitment, going to school
full-time, three times a week after school. And also having so many
observations by your peers and also what you need for your credential, as
well as TFA requires us to be part of this professional learning community
at least two times every month. So it is just all stacked together it is really
hard, on top of grading, lesson planning as a first year teacher with
nothing that you have to go back to. (TFA 1, Interview, p. 3)

Apart from the English learners and special ed. population, causing some
of my strengths and weaknesses is the same thing, that I am a pushover. I
am already thinking about it for next year, being more consistent with my
classroom control. (DI 2, Interview, p. 1)

There are some students whose imaginations and interests I have great
difficulties capturing. [...] It is perhaps because my assignments do not
intrigue them as much as the other half of my classes or because I do
present my assignments in a clear enough manner that is accessible to
them. If I can get these students to understand and become curious of the
material and assignments, I believe this great hurdle can be overcome
and all students will learn effectively. (U 1, Interview, p. 1)

I think my challenge is more a personal challenge, and that is for me to be
optimistic and keep my spirits up because first year teachers go through
this disillusionment phase, and you need to step out of it sometimes and
do something you look forward to. [...] I think the challenge for me is not
the day-to-day struggles but personal because I want so much for my
students and I have very high expectations and I want them to excel and
not just settle for mediocrity so it’s the battle between you wanting them to
do better and me pushing them versus them just doing what they are used
to doing and feeling frustrated. I think what causes my strength and
difficulties are my own high expectations for all my students. (U 2,
Interview, p. 2-3)

Apart from teacher DI 1, the beginning teachers show varying degrees of
perceived challenges, which could influence their efficacy.
Though all six teachers are in urban, low-performing schools serving a
diverse student population, the answers of the two U program teachers suggest
that they reflect more deeply on the challenges within themselves than do the

114
other interviewed teachers. This is apparent in their statements about failing to
capture student interest in the material and in assignments, possibly not being
clear enough, having high expectations and not settling for mediocrity, and
wanting a lot for the students. Such assertions show inner conflict about not
being as successful as they would like to be at the end of their first year. In
contrast, teacher TFA 2 provides a rather simplistic notion that behavior
management is the only challenge to overcome, and TFA 1 talks mainly about
the time commitment as an issue. DI 1 looks at the preexisting pattern of bad
work habits and low parent involvement, as well as students not engaged in high
levels of thinking. Teacher DI 2 cites the demographics of the school as a main
challenge, whereas her own inconsistency is the only “changeable” factor
contributing to challenges.
When describing strengths that could possibly influence efficacy, TFA 1
refers mainly to personal attributes: “Some strengths are my ability to multitask
and keep a clear, focused agenda” (Interview, p. 1). TFA 2 identifies a teaching
skill: “I’d say that my strength would be my ability to scaffold and differentiate
material” (Interview, p. 1). Noticeably, neither TFA teacher talks about his or her
subject matter competency, a strong factor for selection in the TFA program,
while DI 1 and U 2 clearly address this strength: “I think my strengths are the
subject, that I know the subject, that I am comfortable with the subject and I am
comfortable teaching” (DI 1, Interview, p. 1); and, “I have always known that I

115
know my science but there is a difference between what I know and what I can
deliver” (U 2, Interview, p. 1). Other strengths listed are high expectations for
students and an unwillingness to settle (U 2), and a good rapport with the
students (DI 2). In survey items corroborating the data to this interview question,
the traditionally prepared teachers (U teachers) scored higher than the
alternatively prepared teachers, possibly suggesting a more reflective teaching
practice and a more holistic view of their skills. The interviews are congruent with
this, showing examples of how much more analytically and reflectively the U
teachers approach their classroom teaching. This is brought forth in the data
responses, as described above. While U1 and U 2 did not score highest overall
in their quantitative responses, they were able to identify more variables and
reflect more deeply in their qualitative interview data, providing a greater degree
of evidence for efficacy and level of analysis in their work. Again, an
inconsistency is apparent between the quantitative and qualitative data.

Comfort Level of Being Evaluated by Administrator
The third and last question pertaining to efficacy on the survey aligned
mainly with item 20 on the survey, but also addressed underlying beliefs about
how well the teachers are performing in their classroom at this point. All
interviewed teachers from the alternative programs (TFA and DI) feel

116
“comfortable” or “very comfortable”. This conclusion is the result of many
observations and evaluations they have experienced in their first year.
As a new teacher at my school we are all supposed to be viewed by three
administrators and I have had three administrators come to my room at
least two times, so I am actually feeling very comfortable. (TFA 1,
Interview, p. 1)

I am really comfortable. I feel like I have been observed so much since I
started teaching that it’s like I don’t really care who comes in, it’s just
another pair of eyes watching me. (TFA 2, Interview, p. 1)

I have no problem. I have been stulled for over 15 years. Not necessarily
under that term, but evaluated. (DI 1, Interview, p. 2)

I am not nervous about it at all; I have been stulled twice, you know, first
semester and second semester, and I am not nervous about it. (DI 2,
Interview, p. 2)

Data from the U interviews suggests that evaluations are seen more as a
disruption and not helpful, and therefore not valuable. This is congruent with the
low scores of the U teachers on survey item 20, and gives some reasons behind
the low score of the traditionally prepared teachers. However, only U 1 states not
being comfortable with the process for this reason. U 2 states that although the
process can be disruptive, she is comfortable with being evaluated. Below are
some factors given by U 1 for her attitudes towards evaluation.
Because my assistant principal is distant and does not communicate well
with me, I never feel comfortable being stulled/ evaluated by her. She has
observed me twice, and each time she left halfway through the period and
did not discuss with me what her thoughts were. Besides greeting each
other once in a while if we happen to bump into each other in the hallway,
we do not exchange words often at all. I do not feel that she serves a
purpose for me, and I am completely unaware of anything constructive
that she does for me. (U 1, Interview, p.1-2)

117

The data may suggest that both U teachers, but especially U 1, seem more
concerned with the purpose of the process than with simply stating an opinion
about it. If the stull process is not perceived as helpful, then it is a disruption, and
thus a teacher may not feel comfortable about it. Possibly the U teachers, and
especially teacher U 1, expect more of a connected and integrated evaluation
(“does not communicate”, “did not discuss her thoughts”, “left halfway through the
period”) in order for the process to be more meaningful in their own further
professional growth. A gain for the teacher is missing from the process, and
therefore the administrator “serves no purpose for me”. This is a very interesting
interpretation, as teacher U 1 is apparently expecting a useful and constructive
dialogue about her teaching to help her development, and is not getting this
feedback.

Summary and Interpretation of Findings RQ 1 a
The data collected to answer research question 1 a consisted of surveys
and interviews given to first-year teachers of science serving students in a large
Southern California area. The communities served by these teachers are
considered high-diversity and/or low-income. Both surveys and interviews were
conducted in order to measure the beginning teachers’ sense of efficacy.

118
Teacher efficacy is a term to describe a teacher’s judgment about his/her
ability to positively affect student outcomes, especially with students that are
considered difficult to teach. Literature relates this efficacy belief to students’ own
sense of efficacy (Anderson et al., 1988), motivation (Midgley et al., 1989), and
perhaps most importantly, to student achievement (as early as Armor et al.,
1976). A high sense of efficacy will also positively impact a teacher’s effort,
commitment, persistence, and resiliency (studies cited in Tschannen-Moran &
Woolfolk-Hoy, 2001). Teacher efficacy is related to student achievement, most
likely through a causal chain effect, with evidence for this chain published by
McLaughlin and Marsh (1978). These researchers found that teacher efficacy
influences teacher behavior, which influences student efficacy and behavior, and
through this chain impacts student outcomes. Ashton and Webb (1986) built
further on this chain by providing evidence that general teaching efficacy was
related to math scores, while personal teaching efficacy impacted language
scores through teachers’ instructional practices.
The hypotheses for this part of the study were that a measurable
difference in teacher self-efficacy among beginning alternatively and traditionally
prepared teachers exists, with traditionally prepared teachers showing greater
self-efficacy. Further, the hypothesis was that this difference will be observable in
the quantitative surveys as well as in the qualitative interviews. This hypothesis is
supported by existing research by Darling-Hammond et al. (2002), which

119
examined how prepared teachers from different preparation programs feel when
entering the profession. They found that traditionally prepared teachers feel more
prepared and show a higher efficacy than do teachers that did not go through a
traditional program.
However, the quantitative data collected suggests that the teachers from
alternative programs (TFA and DI) felt a greater sense of efficacy than the
traditionally prepared teachers from the U program in two out of three clusters.
While this may be related to the difference in the sample size of the populations,
it can also be interpreted that the interview responses by U teachers had more
depth and showed a greater insight into their teaching, suggesting that they
actually had a greater understanding and sense of efficacy than solicited by the
surveys. Additionally, the traditionally prepared teachers show higher efficacy
when it comes to analyzing the reasons for their difficulties and their strengths in
their practice, and the interviews showed a more reflective approach in the
U teachers’ answers.
As discussed above, the surveys suggest that the highest efficacy beliefs
were held by TFA teachers; however, the interviews did not support this
conclusion. Alternative teachers also defined teacher success and effectiveness
as based on standardized test performance alone. In contrast, interviews with the
two U teachers showed a deeper level of reflection and impact and the teachers
looked for a broader definition of success. This focus on test scores within the

120
TFA program is emphasized and supported by studies such as one by
Glazerman et al. (2005) that cites an increase in mathematics test scores and no
impact on reading scores when students are taught by TFA corps members, and
solicits this finding alone to state that TFA is indeed reducing inequities in
education.
Different perceptions on the profession became clear when teachers
responded to being evaluated by an administrator. The two U teachers did not
perceive this process as helpful, seeing it as a distraction or a disturbance to
their routine, and therefore stated they were not as comfortable with it as were
the other teachers. The alternative teachers did not analyze the stull evaluation in
this way, and thus felt more comfortable. This response also led them to score
higher on the survey than the U teachers. However, given the reasons identified
by the U teachers, it is possible that this outcome can be attributed to a sampling
error, due the small sample population of 11 teachers in the survey, and 2 in the
interviews. It may also be influenced by differences in the developmental stages
of the teachers: As the U teachers had gone through a period of formal student
teaching and experienced formal supervision by an expert in their content area
and in pedagogy; they may have a different perspective on observation and
analysis of classroom teaching. Further research is therefore needed in this area
to confirm to reject these findings.

121
In conclusion, the quantitative data shows a higher efficacy belief for the
alternatively prepared teachers than the U teachers on two of three clusters, but
the survey scores are not supported by the interview findings. This result may be
rooted in deeper levels of reflection and feelings of responsibility by the U
teachers, as expressed in the interviews, and it may be congruent with
Goldhaber and Brewer’s (2001) conclusion that further studies are needed in
order to define who is being “screened out” (p. 84) of the teaching profession by
the hurdle of licensure. Alternative teachers from the TFA program demonstrate
a strong focus and success in raising test scores, but do not show a broader view
of successful education for urban students. Especially in the case of science
teachers, high achieving undergraduate science majors may not feel that they
want to teach for a long time. Thus, a short commitment such as TFA fulfills their
need, offering the opportunity to attend graduate school afterwards on a
scholarship. By defining effective teaching on test achievement alone, alternative
teachers view their responsibility as helping students excel on standardized
measures and thus dispelling the label of low performance. Traditionally
prepared teachers focus on their students’ success in their studies overall,
making them interested in the subject matter, and turning them into resourceful,
college-bound students with positive study and work habits.
Thus, the hypotheses of the traditionally prepared teachers showing
greater efficacy has to be partially rejected, based on the survey data, but it

122
seems supported by the interview data. Further studies including bigger sample
populations and a revised interview protocol to solicit more data on this matter
are needed to explore the impact of credentialing route on self-efficacy.

Data Collected for RQ 1b: How do differently prepared (traditional vs.
alternative) beginning, urban, secondary science teachers compare
in their outcome expectancy?

This research question focused on establishing whether there is a
difference in outcome expectancy beliefs of the studied population according to
its preparation format. In order to answer this question, survey data was
collected, and interviews were administered with three different preparation
programs: one university-based, traditional program, whose beginning teachers
had experienced pedagogical coursework and student teaching before entering
as first-year teachers; and two alternative programs, whose beginning teachers
had no or minimal preparation before entering the profession, but received
pedagogical coursework while they are teaching. The surveys were administered
in late spring (April to May 2008), and the interviews were conducted between
May and July 2008.

123
Findings from the Surveys
The survey used was a modified version of the STEBI instrument,
developed and validated by Riggs and Enochs (1989). From this survey,
questions 1, 3, 6, 8, 9, 10, 12, 13, 14, 15, 19, and 24 are used to establish
outcome expectancy. The items are clustered in three groups to align with
interview questions pertaining to outcome expectancy beliefs.

Cluster 1: Item 1, 3, 10, 14, and 15.
The items in cluster 1 attribute student success in science to the teacher
in some general way, or to specific teacher attributes, methods, or teaching
practice. Item 1 addresses the notion of responsibility for student success, which
can be attributed to either the teacher’s efforts in general in the classroom, or to
other outside factors. Although teachers from the alternative TFA program were
scoring highest (3.68) on this item, this question actually shows a below average
3

belief for TFA in comparison to the other outcome expectancy items. The Δ
between the three programs is 0.41 for this question. All programs show answer

3
Note: As established above in Chapter 4 Introduction, “the average (mean) for an answer
regarding efficacy for U is 3.95, for DI is 3.98, and for TFA is 4.04[...]” Accordingly, efficacy scores
are ranked as averages, above or below average for the purpose of this analysis.

124
scores
4
ranging from 1-5, therefore showing a high intragroup variance. Of the
three programs, the DI group shows the lowest mean for this question (3.27).
Mean answer scores above 3 for all groups suggest that all beginning teachers
attribute student success to teachers’ efforts in the classroom.
Item 3 again addresses the cause of student success, this time linking it to
effectiveness in teaching approach, as opposed to general teacher efforts
(item 1). This question shows an above average score for all programs, with TFA
and U both scoring highest (4.00), and DI scoring 3.73, resulting in a Δ of 0.27,
the second lowest intergroup difference on the outcome expectancy survey. The
answers for TFA and U are in the range from 3-5, for DI from 2-5. The data
suggests that the new teachers from all groups believe an effective teaching
approach is an important factor in student success with a mean score above 3.
Item 10 addresses the factor of attention in the classroom as a
contributing factor in student progress. All groups scored above average, with U
scoring highest with an average of 3.91, TFA second with 3.84, and DI with 3.40.
The Δ for this item is 0.51, showing below average intergroup variance. The
answer scores for U ranged from 3-5, for TFA and DI from 2-5. The data
therefore suggests that teachers from all three groups, but especially the U and

4
The answer score range states which answers teachers from this group gave (1= Strongly
Agree, 2= Agree, 3= Undecided, 4= Disagree, 5= Strongly Disagree).

125
TFA teachers, see teacher attention as an important factor in student progress,
scoring a mean above 3.
Item 14 elicits teacher beliefs about teacher effectiveness being the
determining (not merely a contributing) factor in student achievement. All groups
show an above average score for this item, with TFA scoring highest at 3.80, U
second at 3.45, and DI third at 3.27. This data results in a Δ for the groups of
0.53. The intergroup variance is therefore less than average for this item. Within
the groups, the answers for all groups ranged from 2-5, showing a great
intergroup variance in scores. Teachers from all groups attribute student
achievement to effectiveness in science teaching by scoring an average above 3.
The last cluster item, 15, explores teacher beliefs about the link of student
interest in science and the teacher. Only TFA scored below their average, but still
the highest at 3.64; U scored second and above average at 3.55; DI third and
above average at 3.40. These numbers result in the lowest intergroup variance
for outcome expectancy with a Δ of 0.24, well below average. TFA showed an
answer score range of 2-5, DI and U a range of 2-4. The data therefore suggests
that teachers from all groups link student interest in science to the teacher, as all
groups scored above 3.
In summary, the alternative program, TFA, scored highest in its` teachers’
outcome expectancy beliefs on four out of five items. All items show above
average scores for outcome expectancy from all programs and they show a

126
greater difference Δ in the answer choices among the groups. This information
suggests that all beginning teachers will attribute student success to themselves
in some way, but there is a discrepancy in how much responsibility for student
failure they attribute to themselves. Thus, they accept the responsibility for their
students’ science achievement in varying degrees, with the TFA teachers
claiming responsibility at the highest rates. The interview question will elaborate
on this finding.

Cluster 2: Item 6, 12, and 19.
The items in this cluster are negatively phrased, examining the cause for
low achievement. Each item uses some form of the word “achievement.” Item 6
is a negatively worded version of item 3, and tries to find the cause for
underachievement in science. All groups scored noticeably lower, with U scoring
above their average on this item (3.55); TFA and DI below their average, with
3.36 for TFA, and 2.60 for DI. The Δ for this item is 0.95, which is above average.
The answer score range for TFA is 2-5, and 1-5 for U and for DI. Accordingly,
beginning teachers trace success to teacher effectiveness to a higher degree
than they assign responsibility for underachievement of students to the
effectiveness of the teacher. TFA and U both scored above 3, which signifies
“Undecided”, thus attributing underachievement in general to the ineffectiveness

127
of the teacher. The average score of DI teachers, however, did not attribute
teacher inefficiency to underachievement.
Item 12 questions the teachers’ beliefs about whether teacher effort is a
contributing factor in student achievement. The question shows inter- and
intragroup variance: U scored highest and above their average at 3.64; TFA
scored second and below their average at 3.60; DI scored third and well below
their average at 2.60. The resulting Δ is therefore 1.04. The answer scores for U
ranged from 2-5; for TFA from 1-5; and, for DI from 1-4. The data shows that TFA
and U teachers hold a greater belief that teacher effort matters in students’
achievement, as both scored above 3; DI teachers do not agree to this, scoring
below 3. This is notable, as it produces one of the lowest scores on the survey,
suggesting that the majority of DI teachers do not contribute student achievement
to teacher effort.
Item 19 investigates teacher beliefs about effective teaching having an
influence on students with low motivation in raising achievement. All groups
scored above average, with TFA scoring the highest again at 4.04, U second at
3.55, and DI third at 3.07 and just above average (3.06). The resulting Δ of 0.97
shows an above average intergroup variance. Within the three groups, a large
range of answer choices can be observed as well: TFA and U range from 2-5, DI
from 1-5. With all groups scoring above 3, teachers from all groups believe that

128
teachers are able to positively influence students with low motivation, although
TFA teachers believe in this influence at a much greater rate.
In summary, the groups showed a great range of answer scores when
asked to link low achievement to teacher or classroom practices. While the
beginning teachers saw the teacher as responsible for student success, they
traced student failure to themselves less readily. The traditional U program
teachers are willing to accept responsibility for low achievement at a higher rate,
as indicated by scoring highest on two out of three items in this cluster. The
second interview question addressing outcome expectancy will elicit responses
related to the perceived causes for low student achievement in science from the
beginning teachers.

Cluster 3: Item 8, 9, 13, and 24.
Cluster 3 for outcome expectancy items on the survey addresses the
notion that good teaching or the teachers themselves can overcome all other
factors in making students successful in science. Two items are positively
worded, and two are negatively worded. Item 8 tries to draw out beliefs about the
degree to which a student’s background determines his or her academic
achievement, or whether background can be overcome by good teaching (notice
the wording “good” instead of “effective,” with “good” a much broader term).
Scores on this positively worded item were high and above average. TFA

129
teachers scored highest (4.20) and above average, U second with 3.73 also
above average, and DI third with 3.20 also above average, for an above average
Δ of 1.00. The answers for all groups ranged from 2-5, thus showing great inter-
and intragroup variance. As indicated by the above 3 scores, the data suggests
that more teachers from all programs believe that good teaching is more
important than student background as determinants of academic success.
Item 9 contained the word “blame” in connection with low science
achievement; the item listed “teachers,” not effective teaching, and used the term
“generally,” thus making it unspecific and negative. The item scored well below
average for outcome expectancy for all groups. TFA scored highest, at 3.20,
followed by DI with 2.40, and U with 2.36. The resulting Δ is an above average
0.84. The answer range for TFA was 2-5, and for DI and U, 1-5, again showing
great inter- and intragroup variance. Only TFA teachers assigned blame to
teachers by scoring above 3; both DI and U did not agree to this general
statement.
The next item, item 13, used the word “general” and “teacher,” with other
non-specific descriptors, such as attention, effort, or effectiveness. The item
shows the second greatest intergroup difference (Δ = 1.09), with TFA scoring
highest and above their average at 4.04, DI second and above their average at
3.27, and U third and much below their average with 2.91. The answer scores
ranged for TFA from 3-5, for DI from 2-5, and for U from 1-5, thus showing a

130
great intragroup variance for DI and U as well. The data shows that, by scoring
above 3, both TFA and DI attribute achievement more generally to the teacher,
whereas U teachers are less likely to do so.
Item 24 shows the greatest intergroup variance of any item, with a Δ of
1.17. Its intent was to draw out the general connection between teachers with
good abilities and success for all students (note the word “kids” instead of
“students” in the wording of the item). TFA scored highest in this negatively
worded item, but below its group average at 3.64. U scored second, much below
average at 2.64, and DI third and below average at 2.47. Within the TFA group,
answers ranged from 2-5, within DI and U from 1-4. This result shows large
intragroup variance again. Only TFA teachers scored on average above 3, thus
establishing the connection of a good teacher’s ability to reach all students.
Teachers from both other programs do not believe that good teachers can help
all students.
In summary, this cluster shows the highest score of the TFA prepared
teachers on all 4 items, suggesting that TFA teachers are more readily linking the
cause for students’ success and achievement to the teacher. The positively
worded items, items 8 and 13, showed an above average score, and the
negatively worded items, 9 and 24, a below average score. Therefore, this
cluster shows a great range of answer scores from the programs, and most

131
beginning teachers do not subscribe to the notion of being able to bring success
to all students, especially not when the items were negatively worded.
Below is a summary table of answer data from the outcome expectancy
part of the survey, and a graphical representation of the outcome expectancy
data. Detailed frequency tables and graphs for each question can be found in the
appendix. It should be noted that the difference between programs for each item
is much greater for the teachers’ outcome expectancy beliefs than for self-
efficacy (average Δ of 0.69, vs. average Δ of 0.09 for self-efficacy). Also, both the
DI and U programs show scores below 3 (“Undecided”) on some items, thus
indicating that the majority of their teachers do not have a high outcome
expectancy belief on these items. For DI, these items are 6, 9, 12, and 24; for U
they are 9, 13, and 24. Teachers from the TFA program scored above 3 on all
items.
When comparing the OE Scale and the TE Scale, it is notable that the
majority of answers for the survey items are found in the higher numbered
answer choices. It is also notable that the DI answers show a more even
distribution, noting a more scattered response pattern. The DI teachers show a
great range in their choices on outcome expectancy items.

132
Table 5: Outcome Expectancy Survey Statistics Overview
Statistics
Group
Membership Q1 Q3 Q6 Q8 Q9 Q10 Q12 Q13 Q14 Q15 Q19 Q24
TFA N Valid 25 25 25 25 25 25 25 25 25 25 25 25
Missing 0 0 0 0 0 0 0 0 0 0 0 0
Mean 3.68 4.00 3.36 4.20 3.20 3.84 3.60 4.04 3.80 3.64 4.04 3.64
Std.
Deviation
.988 .289 .907 .707 1.000 .800 .816 .611 .645 .700 .611 1.036
DI N Valid 15 15 15 15 15 15 15 15 15 15 15 15
Missing 0 0 0 0 0 0 0 0 0 0 0 0
Mean 3.27 3.73 2.60 3.20 2.40 3.40 2.60 3.27 3.27 3.40 3.07 2.47
Std.
Deviation
1.100 .961 1.121 1.014 1.056 1.056 1.121 .884 1.033 .737 1.100 .834
U N Valid
11 11 11 11 11 11 11 11 11 11 11 11
Missing 0 0 0 0 0 0 0 0 0 0 0 0
Mean 3.45 4.00 3.55 3.73 2.36 3.91 3.64 2.91 3.45 3.55 3.55 2.64
Std.
Deviation
1.128 .447 1.128 .905 1.120 .701 1.206 .944 .934 .688 .820 .924

Findings from the Interviews
The researcher conducted the interviews in May, June, and July 2008 with
the teachers selected as described above. The interviewed teachers will be
referred to by their program and number 1 or 2, e.g., TFA 1 and DI 2, as
introduced earlier. The interview had three parts: three questions about the

133
teachers’ efficacy beliefs, previously discussed, three questions about their
outcome expectancy beliefs, which will be discussed here, and two questions
about their program, in order to provide background information.
The themes for coding the outcome expectancy questions on the interview
were (1) perceived factors that help students learn science, (2) perceived causes
of why students are not successful in science, and (3) the interviewed teachers’
opinions about whether good teachers can always help all students learn science
successfully.

Perceived Factors Helping Students to Learn Science and Explanation
The coding aligns this question with outcome expectancy survey items 1,
3, 10, 14, and 15. The following interviews show that the majority of the
interviewed teachers perceive that the factors helping students learn science are
located in the classroom or in something that the teacher does or can do.
I think it is really just breaking it down and scaffolding the learning. There’s
so many intricate and confusing, complicated things in science I guess,
that are based on prior knowledge. And the first thing I always do is find
out what prior knowledge they know and especially in ICS class because
most of it is review of middle school physical and life science. [...]
Identifying the key words, [...] as what we’ll be using in the unit and
making a long term plan of what we will be doing in this unit so the kids
know where they are going. (TFA 1, Interview, p. 2)

I think hands-on activities, things that are engaging because the more they
are able to deal with models and manipulate things, like visual pictures or
making models, I think it is a lot easier for them to process especially
when it is a really abstract concept. And also just practicing applying real-

134
life every day situations to classroom science problems in general really
helps them. (TFA 2, Interview, p. 2)

I think that you somehow have to make it real to them. Why do they need
this, you know. In this age of computers and the internet, you can go to
the internet and look up everything really quick. They don’t know why they
should learn biology, why they should learn anything about mitosis. You
have to somehow make it relevant and have them buy into it in some way.
You have to reach them in that way. (DI 2, Interview, p. 2)

I think the most important first step is to put the academic material into a
context that is meaningful for the students. This captures their attention,
their imagination, and the thinking and the questions and the desire to
explore their curiosities begin. (U 1, Interview, p. 2)

From the data it is clear that these four teachers believe there is something they
do or can do in the classroom that will help the students learn science. They
describe methods and strategies they use, and the reasons why they think these
strategies are important. This belief will positively impact their outcome
expectancy, as they have an influence on the factors described. The other two
teachers (DI 1 and U 2) describe the factors that are more innate and less in their
control as teachers: interest (DI 1) and curiosity (U 2). They perceive that these
qualities either lie inside the students or they do not: the students either are
interested and curious or not, not as a trait they can foster. Data supporting these
points is presented below. The first quote has a negative connotation, stating a
lack of interest as the reason for a lack of success in science, whereas the
second quote focuses on the students possessing a quality that makes them
stand out positively in class.

135
I think the kids have to be interested in the subject [...] I just didn’t find one
student out of my 125 students that were interested. And I don’t know
what they were interested in, because I tried to figure out what aspects of
science they could be interested in. [...] I think that they need to be more
interested in it [science]. (DI 1, Interview, p. 2)

Curiosity. I think that is the factor that makes some of my students stand
out in my classes, especially in my honors classes. Science is not about
someone just teaching you the material and you absorbing it but it is more
about you wanting to find out why things are the way they are and you are
asking questions. Also, asking yourself how to challenge an idea and
putting me on the spot. Sometimes kids do that. [...] The gift of curiosity is
definitely a factor that pushes kids to find answers for their own questions.
The desire to learn and figure it out. Especially in science that makes
some kids stand out. (U 2, Interview, p. 3)

It can be interpreted from this data that DI 1 has a much more negative
perception and focuses on the weaknesses and what the students are lacking,
while U 2 expresses a positive view, focusing on the assets some students bring
to class. This sentiment differs significantly from the notion that the teacher is
responsible for making work interesting or making the students curious, which is
expressed by DI 2 and U 1 in the next section (2) on causes for students not
being successful.

136
Perceived Causes for Students Not Being Successful in Science
This second question coding theme aligns with outcome expectancy
survey items 6, 12, and 19. This interview question is negatively worded, as in
the survey items, and asks the teachers to describe what, in their opinion,
hinders student success. When worded negatively, only teachers DI 2 and U 1
attribute the cause of the underachievement of students in science to some
degree with the teacher.
Well, if they are not successful in science there are a lot of reasons. You
haven’t made it real or valuable to them in any way or related it to anything
they know. Besides the difficulties they have as English learners, and the
children that have IEPs and other issues, I think you have to make it
interesting, but that can be hard, because a lot of kids are not interested in
anything besides their family, friends, boyfriend or girlfriend and their
videogames. (DI 2, Interview, p. 2)

I think the biggest hurdle for successful learning in science is the failure to
capture the students’ imagination, which can be caused by unclear
directions by the teacher, low basic skills of students, poor home support,
and various other distractions/obstacles. If students do not feel that
science is interesting or important to their lives, they will not be motivated
to learn it. (U 1, Interview, p. 2)

Although both teachers also cite factors outside their influence, such as
“kids are not interested” (DI 2), or “low basic skills, poor home support” (U 1),
they see at least some cause with the teacher. These responses contrast to the
other four interviewed teachers, who perceive a lack of reviewing (TFA 1), poor
attendance (TFA 1), low math skills (TFA 2), previous failures (TFA 2, U 2), lack
of interest (DI 1), or students’ general preoccupation with non-academic issues

137
(DI 1) as the main causes for unsuccessful science students. Even though
several teachers are reporting outside factors as obstacles to student success,
they were optimistic in their ability to overcome these factors, as discussed in the
next section.

Opinion Whether Good Teachers Can Always Help All Students Learn
Science Successfully
This interview question aligns especially with the general items 8, 9, 13,
and 24 from the survey. It was purposely kept general in order to draw out
possible polarizations in opinions. The interviewed teachers show the complete
range of expected answers, as shown below.

Strong positive opinions, thus high outcome expectancy:

Yes, I do. It doesn’t matter what subject you teach, if you have some sort
of structure in your classroom and you have a really effective engaging
way of teaching and capturing the students interest I think students can do
well in science. (TFA 2, Interview, p. 2)

I think a good teacher will do everything at his disposal to ensure that all
students will be curious about learning science, motivating them to learn,
but also ensure that students of various backgrounds have access to
learning the material. A good teacher takes into account the background
of his students when creating assignments. A good lesson plan involves
an engaging assignment as well as good management of classroom
behavior of students and assistance for students who have certain needs.
This is the crux of all the effort that teachers need to focus on, and
therefore, yes, more teacher effort into science teaching will produce
higher student achievement. A low academic background can be
overcome by good teaching, which entails good management and
activities that engage all students. There are various methods teachers

138
can use to overcome the vastly diverse academic backgrounds of his
students. (U 1, Interview, p. 2)

Positive Opinion:
I guess the word successfully is something each teacher would interpret
for themselves but for me I had some students that came in and said “Mr, I
can’t do science, I have never been good in science, I hate science and I
just hope I can get a C in the class”, and what I would like to do for them is
not just introduce them to the science but to what the science can do for
them, and I can get them engaged, I find that successful. [...] I think it is
possible. [...] And I think it comes down to management and instructional
strategy, not necessarily materials and equipment. (TFA 1, Interview, p. 2)

Somewhat negative opinions and negative opinions, thus low outcome
expectancy:

I would like to think so, but I don’t know if it is just me being realistic, but I
have learned that I could be the best teacher I can possibly be, that
everybody raves about, but I am not god, and I am not going to be able to
save everyone and reach out to everyone. I would like to be able to do so,
but I know that there are some students that I will connect better with than
others and I think it is ridiculous of me to think that good teachers can
always help all students to learn science. [...] (U 2, Interview, p. 4)

I think good teachers can be fabulous teachers and depending on what
the students have going on in their lives and what their interests are. I am
trying to figure out what I am not interested in, because I like a lot of
things. Whatever I am not interested in, I can have the most fabulous
teacher but I am not interested, I don’t care. So, a teacher can be great,
the best one in the nation, and not going to affect that student. (DI 1,
Interview, p. 3)

I am going to say no. I think a good teacher will continue to try, and do
everything in their arsenal that they can think of to help them learn. [...] (DI
2, Interview, p. 3)

139
The data gathered for this question shows a wide range of responses for
the teachers’ beliefs on outcome expectancy. The positive answers focus mainly
on what the teacher can do in the classroom, implying that if the teacher fulfills
this action, success will come to all students. The negative answers range from
failing to reach every student, due to not “connecting” with all students-- and
therefore representing an issue between student and teacher (U 2)— to
uninterested students and the teacher unable to do anything about this fact, no
matter what (DI 1 and 2).
A point of discussion needs to be the definition of “good teacher”: is a
good teacher not one that has success with all students? It seems that all
interviewed teachers operate on the notion that they indeed are good teachers,
but at the same time they are not necessarily able to help all students. This is
clear from quotes such as “I can get them engaged” (TFA 1), and “I could be the
best teacher I can possibly be... and I am not going to be able to save everyone”
(DI 1). The most reflective and theoretically grounded answer was provided by U
1, again emphasizing the fact of a higher degree of reflection and the ability to
connect theory and practice: The teacher mentions ensuring that students are
curious, making curriculum accessible to all students, using student background
to develop assignments, good management, assistance for students with special
needs, and engaging activities along with varying methods to overcome the
challenges of different student backgrounds. In comparison, the other teachers

140
giving specific reasons for success focus on less variables in their answers:
structure and engagement (TFA 2), and engagement, management and strategy
(TFA 1).
Once more, this is at least a partial contradiction to the clear conclusions
drawn from the quantitative data, as it suggests one TFA and one U teacher as
having the greatest positive outcome expectancy belief, and both DI teachers as
well as one U teacher as having the lowest or most negative beliefs. Therefore,
this either suggests that the outcome expectancy of beginning science teachers
does not correspond with their preparation program or that the sampling of
interviewed teachers led to inconclusive results, and a sampling error resulted.

Summary and Interpretation of Findings RQ 1 b
Data collected to answer research question 1 b was comprised of surveys
and interviews given to first-year teachers of science serving students in a large
Southern California area with high-diversity and/or low-income communities.
Both surveys and interviews were conducted to measure the beginning teachers’
outcome expectancy beliefs. The student outcome expectancy questions are
based on the perception of the locus of control, asking teachers if they believe
that they truly have the influence to impact student learning or if they see this as
outside their area of influence.

141
The hypotheses for this part of the study were the existence of a
measurable difference in teacher outcome expectancy among beginning
alternatively and traditionally prepared teachers, with traditionally prepared
teachers showing higher outcome expectancy. This difference was expected to
be observed in the quantitative surveys as well as in the qualitative interviews.
While the U program teachers scored higher than the alternative DI
teachers, they scored lower than the alternative TFA teachers on the survey
items. Thus, the hyotheses are not supported by the quantitative survey data.
The U teachers scored higher on the negatively worded survey items, thus they
appear more ready to accept responsibility in the case of students not being
successful. At this point, it is important to look at the recruitment pool of the three
programs to possible make more sense of this discrepancy and of the high
scores of the TFA teachers.
The TFA program is specifically recruiting teachers for two years into the
profession that are demonstrating leadership through perseverance, ability to
influence and motivate others, and understanding of and desire to work towards
vision of TFA (all children have the opportunity to attain an excellent education).
It is explicitly stated that applicants need to be “promising leaders” and have a
desire to “eliminate educational inequity” (Retrieved on 02/14/2009 from
http://www.teachforamerica.org/mission/index.htm). This specific recruitment goal
may skew the quantitative survey data especially, as TFA recruits are exposed to

142
an intensive one week preparation in which the goals of TFA are strongly
emphasized. However, it seems that the interviews expose a lack of depth of
reflection and understanding of the components needed to eliminate educational
inequity, resulting in less variables in their answer choices when compared to the
U teachers, who showed a richer theoretical background in their interview
answers.
The survey data further shows much lower average outcome expectancy
scores than efficacy scores; TFA teachers on average scored 0.21 points lower,
U teachers 0.59 points lower, and DI teachers 0.92 points lower. This data
suggests that while teachers perceive themselves as effective, they also have
doubts about the extent of their influence on student achievement. The
qualitative data also does not show a correlation to preparation route. On some
items, one interviewed teacher from a program showed greater outcome
expectancy in comparison to the other programs, but the second interviewed
teacher showed much lower outcome expectancy. Therefore, the hypotheses
cannot be supported by either data sample.
According to Bandura (1977), performance accomplishments are the
major source contributing to outcome expectancy. The discrepant current
findings are congruent with findings by Raudenbush and Stephen (1990), who
concluded that teacher preparation did not have an impact after controlling for
student engagement.

143
However, the incongruent findings in the current study are incompatible
with the very clear findings of Darling-Hammond (2001, 2004) and Darling-
Hammond et al. (2002), which state that teachers from quick entry programs
such as TFA are generally less successful with urban students. Another study
done by Laczko-Kerr and Berliner (2002) also contradicts the current study’s
findings, having found traditionally prepared teachers to be more effective with
students than alternatively credentialed teachers.
Notably, the current study did not measure student outcomes, relying
instead on self-reported teacher efficacy and outcome expectancy as a
measurement of teaching success. As perceptions do not always reflect actuality,
this may be one of the reasons for the inconsistency between existing studies
and the current study. Another reason discussed above may be that the current
study’s focus on perception of the teachers may skew the data in favor of the
TFA program, which recruits candidates with strong convictions that they can
change the inequalities in urban education.

144
Data Collected for RQ 2: After the first year of teaching, how do the
retention rates of this teacher population from traditional and
alternative programs compare?

This research question reviewed the program retention data for urban
science teachers. In order to answer this question, three data sources were
used. The first data source consists of surveyed first-year teachers themselves,
identifying their intentions about staying in urban teaching or teaching on the
survey cover. The individual programs also reported data on their (expected)
retention of first-year teachers to the primary researcher, which is the second of
the data sources presented below. A third data source is published data about
the retention of science teachers in California and nationally. This research data
was reviewed and compared to the generated data.

Findings from the Self-Reported Retention Intention on the Survey Cover
For the purpose of establishing intention (commitment) to urban teaching,
the survey cover sheet included a question with several answer choices
regarding retention expectations. In order to analyze this data quantitatively, the
following values were attached to the answer choices by the researcher: “I will
continue teaching in an urban setting as long as I am able” = 4; “I will continue

145
teaching in a different setting as long as I am able” = 3; “I will continue teaching
until something better comes along”=2; and, “I am planning to leave teaching as
soon as possible” = 1. One surveyed teacher in each program chose not to
answer, or checked more than one answer choice, and they are listed in a
separate column as “Not Valid/ No Answer.” An average was established for
each program from all valid answer choices. The data was converted to
percentages and is shown below in Table 6 and in Figure 7. The raw data (not
converted) is shown in Appendix C.
The programs show a difference in their average of Δ = 0.4. In the TFA
program data, all four answer choices were present, with two people expressing
that they wanted to leave the profession as soon as possible. In the DI and U
programs, teachers did not express the wish to leave the profession as soon as
possible. Only the first three answer choices were checked. The data shows an
overall strong commitment to the teaching profession in all three groups, as the
mean scores were above 3 out of 4. However, in the DI program, this information
does not necessarily indicate a commitment to urban teaching, as more surveyed
teachers answered that they will continue teaching in a different setting or until
something better comes along (8 teachers) than teachers expressing that they
will continue teaching in an urban setting (6 teachers). In the other programs,
most teachers chose the answer that they will continue teaching in an urban

146
setting than all of the other answers combined. When averaged, the U program’s
teachers show the highest commitment score to urban teaching.
It has to be said that due to the small population sample of U teachers,
and the discrepancy of sample sizes between the programs, the data may
present a skewed picture of retention. However, as all participants from each
program were included in this data collection, the data provides an accurate
picture of the expected (self-reported) retention in these programs for this year.
For further analysis, the data was converted to percentages instead of raw
data numbers. The purpose was to calculate the percentage of each group that
chose each option. The table and graph below show this data. Data presented as
percentages highlights an observable difference between the programs,
In this survey, teachers from the U program are 25% more likely to think
of staying in urban teaching as long as they are able than are DI teachers, and
13% more likely than TFA teachers. However, all programs show that the
majority of their teachers consider staying in urban teaching. One third, or 33% of
TFA teachers, sees teaching as a temporary solution, whereas less than one
quarter (22%) of U teachers think this way. It is notable that while TFA only asks
its corps members to commit to teaching for two years so many (54%) consider
urban teaching for their future, as indicated by their selection of choice 1: “I will
continue teaching in an urban setting as long as I am able.”

147
Table 6: Percentage of Teachers Choosing Each Self-Reported Option
4 3 2 1
TFA 54% 4% 33% 8%
DI 42% 29% 29% 0%
U 67% 11% 22% 0%

Figure 4: Percentage of Teachers Choosing Each Self-Reported Option
0%
10%
20%
30%
40%
50%
60%
70%
Percentage of
Group
4 3 2 1
Answer Choice
Percentage of Self-Reported Intention per Answer Choice
TFA
DI
U

Findings from the Reported Retention for Each Program
Each program was contacted in July to determine how many of its first
year teachers would be returning to urban teaching, moving to a different setting,
or leaving the profession. As this data was collected in July, these numbers may
change, but when reported they resembled the most accurate account. However,
the data represents a program- reported (subjective) measure of retention. The

148
Table 7 and Figure 8 below show the programs’ reported data for the school year
2008/2009. Notably, the sample size is larger than the surveys, as some
teachers did not participate in the survey.
According to this data, the DI program reported 100% retention from year
1 to year 2 for its 36 science teachers. This is to be expected, as the teachers
form this program have not yet earned their credential, but need to complete a
second year of teaching in order to receive a California Preliminary Teaching
Credential. The U program reported 3 of its 12 (25%) teachers moving to a
suburban teaching position, and TFA reported 2 teachers out of 39 (5 %) moving
to another area with intentions to continue teaching; they are therefore
considered “movers” as well. Both, the U program and TFA program movers
align with answer choice 3 on the self-reported survey (“Will continue teaching in
another setting”). The U teachers were already in possession of a Preliminary
Teaching Credential as well as a Masters’ Degree in Education at the end of the
first year, while the TFA teachers did not possess a Preliminary Credential yet,
like the DI teachers.
The lowest percentage for retention after the first year is shown by
teachers from the U program, followed by TFA. As of early August 2008, the
programs still reported these numbers as accurate reflections of their retention of
beginning science teachers. . A possible reason for the high retention in the
alternative programs is the fact that in order to receive a California Preliminary

149
Teaching Credential, a second program year is needed. Therefore, leaving the
profession after the first year is less likely, while the traditional teachers do not
only possess a credential already, but also a graduate degree (M.Ed.) at the end
of their first year.
This data shows an inconsistency with the data from the first data source,
the survey collection. This creates a discrepancy in the collected data, and again
further research is needed, along with a greater and possibly longitudinal
comparative study of retention in alternative and traditional programs in order to
establish a possible long-term effect of credentialing route on retention
Table 7: Program Reported Retention of Science Teachers Entering their
Second Year in 2008/2009
Total Staying Urban Moving Leaving
TFA 39 (100%) 37 (97%) 2 (3%) 0
DI 36 (100%) 36 (100%) 0 (0%) 0
U 12 (100%) 9 (75%) 3 (25%) 0

150
Figure 5: Program Reported Retention of Science Teachers Entering their
Second Year in 2008/2009 in Percent
0%
20%
40%
60%
80%
100%
Percentage of
Total Group
Staying Urban Moving Leaving
Retention and Attrition
Reported Program Retention in Percent for 2008/2009
School Year
TFA
DI
U

Findings from National and California State Retention Data for Teachers,
Secondary Teachers, and Teachers of Science
National retention data for this comparison was gathered from the Teacher
Follow-Up Survey, presented in a report from the National Center for Education
Statistics (2007). The retention data for the state of California was gathered from
the NTAR data presented in a study by Reed, Rueben, and Barbour (2006). As
not all data segregates “movers” from “leavers”, for this part of the data analysis
the data of movers and leavers will be combined. This may result in a higher

151
attrition rate than if the data were segregated. Table 8 and Figure 9 represent the
national, state, and program data in percentages.
The comparison data shows both alternative programs retaining science
teachers at a higher rate than the national average, as well as higher than the
California secondary teacher average, with the DI program showing 100%
retention. The U program’s retention of teachers in urban settings seems
problematic, as it is below any other retention rate. This data may be due to
several factors, including the very small sample population: the U program had
12 first-year teachers, versus 39 in TFA, and 36 in the DI program. Further, as
the U program offers a preliminary credential and a masters’ degree in education
at the end of this first year, the teachers completing this program may have
different employment opportunities already, while the alternative programs’
teachers are still working towards their credential.
It has to be noted that neither the national nor state data is taking the
school setting (urban, high minority enrollment) into consideration. This may
contribute to a difference in the data, as research (Johnson et al, 2005: National
Data, Reed et al., 2006: CA Data) has shown a lower retention rate in urban
schools serving communities with high minority-enrollment and/or low socio-
economic status. All teachers in the current study populations serve in such
schools. The reason for using the TFS and NTAR data presented is the
segregation by subject area: This allows a comparison of science teachers only.

152
Table 8: Retention Data of Teachers in Comparison, Percentage

Data Source Stayers Movers and Leavers Combined
All Teachers 83.5% 16.5%
2004-2005 (TFS)
Natural Sciences 88.5% 11.5%
2004-2005 (TFS)
All Teachers 94% 6%
California 2001-2002
Secondary Science 77.3% 22.7%
California 2001-2002
TFA Program 97% 3%
2008-2009
DI Program 100% 0%
2008-2009
U Program 75% 25%
2008-2009

Figure 6: Retention Data of Teachers in Comparison, in Percent
0.00%
20.00%
40.00%
60.00%
80.00%
100.00%
Stayers Movers and Leavers
Combined
Comparison of Published and Generated Retention Data, Stayers vs.
Movers/Leavers, Percentages
2004-2005 Total TFS
2004-2005 Natural Sciences TFS
California All Teachers 01/02
California Secondary Teachers 01/02
TFA Program 08/09
DI Program 08/09
U Program 08/09

153
Summary and Interpretation of Findings RQ 2
Certain aspects of teacher education programs seem to equip graduates
to be more successful in urban, high-poverty, and low-performing settings,
according to Darling-Hammond (2006). Preparing teachers that are effective in
producing high student achievement, are able to teach a diversity of students,
and who will persist through the challenges of teaching as a professional career
without losing the desire to reflect continually on and improve their practices are
some of the desired outcomes of teacher preparation, according to these
findings. As it seems warranted to hire only well-qualified teachers by studies
linking teacher qualification to high student outcomes, still many hard-to-staff
schools are having trouble recruiting well-qualified candidates for their open
positions. Therefore, alternative credentialing provides a pool of candidates that
is willing to teach in these settings. The literature on alternative credentialing is
polarized and entrenched, as some strongly advocate against the harmful
consequences of hiring alternatively qualified teachers (Darling-Hammond 2006),
whereas others find positive outcomes in programs such as Teach for America
(TFA), whose sense of mission is to reform and equalize education while
achieving high student outcomes (Glazerman, 2005, on TFA).
The current study examined retention data from the surveyed teachers,
reported retention by the programs, and for comparative measure, national, and
California data. The hypothesis for this study was that the retention of

154
traditionally prepared teachers in urban teaching will be greater at the end of the
first year. This hypothesis was influenced by the intention of TFA to provide
teachers for only two years, and retention not being an expressed focus. In the U
program, however, it is a focus to provide dedicated teachers to urban teaching
in order to promote equal access to high quality teachers in these schools over
time. This hypothesis was grounded in research studies by Kirby, Berends, and
Naftel (1999), Shen (1997), and a dissertation study conducted by Inglie (2007)
in Florida.
This hypothesis was not conclusively supported by the data collected. In
fact, the U program lost 25% of teachers when they moved into suburban
teaching assignments. However, the intention, as stated on the survey by the
first-year teachers themselves, shows a much greater rate of U teachers
expecting to stay in urban teaching as long as they are able (67% of U teachers,
vs. 54% of TFA and 42% of district intern teachers). Therefore, it may be
possible that over a longer time, the retention in urban teaching of U teachers
may be higher. The fact that the survey self-reporting on retention was
anonymous may increase its reliability, whereas the programs’ reported retention
may be subjected to each program’s goals.
The higher loss of teachers from the U program after the first year may be
due to the program’s award of a preliminary credential and a masters’ degree in
education. This higher degree may be a factor in securing a higher-paying

155
teaching position in a less challenging setting. An encouraging fact to be noticed
in the retention data is that no first-year teachers left the profession, but rather
moved to different teaching assignments. The U program shows a retention rate
comparable to the secondary teacher data from California. More research with
greater data samples of alternative and traditional teachers specifically in urban
schools serving a high percentage of minority students are needed. Possibly
longitudinal studies beyond the first year are needed to explore this issue, and
studies that disaggregate retention data by subject area, school community
served and years of teaching experience will further the understanding of the
complexities of teacher retention.

156
Chapter V: Summary, Conclusions, and Implications
Summary
The presented study examined the potential difference in efficacy,
outcome expectancy, and retention of beginning urban science teachers. The
issue of alternative credentialing is discussed in current research literature, and
the findings on the performance of alternatively prepared teachers are disputed
(Lazcko-Kerr & Berliner, 2002). However, the current issues of difficult-to-staff
urban schools, continuing high teacher turnover, and recent budget cuts to
education in California seem indicators that alternative credentialing will continue
to be practiced in an effort to fill teaching positions in California. Therefore, more
research in the area of possible impacts of alternative credentialing is warranted.
This study used a mixed-methods and two-tiered approach to establish
possible differences in efficacy and outcome expectancy between alternative and
traditional program first-year science teachers. All teachers in the three studied
programs were given a 5-point Likert scale survey. The total N was 51, 40
alternative and 11 traditional program teachers. Interviews were conducted with
two teachers from each program, which showed representative scores according
to their surveys, therefore resulting in an N of 6 first-year teachers: 4 from
alternative programs and 2 from a traditional program.
Analyzed findings suggest that the alternative teachers have higher
efficacy beliefs in their survey responses, but a narrower view of teaching

157
success, effectiveness, and their role as educators as expressed in their
interviews. Alternatively credentialed teachers from the TFA program had an
average score of 4.04 for teacher self-efficacy (TE), and 3.75 for outcome
expectancy (OE) on the surveys. Alternatively credentialed DI teachers had an
average score of 3.98 and 3.06, respectively. The teachers from the TFA
program therefore showed higher scores in both areas. Traditionally prepared
teachers from the U program had average scores of 3.95 (TE) and 3.39 (OE).
The quantitative survey data thus suggests that the U teachers have the lowest
self-efficacy beliefs, and their outcome expectancy belief expectations are
between the TFA and DI teachers.
However, the interview data revealed that the TFA teachers used
benchmark standardized assessment as the main variable to establish their own
efficacy, and their outcome expectancy is greatly influenced by standardized test
results (formal classroom exams, periodic assessments and CST scores) as well.
If a student scoring “Proficient” on standardized assessment is a successful
student, then the teacher can view themselves as having the greatest impact on
achievement. Conversely, if teachers do not perceive themselves as successful
unless they are able to turn their students into college-bound and resourceful
individuals, they may not perceive themselves as similarly efficacious. This is
expressed in the interviews with the U teachers especially, and the number of

158
variables they provided in the interviews, determining teacher success in their
view.
This data did not conclusively support the hypotheses of the researcher
that traditionally prepared teachers would exhibit higher efficacy and outcome
expectancy beliefs. As the sample population in general, and especially of
traditional teachers, was small, descriptive statistics were used to compare the
gathered data. The small population also may have caused the kurtosis and
skewness in the survey data of traditionally prepared teachers. More research is
needed to explore self-efficacy and outcome expectancy in beginning teachers
comparatively, including a greater sample population especially of traditionally
prepared teachers, in order to examine this research question further and
establish whether the credentialing route has an impact on self-efficacy and
outcome expectancy or if it indeed does not.
The findings in the current study are congruent with some of the findings
of Goldhaber and Brewer (2001), which suggested that (some) alternative
credentialing routes recruit differently motivated individuals which otherwise may
have been screened out of the profession by traditional ways into the profession.
This difference in recruitment may have influenced the data in the current study
greatly, as TFA attempts to recruit promising leaders, which have already
established a record of achievement and are therefore mostly very self-confident
individuals. Therefore, these teachers may show high scores on the survey

159
items, which did not require a greater depth of analysis. However, in the
interviews, the grounding of the traditional U teachers in educational theory
became evident, as they were able to support their efficacy beliefs and analyze
them in much greater depth than their alternatively prepared counterparts.
The retention data analyzed consisted of data collected from the surveys
and from the programs, as well as national and state data gathered from
published sources: the Teacher Follow-Up Survey (TFS) from 2004/ 2005, and
the Public Policy Institute of California (Reed et al, 2006). In comparison to
published sources, including national and state data, the traditional U teachers
had the lowest retention rate between the first and second year of teaching as
reported by their program. However, more teachers (67%) in the traditional
program expressed plans to stay in urban teaching as long as they are able,
suggesting that a more longitudinal study might possibly reverse these trends.
Further, the fact of credentialing has to be considered as well when
interpreting the retention data: only the traditional U teachers were already in
possession of a teaching credential and a Masters’ degree, while the alternative
candidates still have to work towards their credential in their second year of
teaching. This potential reversal of outcomes in a longitudinal study is also
possible because the TFA program specifies an expected two-year commitment
from its teachers. This may suggests a need for further research with greater

160
sample populations from each group, but especially from traditionally prepared
teachers, and a possibly longitudinal study to establish long-term retention.

Recommendations and Implications
The data gathered in this study provides empirical evidence for some
observable differences in efficacy and outcome expectancy in urban science
teachers from alternative and traditional programs. Clearly, the district intern
teachers show lower efficacy and outcome expectancy than TFA teachers, which
establishes the possibility of great differences between different alternative
programs. Data generated in the current study regarding retention shows two
different trends: surveys of traditional U teachers showed a stronger commitment
to urban teaching, but they also showed the lowest comparative retention in
urban teaching settings after completion of their first year.
Research studies regarding efficacy development suggest that
collaboration and supervisory attention is a major factor in the development of
high efficacy beliefs in beginning teachers (Chester & Beaudin, 1996). The
findings in the current study are compatible with this finding, as TFA alternative
credentialed teachers receive the most collaboration and supervisory hours
during their first year of teaching. Besides having to fulfill the same evaluatory
and professional development procedures at their school site, they also meet
three times each week for their pedagogy coursework, and for a six hour

161
collaborative meeting twice a month on Saturday. Further, they are observed
regularly, at least once a week by other TFA teachers and the TFA program
supervisors.
DI teachers attend only one meeting each week, lasting four hours. There
are no regular observations by DI personnel for beginning DI teachers apart from
the school site evaluation process. The traditionally prepared teachers only met
for their masters’ course requirement on a weekly basis for approximately four
hours. They are observed, although not as frequently as during their student
teaching period in the previous year. They also take part in BTSA, as they hold
preliminary credentials and California requires an enrollment in an induction
program for these teachers. The alternative teachers are not part of BTSA. Thus,
taking into consideration only the collaboration hours and supervisory support
received during the first year of teaching, the higher efficacy and outcome
expectancy of TFA teachers may be explained, in accordance with the Chester
and Beaudin (1996) and similar studies.
Therefore, this data suggests that the TFA program provides an example
of an alternative program whose teachers exhibit high efficacy and outcome
expectancy than teachers from the DI program. The comparative findings of TFA
and U teachers show a discrepancy, as the quantitative data suggested a higher
efficacy and outcome expectancy for TFA, but the interviews showed a higher
efficacy and outcome expectancy from the U teachers based on their greater

162
differentiation of variables and greater reflection. Further, the data suggests that
BTSA support is not equivalent to the support provided by TFA in the study’s
sample population, and the BTSA program may be well advised to consult the
TFA program to determine how its support could potentially be modified to result
in greater teacher self-efficacy and outcome expectancy. However, as mentioned
in a literature review by the Harvard Graduate School of Education in 2005, a
comment about perceived efficacy and actual efficacy is necessary:
It is important to acknowledge that there is a difference between a
teacher’s sense of efficacy and her actual efficacy, as reflected in various
formal or informal measures of student performance. Some teachers may
believe they are effective when they are not; others may doubt their
success with students, even though available measures of student
performance suggest that they are effective. However, given that few
teachers are dismissed before achieving tenure, it is a teacher’s own
sense of efficacy—presumably informed by evidence of students’
performance—that figures into retention. (p. 21)

The current study collected self-reported data through surveys and interviews,
and due to its scope did not connect self-reported scores with actual classroom
observations. While it is possible to believe one is effective, standardized
observation protocols may show a different picture. Thus it is possible that the
discrepancy between perception and reality caused the discrepancy between
quantitative and qualitative data found in this study.
Schools looking to fill science vacancies can use the data collected in this
study to make important hiring decisions, keeping in mind that TFA does not
require a long-term commitment above two years from its teachers. In

163
accordance with this structure, more TFA teachers (33%) showed that they view
teaching as a temporary solution. Thus, if schools choose to fill vacancies with
TFA teachers, they will be more likely to hire a new teacher with high efficacy
and outcome expectancy perceptions, but will most likely have to rehire for the
same position in two years.
Educational policy makers can use this data in order to make some
important decisions regarding the direction of education, perhaps examining
definitions of success for beginning teachers and the focus on test scores
emphasized by TFA teachers as the definition of effectiveness in teaching. If this
focus is indeed to be encouraged, as research studies suggest (e.g., Goertz &
Duffy, 2003), components of TFA teacher recruitment and preparation should be
incorporated into teacher education state-wide. Indeed, it appears that high-
stakes testing in science is an increasingly important factor in determining school
and teacher success and effectiveness as evidenced by the decision to have the
school year (2007/2008) be the first year in California that 8th Grade science
scores on the CST assessment would be factored into the API ranking of a
school. This focus on standardized testing is in contrast to previous movements
supported by findings of Rodger Bybee (2004), which emphasized inquiry in
science teaching and more authentic assessments, and promoted a move away
from traditional test-based approaches. Legislators should play a role in this
decision of the future emphasis in science teaching, as they decide whether to

164
continue the practice of alternative entry into the teaching profession in California
or not, and consider the status of high-stakes testing in science.
Professional development programs in urban schools may be advised by
the collected data to conduct further investigations into the reasons for new
teacher attrition, and, together with district administrators develop programs to
stop this attrition. The data presented showed that even after the first year not all
teachers are retained in the urban settings into which they were hired, and the
commitment to urban teaching is not well established, even for teachers from a
traditional teacher preparation program. Studies by Ingersoll (2001), which
explain some factors for attrition, need to be taken into more serious
consideration if districts wish to retain teachers at higher rates than were found in
this study.

Suggested Areas for Further Research
A possible area of further research arising from this study is a larger study,
including a bigger population size for surveys and interviews, and the
involvement of more teacher preparation programs of secondary science
teachers. This is crucial, as the current study found an inconsistency between the
quantitative and qualitative data. This research can either be conducted in the
same Southern California urban area, or could examine data from different
areas. This further study would also help to establish statistical significance by

165
using a larger population, and provide different levels of analysis for gathered
data; the same modified survey and interview questions can be used in order to
collect this data, or they can be modified to prevent the same inconsistencies
between quantitative and qualitative data found in the current study.
Another area of future research would be to expand this study into a
longitudinal study, following beginning teachers over several years, surveying
and interviewing them at different times in their teaching careers. This extended
study could establish the longer term effects of alternatively prepared science
teachers, and give a better impression of their long-term retention. The study can
also include an observation protocol to examine the differences between
perceived and actual self-efficacy and outcome expectancy.
Different research directions may be an efficacy study for secondary
teachers of subject areas other than science. Similar to science teachers,
teachers of mathematics and special education show high attrition rates, and
may be important to study. In urban low-performing schools, all teachers,
including veteran teachers, may be interesting to study, regarding their efficacy
and outcome expectancy beliefs. This work could be done in a comparison study
of teachers from low and high performing schools, trying to establish further
connections of efficacy, outcome expectancy, and student performance in
California.

166
Finally, more research regarding the factors for high attrition and low
commitment of urban science teachers is necessary, as the current study
demonstrates high attrition to be a fact in urban education through a 25% moving
rate from urban to suburban teaching in the traditional program. In order to
explore attrition factors, longitudinal studies of newly hired teachers are needed,
which would include exit surveys and/or interviews of teachers leaving urban
science positions and soliciting their reasons for doing so. These studies may
give urban school districts insight into what they need to change in order to retain
urban science teachers in higher numbers and thus to build an experienced core
of teachers.

Conclusion
The current study compared teachers from different credentialing routes in
their self-efficacy, outcome expectancy, and retention. This study was
accomplished through survey and interview data collection and analysis of two
alternative and one traditional teacher preparation program in a large, urban
Southern California area.
The survey data showed a higher teacher efficacy belief in both alternative
programs when compared to the traditional program, and a higher outcome
expectancy belief in one alternative program when compared with the traditional
program. The interview data offered insight that this higher efficacy belief may be

167
due to a narrower focus on standardized test achievement and on what effective
teaching entails, thus resulting in higher self-efficacy of the traditionally prepared
teachers due to more variables found in the interview. Further, the traditionally
prepared teachers showed a deeper level of reflection about their classroom
practice. The reported retention data of the alternative programs shows a higher
retention from year 1 to year 2 of their teaching candidates when compared to
the traditional university program. However, in the surveys the traditional
program’s teachers expressed a higher long-term commitment to urban teaching.
They stated at the highest percentage that they would be staying in urban
teaching as long as they are able. In comparison to published state and national
retention data, the data collected in this study shows that the two alternative
programs did better in retaining teachers than the comparative data, and the
traditional program did worse.
In conclusion, an effect on self-efficacy or outcome expectancy cannot be
conclusively traced back to the credentialing route, nor can the credentialing
route be established as a definitive factor on retention; however, a positive
impact of one alternative program over the other alternative program in teacher
efficacy beliefs in the short-term retention of beginning urban science teachers
can be observed.

168

References

Abdal-Haqq, I. (1998). Professional development schools: Weighing the
evidence. Thousand Oaks, CA: Corwin Press.

Allen, M. (2003). Eight questions on teacher preparation: What does the research
say? A summary of findings. Education Commission of the States, Full
text retrieved on 01/15/2008 from:
http://www.ecs.org/ecsmain.asp?page=/html/educationIssues/teachingqua
lity/tpreport/index.asp.

Amobi, F. (2007) Making it real: Promoting novice teachers’ critical perspectives
on educational issues through structured debates. Journal of Teaching
and Learning, 4(2), 2006. 418 05.

Anderson, R., Greene, M., & Loewen, P. (1988). Relationships among teachers’
and students’ thinking skills, sense of efficacy, and student achievement.
Alberta Journal of Educational Research, 34,148-165.

Anthony, T., & Kritsonis, W. (2007). A mixed methods assessment of the
effectiveness of strategic ementoring in improving the self-efficacy and
persistence (or retention) of alternatively certified novice teachers within
an inner city school district. Doctoral Forum: National Journal For
Publishing and Mentoring Doctoral Student Research, 4(1),

Armor, D., Conry-Osegura, P., Cox, M., King, N., McDonnell, L., Pascal, A., et al.
(1976). Analysis of the school preferred reading program in selected Los
Angeles minority schools. Santa Monica, CA: RAND Corporation.

Ashton, P., & Webb, R. (1986). Making a difference: Teacher’s sense of efficacy
and student achievement. New York: Longman.

August, D., & Hakuta, K. (1997). Improving schooling for language-minority
children. Washington, DC: National Academies Press.

169
Bandura, A. (1977, March). Self-Efficacy: Toward a unifying theory of behavioral
change. Psychological Review, 84(2), 191-215.

Bandura, A. (1986). Social foundations of thought and action: A social cognitive
theory. Englewood Cliffs, NJ: Prentice Hall.

Bandura, A. (1997). Self-Efficacy: The exercise of control. New York: Freeman.

Bartell, C. (2005). Cultivating high quality teaching through induction and
mentoring. Thousand Oaks, CA: Corwin Press.

Begle, E. (1979). Critical variables in mathematics education. Washington, DC:
Mathematical Association of American and National Council of
Teachers of Mathematics.

Bennett, C. (2001, Summer). Genres of research in multicultural education.
Review of Educational Research, 71, 171-217.

Berliner, D., & Biddle, B. (1995). The manufactured crisis. New York: Perseus
Books.

Berry, B., Hoke, M., & Hirsch, E. (2004, May). NCLB: HIGHLY QUALIFIED
TEACHERS- The search for highly qualified teachers. Phi Delta
Kappan, 85(9).

Bey, T., & Holmes, T. (Eds.). (1992). Mentoring: Contemporary principles and
issues. Collected Works. Reston, VA: Association of Teacher
Educators.

Boccia, J. (1991, April). Beginning teachers speak out: A study of professional
concerns in the first three years of teaching. Part II, elementary
teachers. Paper presented at the Annual Meeting of the American
Educational Research Association. Chicago, IL.

Brouwer, N., & Korthagen, F. (2005). Can teacher education make a difference?
American Educational Research Journal, 42(1), 153-224.

Bullough, R.V., Jr. (1989). First-year teacher: A case study. New York: Teachers
College Press.

170
Bybee, R. (2004). “Scientific Inquiry and the Nature of Science: Implications for
Teaching, Learning and Teacher Education:. In: Scientific Inquiry
and the Nature of Science, Flick and Lederman (eds), p.1-14.
Springer Books, Netherlands.

Campbell, D., & Fiske, D. (1959, March). Convergent and discriminant validation
by the multitrait- multimethod matrix. Psychological Bulletin, 56(2),
81-105.

Carlson, H. (1999, October). From practice to theory: A social constructivist
approach to teacher education. Teachers and Teaching, 5(2), 203-
218.

Cattell, R. (1966, April). The screen test for the number of factors. Multivariate
Behavioral Research, 1(2), 245- 276.

Chester, M., & Beaudin, B. (1996). Efficacy beliefs of newly hired teachers in
urban schools. American Educational Research Journal, 33(1),
233-257.

Chin, E., & Young, J. (2007). A person-oriented approach to characterizing
beginning teachers in alternative certification programs.
Educational Researcher, 36(2), 74-83..

Corcoran, E. (1981). Transition shock: The beginning teacher's paradox. Journal
of Teacher Education, 32(3), 19-23.

Corley, E. (1998, October), First-year teachers: Strangers in strange lands.
Paper presented at the annual meeting of the Midwestern
Educational Research Association Chicago, IL.

Cummins, J. (2000). Language, power and pedagogy. Tonawanda, NY:
Multilingual Matters Publishing.

Darling-Hammond, L. (1997). The right to learn: A blueprint for creating schools
that work. Somerset, NJ: Jossey Bass.

Darling-Hammond, L. (2000). How teacher education matters. Journal of Teacher
Education, 51(3), 166-173.

171
Darling-Hammond, L. (2004). Inequality and the right to learn: Access to qualified
teachers in California’s public schools. Teachers College Record,
106(10), 1936-1966.

Darling-Hammond, L. (2006). Securing the right to learn: Policy and practice for
powerful teaching and learning. Educational Researcher, 35(17),
13-24.

Darling-Hammond, L., Chung, R., & Frelow, F. (2002). Variation in teacher
preparation. Journal of Teacher Education, 53(4), 286-302.

Darling-Hammond, L., & Sykes, G. (2003). Wanted: A national teacher supply
policy for education: The right way to meet the "Highly Qualified
Teacher" challenge.[Electronic version]. Education Policy Analysis
Archives, 11(33).
Day, H.P. (1959). Attitude changes of beginning teachers after initial teaching
experience. Journal of Teacher Education, 10(3), 326-328.
Decker, P. T., Mayer, D. P., & Glazerman, S. (2004). The effects of Teach for
America on students: Findings from a national evaluation.
Mathematica Policy Research. Cambridge, MA.

Delpit, L. (2006). Lessons from Teachers. Journal of Teacher Education, 57(3),
220-231.
Dewey, J. (1933). How we think: A restatement of the relation of reflective
thinking to the educative process. Boston: DC Heath and Company.
Druva, C. A., & Anderson, R. D. (1983). Science teacher characteristics by
teacher behavior and by student outcome: A meta-analysis of research.
Journal of Research in Science Teaching, 20(5), 467-479.
Earley, P. (2000). Finding the culprit: Federal policy and teacher education.
Educational Policy, 14(1), 25-39.
Escamilla, K., Shannon, S., Carlos, S., & Garcia, J.(2003). Breaking the code:
Colorado’s defeat of the anti-bilingual education initiative
(Amendment 31). Bilingual Research Journal, 27(3), 357-382.

172
Ferguson, R. (1990). Racial patterns in how school and teacher quality affect
achievement and earnings. Paper intended for publication in the
journal Challenge: A Journal of Research on Black Men.
Ferguson, R. (2003). Teachers’ perceptions and expectations and the Black-
White test score gap. Urban Education, 38(4), 460-507.
Ferguson, P., & Womack, S. T. (1993). The impact of subject matter and
education coursework on teaching performance. Journal of Teacher
Education, 44, 55-63.
Fetler, M. (1999, March). High school staff characteristics and mathematics test
results. Education Policy Analysis Archives, 7. Retrieved on
04/12/2007 from: http://epaa.asu.edu.
Gaede, O. (1978). Reality shock: A problem among first-year teachers.
Clearinghouse, 51, 405-409.
Garcia, E. (1992). Bilingualism and schooling in the United States. International
Journal of the Sociology of Language, 1, 155-156.
García, G. (1991). Factors influencing the English reading test performance of
Spanish-speaking Hispanic children. [Electronic version]. Reading
Research Quarterly, 26(4), 371-392.
Gibson, S., & Dembo, M. (1984, August). Teacher efficacy: A construct
validation. Journal of Educational Psychology, 76(4), 569-82.

Goddard, R., Hoy, W., & Woolfolk-Hoy, A. (2000). Collective teacher efficacy: Its
meaning, measure, and impact on student achievement. American
Educational Research Journal, 37, 479-507.

Goldhaber, D., & Anthony, E. (2003). Teacher quality and student achievement.
New York: ERIC Clearinghouse on Urban Education.

Goldhaber, D., & Brewer, D. (2000). Does teacher certification matter?
Educational Evaluation and Policy Analysis, 22(2), 129-145.

Greene, J. (1998, March). A meta-analysis of the effectiveness of bilingual
education. University of Texas, Austin: The Tomas Rivera Policy
Institute,

173
Grissmer, D., Flanagan, A., & Williamson, S. (1998). Why did the Black-White
score gap narrow in the 1970s and 1980s? In The Black-White Test
Score Gap. Washington, DC: Brookings Institution Press.

Haberman, M. (1991). Can cultural awareness be taught in teacher education
programs? Teaching Education, 4, 25-32.

Harber, C. (2005, November). Reflections on the new conservatisms.
Educational Review, 57(3), 295-297.

Hawk, P., Coble, C., & Swanson, M. (1985). Certification: It does matter. Journal
of Teacher Education, 36(3), 13-15.

Haycock, K. (1998). Good teaching matters...a lot. OAH Magazine of History,
13(1), 61-63.

Hedges, L., & Nowell, A. (1998). Black-White test score convergence since 1965.
In The Black-White Test Score Gap. Washington, DC: Brookings
Institution Press.

Henson, R., Kogan, L., & Vacha-Haase, T. (2001). A reliability generalization
study of the teacher efficacy scale and related instruments.
Educational and Psychological Measurement, 61(3), 404-420.

Hogben, D., & Petty, M. (1979). Early changes in teacher attitude. Educational
Research, 21(3), 212-219..

Hoy, W., & Woolfolk, A. (1990). Socialization of student teachers. American
Educational Research Journal, 27, 279-300.

Huck, S. (2004). Reading statistics and research (4th ed.). Boston: Allyn and
Bacon.

Ingersoll, R. (2001) Teacher turnover and teacher shortages: An organizational
analysis. American Educational Research Journal, 38, 499-534.

Ingersoll, R., & Smith, T. (2003, May). The wrong solution to the teacher
shortage. Educational Leadership, 60(8).

Jick, T. (1979). Mixing qualitative and quantitative methods: Triangulation in
action. Administrative Science Quarterly, 24(4), 602-11.

174
Johnson, B. (1997). Examining the validity structure of qualitative research.
Education, 118(3), 282-292.

Johnson, B. (2007). Educational research: Quantitative, qualitative, and mixed
approaches. Boston: Allyn & Bacon.

Johnson, R. (2002). Using data to close the achievement gap. Thousand Oaks,
CA: Corwin Press.

Jordan, H., Mendro, R., & Weerasinghe, D. (1997, July). Teacher effects on
longitudinal student achievement. Paper presented at the Sixth
Annual National Evaluation Institute sponsored by CREATE,
Indianapolis.

Katz, M. (Ed). (1999). Justice and caring: The search for common ground in
education. Professional Ethics in Education. New York: Teacher’s
College Press.

Klitgaard, R., & Hall, G. (1974). Are there unusually effective schools? Journal of
Human Resources, 10(3), 90-106.

Krashen, S., & Biber, D. (1988). On course: Bilingual education’s success
in California. Los Angeles: California Association for Bilingual
Education.

Krashen, S. (1999). Bilingual education: Arguments for and (bogus) arguments
against. In Round Table on Language and Linguistics. Washington,
DC: Georgetown University Press.

Kuhlman, E, & Hoy, W. (1974). The socialization of professionals into
bureaucracies: The beginning teacher in the school. Journal of
Educational Administration, 12(2),18-27.

Labaree, D. (1997). Public goods, private goods: The American struggle over
educational goals. American Educational Research Journal, 34(1),
39-81.

175
Laczko-Kerr, I., & Berliner, D. (2002). The effectiveness of "Teach for America"
and other under-certified teachers on student academic
achievement: A case of harmful public policy. Education Policy
Analysis Archives, 10(37),

Ladson-Billings, G.. (1994).The dreamkeepers: Successful teaching for African
American students. San Francisco: Jossey-Bass.

Ladson-Billings, G. (1995). Toward a theory of culturally relevant pedagogy.
American Educational Research Journal, 32(3), 465-491.

Ladson-Billings, G. (1997, December). It doesn’t add up: African American
students’ mathematics achievement. Journal for Research in
Mathematics Education, 28(6), 697-708.

Ladson-Billings, G. (2006). From the achievement gap to the education debt:
Understanding achievement in US schools. Educational
Researcher, 35(7), 3-12.

Landon Beyer, E. (2001). The value of critical perspectives in teacher education.
Journal of Teacher Education 2001, 52, 151-163.

Lankford, H., Loeb, S., & Wyckoff, J.. (2002) Teacher sorting and the plight of
urban schools: A descriptive analysis. Educational Evaluation and
Policy Analysis, 24(1), 37-62.

Lee, H. (2005, August). Understanding and assessing preservice teachers’
reflective thinking. Teaching and Teacher Education, 21(6), 699-
715.

Lee, J. (2002). Racial and ethnic achievement gap trends: Reversing the
progress toward equity? Educational Researcher, 31(1), 3-12.

Lieberman, A., & Miller, L. (1990, June). Restructuring schools: What matters
and what works. Phi Delta Kappan, 71(10), 759-764.

Ligana, J. (1970). What happens to the attitudes of beginning teachers? Danville,
IL: Interstate Printers & Publishers.

Lipton, L., & Wellman, B. (2003). Making mentoring work: An association for
supervision and curriculum development action tool. Alexandria,
VA: ASCD.

176

Lucas, C. (1997). Teacher education in America. New York: Macmillan.

Manley, M., Siudzinski, L., & Varah, L. (1989). Easing the transition for first-year
teachers. NASSP Bulletin, 73(515), 16-21.

Marvel, J,.Lyter, D. M., Peltola, P. Strizek, G. A., Morton, B. A., & Rowland, R.
(2007). Teacher attrition and mobility: Results from the 2004-05
teacher follow-up survey. NCES, 307.

McGrath, J.E. (1982). Dilemmatics: The study of research choices and dilemmas.
In McGrath, J.E., Martin, J., & Kulka, R.A. (Eds.), Judgement calls
in research (pp. 69-102). Beverly Hills, CA: Sage.

McLaughlin, M., & Marsh, D.(1979). Staff development and school change. In
New Demands, new realities, new perspectives. New York:
Teachers College Press, Columbia University.

Midgley, C., Feldlaufer, H., & Eccles, J. (1989). Change in teacher efficacy and
student self- and task-related in mathematics during the transition
to junior high school. Journal of Educational Psychology, l81, 247-
258.

Monk, D. (1994). Subject area preparation of secondary mathematics and
science teachers and student achievement. Economics of
Education Review, 13(2), 125-145.

Morey, A., Bezuk, N., & Chiero, R. (1997, January). Preservice teacher
preparation in the United States. Peabody Journal of Education,
72(1), 4-24.

Mulholland, J., & Wallace, J. (2001, February) Teacher induction and elementary
science teaching: Enhancing self-efficacy. Teaching and Teacher
Education, 17(2), 243-261.

Murnane, R., & Philips, B. (1981, August). Learning by doing, vintage, and
selection: Three pieces of the puzzle relating teaching experience
and teaching performance. Economics of Education Review, 1(4),
453-465.

Murnane, R., Singer, J., Willett, J., Kemple, J., & Olsen, R (Eds.). (1991). Who
will teach?: Policies that matter. Boston: Harvard University Press.

177

Murray, F. (2000). The role of accreditation reform in teacher education.
Educational Policy, 14(1), 40-59.

Nieto, S. (2005). Public education in the twentieth century and beyond: High
hopes, broken promises, and an uncertain future. Harvard
Educational Review, 75(1), 43-64.

Noguera, P. (2003).City schools and the American dream: Reclaiming the
promise of public education. New York: Teachers College Press.

Oakes, J.(2002). Education inadequacy, inequality, and failed state policy: A
synthesis of expert reports prepared for Williams v. State of
California, 1-30.

Oakes, J. and IDEA. (2002).Concept 6 and busing to relieve overcrowding:
Structural inequality in California schools. (Rep.). Los Angeles:
UCLA, IDEA.

Oakes, J., Franke, M., Quartz, K., & Rogers, J. (2002). Research for high-quality
urban teaching: Defining it, developing it & assessing it. Journal of
Teacher Education, 53(3), 228–234.

Oakes, J., & Lipton, M. (1999). Teaching to change the world. Boston: McGraw
Hill.

Odell, S.J., Loughlin, C.E., & Ferraro, D.P. (1987). Functional approach to
identification of new teacher needs in an induction context. Action
in Teacher Education, 8(4), 51-57.

Olsen, B., & Anderson, L. (2006). Investigating early career urban teachers’
perspectives on and experiences in professional development.
Journal of Teacher Education, 57(4), 359-377.

Orfield, G., & Yun, J. (1999). Resegregation in American schools. The civil rights
project. Boston: Harvard University Press.

Ormrod, J. (2006). Educational psychology. Columbus, OH: Prentice Hall.

Pajares, F. (1996). Self-Efficacy beliefs in academic settings. Review of
Educational Research, 66(4), 543-578.

178
Patton, M. (2002).Qualitative research and evaluation methods. Thousand Oaks,
CA: Sage.

Quartz, K. (2003). Too angry to leave. Journal of Teacher Education, 54(2), 99-
111..

Ramirez, D., Yuen, S., & Ramey D. (1991). Executive summary, final report:
Longitiudinal study of structured English immersion strategy. San
Mateo, CA: Aguirre International.

Reed, D., Rueben, K., & Barbour, E. (2006). Retention of new teachers in
California. San Francisco: Public Policy Institute of CA.

Riggs, I., & Enochs, L. (1989, March). Toward the development of an elementary
teacher's science teaching efficacy belief instrument. Paper
presented at the annual meeting of the National Association for
Research in Science Teaching, San Francisco.

Rivkin, S., Hanushek, E., & Kain, J. (2005). Teachers, schools, and academic
achievement. Econometrica, 73(2), 417-458.

Rosenholtz, S. (1986). The organizational context of teaching. Learning to
Teach. Champaign-Urbana, IL: University of Illinois at Champaign-
Urbana.

Ross, J. (1992). Teacher efficacy and the effects of coaching on student
achievement. Canadian Journal of Education, 17(1), 51- 65.

Ross, J. (1994, July). The Impact of an inservice to promote cooperative learning
on the stability of teacher efficacy. Teaching and Teacher
Education, 10(4), 381-394.

Ross, J. (1998).The antecedents and consequences of teacher efficacy. In J.
Brophy (Ed.), Advances in Research on Teaching. Vol. 7 (pp. 49-
74). Greenwich, CT: JAI Press.

Sanders, W., & Rivers, J. (1996, November). Cumulative and residual effects of
teachers on future student academic achievement. (Research
Progress Rep.). Knoxville, TN: University of Tennessee.

179
Sanders, W., Saxton, A., & Horn, S. (1997). The Tennessee Value-Added
Assessment System: A quantitative outcomes-based approach to
educational assessment. In J. Millman (Ed.), Grading teachers,
grading schools. Newbury Park, CA: Sage Publications.

Sanders, W., Wright, P., & Horn, S. (1997, April). Teacher and classroom context
effects on student achievement: Implications for teacher evaluation.
Journal for Personnel Evaluation in Education, 11(1), 57-67.

Schlechty, P., & Vance, V. (1981). Do academically able teachers leave
education? The North Carolina case. Phi Delta Kappan, 63, 106-
112.

Schlechty, P., & Vance, V. (1983). Recruitment, selection, and retention: The
shape of the teaching force. The Elementary School Journal, 83, 4.

Schoeneberger, M., & Russell, T. (1986). Elementary science as a little added
frill: A Report of two case studies. Science Education, 70(5), 519-
538.

Schorr, J. (1993, December). Class action: What Clinton’s National Service
Program could learn from Teach for America. Phi Delta Kappan, 72
(4), 315-318.

Shaw, G. (1944). Everybody’s political what’s what? London: Constable and
Company.

Shen, J. (1997, November-December). Teacher retention and attrition in public
schools: Evidence from SASS 91. Journal of Educational Research,
91(2), 81-88.

Shulman, L. (1986). Paradigms and research progams in the study of teaching: A
contemporary perspective. In Wittrock, M.C. (Ed.), Handbook of
research on teaching. (3rd ed., pp.3-36). New York: Macmillan.

Smith, M.S., & O'Day, J. (1991). Systemic school reform. In Fuhrman, S.H., &
Malen, B. (Eds.), The politics of curriculum and testing: The 1990
yearbook of the Politics of Education Association (pp. 233-268).
Briston, PA: Falmer Press.

Smith, S., & Brownell, M. (1993, Summer). Understanding special education
teacher attrition: A conceptual model and implications for teacher

180
educators. Teacher Education and Special Education, 16(3), 270-
82.

SPSS. (2007). SPSS Base 16.0 User's Guide. Chicago, IL.

Stake, R., & Easley, J. (1978). Case Studies in Science Education. National
Science Foundation, Washington DC: Government Printing Office

Stanford Working Group. (1993, June). Stanford Working Group on federal
education programs for limited-English proficient students: A
blueprint for the second generation. Stanford, CA: Stanford
University.
Stoddart, T.(1990). Los Angeles Unified School District intern program:
Recruiting and preparing teachers for an urban context. Peabody Journal
of Education, 67.

Strong, T. (2006, February). Bringing memory forward: A method for engaging
teachers in reflective practice on narrative and memory. Reflective
Practice, 7(1).

Thompson, G. (2007). Up where we belong: Helping African American and Latino
students rise in school and in life. Indinapolis, IN: Jossey-Bass.

Tschannen-Moran, M., & Woolfolk-Hoy, A. (1998).Teacher efficacy: Its meaning
and measure. Review of Educational Research, 68(2), 202-248.

Tschannen-Moran, M., & Woolfolk-Hoy, A. (2001, October). Teacher efficacy:
Capturing an elusive construct. Teaching and Teacher Education,
17(7), 783-805.

Valencia, R.(1997) The plight of Chicano students: An overview of schooling
conditions and outcomes. In Valencia (Ed.). Chicano school failure
and success: Research and policy agendas for the 1990 's (pp. 3-
26). London: The Falmer Press.

Veenman, S. (1984). Perceived problems of beginning teachers. Review of
Educational Research, 54(2), 143-178.

Vogt, W. P. (Ed.) (1999). Dictionary of statistics & methodology. Thousand Oaks,
CA: Sage.

181
Webb, P. (1999). The use of language in reflective teaching: Implications for self-
understanding. Paper presented at the annual meeting of the
American Educational Research Association, Montreal, Canada.

Weinstein, R., Gregory, A., & Strambler, M. (2004, September). Intractable self-
fulfilling prophecies fifty years after Brown v. Board of Education.
American Psychology, 59(6), 511- 520.

Willig, A. (1985). A meta-analysis of selected studies on the effectiveness of
bilingual education. Review of Educational Research, 55(3), 269-
317.

Willower, D., Eidell, T., & Hoy, W. (1967). The school and pupil control ideology.
University Park, PA: Pennsylvania State University Press.

Wilson, S.M., Floden, R.E., & Ferrini-Mundy, J. (2001, February). Teacher
preparation research: Current knowledge, gaps, and
recommendations (Document R-01-3). Seattle, WA: Center for the
Study of Teaching and Policy, University of Washington.

Wong, H., & Wong, R.(1998). The first days of school. Mountain View, CA: .
Harry K. Wong Publications.

Woody, E. L., Buttles, M., et al. (2004). Voices from the field: Educators respond
to accountability. Berkeley, CA: Policy Analysis for California
Education.

Woolfolk-Hoy, A., & Spero, R. (2005, May). Changes in teacher efficacy during
the early years of teaching. Teaching and Teacher Education,
21(4), 343-356.

Wright, B., & Tuska, S. (1968). From dream to life in the psychology of becoming
a teacher. School Review, 76, 253-293.

Zeichner, K., & Grant, C. (1981, October). Biography and social structure in the
socialization of student teachers: A re-examination of the pupil
control ideologies of student teachers. Journal of Education for
Teaching, 7(3), 298-314.

Zeldin, A., & Pajares, F. (2000). Against the odds: Self-Efficacy beliefs of women
in mathematical, scientific, and technological careers. American
Educational Research Journal, 37(1), 215-246.

182
Appendices
Appendix A: Teacher Efficacy Survey Data
1. Teacher Efficacy Scale Frequency Tables p. 182
2. Teacher Efficacy Scale Graphs p. 183
Appendix B: Outcome Expectancy Survey Data
3. Outcome Expectancy Frequency Tables p. 184
4. Outcome Expectancy Graphs p. 185
Appendix C: Retention Data Collected
3. Retention Data Table p. 186
4. Retention Data Graph p. 186
Appendix D: Instrumentation of Study
5. Survey: Modified STEBI p. 187
6. Interview Plan p. 190
Appendix E: IRB Traditional Program
7. IRB Exemption Letter p.191

183
Teacher Efficacy Scale (Frequency Table)

Questions: 2,4,5,7,11,16,17,18,20,21,22,23

Groups: TFA, DI, U

Statistics
Group Membership Q2 Q4 Q5 Q7 Q11 Q16 Q17 Q18 Q20 Q21 Q22 Q23
TFA N Valid 25 25 25 25 25 25 25 25 25 25 25 25
Missing 0 0 0 0 0 0 0 0 0 0 0 0
Mean 4.48 3.48 3.48 3.80 4.40 3.92 4.40 3.88 4.08 4.08 4.64 3.88
Std. Deviation .510 .586 .823 .816 .707 .862 .500 1.092 1.038 .759 .490 .781
DI N Valid 15 15 15 15 15 15 15 15 15 15 15 15
Missing 0 0 0 0 0 0 0 0 0 0 0 0
Mean 4.20 3.33 3.40 4.07 4.13 3.87 4.40 4.07 4.13 4.00 4.53 3.60
Std. Deviation .941 .724 .910 .799 .834 .640 .828 .961 .640 .535 1.060 .828
U N Valid 11 11 11 11 11 11 11 11 11 11 11 11
Missing 0 0 0 0 0 0 0 0 0 0 0 0
Mean 4.36 3.36 3.27 3.91 4.45 4.00 4.18 4.18 3.45 4.09 4.55 3.64
Std. Deviation .505 .924 .647 1.136 .522 .775 .405 .603 1.293 .539 .522 .674

184
TE Comparison Graph:
Comparison by Scale Means and by Scale Question
Clustered by Group Membership

185
Outcome Expectancy Scale (Frequency Table)
Questions: 1, 3, 6, 8, 9, 10, 12, 13, 14, 15, 19, 24
Groups: TFA, DI, UCLA

Statistics
Group Membership Q1 Q3 Q6 Q8 Q9 Q10 Q12 Q13 Q14 Q15 Q19 Q24
TFA N Valid 25 25 25 25 25 25 25 25 25 25 25 25
Missing 0 0 0 0 0 0 0 0 0 0 0 0
Mean 3.68 4.00 3.36 4.20 3.20 3.84 3.60 4.04 3.80 3.64 4.04 3.64
Std. Deviation .988 .289 .907 .707 1.000 .800 .816 .611 .645 .700 .611 1.036
DI N Valid 15 15 15 15 15 15 15 15 15 15 15 15
Missing 0 0 0 0 0 0 0 0 0 0 0 0
Mean 3.27 3.73 2.60 3.20 2.40 3.40 2.60 3.27 3.27 3.40 3.07 2.47
Std. Deviation 1.100 .961 1.121 1.014 1.056 1.056 1.121 .884 1.033 .737 1.100 .834
U N Valid 11 11 11 11 11 11 11 11 11 11 11 11
Missing 0 0 0 0 0 0 0 0 0 0 0 0
Mean 3.45 4.00 3.55 3.73 2.36 3.91 3.64 2.91 3.45 3.55 3.55 2.64

Std. Deviation 1.128 .447 1.128 .905 1.120 .701 1.206 .944 .934 .688 .820 .924

186
OE Comparison Graph:
Comparison by Scale Means and by Scale Question
Clustered by Group Membership

187
Retention Data
Self-Reported Retention Data Table

Program
Answer
Choice
4
Answer
Choice
3
Answer
Choice
2
Answer
Choice
1
Valid
Answers
Average
(Mean)
No
Answer/
Not Valid
TFA 13 1 8 2 24 3.04 1
DI 6 4 4 0 14 3.14 1
U 6 1 2 0 9 3.44 1

Self-Reported Retention Data Averages, Figure
2.8
2.9
3
3.1
3.2
3.3
3.4
3.5
Group Average
TFA DI U
Program
Retention Average per Group Self-Reported

188
Instrumentation
1. Survey for New Science Teachers, Modified STEBI

Cover Sheet: Background of Surveyed Teachers

Male____ Female ___

Type of Certification: ___ Alternative: ___TFA ___ District Intern

___ Traditional: ___ UCLA ___ USC

Months of Teaching Experience: _______________

Date of Survey: ______/________/ 2008

Which statement describes you best?

____ I will continue teaching in an urban setting as long as I am able.

____ I will continue teaching in a different setting as long as I am able.

____ I will continue teaching until something better comes along.

____ I am planning to leave teaching as soon as possible.

If you may be contacted for a follow-up interview, please indicate by
providing either your phone number or email address below:

Teacher Name: ________________________ _____ Decline to State

(______)_______________or ____________@_________________

The interviews will be audiotaped, approximately one hour in length and
take place in the next 2 months.

189
Survey: Modified Science Teaching Efficacy Belief Instrument

Please indicate the degree to which you agree or disagree with each statement
below by circling the appropriate letters to the right of each statement.

SA= Strongly Agree
A= Agree
UN= Uncertain
D= Disagree
SD= Strongly Disagree

1. When a student does better than usual in science, it is
often because the teacher exerted a little extra effort.

SA A UN D SD
2. I am continually finding better ways to teach science. SA A UN D SD
3. When the science grades of students improve, it is
often due to their teacher having found a more effective
teaching approach.

SA A UN D SD
4. I know the steps necessary to teach science concepts
effectively.

SA A UN D SD
5. I am not very effective in monitoring science
experiments.

SA A UN D SD
6. If students are underachieving in science, it is most
likely due to ineffective science teaching.

SA A UN D SD
7. I generally teach science ineffectively. SA A UN D SD
8. The inadequacy of a student’s science background can
be overcome by good teaching.

SA A UN D SD
9. The low science achievement of some students cannot
generally be blamed on their teachers.

SA A UN D SD
10. When a low-achieving child progresses in science, it is
usually due to extra attention given by the teacher.

SA A UN D SD
11. I understand science concepts well enough to be

190
effective in teaching science. SA A UN D SD
12. Increased effort in science teaching produces little
change in some students’ science achievement.

SA A UN D SD
13. The teacher is generally responsible for the
achievement of students in science.

SA A UN D SD
14. Students’ achievement in science is directly related to
their teacher’s effectiveness in science teaching.

SA A UN D SD
15. If parents comment that their child is showing more
interest in science at school, it is probably due to the
child’s teacher.

SA A UN D SD
16. I find it difficult to explain to students why science
experiments work.
SA A UN D SD
17. I am typically able to answer students’ science
questions.
SA A UN D SD
18. I wonder if I have the necessary skills to teach
science.
SA A UN D SD
19. Effectiveness in science teaching has little influence
on the achievement of students with low motivation.
SA A UN D SD
20. Given a choice, I would not invite the principal to
evaluate my science teaching.
SA A UN D SD
21. When a student has difficulty understanding a science
concept, I am usually at a loss as to how to help the
student understand it better.

SA A UN D SD
22. When teaching, I usually welcome student questions. SA A UN D SD
23. I do not know what to do to turn students on to
science.

SA A UN D SD
24. Even teachers with good science teaching abilities
cannot help some kids to learn science.

SA A UN D SD

191
2. Interview Protocol
Section 1: Questions about Teacher Efficacy
(Opinion and Emotion Questions)

1. In your opinion, how effective are you in your teaching so far and how do
you know?

2. What are challenges in your teaching so far and how do you know? What
do you think causes your strengths or difficulties?

3. How comfortable do you feel about being stulled/ evaluated at this point?
5

Section 2: Questions about Outcome Expectancy
(Opinion Questions)

4. Generally, what are the most important factors that help students learn
science? Explain why you think so.

5. If students are not successful in science, what do you think could be some
causes?

6. Do you think good teachers can always help all students learn science
successfully?
6

Section 3: Question about Preparation Program
(Opinion and Experience Question)

7. How well do you feel prepared by your program to teach science to urban
populations of students?

8. What are the strengths of your preparation program and what could be
improved?

5
Probing Questions:
Are you easily able to improve your practice?
Do you know how to effectively monitor experiments in your classroom?
Are you confident that you can answer students’ science questions?
How well do you know science?
Do you think you have the skills you need to successfully teach science?

6
Probing Questions:
When students do not perform well in science, do you think it is usually due to the teacher?
If a teacher puts more effort into their science teaching, do you think it will produce higher student
achievement?
Do you think that a student’s low academic background can be overcome by good teaching?

192
University (Traditional) Program IRB Exemption

Abstract (if available)

Abstract The purpose of the multi-tiered study presented is to compare the effect of credentialing route on the self-efficacy, outcome expectancy, and retention of beginning urban science teachers serving students in a large urban school district in Southern California. Candidates from one traditional, university-based teacher education program and from two alternative programs, the Teach for America and District Intern Programs, were surveyed and interviewed during the second semester of their first year of teaching. To determine the potential of a difference in self-efficacy and outcome expectancy, the study gave teachers a modified version of the Science Teachers' Efficacy Belief Instrument (STEBI), developed and validated by Riggs and Enochs (1989). Two representative candidates from each program were then interviewed in order to probe for deeper understanding of possible sources of their efficacy and outcome expectancy. The final part of the study is an evaluation of retention data from the three programs, each to triangulate this information with data collected from the surveys, and comparing these retention rates with published data. The study provides data on unresearched questions about traditionally and alternatively credentialed science teachers in urban settings in California.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Urban teacher persistence: self-efficacy, affect, and values

PDF

Beginning teachers' perceptions of effective practices used by their mentor

PDF

No teacher left behind: an examination of beginning teachers' preconceptions and attitudes about induction

PDF

Factors influencing teachers' differentiated curriculum and instructional choices and gifted and non-gifted students' self-perceptions

PDF

The impact of induction: a study of program designs and teacher quality outcomes

PDF

What are the relationships among program delivery, classroom experience, content knowledge, and demographics on pre-service teachers' self-efficacy?

PDF

What is the relationship between early childhood teachers' training on the development of their teaching self-efficacy?

PDF

Preservice teacher preparation for engineering integration in K-5

PDF

The effect of reading self-efficacy, expectancy-value, and metacognitive self-regulation on the achievement and persistence of community college students enrolled in basic skills reading courses

PDF

Evaluation study: building teacher efficacy in K8 computer science integration

PDF

"Thinking in the middle" what instructional approaches are employed by urban middle school teachers to effect changes in African American students' learning outcomes and performances?

PDF

Developing a computer science education program: an innovation study

PDF

Leadership and implementation of 1:1 technology: considering teacher self-efficacy in the implementation process

PDF

Urban teacher residencies and their impact on teacher retention in high-needs school settings

PDF

The effect on teacher career choices: exploring teacher perceptions on the impact of non‐instructional workload on self‐efficacy and self‐determination

PDF

A comparative study of novice and veteran teachers in response to principal initiated efforts to cultivate a positive school culture

PDF

Confidence is key: peer observations and online teacher self-efficacy in higher education

PDF

Latinas’ self-efficacy in computer science and potential factors undergirding their perspective

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

A quantitative study on southeast Asian and Latino student's perceptions of teachers' expectations and self-efficacy

Asset Metadata

Creator Klein, Nina (author)

Core Title A comparative study of self-efficacy, outcome expectancy, and retention of beginning urban science teachers

School Rossier School of Education

Degree Doctor of Education

Degree Program Education

Publication Date 05/14/2009

Defense Date 03/23/2009

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag alternative credentialing,beginning,credentialing route,district intern,OAI-PMH Harvest,outcome expectancy,retention,science teachers,self-efficacy,teacher preparation,TFA,traditional credentialing

Place Name California (states)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Pensavalle, Margo (committee chair), Freking, Frederick W. (committee member), Ragusa, Gisele (committee member)

Creator Email nina.klein@usc.edu,nxk0726@lausd.net

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m2202

Unique identifier UC172060

Identifier etd-Klein-2815 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-248012 (legacy record id),usctheses-m2202 (legacy record id)

Legacy Identifier etd-Klein-2815.pdf

Dmrecord 248012

Document Type Dissertation

Rights Klein, Nina

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu