University of Southern California Dissertations and Theses

Using mastery learning to address gender inequities in the self-efficacy of high school students in math-intensive STEM subjects: an evaluation study

(USC Thesis Other)

Using mastery learning to address gender inequities in the self-efficacy of high school students in math-intensive STEM subjects: an evaluation study

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Running head: ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

1

USING MASTERY LEARNING TO ADDRESS GENDER INEQUITIES IN THE
SELF-EFFICACY OF HIGH SCHOOL STUDENTS IN MATH-INTENSIVE STEM
SUBJECTS: AN EVALUATION STUDY

by

Sally D Mingarelli

_____________________________________________________________________

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

December 2019

Copyright 2019 Sally D Mingarelli

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

2
DEDICATION
First and foremost, this dissertation is dedicated to my daughter, Laynee. Laynee is a
fierce and forthright four-year-old girl, and I don’t want the world to take that away from her. I
want her to know that she has every option available to her and that her dreams should not be
tied to her gender. I love you, Laynee, and this dissertation is for you.
This work is also dedicated to my husband, Raf, for his continuous, loving support. Not
only did he do a lot of single parenting while I pursued this work, but he has always made clear
that my drive and my intellect are what he appreciates the most about me. I am so grateful for a
marriage that has pushed me to continue to set and achieve ambitious goals.
Lastly, for every friend, colleague, teacher, and family member who has encouraged me,
celebrated me, comforted me, and supported me in this pursuit and in all others . . . I am forever
grateful. I am so lucky to have always been surrounded by love from people who lift me up.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

3
ACKNOWLEDGEMENTS
I would like to gratefully and humbly acknowledge the support and guidance from my
committee members, Dr. Monique Datta and Dr. Ravneet Tiwana; my assistant chair, Dr. Adrian
Donato; and my awe-inspiring chair, Dr. Darline Robles. I am thankful for the time, attention,
and care that they provided to me and my work throughout this process. I am particularly
inspired by Dr. Robles’ career in diversity, equity, and inclusion and her insights into my
problem of practice moved my dissertation in important directions.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

4
TABLE OF CONTENTS
Dedication 2
Acknowledgements 3
List of Tables 6
List of Figures 8
Abstract 10
Chapter 1: Introduction 11
Introduction to the Problem of Practice 11
Organizational Context and Mission 11
Organizational Goal 12
Related Literature 14
Importance of the Evaluation 15
Description of Stakeholder Groups 16
Stakeholder Groups’ Performance Goals 17
Stakeholder Group for the Study 18
Purpose of the Project and Questions 18
Methodological Framework 19
Definitions 19
Organization of the Dissertation 20
Chapter 2: Review of the Literature 22
Gender and STEM Self-Efficacy 22
Promising Practices for Gender Equity in STEM Classrooms 27
Role of Stakeholder Group of Focus 32
The Clark and Estes’ (2008) Gap Analytic Framework 32
Stakeholder Knowledge, Motivation, and Organization Influences 33
Conclusion 58
Chapter 3: Methodology 60
Introduction 60
Participating Stakeholders 61
Quantitative Data Collection and Instrumentation 66
Qualitative Data Collection and Instrumentation 70
Data Analysis 72
Credibility and Trustworthiness 74
Validity and Reliability 76
Ethics 78
Limitations and Delimitations 80
Chapter 4: Results and Findings 83
Introduction 83
Participating Stakeholders 84
Data Validity 86
Research Question 1: To What Extent is the Organization Meeting its Goals? 87
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

5
Research Question 2: Knowledge, Motivation, and Organizational Results 111
and Findings
Summary 158
Conclusion 163
Chapter 5: Solutions and Integrated Implementation and Evaluation Plan 164
Organizational Context and Mission 164
Organizational Goal 165
Description of Stakeholder Groups 166
Goal of the Stakeholder Group for the Study 167
Purpose of the Project and Questions 169
Introduction and Overview 169
Recommendations for Practice to Address KMO Influences 171
Integrated Implementation and Evaluation Plan 190
Strengths and Weaknesses of the Approach 213
Limitations and Delimitations 215
Future Research 216
Conclusion 218
References 220
Appendices 235
Appendix A: Existing Data 235
Appendix B: Survey Items 238
Appendix C: Interview Protocol 242
Appendix D: Sample Survey Items Measuring Kirkpatrick Levels 1 and 2 246
Appendix E: Sample Survey Items Measuring Kirkpatrick Level 3 Drivers 248
Appendix F: Sample Blended Evaluation Items Measuring Kirkpatrick Levels 249
1, 2, 3, and 4
Appendix G: Sample Teacher Dashboard Using Levels 1, 2, and 3 Driver 251
Evaluation Data
Appendix H: Sample Administrator Dashboard Using Level 3 Behaviors, 252
Level 4 Results and Institutional Data

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

6
LIST OF TABLES
Table 1. Hawaii School Diversity Data from INDEX 12
Table 2. Organizational Mission, Global Goal and Stakeholder Performance Goals 17
Table 3. Knowledge Influences, Types, and Assessments for Knowledge Gap Analysis 38
Table 4. Assumed Motivation Influences and Assessments for Motivation Gap Analysis 46
Table 5. Assumed Organizational Influences and Assessments for Organizational Gap 52
Analysis
Table 6. Distribution of Study Participants by Race/Ethnicity and Gender 84
Table 7. Pseudonyms, Race/Ethnicity, and Self-Efficacy Data for Interview Subjects 86
Table 8. Organizational and Stakeholder Goals Evaluated in this Study 88
Table 9. Percentage of Students Demonstrating Mastery and Percentage of Students 91
Setting Goals Aligned to Particular Competencies on the Results from
Student Self-Assessment
Table 10. Mean of Likert Scale Responses to Three Self-Efficacy Questions by Gender 94
and the Difference Between Male and Female Means
Table 11. Science Enrollment Data by Gender for the Two Years Following Completion 109
of Regular Chemistry
Table 12. Assumed Knowledge, Motivation, and Organizational Influences Evaluated 112
in this Study
Table 13. Relationship Between Responses to Survey Items 13 and 14 126
Table 14. Comparison of Percent Agreement on Survey Items 15 and 16 for All 141
Respondents, Female Respondents, and Interview Subjects
Table 15. Degree of Validation of Assumed Influences (V = validated, PV = partially 159
validated)
Table 16. Summary of Results and Findings, Reported as Identified Assets and 161
Validated Influences
Table 17. Hawaii School Diversity Data from INDEX 165
Table 18. Organizational Mission, Global Goal and Stakeholder Performance Goals 168
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

7
Table 19. Summary of Knowledge Influences and Recommendations 172
Table 20. Summary of Motivation Influences and Recommendations 178
Table 21. Summary of Organization Influences and Recommendations 184
Table 22. Outcomes, Metrics, and Methods for External and Internal Outcomes 193
Table 23. Critical Behaviors, Metrics, Methods, and Timing for Evaluation 195
Table 24. Required Drivers to Support Critical Behaviors 198
Table 25. Evaluation of the Components of Learning for the Program 204
Table 26. Components to Measure Reactions to the Program 205
Table 27. Faculty Dashboard with Kirkpatrick Level 1, Level 2, and Level 3 Drivers 209
Evaluation Data
Table 28. Administration and Faculty Dashboard with Kirkpatrick Levels 3 and 4 210
Delayed Evaluation Data and Additional Institutional Data

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

8
LIST OF FIGURES
Figure 1. Interactions between knowledge and motivation within the organizational 55
models and settings
Figure 2. Likert scale responses by gender on Q1 from institutional self-efficacy 95
data set (2018)
Figure 3. Likert scale responses by gender on Q2 from institutional self-efficacy 96
data set (2018)
Figure 4. Likert scale responses by gender on Q3 from institutional self-efficacy 96
data set (2018)
Figure 5. Likert scale responses by gender to survey item 10 about laboratory 97
self-efficacy
Figure 6. Likert scale responses by gender to survey item 11 about mathematical 98
self-efficacy
Figure 7. Pre- and post- self-efficacy mean scores for students identifying as 100
non-binary or prefer not to answer
Figure 8. Pre- and post- self-efficacy mean score comparisons for male and female 101
respondents
Figure 9. Enrollment statistics by gender in physics and AP biological science courses 105
in the fall following participation in regular chemistry
Figure 10. Enrollment statistics by gender in physics and AP biological science courses 107
one full year after completing regular chemistry
Figure 11. Enrollment statistics by gender of chemistry students taking honors or 110
AP physics in the two years following completion of regular chemistry
Figure 12. Distribution of Likert scale responses to procedural knowledge survey 115
items 1, 2, 3, 4, and 19 about the skills and habits of mind required for
STEM inquiry
Figure 13. Distribution of Likert scale responses to metacognitive knowledge survey 119
items 5, 8, 9, and 25 about self-reflection and laboratory self-regulation
Figure 14. Likert scale responses by gender to metacognitive knowledge survey item 124
13 about gendered STEM beliefs

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

9
Figure 15. Likert scale responses by gender to metacognitive knowledge survey item 125
14 about gendered STEM beliefs
Figure 16. Comparison of pre- and post- self-efficacy data for all females (n = 73) and 131
female interview subjects (n = 12)
Figure 17. Number of interview subjects connecting self-efficacy to each Bandura 136
(1986) influencer and separation of the factors independent from or
connected to the pilot course
Figure 18. Distribution of Likert scale responses to motivation survey items 17, 18, 137
and 19 about success attributions
Figure 19. Distribution of Likert scale responses to motivation survey items 15 and 16 140
about failure attributions
Figure 20. Percent agreement to attributional survey item 16 disaggregated by 145
race/ethnicity
Figure 21. Distribution of Likert scale responses by gender to motivation survey items 146
21, 22, 23, and 33 about mastery goal orientation
Figure 22. Distribution of Likert scale responses by gender to motivation survey 149
items 20, 24, and 32 about performance goal orientation
Figure 23. Likert scale responses by gender to survey item 12 about science anxiety 154
Figure 24. Percent agreement statistics by race and ethnicity on survey item 26 about 155
feelings of belonging

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

10
ABSTRACT
This study sought to explore the knowledge-based, motivational, and organizational root causes
of gender inequities in self-efficacy and gender gaps in participation in math-intensive STEM
courses at the Hawaii School. Clark and Estes’ (2008) gap analysis provided the conceptual and
methodological framework for this study. Using an explanatory sequential mixed methodology,
the relationships between goal orientation, metacognition, and self-efficacy were particularly
investigated within the cultural setting of a competency-based pilot course in chemistry. Results
from surveys, interviews, and document analysis verified nine influences on the problem of
practice in the areas of procedural and metacognitive knowledge, goal orientation, attributions,
self-efficacy, and cultural models and settings. The verified influences were utilized in the
selection of evidence-based recommendations for solutions and the creation of an integrated
implementation and evaluation plan using the New World Kirkpatrick Model (Kirkpatrick &
Kirkpatrick, 2016). The suggested program in Chapter 5 informs the next round of pilot courses
in the Hawaii School’s change initiative towards competency-based instruction and mastery
assessment practices.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

11
CHAPTER 1
INTRODUCTION
Introduction to the Problem of Practice
Gender inequity in self-efficacy in math-intensive, high school STEM subjects is a
significant problem of practice for high school educators. Self-efficacy is a student’s
understanding of his or her own competence on specific tasks in a particular subject area
(Bandura, 1986; Peters, 2013). This self-perception is found to influence the level of interest one
will have in a subject as well as the amount of persistence and resilience one will demonstrate
while learning it (Bandura, 1986; Peters, 2013; Sadler, Sonnert, Hazari, & Tai, 2012). High
school boys are found to have overall higher academic self-efficacy than girls, with discipline-
specific differences in chemistry, physics, and mathematics (Pajares & Miller, 1994; Uitto,
2014). This problem is important to address because STEM self-efficacy deficits in girls are
related to both performance and participation gaps in math-intensive subjects in high school; and
unaddressed self-efficacy gaps are thought to contribute to unequal gender distributions in math-
intensive STEM college majors and careers (Bottia, Stearns, Mickelson, Moller, & Parker, 2015;
Elster, 2014; Reilly, Neumann, & Andrews, 2015; Wang & Degol, 2016).
Organizational Context and Mission
The Hawaii School (a pseudonym) is a large, independent, K-12, day school in the state
of Hawaii, with a mission to cultivate within each student the capacity to collaborate,
communicate, create, think critically, empathize, embrace challenge, engage with a global
perspective, and honor self and place. The academic mission has a specific additional focus on
developing students’ skills and habits of mind for engagement in authentic inquiry. The Hawaii
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

12
School has a well-regarded high school, described as a thought leader and an innovation
incubator in the independent school world.
The high school is comprised of 1700 students (48.2% male and 51.8% female) and 168
faculty members. A Principal, two Assistant Principals, and a team of eight Student Deans lead
the high school, and the entire K-12 school is led by a senior administrative team of 17 people
including the school President, Vice Presidents, division directors, and team leaders. Table 1
shows self-reported diversity data for the 2018–2019 school year, which indicates the racial and
ethnic demography of the students, faculty, and school leadership (indexgroups.org, 2018).

Table 1
Hawaii School Diversity Data from INDEX

African
American
/ Black Latinx
Asian
American
Native
American
Multiracial
American
Pacific
Islander
American
White Non-
Latinx
American
Unsure /
Unreported
Students 0.34% 0.4% 22.2% 0.06% 21.3% 9.4% 8.3% 38.0%
Faculty 0.3% 2.2% 26.2% 23.8% 1.5% 46.0%
High School
Leaders
18.2% 9.1% 72.7%
Senior Leader
Team
41.2% 23.5% 35.3%

Organizational Goal
Although the Hawaii School enjoys a high status both locally and nationally and its
students achieve impressive college admissions outcomes, the school is not immune to the
national trend of underrepresentation of girls in upper-level, math-intensive STEM courses
(College Board, 2014). In the 2017–2018 school year, overall enrollment in all AP courses in
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

13
calculus, computer science, chemistry and physics was 47.6% female and 52.4% male (n = 519).
Larger gender disparities existed in the highest levels of calculus and physics, with 64% male
students and 36% female students in AP Calculus BC (n = 44), and 96% male students with only
4% females in AP Physics C (n = 26). In addition, AP Computer Science was 76% male
students and 24% female students (n = 34) and a course in engineering projects was 73% male
and 27% female (n = 11). Given the confirmed participation gap, the organizational goal of the
Hawaii School is to achieve improved gender equality, as defined by a gender representation that
approaches the demography of the student body (48.2% male and 51.8% female), in participation
in math-intensive STEM courses in the high school by the fall semester of 2021.
While there are many possible pathways to addressing this problem, from the intersecting
motivational constructs of goal orientation and self-efficacy it can be argued that girls’
motivation to participate in math-intensive STEM courses could be enhanced if: (1) they were to
engage in these classes with mastery goals rather than performance goals; and (2) if the course
were designed to provide opportunities for girls to cultivate their self-efficacy through
developing the skills and habits of mind to engage in authentic inquiry (Lau & Roeser, 2008;
Lewis, 2018; Martin & Elliot, 2016; Yough & Anderman, 2006).
Currently, the Hawaii School is a leader in its commitment to the work of the Mastery
Transcript Consortium (MTC, n.d.), a consortium of over 300 independent and public schools
that are working together to disrupt the traditional transcript and create curriculum, pedagogy,
and assessments that promote mastery over performance goals. Along with other MTC member
schools, the Hawaii School is actively working to reshape its curriculum to highlight desired
skills and competencies over factual knowledge in all content areas. These changes are
redefining what it means to be “good at STEM,” which could have a positive impact on girls’
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

14
self-efficacy, and ultimately their participation, in math-intensive STEM courses (Dembo &
Eaton, 2000; Flowers III & Banda, 2016; Leaper, Farkas, & Brown, 2012). Given the research in
support of mastery learning as a tool to enhance self-efficacy and the context of the Hawaii
School as a leading school in the MTC, a shorter-term organizational goal was to evaluate and, if
necessary, improve gender equity in self-efficacy in math-intensive STEM pilot courses aligned
to the new competency-based curriculum and mastery assessments in the high school by the end
of the spring semester of 2019.
Related Literature
Since the origin of social cognition theory in Bandura’s work in 1977, researchers have
sought to validate his proposed connection between self-efficacy and student outcomes. An
early study on academic self-efficacy revealed significant gender inequities in self-efficacy in
mathematics (Pajares & Miller, 1994). In the last 20 years, several studies have explored how
self-efficacy influences girls’ affect and performance in math-intensive STEM subjects as well as
their participation in related STEM careers (Bottia et al., 2015; Britner, 2008; Kitts, 2009;
Patterson & Johnson, 2017; Sadler et al., 2012; Uitto, 2014; Zeldin, Britner, & Pajares, 2008).
Gender inequities in self-efficacy are causally related to performance gaps in high school
physical sciences and mathematics. This nationwide performance gap has persisted from the
1960s until the present day; and while it first manifests in middle school, it grows larger in high
school (Burkam, Lee, & Smerdon, 1997; Hedges & Nowell, 1995; Reilly et al., 2015). Research
has shown that self-efficacy beliefs are the most reliable predictor of student performance in high
school science (Bandura, 1986, 1998; Britner, 2008). Reduced performance leads to a decrease
in the desire to pursue these fields. It has been shown that gender gaps in participation in
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

15
Physical Science, Engineering, and Mathematics (PSEM) at the college level begin with gaps in
the intent to pursue math-intensive STEM in high school (Bottia et al., 2015; Sadler et al., 2012).
Women are still largely underrepresented in undergraduate PSEM majors (National
Science Foundation [NSF], 2016). NSF statistics reveal that even though 57.2% of all college
graduates in the United States in 2014 were female, the percentage of females earning bachelor’s
degrees in PSEM fields was disproportionately low. In 2014, females earned 42.8% of the
bachelor’s degrees in mathematics, 39.7% in physical sciences, 19.8% in engineering, and 18.1%
in computer science (NSF, 2016). Gender inequities in self-efficacy in math-intensive STEM
fields in high school contribute to participation gaps in PSEM college majors and careers. The
underrepresentation of women in these careers leads to a lack of diversity in the perspectives
brought to problem solving and innovation in these fields.
Importance of the Evaluation
It was important to evaluate the Hawaii School’s performance in relationship to its goal
of improved gender equality in enrollment in math-intensive stem courses, with a close look at
the effectiveness of the competency-based pilot courses at narrowing possible gender self-
efficacy gaps. Women comprise half the pool of potential science talent in the United States and
their inclusion in PSEM careers would bring a broader range of perspectives to this work
(Britner, 2008; Li, 2002). The greater inclusion of women in these careers would not only help
to meet the demand for more workers in these fields, but it would also increase workforce
diversity (Sadler et al., 2012). Closing the gender participation gap in PSEM careers is
considered crucial to sharpening the United States’ competitiveness in research and innovation
(NSF, 2002).
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

16
Science and math self-efficacy developed in high school is a crucial determining factor in
the decision to pursue PSEM careers (Bottia et al., 2015; Uitto, 2014). As such, more equitable
instructional techniques and classroom climates need to be created to cultivate the self-efficacy
of high school girls in math-intensive STEM subjects (Bottia et al., 2015; Britner, 2008). Equity
is different from equality. While equal instruction uses identical teaching methodologies for all
students, equitable instruction seeks to ensure the success of all constituent groups by providing
differentiated support (Espinoza, 2007). Converting to a mastery-based curriculum that is
student-centered, individualized, and based upon improving competencies rather than competing
for performance outcomes could be a step closer to gender equity in STEM courses (Lau &
Roeser, 2008; Lewis, 2018; Martin & Elliot, 2016; Yough & Anderman, 2006). It was the hope
of the Hawaii School that the work to convert to a mastery curriculum could foster STEM self-
efficacy in high school girls, and show promise as a pathway towards a more gender equitable
STEM workforce.
Description of Stakeholder Groups
The key stakeholder groups involved in meeting the goals of improved gender equality in
participation in math-intensive STEM courses, and greater gender equity in self-efficacy, include
the Hawaii School administration, faculty, and students. The school administration will continue
to be responsible for connecting the faculty with research on gender-based best practices in
STEM education and providing time and guidance for faculty to innovate their pedagogy to be
more inclusive as they make the shift towards mastery curriculum. The STEM faculty members
will keep working to create curriculum, pedagogy, and assessments that are more gender
inclusive, with a focus on techniques that will enhance girls’ STEM self-efficacy. The students
will need to keep living the school’s mission by developing the skills and habits of mind to
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

17
perform STEM inquiry, and engaging in self-reflection and targeted goal setting to promote their
continued success in STEM courses.
Stakeholder Groups’ Performance Goals
Table 2 articulates the organizational goal and cascading stakeholder goals in the context
of the organization’s mission.

Table 2
Organizational Mission, Global Goal and Stakeholder Performance Goals
Organizational Mission
The mission of the Hawaii School is to cultivate within each student the capacity to collaborate,
communicate, create, think critically, empathize, embrace challenge, engage with a global perspective,
and honor self and place. The academic mission has a specific additional focus on developing students’
skills and habits of mind for engagement in authentic inquiry.
Organizational Performance Goals
Long Range Goal: The goal of the Hawaii School is to achieve improved gender equality, as defined by a
gender representation that approaches the demography of the student body (48.2% male and 51.8%
female), in participation in math-intensive STEM courses in the high school by the fall semester of 2021.
Supportive Shorter Term Goal: The shorter term goal of the Hawaii School is to evaluate and, if
necessary, improve gender equity in self-efficacy in math-intensive STEM pilot courses aligned to the
new competency-based curriculum and mastery assessments in the high school by the end of the spring
semester of 2019.
Stakeholder 1 Goal Stakeholder 2 Goal Stakeholder 3 Goal
School Administration:
By the fall semester of 2019, the
school administration will have
provided 100% of high school
STEM faculty members with
professional development in the
area of gender-inclusive STEM
best practices for pedagogy and
assessment, with a focus on
techniques for enhancing students’
self-efficacy.
STEM Faculty Members:
By the spring semester of 2021,
the Hawaii School STEM faculty
members will have critically
assessed 100% of their curriculum
for gender inclusivity and will
have responded to all areas for
improvement with meaningful
changes in pedagogy and
assessment, with a focus on
techniques for enhancing
students’ self-efficacy.
Students:
By the spring semester of
2019, 100% of students who
participated in the competency-
based STEM pilot courses will
be able to self-assess their
inquiry skills and habits of
mind, and create action plans
to ensure their growth in these
areas and support their success
in future STEM courses.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

18
Stakeholder Group for the Study
While the joint efforts of all stakeholder groups will be required to achieve greater gender
equality in participation in math-intensive STEM courses, the STEM self-efficacy of the students
is thought to be a crucial determining factor in reaching this goal. Therefore, the stakeholder
group of focus for this study was the Hawaii School students. The school’s mission is a call to
students to develop their skills in inquiry and adopt the habits of mind for persistence and
resourcefulness in their pursuit of rigorous studies. Self-efficacy, or the belief in one’s own
ability, determines the resilience a student will have in the face of setbacks, and it is a predictor
of the interest a student will have in a subject and the persistence she will demonstrate in pursuit
of her learning goals (Bandura, 1986; Peters, 2013; Sadler et al., 2012). Focusing on students as
the stakeholder group for this study allowed for an evaluation of the factors that influence the
development of girls’ self-efficacy in math-intensive STEM subjects.
Purpose of the Project and Questions
The purpose of this project was to evaluate the degree to which the Hawaii School is
meeting its goals of gender equality in math-intensive STEM course enrollment and gender
equity in self-efficacy in competency-based pilot courses. The analysis focused on the
knowledge, motivation, and organizational elements related to achieving the organizational
goals. While a complete performance evaluation would have focused on all stakeholders, for
practical purposes the stakeholder focused on in this analysis was the Hawaii School students.
As such, the questions that guided this study were the following:
1. To what extent is the organization meeting its goals?
2. What are the knowledge, motivation, and organizational elements related to the
Hawaii School’s goal to achieve improved gender equity in self-efficacy in math-
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

19
intensive STEM courses aligned to the new competency-based curriculum and
mastery assessments in the high school by the end of the spring semester of 2019?
3. What are the recommendations for organizational practice in the areas of knowledge,
motivation, and organizational resources that may be appropriate for solving the
problems of gender inequities in self-efficacy and gender gaps in participation in
math-intensive STEM courses at the Hawaii School?
Methodological Framework
This study employed a mixed methodology for data gathering, and Clark and Estes’
(2008) gap analytic, theoretical framework was applied for data analysis. Clark and Estes (2008)
propose that gaps can be understood through a careful analysis of knowledge-based (K),
motivational (M), and organizational (O) root causes; and that solutions to organizational
problems lie in specific KMO domains. For this study, a mixed methodology was utilized to
discover not only if a competency-based curriculum and mastery assessment structure were
correlated with improved self-efficacy for girls, but also exactly how they did so through their
interaction with other knowledge, motivation, and organizational factors. This depth of
understanding is crucial for the continuing work of teachers and administrators to close the
gender self-efficacy gap in math-intensive STEM subjects and improve equality in participation
in these courses.
Definitions
Competency: A specific skill or habit of mind that can be observed and assessed.
Competency-based: Any curriculum, pedagogy, or assessment that is structured around
improving student competencies as opposed to being driven by content acquisition.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

20
Mastery: As an adjective, mastery describes any curriculum, pedagogy, or assessment
that is designed to support a student’s evolving progress toward developing competencies (as
such, mastery is often utilized as a synonym for competency-based). As a noun, when a student
achieves mastery they have reached the highest defined level of performance on a competency.
Mastery Transcript Consortium (MTC): A consortium of over 300 public and private
schools that is advocating for colleges to accept an alternative transcript. This transcript would
highlight a student’s mastery of competencies, rather than reporting A-F grades in discipline-
specific courses.
Messy data: Messy data describes actual real-world data that is unstructured,
heterogeneous, and could include either missing or errant data points. Traditional high school
lab curricula avoid presenting students with messy data, while authentic inquiry experiences
require the skills to sort, organize, and interpret messy data.
PSEM: Acronym for Physical Science, Engineering, and Mathematics. Useful for its
exclusion of the biological sciences and psychology, where underrepresentation of women is not
a problem.
STEM: Acronym for Science, Technology, Engineering, and Mathematics. This broad
term includes all branches of science, however the phrase “math-intensive STEM” is meant to
exclude biological sciences and psychology.
Organization of the Dissertation
Five chapters are used to organize this dissertation. This chapter provided the reader with
the key concepts and terminology commonly found in a discussion about gender inequities in
math-intensive STEM self-efficacy, performance, and participation and introduced research in
support of a competency-based curriculum as a tool for improving equity. The organization’s
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

21
mission, goals, and stakeholders, as well as a review of the evaluation framework and the study’s
guiding questions, were provided. Chapter 2 provides a review of current literature surrounding
the scope of the study. Topics of self-efficacy development, gender performance and
participation gaps in math-intensive STEM fields, gender inclusive best practices for teaching,
and promising practices in competency-based learning and mastery assessment will be
addressed. Chapter 3 details the methodological choices surrounding participants, data
collection, and analysis. In Chapter 4, the data and results are described and analyzed.
Chapter 5 provides context-specific recommendations, based on data and literature, for
addressing gender gaps in self-efficacy and participation, as well as recommendations for an
implementation and evaluation plan.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

22
CHAPTER 2
REVIEW OF THE LITERATURE
Chapter 2 outlines possible root causes of gender gaps in self-efficacy in math-intensive,
high school STEM subjects and current research on gender-inclusive promising practices. The
first section focuses on an introduction to Bandura’s (1977) concept of self-efficacy, with a close
investigation into each of the four factors that affect self-efficacy development. The second
segment explores current best practices for gender-inclusive teaching. That section begins with a
broad look at gender-inclusive STEM curriculum, pedagogy, and assessment techniques, and
concludes with an introduction to the Mastery Transcript Consortium. The chapter ends with a
review of a methodological gap analysis framework, which investigates possible knowledge-
based, motivational, and organizational root causes for the gender inequities in students’ self-
efficacy in math-intensive, high school STEM subjects and the underrepresentation of girls in
these courses (Clark & Estes, 2008).
Gender and STEM Self-Efficacy
In the first articulation of his theory of self-efficacy, Bandura (1977) outlined four key
influences on an individual’s efficacy: performance accomplishments, vicarious experience,
verbal persuasion, and emotional arousal. In 1986, Bandura changed the name of three of these
factors, resulting in the list of four self-efficacy influences that is still in use today: (1) mastery
experiences, (2) vicarious experiences, (3) social persuasions, and (4) physiological states. In
this section each of these factors will be investigated closely to establish a deep understanding of
the construct of self-efficacy and how it ultimately relates to gender gaps in performance and
participation in math-intensive STEM.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

23
Applicability of Self-Efficacy
The motivational theory of self-efficacy, first developed in 1977 by Albert Bandura,
seeks to explain how learning behaviors are impacted by particular environmental and internal
states. Self-efficacy is a person’s perception of their ability to perform a particular task, and
there is a direct connection between this perception and persistence (Bandura 1977, 1986).
Students with low self-efficacy tend to avoid difficult tasks and give up more quickly, and as a
result there is also a connection between self-efficacy and performance (Bandura, 1998). The
enduring gender gaps in participation and performance in math-intensive STEM fields has led
researchers to investigate possible gender disparities in self-efficacy (Kitts, 2009; Pajares &
Miller, 1994; Uitto, 2014; Zeldin et al., 2008; Zeldin & Pajares, 2000). In their 1994 study,
Pajares and Miller set out to show that self-efficacy is the most powerful predictor of
performance over other theories of learning and motivation. In addition to verifying this
powerful link between self-efficacy and performance, Pajares and Miller (1994) linked self-
efficacy to gender, finding that males have higher mathematical self-efficacy. Uitto (2014)
confirmed this self-efficacy gender gap and found that high school boys have overall higher
academic self-efficacy than girls, with discipline-specific differences in math, chemistry, and
physics.
There are gender differences both in the inherent self-efficacy of high school boys and
girls in math-intensive STEM subjects and in their routes to attaining higher self-efficacy in
these fields. Bandura (1986) argued that mastery experiences should be the most significant of
the four influences on self-efficacy development, however his theory was not specific to the
efficacy of individuals in underrepresented minority groups. Zeldin and Pajares (2000)
conducted a series of interviews with women in math-intensive STEM careers to explore how
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

24
their self-efficacy was nurtured in their youth. A similar set of interviews was conducted with
men in math-intensive STEM careers. The researchers found that men in STEM, when reflecting
back on their motivation in school, found mastery experiences to be the biggest influence on
their self-efficacy beliefs (Zeldin et al., 2008). Meanwhile, the career women in STEM reflected
that their female role models and social persuasions, in the form of encouragement from family,
teachers, and peers, were the greatest contributing factors to their self-efficacy development
(Zeldin & Pajares, 2000). In the sections that follow, each of Bandura’s (1986) proposed four
factors that impact self-efficacy development will be further explored in the specific context of
girls in math-intensive, high school STEM classrooms.
Mastery experiences. As girls have mastery experiences in their STEM classes, through
the development of their competence in the knowledge and skills required for success, they
cultivate a stronger sense of science identity. Bandura (1986) argued that these experiences of
mastery would have the most profound influence on a person’s self-efficacy because people’s
perceptions of their knowledge and skills and their ability to activate strategies drive their
understanding of their capacity. Flowers III and Banda (2016) agreed that a person’s sense of
their competence with content knowledge and inquiry skill sets is crucial in the development of a
science identity. However beyond just these competencies, cultivating a science identity also
requires recognition, or the sense that one will be able and welcome to contribute to the scientific
community (Flowers III & Banda, 2016). Underrepresented groups, like girls in math-intensive
STEM, will require more than just mastery experiences in the classroom to develop their
identities as scientists. They will also need cues that women are welcome in the world of STEM
(Bandura, 1986, 1998; Flowers III & Banda, 2016; Kekelis, Ryoo, & McLeod, 2017).
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

25
Vicarious experiences. Vicarious experiences, particularly seeing oneself reflected in a
role model, are of particular importance in fostering the self-efficacy of girls in math-intensive
STEM disciplines. It is especially meaningful for girls to have women as mentors and role
models in these fields, as these women can clarify for girls what STEM careers can offer to them
and they can begin to reframe the idea of who is welcome in the world of STEM (Kekelis et al.,
2017). Role models need not always come in the form of an adult female, as peer mentoring can
also have a significant impact on girls’ inclusion in STEM. Leaper et al. (2012) found that
adolescent girls’ motivation in their math and science courses increased with their participation
in female study groups and affinity groups for girls in STEM. The mindful cultivation of
environments that highlight the contributions of females in math-intensive STEM fields can also
supports girls’ ability to see themselves in these careers (Ramsey, Betz, & Sekaquaptewa, 2013).
Social persuasions. Supportive feedback and encouragement from teachers, family, and
peers can bolster girls’ self-efficacy in math-intensive STEM courses, while ingrained gender
stereotypes can diminish their STEM self-efficacy. Girls in STEM fields are often exposed to
unsupportive and unwelcoming academic environments that can cause them to: recognize the
scarcity of women, link STEM with men, and reinforce stereotypes about inherent and gendered
STEM ability (Leaper et al., 2012). Stereotype threat describes the possibility that girls will
inadvertently reinforce stereotypes by underperforming due to their diminished self-efficacy
(Schuster & Martiny, 2017). Successful interventions to combat stereotype-consistent outcomes
include explicitly teaching girls about gender equality and feminism, and putting learning
environments through a process of stereotype inoculation (Leaper et al., 2012; Ramsey et al.,
2013). An inoculated environment has worked to eradicate the negative social persuasions for
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

26
members of an underrepresented group and places an intentional focus on providing supportive
feedback and encouragement (Ramsey et al., 2013).
Physiological states. STEM anxiety is more prevalent among girls than boys, and this
negative physiological feedback is in direct conflict with their self-efficacy development. It is
more common for girls to express that “science is too hard for me” (Kitts, 2009, p. 160). Even
girls achieving results equal to or greater than boys can suffer from greater anxiety in math-
intensive STEM courses (Britner, 2008). These feelings of anxiety are in direct conflict with
their positive mastery experience, and their negative physiological state can impede motivation
and persistence (Bandura, 1986; Britner, 2008). Stereotype or identity threat heightens the
feelings of stress in the face of new and complex tasks, however it has been found that girls who
are taught that everyone experiences epistemic emotions during challenges are more likely to
persist (Lee, Alston, & Kahn, 2015; Ramsey et al., 2013; Schuster & Martiny, 2017). If girls are
experiencing unregulated epistemic emotions, ingrained gender stereotypes, lack of female role
models, limited family or community support, or a shortage of opportunities for mastery
experiences that create a healthy science identity, their STEM self-efficacy will be impacted and
their performance in STEM classes will likely suffer.
Self-Efficacy and Performance
Gender inequities in self-efficacy are causally related to performance gaps in high school
physical sciences and mathematics. This pervasive gender performance gap has been observed
in NAEP testing data from the 1960s until the present day (Burkam et al., 1997; Hedges &
Nowell, 1995; Reilly et al., 2015). While early studies sought to establish a biologically-based
difference in aptitude between girls and boys as the reason for the gap, research following
Bandura’s development of self-efficacy theory has continued to conclude that self-efficacy is the
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

27
most reliable predictor of performance (Bandura, 1986, 1998; Britner, 2008; Pajares & Miller,
1994; Reilly et al., 2015). While many researchers agree that self-efficacy gaps give rise to
performance gaps, Bandura (1986) argued that performance also has a powerful impact on self-
efficacy. The possible negative feedback loop between performance and self-efficacy can
contribute to decreased motivation for girls in STEM courses and their lack of persistence in
pursuing these fields. The leaky pipeline of female participation in math-intensive STEM is
related, in part, to this negative feedback loop between performance and self-efficacy
(DiBenedetto & Bembenutty, 2013).
Self-Efficacy and Participation
The gender self-efficacy gap in math-intensive STEM disciplines in high school is a
crucial contributing factor in the existing participation gaps in physical science, mathematics,
and engineering (PSEM) majors in college and in the pursuit of related careers. Research has
shown that students’ experiences with science and math in high school are crucial in determining
their career affinity, and gender gaps in participation at the college level begin with gaps in the
intent to pursue STEM in high school (Bottia et al., 2015; Uitto, 2014). The persistent and
dramatic underrepresentation of women in PSEM careers creates a sense of urgency for
addressing gender self-efficacy gaps in math-intensive STEM classes in high school.
Promising Practices for Gender Equity in STEM Classrooms
The interconnections between performance, participation, and self-efficacy for girls in
math-intensive STEM disciplines in high school inspires a closer look at how teaching and
learning is approached in these fields. In this section, gender-inclusive best practices for math-
intensive STEM courses will be explored through the lenses of curricular design, pedagogical
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

28
choices, and assessment structures. This section will conclude with an introduction to the
promising work of the Mastery Transcript Consortium (MTC).
Gender-Inclusive Teaching and Learning
The structure of K-12 education has been identified as both a root cause of the gender
participation gap in STEM and a promising area in which to make changes that can increase
equity and inclusion. Kanny, Sax, and Riggers-Piehl (2014) defined possible structural barriers
to gender equity in K-12 education as schools, teachers, curriculum, pedagogy, assessments, and
classroom environments. A multicultural framework for improving equity would embrace the
characteristics of underrepresented groups and structure learning environments to celebrate their
strengths and suit their needs (Aragón, Dovidio, & Graham, 2017). As such, equitable
instruction has significant overlap with individualized instruction (Chetcuti, 2008). Research on
best practices for STEM education has argued for over 20 years that student-centered
classrooms, which provide hands-on learning, opportunities for collaboration, and real-life
applications, will improve gender equity and inclusion (Burkam et al., 1997). In the sections that
follow, K-12 STEM education will be further discussed in terms of the curricula, pedagogies,
and assessments that promote gender equity in self-efficacy, performance, and participation.
Equitable curriculum. A lab-based high school STEM curriculum, that exposes
students to multiple STEM disciplines and allows them to authentically engage in research,
promotes increased gender equality in the future participation in STEM fields. In a longitudinal
study, Kang and Keinonen (2017) discovered a meaningful connection between exposure to the
real scientific process in high school and the future pursuit of PhDs in science. Inquiry-based
learning (IBL) shifts the focus of a curriculum away from just content to the actual skills and
habits of mind of scientists (Kang & Keinonen, 2017). In fact, broad exposure in high school to
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

29
a variety of STEM disciplines and modes of inquiry was found to be a more powerful influencer
on the selection of a college STEM major than a student’s actual achievement in those courses
(Wang, 2013). While an inquiry curriculum and breadth of STEM learning can increase
motivation to participate for all students, girls are even more likely to be engaged when their
STEM curriculum feels relevant and connected to real science (Fredricks, Hofkens, Wang,
Mortenson, & Scott, 2018). Equitable teaching, however, is about more than just what is taught.
Pedagogical choices are also a crucial factor in creating gender inclusive classrooms.
Equitable pedagogy. Relational pedagogies, which foster communication and
collaboration both between students and teachers and within student work groups, encourage the
growth of girls’ STEM self-efficacy. Strong relationships with teachers and peers are often more
important for girls than boys in math-intensive STEM classes, as a sense of connection and
collaboration are particularly motivating for underrepresented groups (Fredricks et al., 2018). To
create a classroom that cultivates this environment for female students, teachers need a high level
of emotional intelligence and skills in both instructional and relational pedagogies (Darby, 2005;
Demetriou & Wilson, 2009). Pedagogies that create opportunities to relate and collaborate can
lead to gains in content knowledge, critical thinking, and intellectual risk-taking for all students
that are underrepresented in math-intensive STEM fields, including girls (Goeden, Kurtz,
Quitadamo, & Thomas, 2015; Lundeberg & Moch, 1995). To support the gains that can be made
for girls by teaching with relational pedagogies, gender equitable assessment practices should
focus on individualized growth rather than competition.
Equitable assessment. Assessment philosophies that emphasize mastery over
performance can have a positive impact on girls’ STEM self-efficacy. Performance goals
emphasize the appearance of intelligence and comparisons to the performance of others, and this
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

30
orientation is linked to high levels of student anxiety (Simon, Aulls, Dedic, Hubbard, & Hall,
2015). Mastery goals instead focus on individual growth on learning outcomes (Senko &
Tropiano, 2016; Simon et al., 2015). Just as individualized instruction is considered more
equitable, so too are assessment structures that focus on the individual (Chetcuti, 2008;
DiBenedetto & Bembenutty, 2013). It is clear that what is taught, how it is taught, and how it
gets measured are important factors in closing the self-efficacy gap for girls in math-intensive
STEM courses in high school. Promising practices for gender inclusive classrooms include:
(1) generating curricula that focus more on cultivating competence in authentic STEM skills and
less on content delivery; (2) building pedagogies that are relational and student-centered; and
(3) utilizing assessments that celebrate mastery over performance (Chetcuti, 2008; Darby, 2005;
Demetriou & Wilson, 2009; DiBenedetto & Bembenutty, 2013; Fredricks et al., 2018; Goeden et
al., 2015; Kang & Keinonen, 2017; Senko & Tropiano, 2016; Simon et al., 2015; Wang, 2013).
Competency-based teaching and learning practices provide a promising framework not only for
improved gender equity in STEM, but for better overall learning outcomes for all students.
Competency-Based Education
Competency-based education (CBE) refers to the curricula and pedagogies that
specifically emphasize the development of skills over distribution of content. These learning
environments are active, student-centered environments where student achievement is defined by
what they can do over what they know. Increasingly more research indicates that the historical
hyperfocus on content in high school is, in fact, detrimental to student learning (Curry &
Docherty, 2017; Renshaw, 2014; Senko & Tropiano, 2016). Renshaw (2014) argued that upper
level students struggle to perform scientific inquiry processes because their introductory classes
overemphasized content. In addition to this change in focus towards skills, CBE frameworks
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

31
recognize the reality that students will achieve mastery of competencies on individualized
timelines. This notion is in stark contrast with the ubiquitous practice in schools of relating time
spent in a course to presumed competence (Curry & Docherty, 2017).
Developing a CBE model within an educational institution is a massive undertaking,
which requires recalibrating all course design and graduation requirements to align with a set of
specific schoolwide competencies. However, if done well, CBE can create the landscape for
students and teachers to be true partners in an educational journey that focuses beyond just
acquisition of knowledge and skills, towards deeper learning through application and transfer
(Curry & Docherty, 2017). The member schools of the Mastery Transcript Consortium are
committed to the work of developing CBE frameworks on their campuses and advocating for a
transcript that accurately reflects this educational model.
The Mastery Transcript Consortium
The Mastery Transcript Consortium (MTC) is a group of over 300 independent and
public schools that are committed to the work of rethinking how teaching and assessment should
be done to promote better outcomes for all students. The MTC is an advocacy group that is
working with college admissions offices on behalf of schools to normalize a competency-based
transcript, so that schools are empowered to fully embrace a CBE model. The work of the MTC
is about improving learning for all students, but it could also offer opportunities to address self-
efficacy deficits for girls in math-intensive STEM subjects.
The focus on mastery is found to promote high self-efficacy, positive affect, and
increased interest, by cultivating in students the strategies and behaviors that facilitate learning,
such as self-regulation or cooperation (Senko & Tropiano, 2016). A mastery orientation towards
learning creates increased opportunities for students to develop a growth mindset, viewing their
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

32
ability as malleable and setbacks as part of the learning process towards developing
competencies (Dweck, 1986; Senko & Tropiano, 2016). Vincent-Ruz and Schunn (2017) found
that girls have a higher need to understand their competencies in order to achieve strong learning
outcomes in STEM. The relationship between mastery orientation and self-efficacy will be
further explored in the sections that follow, as the problems of gender inequities in self-efficacy
and related gender gaps are explored through the lens of knowledge-based, motivational, and
organizational root causes.
Role of Stakeholder Group of Focus
The stakeholder group of focus for this study is the Hawaii School students, with a
specific focus on female students enrolled in math-intensive STEM courses. While other
stakeholders in the organization are poised to play a role in improving self-efficacy for girls in
these disciplines, the most compelling outcomes of this promising practice research will involve
noticeable changes in the students’ behaviors and beliefs. In addition, self-efficacy development
relies upon both environmental and internal factors, and only through a study of the students will
the internal factors be accessible (Bandura 1977, 1986, 1998). In the Clark and Estes (2008) gap
analysis that follows, the knowledge-based and motivational barriers to gender equity in self-
efficacy in math-intensive STEM courses will be studied alongside the broader organizational
and cultural barriers that prevent their participation.
The Clark and Estes’ (2008) Gap Analytic Framework
Clark and Estes (2008) presented a gap analysis framework for investigating the root
causes of organizational performance problems. The authors argued there are only three
categories for the possible causes of performance gaps: (1) people’s knowledge and skills,
(2) people’s motivation, and (3) organizational barriers. Knowledge gaps can include from what
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

33
a particular stakeholder group does not know or cannot do, to their inability to monitor their own
understanding and regulate their learning behaviors. Krathwohl (2002) identified these
knowledge types as declarative, procedural, and metacognitive. Motivation gaps arise in
different phases of task completion, beginning with the choice to do a task and then sticking with
the task until the end and ensuring a quality outcome. Stakeholders’ motivation can falter in
their active choice, persistence, or mental effort; and the theories of self-efficacy, attributions,
and goal orientation can help to illuminate why these various motivation hurdles exist
(Anderman & Anderman, 2006; Bandura, 1998; Clark & Estes, 2008; Pajares, 2006; Yough &
Anderman, 2006). Clark and Estes’ (2008) final category for the root causes of performance
problems is organizational barriers, which include unsupportive policies or procedures, lack of
resources, or cultures that subvert organizational goals. This section explores the literature on
the potential knowledge-based and motivational influences on high school girls’ self-efficacy
gap in math-intensive STEM subjects, along with the organizational barriers to equity in girls’
performance and participation in these courses.
Stakeholder Knowledge, Motivation, and Organization Influences
Knowledge and Skills
Knowledge is categorized into four types: factual, conceptual, procedural, and
metacognitive (Krathwohl, 2002). Factual and conceptual knowledge are two types of
declarative knowledge, representative of the things that people know and understand; however,
factual knowledge is discrete and conceptual knowledge is inherently interconnected
(Krathwohl, 2002). Rueda (2011) referenced Bloom’s taxonomy of cognitive processes to
explain that conceptual knowledge requires an understanding of the interrelationship between
facts and enables students to perform higher level cognitive functions like classifying,
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

34
categorizing, and modeling. The third knowledge type, procedural knowledge, is defined as
knowing how to do something (Rueda, 2011). Procedural knowledge includes discipline-
specific modes of inquiry, including the skills, algorithms, methods, and techniques for solving
problems and investigating questions (Krathwohl, 2002). Metacognitive knowledge, defined in
1977 by John Flavell, refers to thinking about thinking (Baker, 2006). More specifically,
metacognition includes both the knowledge of and the control of cognition, and it requires the
ability to self-reflect and self-regulate (Baker, 2006; Krathwohl, 2002).
The method by which knowledge is constructed has long been a topic of great debate by
educational theorists. Piaget’s theory of cognitive science and Vygotsky’s sociocultural theory
have disagreed as to whether the process of learning is inherently internal or social (Lourenço,
2012). However, both behaviorists and social cognitive theorists believe that there is a direct
relationship between learning and the environment (Bandura, 2005; Daly, 2006; Denler, Wolters,
& Benzon, 2006). Social cognitive theory (SCT) describes a triadic reciprocity of interaction
between cognitive, behavioral, and environmental factors that influence a person’s learning and
achievement (Denler et al., 2006). Unlike behaviorism, which focuses solely on the interactions
between behavior and environment, SCT assigns greater agency to the learner and focuses on
self-reflective and self-regulatory practices (Bandura, 2005; Daly, 2006). In the problem of
underrepresentation of girls in math-intensive STEM courses in high school, the social cognitive
triad can be defined by the interactions between the student’s knowledge and skills (cognition)
and their school and home environments (environment), which ultimately influence their
decisions about participation (behavior) (DiBenedetto & Bembenutty, 2013).
Factual and procedural knowledge. One knowledge influence that limits girls’
participation in math-intensive STEM is a misunderstanding of exactly what scientists do and
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

35
how they do it. These misconceptions represent a combination of declarative and procedural
knowledge gaps, and filling in these deficits will require doing a better job teaching girls about
scientists themselves and the modes of inquiry that they use in their work (Bottia et al., 2015;
Hong & Lin-Siegler, 2012; Kang & Keinonen, 2017; Lundeberg & Moch, 1995; Wang, 2013).
Girls’ knowledge of what scientists actually do. Wang (2013) found that exposure to a
wide variety of math and science courses was more important for girls’ continued participation in
STEM than their actual achievement in those courses. In a longitudinal study that followed
6,300 survey respondents from high school into their undergraduate years, Wang (2013)
concluded that girls are more likely to persist in STEM when: (1) they are more deeply
acquainted with specific STEM sub-disciplines and (2) they understand what scientists,
mathematicians, and engineers actually do for a living. Bottia et al. (2015) also found that
exposure to authentic STEM experiences positively impacts girls’ decisions to continue in
STEM. In a longitudinal study following 12,000 students from middle school through college,
the researchers found that participation in enrichment programs that helped students to
understand careers in math and science was ultimately correlated with declaration of STEM
majors (Bottia et al., 2015).
In addition to having factual knowledge about STEM disciplines and career paths, it
helps for students to know that scientists throughout history encountered struggles as they sought
to make their break-through discoveries (Hong & Lin-Siegler, 2012). In their empirical study,
Hong and Lin-Siegler (2012) randomly assigned 271 tenth grade students to be taught about
famous scientists using either performance-oriented (PO) background information or struggle-
oriented (SO) background information. Students in the SO group were found to better identify
with the scientists and were more likely to report seeing themselves as future scientists (Hong &
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

36
Lin-Siegler, 2012). Additionally, Kang and Keinonen (2017) argued that pedagogies that mimic
the actual scientific process, by allowing students to construct scientific knowledge through
experimental design and data analysis, will increase students’ interest in pursuing science.
Lundeberg and Moch (1995) pointed out that science is an inherently collaborative endeavor,
which celebrates the open sharing and debating of ideas and exchanging of lab results to further
scientific knowledge. In a two-year study comparing the performance of 91 female nursing
students in large lectures versus small, group-based, interactive learning programs, the
researchers found increased intellectual risk-taking in the community of collaboration
(Lundeberg & Moch, 1995). Increasing girls’ understanding that science is a collaborative,
inquiry-based process, in which setbacks are both expected and valuable in constructing
scientific knowledge, could potentially help to close the participation gap in STEM courses.
Metacognitive knowledge. In addition to acquiring the declarative and procedural
knowledge of what scientists do and how they do it, girls could benefit from an increased focus
on developing their science metacognition. Science metacognition involves both self-regulation
of one’s conceptual learning and laboratory strategies along with self-reflection about oneself as
a learner of science (Dembo & Eaton, 2000; Flowers III & Banda, 2016; Leaper et al., 2012; Lee
et al., 2015; Mathabathe & Potgieter, 2017).
Girls’ understanding of their science learning and their place in science. Mathabathe
and Potgieter (2017) contended that in a science context, metacognition includes not just
thinking about declarative knowledge but also the monitoring of procedural knowledge
including: planning investigations, applying practical lab skills, and controlling experimental
conditions. At the conclusion of their in-depth case study of two lab groups’ work to complete a
practical final project, the researchers concluded that metacognition is inherently embedded in
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

37
scientific inquiry due to the necessity of constant reflection and regulation of thinking towards
goal oriented outcomes (Mathabathe & Potgieter, 2017). In addition, Dembo and Eaton (2000)
argued that students need to be taught the self-regulatory practices of self-observation and self-
evaluation in order to enhance their performance. Teaching the metacognitive strategies
necessary to recover from inevitable laboratory setbacks could be crucial to retaining girls in
STEM.
For female students in STEM, metacognition has another complex layer — thinking
about their own thinking about being girls in science, perhaps better articulated as their science
identities (Flowers III & Banda, 2016; Leaper et al., 2012). Because girls are underrepresented
in math-intensive STEM fields, identity threat can lead to anxiety, which can lead to impaired
performance or motivation and cause stereotype-consistent outcomes (Lee et al., 2015). Lee et
al. (2015) defined identity threat as the fear of being devalued for your group identity or of
conforming to a negative stereotype. However, Leaper et al. (2012) argued that knowledge of
feminism and a gender-egalitarian mindset can help a girl imagine that the STEM workforce
includes a future place for her, despite the current underrepresentation of women in these fields.
In their survey of 579 girls, the results indicated a positive correlation between knowledge of
feminism and gender-egalitarian beliefs and choice to participate and persist in STEM courses
(Leaper et al., 2012). Flowers III and Banda (2016) noted that a strong metacognitive sense of
her own conceptual knowledge and laboratory competence will be crucial to a girl’s formation of
a positive science identity. The literature on identity formation argues that building a solid
science identity is of particular importance for the persistence and success of those currently
underrepresented in science (Flowers III & Banda, 2016). Closing the participation gap for girls
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

38
in math-intensive STEM courses could require both increased instruction on how to learn science
and how to cultivate a positive science identity.
Table 3 provides information on the organizational mission, organizational and
stakeholder goals, and the knowledge influences discussed in this paper. The table showcases
samples of assessments that were used to study stakeholder’s knowledge.

Table 3
Knowledge Influences, Types, and Assessments for Knowledge Gap Analysis
Organizational Mission
The mission of the Hawaii School is to cultivate within each student the capacity to
collaborate, communicate, create, think critically, empathize, embrace challenge, engage
with a global perspective, and honor self and place. The academic mission has a specific
additional focus on developing students’ skills and habits of mind for engagement in
authentic inquiry.
Organizational Global Goal
Long Range Goal: The goal of the Hawaii School is to achieve improved gender equality, as
defined by a gender representation that approaches the demography of the student body
(48.2% male and 51.8% female), in participation in math-intensive STEM courses in the
high school by the fall semester of 2021.
Supportive Shorter Term Goal: The shorter term goal of the Hawaii School is to evaluate
and, if necessary, improve gender equity in self-efficacy in math-intensive STEM pilot
courses aligned to the new competency-based curriculum and mastery assessments in the
high school by the end of the spring semester of 2019.
Stakeholder Goal (for students)
By the spring semester of 2019, 100% of students who participated in the competency-based
STEM pilot courses will be able to self-assess their inquiry skills and habits of mind, and
create action plans to ensure their growth in these areas and support their success in future
STEM courses.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

39

Table 3, continued
Assumed Knowledge
Influences
Knowledge Type
(i.e., declarative
(factual or
conceptual),
procedural, or
metacognitive) Knowledge Influence Assessment
Students need to know
the particulars of what
a career in STEM is
actually like.
Factual While not pursued in this study (because it
would likely be best learned outside the
cultural setting of the pilot course), this
influence could be assessed using a survey or
quiz meant to reveal the depth and accuracy of
students’ understanding of STEM careers.
Students need to
understand the steps of
inquiry inherent in
STEM disciplines.
Procedural Examples of survey items used in this study (4
point, forced choice Likert scale items):
● “In this class I learned not just about
science but how to do science.”
● “Collaboration with others is one
important step in the process of
scientific inquiry”
● “I know the steps to follow in order to
interpret authentic, messy data sets.”
Examples of interview questions used in this
study:
● “What is the difference between
knowing science and being able to do
science?”
● “Can you describe a specific lab
experience to me?”

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

40
Table 3, continued
Assumed Knowledge
Influences
Knowledge Type
(i.e., declarative
(factual or
conceptual),
procedural, or
metacognitive) Knowledge Influence Assessment
Students need to
develop science
metacognition skills —
to include laboratory
self-regulation and
conceptual self-
reflection.
Metacognitive Examples of survey items used in this study (4
point, forced choice Likert scale items):
● “Once I can identify what I do and do
not understand in this course, I am able
to use that knowledge in new ways.”
● “I can adjust my process in the lab
when my procedure is not producing
useable data.”
Examples of interview questions used in this
study:
● “How did you set goals for yourself in
this class?”
● “What strategies did you use when you
were stuck or frustrated?”
Document Analysis — student work that
demonstrated how students can create action
plans to grow their skills and habits of mind
for engagement in STEM inquiry.
Students need to reflect
on their own beliefs
about how gender roles
relate to science.
Metacognitive Examples of survey items used in this study (4
point, forced choice Likert scale items):
● “There is a relationship between gender
and science ability.”
● “People of my gender are better at
science.”
Example of interview question used in this
study:
● “Describe your sense of your science
identity.”

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

41
Motivation
In addition to gaps in knowledge, motivational issues are considered another possible
root cause for organizational problems (Clark & Estes, 2008). Motivation can be defined as “an
internal state that initiates and maintains goal directed behavior” (Mayer, 2011, p. 39). Pintrich
(2003) described motivation as the combination of energization and direction, answering the
questions of what moves individuals and towards what. Motivational problems can arise during
three phases of task achievement: (1) getting starting, or active choice; (2) sticking with it, or
persistence; and (3) completing it to a high standard, or mental effort (Clark & Estes, 2008). Lee
et al. (2015) argued that the motivational gap for women in STEM occurs in the persistence
phase of task achievement. Unequal representation of girls in math-intensive STEM courses in
high school could be considered an active choice problem in the short term, but over the course
of a girl’s science education it could be looked at as a persistence issue. There are many factors
that influence motivation, however in this study, literature on self-efficacy theory, attribution
theory, and goal orientation theory will be reviewed.
Self-efficacy theory. Self-efficacy is a student’s understanding of his or her own
competence on a particular task in a particular subject area (Peters, 2013). Self-efficacy is
influenced by four key factors: (1) mastery experiences; (2) vicarious experiences or modeling;
(3) social persuasions, ranging from positive feedback to ingrained cultural stereotypes; and
(4) physiological states, particularly stress responses (Bandura, 1998; Pajares, 2006). Self-
efficacy deficits can disrupt motivation in any of the three phases of task achievement
highlighted by Clark and Estes (2008). Those with low self-efficacy avoid difficult tasks, give
up more quickly when they encounter difficulties, and are slower to recover their efficacy in the
face of setbacks (Bandura, 1998).
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

42
Girls’ self-efficacy in math-intensive STEM. In order to maintain motivation, girls
need to believe that they are capable of the mathematical and laboratory tasks required in math-
intensive STEM courses. Uitto (2014) has found that boys have generally higher self-efficacy
than girls, with significant discipline-specific differences in math, chemistry, and physics.
Interview data from 321 students (157 girls, 164 boys), between the ages of 16–19, revealed this
gender disparity in self-efficacy in math-intensive STEM courses. It has also been proposed that
self-efficacy influences may be gender specific. While a boy’s sense of self-efficacy is largely
enhanced via mastery experiences, girls’ self-efficacy is more powerfully built via vicarious
experiences and social persuasions (Zeldin et al., 2008; Zeldin & Pajares, 2000). In 2000, Zeldin
and Pajares interviewed 15 women in math-intensive STEM careers (ages 26–53) to determine
the source of their resilience in both their studies and their career pursuits. In 2008, Zeldin et al.
conducted a similar set of interviews with 10 men in math-intensive STEM careers (ages 24–64).
Their analysis revealed that the most significant efficacy influences varied by gender, with the
men more heavily reliant upon mastery experiences. However, Bandura (1998) believed mastery
experience to be the most influential efficacy-enhancing factor, regardless of gender.
Whatever the rank of mastery experiences among the four factors that influence self-
efficacy, the relationship between mastery and self-efficacy is inherently cyclic. Self-efficacy is
predictive of performance, which in turn, affects self-efficacy. DiBenedetto and Bembenutty
(2013) found that participation and persistence increase in cases where there is a positive
feedback loop where mastery experiences are enhancing self-efficacy, which then causes
improved performance thereby creating additional mastery experiences. However, given the
potentially more significant impact of social persuasions and vicarious experiences on girls’ self-
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

43
efficacy, educators should be mindful of optimizing those efficacy influences for girls in math-
intensive STEM courses (Leaper et al., 2012).
Attribution theory. Attribution theory is based on the human drive to make sense of
events. Anderman and Anderman (2006) described three causal dichotomies that people employ
to make sense of outcomes: locus (internal or external), controllability, and stability over time. It
is important to note that it is irrelevant if a person correctly assigns attributions, because the
power of attribution as an influencer on motivation depends only on what a person believes to be
true (Anderman & Anderman, 2006). When a student believes that ability is malleable and effort
is valuable, that is referred to as a growth mindset and is synonymous with attributing success or
failure to internal, controllable, and unstable causal conditions (Hochanadel & Finamore, 2015).
Girls’ attributions in math-intensive STEM. In the language of attribution theory, for
girls to be motivated in math-intensive STEM courses they need to believe that their
performance is internally caused (success is based on their effort and habits of mind rather than
environmental causes), controllable (they control their engagement and effort rather than
believing their inherent intelligence is fixed), and unstable (conditions of being confused or
experiencing setbacks must be viewed as temporary) (Anderman & Anderman, 2006;
Hochanadel & Finamore, 2015). Increasing girls’ procedural knowledge about modes of
scientific inquiry and the habits of mind of a scientist could help to realign their attributions with
a growth mindset (Kang & Keinonen, 2017; Lundeberg & Moch, 1995). Bauer (2005) defined
scientific habits of mind as “the shared values, attitudes, and skills of the cultural tradition of
science” (p. 1864). These values, attitudes, and skills include the competencies in collaboration,
critical thinking, and embracing challenge that the Hawaii School mission aims to instill in its
students (Bauer, 2005).
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

44
Deemer (2015) warned that the underrepresentation of girls in math-intensive STEM
could work in direct opposition to developing internal, controllable, unstable attributions.
Critical mass theory argues that motivation is influenced by environment, and survey data from
255 female undergraduate students revealed that many women attributed their lack of satisfaction
with their performance in math-intensive laboratory courses to the gender imbalance in the
classroom (Deemer, 2015). These survey results describe an external, uncontrollable, and stable
set of attributions (Anderman & Anderman, 2006). As self-efficacy theory highlights the need
for vicarious experiences and modeling to build efficacy, so too might attributional motivation
rely, in part, upon having a critical mass of girls in math-intensive STEM (Bandura, 1998; Zeldin
& Pajares, 2000).
Goal orientation theory. Goal orientation theory is a social cognitive theory of
achievement motivation used largely to describe academic motivation. Goals can be mastery-
oriented, relating to improvement and benchmarking against one’s own previous performance, or
performance-oriented, which are driven by competition and comparing oneself to others (Yough
& Anderman, 2006). Both goal categories are further subdivided into approach and avoid types,
where mastery goals will be either approaching greater understanding or avoiding
misunderstanding and performance goals will be either approaching competitively superior
outcomes or avoiding seeming incompetent (Yough & Anderman, 2006). Dweck (1986)
suggested a relationship between a fixed mindset and performance goal orientation, and a
correlation between a growth mindset and mastery goal orientation. Furthermore, mastery-
oriented students are said to be more adaptive in the face of setbacks (Lau & Roeser, 2008;
Lewis, 2018). According to Duckworth and Quinn (2009), the construct of grit describes a
person’s perseverance in pursuit of long-range goals. The authors explained how grit allows
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

45
individuals to maintain focus and effort on their goals, even in the moments devoid of positive
feedback (Duckworth & Quinn, 2009). For improved gender equality in participation in math-
intensive STEM courses, girls will need to be guided to make mastery-oriented long-range goals
and develop the grit to stay committed to them.
Mastery goals of girls in math-intensive STEM. From a goal orientation standpoint, it
can be argued that girls’ motivation will be enhanced if they engage in math-intensive STEM
courses with the desire to master the skills and habits of mind necessary for authentic STEM
inquiry. They should not enter into these courses simply to obtain a GPA bump or a transcript
“check mark” for college admissions, nor should they focus on the grade over the actual
learning. Martin and Elliot (2016) have found that personal best goals, a type of mastery goal
where the target is set specifically against one’s own previous achievement, are the most
effective at maintaining engagement and motivation throughout a course. Meanwhile
performance goals have a negative effect on lasting engagement (Martin & Elliot, 2016). In a
survey of 1,160 high school students, it was found that those with mastery goals displayed higher
adaptive engagement, demonstrating better skills in planning, task management, persistence, and
perseverance in the face of setbacks (Martin & Elliot, 2016).
Lewis (2018) argued that mastery goals can actually be further divided into task- and
self- types. In a survey of students from nine sections of general chemistry at a large university,
Lewis (2018) was able to correlate task-specific mastery goals to higher academic achievement.
This result highlighted an important connection between metacognition and goal setting, as task-
specific mastery goals require the learner to first identify a particular deficit area for knowledge
or skill growth (Baker, 2006; Krathwohl, 2002; Lewis, 2018). Classroom environments that
support mastery goals could impact both knowledge and motivation gaps for girls in STEM.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

46
Table 4 provides information on the organizational mission, organizational and
stakeholder goals, and the motivation influences of self-efficacy, attributions, and goal
orientation. The table highlights examples of assessments that were used to study stakeholder’s
motivation.

Table 4
Assumed Motivation Influences and Assessments for Motivation Gap Analysis
Organizational Mission
The mission of the Hawaii School is to cultivate within each student the capacity to
collaborate, communicate, create, think critically, empathize, embrace challenge, engage
with a global perspective, and honor self and place. The academic mission has a specific
additional focus on developing students’ skills and habits of mind for engagement in
authentic inquiry.
Organizational Global Goal
Long Range Goal: The goal of the Hawaii School is to achieve improved gender equality, as
defined by a gender representation that approaches the demography of the student body
(48.2% male and 51.8% female), in participation in math-intensive STEM courses in the
high school by the fall semester of 2021.
Supportive Shorter Term Goal: The shorter term goal of the Hawaii School is to evaluate
and, if necessary, improve gender equity in self-efficacy in math-intensive STEM pilot
courses aligned to the new competency-based curriculum and mastery assessments in the
high school by the end of the spring semester of 2019.
Stakeholder Goal
By the spring semester of 2019, 100% of students who participated in the competency-based
STEM pilot courses will be able to self-assess their inquiry skills and habits of mind, and
create action plans to ensure their growth in these areas and support their success in future
STEM courses.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

47
Table 4, continued
Assumed Motivation
Influences Motivational Influence Assessment
Self-Efficacy —
Students need to
believe they are
capable of the
mathematical and
laboratory tasks
required in math-
intensive STEM
courses.
Examples of survey items used in this study (4 point, forced
choice Likert scale items):
● “I am confident that I am able to perform the laboratory
tasks required in this class.”
● “I am confident that I am able to perform the mathematical
tasks required in this class.”
Examples of interview question used in this study:
● “Tell me a story of a moment in class when you felt like
you could really do science.”
Attributions —
Students need to feel
their performance
depends upon their
sustained effort rather
than believing their
STEM aptitude is
fixed.
Examples of survey items used in this study (4 point, forced
choice Likert scale items):
● “Confusion is temporary. I can find ways to get unstuck.”
● “Persistence is important for success in science.”
● “If I perform poorly on an assessment, it is because I did
not try hard enough to learn the material.”
Examples of interview questions used in this study:
● “How was your persistence related to your learning in this
course?”
● “Did you feel as if you were in control of your ability to
master the material? Explain.”
Goal-Orientation —
Students need to
engage in math-
intensive STEM
courses with the desire
to master the skills and
habits of mind
necessary for authentic
STEM inquiry.
Examples of survey items used in this study (4 point, forced
choice Likert scale items):
● “As I approach class each day, my goal is to gain a deeper
understanding of science.”
● “My goals are about self-improvement.”
● “My goals are about outperforming others.”
Example of interview question used in this study:
● “Describe the ways in which you were able to align your
goals to the course competencies.”

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

48
Organizational Barriers
The final barrier to solving performance problems in an organization is the organization
itself. Clark and Estes (2008) identified the following as organizational gaps: poor work
processes, limited material resources, and dysfunctional interactions or ineffective goal
achievement in work that occurs between departments. The researchers emphasize that these
processes and interactions are shaped by the culture of the organization. Gallimore and
Goldenberg (2001) further explored the idea of organizational culture and articulated the
difference between cultural models and settings. Cultural models are the engrained, often
invisible, set of values that create the lens through which experiences are collectively interpreted.
Alternatively, cultural settings are the “occasions where people come together to carry out joint
activity that accomplishes something they value” (Gallimore & Goldenberg, 2001, p. 48). In a
high school, the important cultural settings include the formal gatherings in classrooms as well as
the informal groups that come together in other spaces on campus or at times outside the school
day. While cultural settings are fairly uniform from school to school, the cultural models of
different institutions will be more varied as they will be tied to the unique mission and history of
each school, along with the demography of its faculty, staff, and students. In this section both
cultural models and settings will be explored in relation to improving gender equity and
inclusion in math-intensive STEM subjects and the potential value of mastery curriculum.
Cultural models and settings for gender equity and inclusion. An organization that
values the inclusion of girls in all STEM fields and believes in the importance of achieving
gender equity in participation, engagement, and performance in these courses espouses the
necessary cultural model for addressing the current gender participation gap. However structural
barriers to gender equity in K-12 education are more often described in terms of the observable
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

49
manifestations of a cultural model, including school and classroom environments; teachers;
pedagogy, curriculum and assessment practices; and modes of peer interaction and collaboration
(Kanny et al., 2014). As such, it appears likely that shifts in organizational culture will begin
with intentional changes to cultural settings in schools, both inside and outside of the classroom
(Gallimore & Goldenberg, 2001).
Gender inclusive classrooms will need to utilize an individualized approach, matching
curriculum, pedagogy, and assessment decisions with the values and strengths of the diverse
students in the class (Aragón et al., 2017; Burkam et al., 1997; Chetcuti, 2008; Darby, 2005;
Demetriou & Wilson, 2009; Kanny et al., 2014; Ramsey et al., 2013; Schuster & Martiny, 2017).
Aragón et al. (2017) argued that the motivations and aspirations of underrepresented groups are
particularly fostered within a classroom environment that delivers intentional cues of acceptance
and belonging. Girls in STEM classes are often exposed to unsupportive and unwelcoming
classroom environments that only serve to deepen their belief in a connection between men and
STEM and reinforce their stereotypes about inherent, gendered STEM ability (Ramsey et al.,
2013). The two competing frameworks for promoting inclusion and addressing inequity are
multiculturalism and colorblindness (Aragón et al., 2017). While a colorblind framework strives
for equality by treating all students the same, multiculturalism strives for equity by creating
educational environments that celebrate the unique sets of values and skills that different groups
bring to the cultural setting. Gender equitable instruction has significant overlap with
individualized instruction, which emphasizes the use of relational pedagogies to meet the need of
each individual student (Chetcuti, 2008; Darby, 2005; Demetriou & Wilson, 2009).
Cultural settings beyond the classroom, particularly those settings that connect girls with
each other or with adult female STEM mentors, are an important additional consideration for
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

50
improving gender equity and inclusion in STEM (Bandura, 1977, 1986, 1998; Kekelis et al.,
2017; Leaper et al., 2012; Ramsey et al., 2013; Zeldin & Pajares, 2000). Self-efficacy deficits
are causally related to both girls’ underperformance and their inequitable participation in math-
intensive STEM, but there are known pathways toward improved self-efficacy (Bandura, 1977,
1986, 1998; Britner, 2008; Pajares & Miller, 1994; Reilly et al., 2015). While Bandura (1986)
argued that mastery experiences are the most crucial contributor to improved self-efficacy,
Zeldin et al. (2008) found that social persuasions and modeling are more significant influences
on girls. Social persuasions could include the support that comes from connecting girls in STEM
with other girls who share their interests, though clubs and teams; and modeling involves
providing girls with opportunities to be mentored by women working in STEM fields (Kekelis et
al., 2017; Leaper et al., 2012).
The mindful cultivation of cultural settings that enhance gender equity and inclusion are
clearly organizational level solutions, requiring allocation of resources, clear and consistent
processes for establishing and maintaining the cultural setting, and accountability measures to
ensure that all stakeholders are meeting goals and benchmarks (Clark & Estes, 2008). One way
in which many independent schools are trying to hold themselves and each other accountable to
more equitable, student-centered teaching is through their participation in the Mastery Transcript
Consortium (MTC).
Cultural models and settings with a mastery orientation. The cultural model in
support of a mastery curriculum in STEM embraces teaching the skills and competencies of
inquiry and research, and it values giving students an authentic opportunity to do science rather
than just learn facts and solve problems (Kang & Keinonen, 2017; Wang, 2013). Cultivating the
cultural setting for a mastery curriculum requires relational pedagogies that encourage students’
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

51
growth in non-cognitive skills like curiosity, resourcefulness, persistence, and resilience (Darby,
2005; Demetriou & Wilson, 2009; Fredricks et al., 2018; Goeden et al., 2015; Lundeberg &
Moch, 1995). In this mastery-oriented cultural setting, assessments will then reflect each
student’s progress towards meeting their mastery goals, emphasizing individualized growth in
skills and competencies over performance-based competition (Lau & Roeser, 2008; Lewis, 2018;
Yough & Anderman, 2006).
Mastery curriculum, pedagogy, and assessment have a positive effect on the affect, self-
efficacy, and engagement of girls in math-intensive STEM (DiBenedetto & Bembenutty, 2013;
Fredricks et al., 2018; Goeden et al., 2015; Kang & Keinonen, 2017; Lundeberg & Moch, 1995;
Senko & Tropiano, 2016; Simon et al., 2015; Vincent-Ruz & Schunn, 2017). While
performance goals tie success to outperforming others, mastery goals emphasize personal
learning and growth as the ultimate goals (Simon et al., 2015). A mastery orientation is found to
enhance metacognition and self-regulation in all students, but this framework is found
particularly beneficial at motivating girls as it interacts directly with their STEM self-efficacy
(DiBenedetto & Bembenutty, 2013; Senko & Tropiano, 2016; Vincent-Ruz & Schunn, 2017).
The inquiry-based nature of mastery pedagogy, which makes learning hands-on, collaborative,
and relevant, has been found to increase girls’ intellectual risk taking and build their leadership
capacity in math-intensive STEM courses (Fredricks et al., 2018; Goeden et al., 2015; Lundeberg
& Moch, 1995). Kang and Keinonen (2017) have found that students who engage in the actual
scientific process in high school are more likely to pursue graduate degrees in math-intensive
STEM fields.
While self-efficacy and goal orientation are motivational influences on students in the
problem of inequitable participation in math-intensive STEM subjects, it is crucial to begin the
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

52
work of addressing these barriers at the organizational level. Students cannot be expected to
adopt a mastery goal orientation within a cultural model that only values performance, and
students cannot enhance their self-efficacy within cultural settings that are not intentional about
creating improved gender equity and inclusion (Bandura, 1977, 1986, 1998; Britner, 2008;
Gallimore & Goldenberg, 2001; Lau & Roeser, 2008; Lewis, 2018; Pajares & Miller, 1994;
Reilly et al., 2015; Yough & Anderman, 2006).
Table 5 provides information on the organizational mission, organizational and
stakeholder goals, and the possible organizational influences of equity, inclusion, and mastery
curriculum. The table presents both cultural models and cultural settings and highlights samples
of assessments that were used to study these components of organizational culture.

Table 5
Assumed Organizational Influences and Assessments for Organizational Gap Analysis
Organizational Mission
The mission of the Hawaii School is to cultivate within each student the capacity to collaborate,
communicate, create, think critically, empathize, embrace challenge, engage with a global
perspective, and honor self and place. The academic mission has a specific additional focus on
developing students’ skills and habits of mind for engagement in authentic inquiry.
Organizational Global Goal
Long Range Goal: The goal of the Hawaii School is to achieve improved gender equality, as defined
by a gender representation that approaches the demography of the student body (48.2% male and
51.8% female), in participation in math-intensive STEM courses in the high school by the fall
semester of 2021.
Supportive Shorter Term Goal: The shorter term goal of the Hawaii School is to evaluate and, if
necessary, improve gender equity in self-efficacy in math-intensive STEM pilot courses aligned to
the new competency-based curriculum and mastery assessments in the high school by the end of the
spring semester of 2019.
Stakeholder Goal
By the spring semester of 2019, 100% of students who participated in the competency-based STEM
pilot courses will be able to self-assess their inquiry skills and habits of mind, and create action plans
to ensure their growth in these areas and support their success in future STEM courses.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

53
Table 5, continued
Assumed Organizational
Influences Organizational Influence Assessment
Cultural Model: The
organization needs to
sustain a culture of gender
equity and inclusion in all
STEM courses.
Example of survey item used in this study (4 point, forced choice Likert
scale items):
● “The Hawaii School culture promotes equal participation in
STEM by male and female students.”
● “When I am in science class I feel as if I belong.”
Examples of interview questions used in this study:
● “What is the Hawaii School culture around girls in science?”
● “Are girls encouraged and included in science the same as
boys? Explain.”
Cultural Setting: The
organization needs to
provide support and
resources for girls to
develop mentoring
relationships with each
other and with
professional women in
STEM.
While not pursued in this study (because it would occur outside the
cultural setting of the pilot course), this influence could be assessed
using interviews to understand the details of the mentoring
opportunities in STEM girls are seeking both on and off campus.
Cultural Model: The
organization needs to
place a high value on
developing all students’
competency in STEM
skills and habits of mind.
Examples of interview questions used in this study:
● “Describe what you believe to be the correlation between the
course competencies and the ability to do science.”
● “What made you feel like these skills mattered?”
Cultural Setting: The
organization needs to
develop and deliver STEM
curriculum that is focused
on mastery of skills over
content-based
performance.
Examples of survey items used in this study (4 point, forced choice
Likert scale items):
● “Compared to traditional grading, the competency-based
method gives me a better sense of what I do and don’t
understand.”
Examples of interview questions used in this study:
● “How would you describe the way that chemistry is taught in
this course?”
● “Describe the way your work is graded.”
● “How did you know when you had mastered a particular
competency?”

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

54
Conceptual Framework
A conceptual framework represents the interactions between the concepts, assumptions,
expectations, and theories that inform the methodological and analysis choices in a research
study (Maxwell, 2013). Merriam and Tisdell (2016) referred to this as a theoretical framework,
which they describe as the supporting structure or scaffolding of a study. The conceptual or
theoretical framework can be narrative or visual, with the goal of highlighting the relationships
between the important concepts, variables, and theories that frame the study (Maxwell, 2013;
Merriam & Tisdell, 2016). In earlier phases of enacting Clark and Estes’ (2008) gap analytic
framework to investigate root causes for a performance problem, influences on the problem are
intentionally developed in isolation and divided into their knowledge, motivation, or
organizational categories. The conceptual framework allows an opportunity to explore the
connections between the influences and to establish a working theory to frame a research
question (Maxwell, 2013).
In earlier sections of this chapter, the knowledge and motivation influences on female
high school students’ inequitable self-efficacy and their underrepresentation in math-intensive
STEM courses were explored. Knowledge and motivation influences were followed by an
exploration of the cultural models and settings active within the organization. In Figure 1, the
relationships between these influences are represented in a visual which serves to highlight the
working theory that framed this research study.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

55

Figure 1. Interactions between knowledge and motivation within the organizational models and
settings

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

56
Figure 1 represents the interactions between knowledge-based and motivational
influences on the underrepresentation of girls in math-intensive STEM; and it shows how
addressing these potential root causes could contribute to the achievement of the organizational
and stakeholder goals at the Hawaii School. The figure also models how the stakeholders’
knowledge and motivation are contained within the broader organizational context of the
institution’s existing cultural models and settings (Gallimore & Goldenberg, 2001). Addressing
the problem of participation gaps in math-intensive STEM will not be possible unless the Hawaii
School values gender equity and inclusion and enacts cultural settings that promote increased
participation for girls, particularly through choices of equitable curriculum, pedagogy, and
assessment (Aragón et al., 2017; Burkam et al., 1997; Chetcuti, 2008; Darby, 2005; Demetriou &
Wilson, 2009; Kanny et al., 2014; Ramsey et al., 2013; Schuster & Martiny, 2017). The
adoption of a mastery curriculum, which emphasizes each student’s individual growth in skills
and competencies rather than promoting a continued culture of competitive performance, is
known to improve girls’ self-efficacy and promote gender equity in participation, engagement,
and performance in STEM subjects (DiBenedetto & Bembenutty, 2013; Fredricks et al., 2018;
Goeden et al., 2015; Kang & Keinonen, 2017; Lundeberg & Moch, 1995; Senko & Tropiano,
2016; Simon et al., 2015; Vincent-Ruz & Schunn, 2017).
If the organizational barriers at the Hawaii School can be addressed through adjustments
to cultural models and settings, then the knowledge-based and motivational influences on the
students in relation to the participation gap would be the next hurdles to overcome in achieving
the institution’s goals. Clark and Estes (2008) suggested that when working with stakeholder
groups, it is crucial to address motivational influences first. As such, girls’ attributions and goal
orientation make an excellent starting place. Girls need to believe that their STEM aptitude is
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

57
not fixed, but rather that they can control their own engagement and performance through
sustained effort (Anderman & Anderman, 2006; Hochanadel & Finamore, 2015; Kang &
Keinonen, 2017; Lundeberg & Moch, 1995). This growth mindset is also a crucial component of
adopting a mastery goal orientation, which measures progress by looking at personal growth in
skills and competencies (Dweck, 1986; Martin & Elliot, 2016; Yough & Anderman, 2006). Both
attributions and goal orientation connect directly to the metacognitive knowledge influences of
self-reflection and self-regulation, as can be seen by the arrows on the figure (Baker, 2006;
Krathwohl, 2002; Lewis, 2018). Mastering these metacognitive skills can lead to improved
STEM performance, a stronger sense of STEM identity, and ultimately an increased STEM self-
efficacy (Bandura, 1977, 1986, 1998; Dembo & Eaton, 2000; Flowers III & Banda, 2016;
Goeden et al., 2015; Lundeberg & Moch, 1995; Mathabathe & Potgieter, 2017). As highlighted
on Figure 1, the metacognitive knowledge influence is also directly connected to the stakeholder
goal of girls’ self-assessment of their inquiry skills and habits of mind, and the creation of action
plans to ensure their growth in these areas and support their related success in future STEM
courses.
Procedural knowledge about the steps of inquiry in math and science is also a crucial
component of performance and, as such, is directly linked to improving girls’ STEM self-
efficacy (Bottia et al., 2015; Hong & Lin-Siegler, 2012; Kang & Keinonen, 2017; Lundeberg &
Moch, 1995; Wang, 2013). The most powerful conclusion to be drawn from Figure 1 is that all
other knowledge, motivation, and organizational influences ultimately connect, either directly or
indirectly, to self-efficacy. A significant amount of research concludes that STEM self-efficacy
is, in fact, the single most important influence on a girl’s choice to pursue STEM fields and her
persistence during the hardships she encounters during that pursuit (Bandura, 1977, 1986, 1998;
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

58
DiBenedetto & Bembenutty, 2013; Leaper et al., 2012; Pajares, 2006; Peters, 2013; Uitto, 2014;
Zeldin et al., 2008; Zeldin & Pajares, 2000). It is for this reason that an arrow is drawn directly
between self-efficacy and the organization’s global goal on Figure 1.
A close study of Figure 1 reveals a color-coding scheme of green, red, and yellow for the
arrows. The red arrows highlight relationships that existed in the literature, but which were not
the driving principles behind the pilot chemistry curriculum. The yellow arrows summarize the
foundational hypothesis of this conceptual framework, that self-efficacy is the most powerful
influencer on girls’ participation and persistence in math-intensive STEM. Grounded in this
hypothesis, the green arrows represent the relationships that were explicitly explored in this
research study. This conceptual framework offered the tentative theory that the cultural setting
of a mastery curriculum and a cultural model that values individualized mastery over competitive
performance can positively impact self-efficacy for girls in math-intensive STEM. More
specifically, this research project explored how cultivating a mastery goal orientation within girls
could improve their metacognition and ultimately impact their self-efficacy.
Conclusion
Chapter 2 explored possible root causes for gender gaps in self-efficacy in math-intensive
STEM subjects for high school students and explored best practices for improved gender
inclusivity in these courses. Bandura’s (1977) motivational construct of self-efficacy was
presented in terms of the factors known to improve girls’ self-efficacy in STEM fields; and the
relationships between these factors and gender equitable curricula, pedagogies, and assessments
were investigated. While there is a known connection between low self-efficacy and
participation gaps, a broader look at all knowledge-based, motivational, and organizational
influences on girls’ underrepresentation in math-intensive STEM fields was explored in the
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

59
second half of Chapter 2 using Clark and Estes’ (2008) gap analytic KMO methodology. A
conceptual framework for this dissertation was then established by highlighting the
interrelationships between the knowledge, motivation, and organization influences. While self-
efficacy is situated in the motivation category, its connections to other influences on the gender
participation gap in math-intensive STEM was clarified in the conceptual framework. Chapter 3
will present the study’s methodological framework that was used to explore the connection
between a mastery goal orientation, as promoted through the cultural setting of a competency-
based pilot course, and self-efficacy for high school girls in math-intensive STEM.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

60
CHAPTER 3
METHODOLOGY
Introduction
Chapter 3 will present the overall research design and the methods for data collection and
analysis that were used to pursue the three research questions that guided this study:
1. To what extent is the organization meeting its goals?
2. What are the knowledge, motivation, and organizational elements related to the
Hawaii School’s goal to achieve improved gender equity in self-efficacy in math-
intensive STEM courses aligned to the new competency-based curriculum and
mastery assessments in the high school by the end of the spring semester of 2019?
3. What are the recommendations for organizational practice in the areas of knowledge,
motivation, and organizational resources that may be appropriate for solving the
problems of gender inequities in self-efficacy and gender gaps in participation in
math-intensive STEM courses at the Hawaii School?
This chapter will first justify the choice of stakeholder group for the study and outline the
sampling and recruitment strategies. Following the discussion of participant recruitment and
selection, is a detailed description of the data collection methodology, a discussion of the
strategy for developing and implementing each measurement instrument, and an overview of the
data analysis techniques that were employed to process the data. The next sections describe the
steps taken to ensure the credibility and trustworthiness of the qualitative data and the validity
and reliability of the quantitative data. The chapter concludes with a reflection on the ethics,
limitations, and delimitations of the study.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

61
Participating Stakeholders
Due to the student-centered nature of the research questions, the stakeholder group for
this study was the Hawaii School students. The current enrollment in the high school at the
Hawaii School is exactly 1,700 students, with a gender breakdown of 51.8% female and 48.2%
male. Of these 1,700 students, there were 201 students (45.9% female and 54.1% male) enrolled
in regular chemistry for the 2018–2019 school year. The group of five teachers who teach the 11
sections of regular chemistry committed to piloting a competency-based curriculum with mastery
assessment in this course for the 2018–2019 school year, in keeping with the Hawaii School’s
investment in the work of the Mastery Transcript Consortium. For the quantitative portion of
this study, the entire population of students enrolled in this regular chemistry pilot course
(n = 201) was invited to participate in the study survey. A response rate of 85.07% was achieved
on the survey with 171 students (45.03% female, 53.22% male, and 1.75% students self-
identifying as non-binary/prefer not to answer) agreeing to participate. Following an explanatory
sequential mixed methods design, a sample of girls (n = 12) was then purposefully selected from
the survey respondents to be interviewed in the qualitative phase of the methodology. Given the
intent of the survey to be a census of the full population, the criteria for the selection of survey
respondents were less rigorous than for the interview participants.
Survey Sampling Criteria
Because the survey was intended to be a census sample, there was only one criterion for
selection for this measure. Because this study was focused on the effects of a new competency-
based curriculum and mastery assessment structure on the presumed KMO influences on the
problem of practice, participation was not considered from students who were not enrolled in the
course for the entire school year.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

62
Criterion 1: year-long enrollment. Students were enrolled in regular chemistry for the
entirety of the 2018–2019 school year. Given the timing of the survey in late March, nearing the
end of the third quarter, insights from students who had not participated fully in the course up
until that point were considered less representative of the full experience. All of the survey
respondents met this criterion.
Survey Sampling and Recruitment Strategy and Rationale
In the quantitative phase of this research, a non-random sampling strategy was employed.
Because the treatment in this study was a pilot curriculum that was being taught to all regular
chemistry students, there was no opportunity to create a control group and therefore no ability to
construct an actual experiment (McEwan & McEwan, 2003). The quantitative data was not
intended to produce generalizable conclusions, but rather to generate responses that could be
disaggregated by gender and then further investigated by a qualitative methodology that
specifically investigated the experiences of girls in the course (Creswell, 2014). As such, it was
wise to use a census technique to gather as much quantitative data as possible from the
population (Johnson & Christensen, 2015).
To achieve a high response rate, Fink’s (2013) advice was considered by making the
survey easy to complete and ensuring that the administration of the survey encouraged the
receipt of timely completed surveys from respondents. Parental consent and youth assent forms
were prepared for all students in regular chemistry (n = 201) in mid-February of 2019. Hard
copies of the forms were distributed to students by the researcher in each of the 11 sections of
regular chemistry over a two-day period. Digital copies of the consent forms were emailed to all
parents and guardians of the students over the same two days. Families were given a three-week
period to review the forms, ask questions, and return the signed forms to the researcher. Surveys
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

63
were then distributed via individualized, emailed invitations to only students with signed parental
consent and youth assent forms. Class time was set aside over a two-day period in early March
for students to complete the survey. The researcher visited each of the 11 sections of this course
again to administer the survey, and she provided a second explanation of the project and
encouraged honest participation.
Forms to allow participation in the survey were completed by 175 of the 201 enrolled
students and their families, generating an 87.06% consent rate. Of those with consent forms
submitted, 171 students completed the survey, representing 85.07% of the full population
participating in this census survey. The survey was administered midway through the second
semester, after the students had been immersed in the mastery curriculum, and the data was
compared with existing data that was collected by the school’s institutional researcher in the fall.
The interview sampling strategy was informed by trends in the results from the study survey and
meaningful categories that arose from comparing these survey results with baseline existing data.
Interview Sampling Criteria
Criterion 1: female. Students are female. The purpose of this study was to explore the
effectiveness of a competency-based pilot curriculum at closing self-efficacy gaps for girls and
to investigate the relationship between self-efficacy and the other presumed KMO influences on
gender gaps in self-efficacy and participation in math-intensive STEM. It was determined that
an in-depth analysis of the experiences of female students in the competency-based chemistry
course would provide the richest data and connect most directly to the research questions.
Criterion 2: range of self-efficacy changes. Students comprised a sample with an even
distribution among the categories for self-efficacy change determined by comparing the results
from the study survey to existing institutional data. One purpose of utilizing the explanatory
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

64
sequential design was to create the possibility for purposeful sampling of interview participants
with maximum variation represented. The four categories for self-efficacy change from the
period of September to March included: girls whose self-efficacy decreased, girls whose self-
efficacy remained static, girls whose self-efficacy increased by less than 30%, and girls whose
self-efficacy increased 30% or more.
Criterion 3: distribution among instructors. Students comprised a sample that was
taught by a variety of chemistry teachers, to ensure that any observed effect was not unique to a
single teacher. It was important that the understanding built from the cumulation of interview
responses was representative of the competency-based curriculum and mastery assessments, not
just the way in which a single teacher enacted these practices.
Criterion 4: maximum variation of entry year into Hawaii School. Students
comprised a sample with maximum variation in entry grade into the Hawaii School. It was
important to acknowledge that the variable of when students entered the school could have an
impact on their experiences with the curriculum. By ensuring maximum variation in this
category, the data set did not inadvertently overrepresent a particular entry grade level.
Criterion 5: maximum variation of race/ethnicity. Students comprised a sample with
maximum variation in race and ethnicity. This criterion was enacted for two reasons: (1) to
enable the possibility of investigating any particular effects of this curriculum aligned to racial or
ethnic identity; (2) to ensure that the understanding of the effects of this curriculum was not
limited to a narrow range of racial or ethnic identities.
Criterion 6: exclusion of researcher’s students. Students were not enrolled in the
researcher’s single section of regular chemistry. Given the relationship between the researcher
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

65
and her own students, the credibility of the interview findings would be questionable due to the
power dynamic and related impact of reflexivity in the conversation.
Interview Sampling and Recruitment Strategy and Rationale
As Merriam and Tisdell (2016) pointed out, nonprobabilistic sampling was appropriate
for this study because generalization was not the goal. Nonprobabilistic techniques include a
variety of purposeful sampling strategies, particularly those guided by principles of maximum
variation, which guaranteed the creation of a sample from which the most insight could be
derived (Merriam & Tisdell, 2016). For this study, the sample was purposefully constructed
based on the principles of maximum variation by recruiting interviewees that fell into a variety
of disparate categories from survey results, institutional baseline data, teacher, entry year into the
Hawaii School, and race/ethnicity. The sample for the second phase of the mixed study derived
from the results of the first measure; this is a design referred to by Johnson and Christensen
(2015) as nested sequential sampling.
On the parental consent and youth assent form, students and their families were able to
indicate their willingness to participate in a follow-up interview and consent to being audio
recorded if they were selected for an interview. One criterion for participation in the interview
was that the respondent identified as female. Of the 171 survey respondents, 77 were female;
however, only 29 of those students and families consented to a possible audio-recorded
interview. For 27 of those 29 students, baseline self-efficacy data existed from a measure,
distributed in early September by the school’s institutional researcher, that was used in the
creation of self-efficacy change categories. The desire to have baseline data about each
interview subject, therefore, eliminated two possible interviewees. In addition, one possible
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

66
interview subject had to be eliminated because she was enrolled in the researcher’s section of
regular chemistry.
The 26 remaining possible interview subjects were sorted into four categories of
chemistry self-efficacy change during the period of September to March: (1) students whose self-
efficacy decreased (n = 7); (2) students whose self-efficacy remained static (n = 6); (3) students
whose self-efficacy increased by less than 30% (n = 6); and (4) students whose self-efficacy
increased 30% or more (n = 7). These categories were determined by calculating the percent
change between the average of three responses on the baseline self-efficacy measure and the
average on two comparable items on the survey administered in March.
Three students from each category were selected to be interviewed, striving for maximum
variation of student ethnicity, teacher, and grade level of entry into the Hawaii School.
Additional survey items were reviewed for outlying or noteworthy responses, and that data was
also factored into selecting maximally diverse interview subjects. Merriam and Tisdell (2016)
suggest that enough subjects have been interviewed when a point of saturation or redundancy is
reached, where no new insights are being gained from additional conversations. After
conducting 12 interviews, the data was saturated as determined by significant typicality on a
number of codes.
Quantitative Data Collection and Instrumentation
This study used a mixed methodology to collect data from a sample of students engaged
in a competency-based pilot curriculum, in order to investigate the proposed KMO influences on
the problem of girls’ underrepresentation in math-intensive STEM courses at the Hawaii School.
The relationship between the mastery curriculum and girls’ self-efficacy in these chemistry
courses at the Hawaii School was also a particular focus of this evaluation study. A comparison
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

67
of two quantitative data sets was used to create categories for the purposeful sampling of
interview participants in the qualitative component of the study (Creswell, 2014). An existing
data set from the fall semester of 2018 was aligned with survey data from the spring semester of
2019 to look for trends in self-efficacy changes that inspired interview sampling categories.
While it was possible to use the study data to establish an apparent correlation between the
mastery curriculum and girls’ self-efficacy in these chemistry courses, this study was not a true
experiment with a control group and random sampling, therefore the results cannot be considered
generalizable (McEwan & McEwan, 2003). Instead, the overarching goal of this mixed methods
study was to gather rich data on the relationship between mastery, self-efficacy, and the other
presumed knowledge-based, motivational, and organizational influences on girls’ unequal
participation in math-intensive STEM courses at the Hawaii School.
Documents and Artifacts
Self-efficacy baseline. Nearly 400 students enroll in chemistry each year, being
recommended or self-selecting for either the honors or regular track. In the 2018–2019 school
year, 201 students were enrolled in regular chemistry. The regular chemistry sub-department
was piloting a new mastery curriculum in the 2018–2019 school year, in large part with the hope
to effect change in the way that the school’s lowest performing science students feel about
themselves as scientists. The Hawaii School utilizes its Institutional Researcher (IR) to engage
in continuous inquiry, particularly around curricular innovations on campus. Given the
chemistry sub-department’s goal, the IR distributed a self-efficacy questionnaire (see Appendix
A) at the beginning of the school year to inform the teachers’ instructional design and
pedagogical choices. Demographic data was collected from students who completed the survey
(n = 194), including: gender, graduation year, ninth grade science course, and entry grade into
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

68
the Hawaii School. The results from the IR’s self-efficacy questionnaire were utilized as
existing institutional data to serve as a baseline for self-efficacy in this study. This baseline was
compared against the self-efficacy questions in the study survey instrument, to assist in the
creation of the sampling categories for interview participants, triangulation of data from all
research measures, and analysis of the progress on the organizational goals (Bowen, 2009).
Additional document analysis. In addition to the institutional data on baseline self-
efficacy described in the previous section, course enrollment data and student work was
evaluated in this study. Given the organizational goal of improved gender equality in enrollment
in math-intensive STEM courses at the Hawaii School, enrollment data was a useful indicator of
any success in approaching this goal for students engaged with the mastery pilot. The 2018–
2019 and 2017–2018 STEM enrollment data for students who took regular chemistry during the
2017–2018 and 2016–2017 school years respectively, were compared against enrollment into
2019–2020 STEM courses by students who participated in the mastery pilot chemistry course
during the 2018–2019 school year (n = 185 of 201 pilot students enrolled in science courses for
2019–2020). Looking at enrollment trends by gender provided additional insights into the
impact of a competency-based curriculum and mastery assessment on students’ choices about
their further STEM pursuits and the overall effects on gender participation gaps.
The stakeholder goal for students in the pilot curriculum was for them to be able to self-
assess their inquiry skills and habits of mind, and create action plans to ensure their growth in
these areas and support their success in future STEM courses. Because the pilot curriculum was
built around attempting to increase students’ metacognition and self-reflection through mastery
assessment, it was possible to analyze student work for evidence of alignment with this
stakeholder goal. In consultation with the teacher team, it was determined that the best sample of
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

69
student work to analyze was the third quarter self-assessments completed by 91.54% of the
students in the pilot course (n = 184).
Surveys
Survey instrument. The survey (see Appendix B) included 33 questions that were
designed to align with the proposed knowledge, motivation, and organizational influences on the
problem of practice, as shown on this study’s conceptual framework (see Figure 1 in Chapter 2).
The analysis plan was carefully considered during survey construction to ensure that the results
would generate meaningful descriptive statistics about the influences (Pazzaglia, Stafford, &
Rodriguez, 2016). For this reason, and because the main goal of this quantitative data was to
generate categories for purposeful interview sampling, the survey was designed with all closed
questions (Fink, 2013). Students used a forced-choice, 4-point Likert scale to represent the
degree to which they agreed or disagreed with statements that probed the following: (1) their
procedural and metacognitive knowledge in a competency-based chemistry course; (2) their self-
efficacy, goal orientation, and attributions within a mastery assessment structure; and (3) their
understanding of the cultural setting of mastery learning in their chemistry classrooms and the
cultural model of gender inclusion in the larger context of the Hawaii School. Review by experts
was the technique used to test early drafts of the survey instrument in order to ensure actual
alignment of the KMO influences and the survey questions (Irwin & Stafford, 2016).
Survey procedures. Surveys were distributed to each consenting student (n = 175 of
201 possible students) in the 11 sections of regular chemistry via a personal invitation to a
Google form linked to their school email address. Of those consenting to take the survey, 171
students (85.07% response rate) completed the survey. Students used their student ID number to
identify themselves on the study survey, and their results were aligned to the existing baseline
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

70
self-efficacy data taken at the onset of the school year by the IR. The seven-digit student ID
numbers were the only identifying information in the raw data set, maintaining student
confidentiality as the researcher performed aggregate analysis. The survey data was collected in
March of 2019, and was utilized to create self-efficacy sampling categories for the qualitative
component of the study that took place throughout the month of April and into early May of
2019. While the comparison of baseline data from the fall semester with the self-efficacy
questions on the spring semester survey was important in generating interview sampling
categories, the aggregate quantitative analysis of the data on all surveyed students’ knowledge
and motivation, and on their interpretations of the organization’s culture, contributed rich
insights to this study’s research questions (Pazzaglia et al., 2016).
Qualitative Data Collection and Instrumentation
The qualitative phase of this study was conducted as a follow up to the quantitative data
analysis, in an explanatory sequential mixed methodology (Creswell, 2014). Rather than simply
looking to see if the curricular intervention appears to be correlated with changes in self-efficacy
for girls, the qualitative portion of the study was designed to investigate how the changes in
curriculum, pedagogy, and assessment contribute to self-efficacy changes. The qualitative
component probed more deeply into this phenomenon within the culture and context of the
Hawaii School by looking at the ways in which participants understood and assigned meaning to
their own learning experiences (Creswell, 2014; Merriam & Tisdell, 2016). The structured
interviews revealed girls’ experiences in the mastery curriculum and their self-efficacy
development, while also investigating the contributions and interactions of other knowledge,
motivation, and organizational factors on the underrepresentation of girls in math-intensive
STEM courses.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

71
Interviews
Interview protocol. The interview protocol (see Appendix C) began with introductions
and an overview of the goal of the interview, and then delivered 12 open-ended questions that
tied directly to potential knowledge-based, motivational, and organizational influences on the
problem of practice. Questions aligned most heavily with Patton’s (2015) categories of
experience & behavior, opinion & values, feeling, and knowledge questions; and the interview
also highlighted questions that originated from Strauss, Schatzman, Bucher, and Sabshin’s
(1981) two categories of hypothetical and ideal position questions (as cited in Merriam &
Tisdell, 2016). Each question was written to connect with a specific KMO influencer on the
conceptual framework including: procedural or metacognitive knowledge, self-efficacy,
attributions, goal orientation, or cultural models and settings. Overall, the interview protocol had
a greater emphasis on questions that dug deeper into the construct of self-efficacy and Bandura’s
(1986) factors that affect self-efficacy development. The interview protocol was a structured
format, to ensure coherence in the data collected from interview to interview. The protocol also
included a list of possible probes after each of the 12 questions. The probes were also scripted,
both to maintain continuity and to ensure that the probes were well aligned with the KMO
influence under investigation in each question.
Interview procedures. Interview subjects (n = 12) were selected from the four self-
efficacy categories that emerged from the comparative analysis of survey results and existing
self-efficacy data on students, with a mindful variation of race/ethnicity, teacher, and entry year
into the Hawaii School. Three girls from each self-efficacy category were interviewed to
achieve saturation, as determined by high typicalities for a variety of codes (Merriam & Tisdell,
2016). Each participant was interviewed one time, in an interview that lasted between 25 and 30
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

72
minutes, during the months of April and May of 2019. This time frame aligned well with the end
of the school year, ensuring that interview participants had completed almost a full year of
instruction in the competency-based pilot course in chemistry. Interviews took place in the
researcher’s office in comfortable chairs overlooking the quad, which provided a safe and
welcoming location for the students (Bogdan & Biklen, 2007). The researcher recorded each
interview and uploaded the recording to NVivo transcription the same day. The interview
transcript was reviewed and corrected, by listening to the audio recording of the interview, and
notes were added to the transcript to reflect the nuances of each dialogue and categories of
student responses. The goal of these post-interview reflections was both to ensure coherent use
of the interview protocol and to make judgements about the level of saturation in the data
(Merriam & Tisdell, 2016).
Data Analysis
Quantitative Data Analysis
For every survey item, the frequencies were calculated for each of the four possible
Likert scale responses. Despite the large sample size (n = 171), overall percent agreement (based
upon respondents who agreed or strongly agreed) and percent disagreement (based upon
respondents who disagreed or strongly disagreed) was used preferentially over means for most
analyses and comparisons. This decision was made because keeping the data in categories of
agreement and disagreement rather than calculating averages is more conceptually aligned with
the ordinal nature of the data (Alkin, 2011; Stafford, 2006). Data was analyzed first for the
whole sample (n = 171), then disaggregated two ways for analysis by gender and by
race/ethnicity. The presentation of the data in Chapter 4 provides results for the aggregate and
highlights meaningful disparities discovered through the disaggregation.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

73
All survey items were considered valid representations of the thoughts of the population,
as no item response rate was lower than 98.83%. For items with lower than 100% response rate,
the percent agreement and percent disagreement calculations factored in the adjusted number of
respondents. No other data cleaning was necessary, given the high response rates on each item.
Survey items were clustered by theoretical construct and analyzed in conjunction with interview
data and document analysis to validate KMO influences on the problem of practice or to discover
assets in the competency-based curriculum.
Document analysis of self-efficacy baseline data was analyzed using the same methods
and descriptive statistics as the study survey. Analysis of enrollment data was somewhat similar,
in that percentages by gender were used to evaluate the differences between male and female
enrollment statistics.
Qualitative Data Analysis
Interview data underwent ongoing analysis both throughout and after the interviewing
process. Reflective reviews of each transcript were conducted shortly after each interview, using
Corbin and Strauss’ (2008) techniques for questioning and constant theoretical comparisons.
The researcher became acquainted with each transcript by reading with questioning to determine
meaning. Because the interview protocol was carefully aligned with the conceptual framework,
the early read-throughs of each transcript applied the lens of aligning interview responses to the
theories of knowledge, motivation, or organization that were being investigated as influences.
Early open coding occurred during this phase as particular phrases from the interview transcripts
were identified as the crucial data from each interview.
Upon completion of all interviews, a process of a priori coding was applied to the phrases
identified in open coding. A priori codes were derived from the conceptual framework and
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

74
aligned to presumed KMO influences. An iterative process was utilized to review the data for
alignment to a priori codes and question the data for evidence of empirical codes that could
produce additional insights. The axial coding process was largely driven by a deductive
approach, guided by the study’s conceptual framework, but the iteration allowed room for
inductive codes to emerge as important.
Once the axial codebook was constructed, the third phase of analysis involved looking for
patterns, codes, and themes that emerged in relation to the conceptual framework and study
questions. One method for coalescing axial codes into patterns was the use of typicality,
identifying the percentage of interview subjects who aligned to particular axial codes. In the
presentation of interview data in Chapter 4, typicality is often used to strengthen an assertion.
Similar to the interview data, document analysis of student work samples also relied upon
typicality calculations from a priori codes aligned to the stated competencies in the competency-
based course.
Given the mixed methodology in this study, the researcher ensured that the qualitative
and quantitative data sets interacted with each other to produce additional insights or to deepen
the evidence for particular assertions.
Credibility and Trustworthiness
The terms credibility and trustworthiness are used to describe the quality of qualitative
research studies. Credibility refers to the quality of the findings, while the idea of
trustworthiness can be applied to both the researcher and the process utilized in the data
collection and analysis. Coherence between different phases of the research process, especially
between the literature review and the data collection and analysis, creates the conditions for a
trustworthy qualitative project (Merriam & Tisdell, 2016). In this process, the mindful alignment
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

75
of the literature review, research questions, conceptual framework, the interview protocol, and
data coding and analysis was an attempt to create this crucial coherence and a trustworthy
process.
In order to be a trustworthy researcher both a carefully constructed process and clearly
articulated positionality are necessary. The researcher must explore her biases while also
reflecting upon the reflexivity created when a human is the primary instrument of data collection
(Maxwell, 2013; Merriam & Tisdell, 2016). The researcher in this study was positioned as a
woman in STEM who is researching girls in STEM. She had a strong desire to improve the
conditions of equity for girls in math-intensive STEM settings, and her passion for this topic
could lead to hopeful interpretations of data in this evaluation study. The researcher was also
positioned in this study as a 13-year veteran teacher with a background in curriculum, pedagogy,
and assessment design. She had years of anecdotal evidence that supported her assumption that
girls benefit from learning environments in STEM fields where they are encouraged to pursue
learning mastery goals rather than competitive performance goals. The researcher needed to be
mindful of avoiding confirmation bias, and it was crucial to do as Maxwell (2013) suggested and
search carefully for discrepant evidence and negative cases. In addition to confronting biases,
the use of an interview required thoughtful management of reflexivity. Reflexivity describes the
researcher’s integrated relationship with the research subjects in qualitative methodologies,
which can lead to the subjects being unduly influenced by the researcher (Merriam & Tisdell,
2016). To address the pitfalls of reflexivity, a careful design process was used to ensure that
interview questions were not leading and the interview protocol was highly structured and
scripted. In addition, one criterion for participation in the interview was that the interview
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

76
subject was not a student of the researcher, ensuring that no personal relationship or power
dynamic existed between interviewer and interviewee.
To ensure the credibility of the research findings, the techniques of triangulation and
saturation were utilized. Triangulation is considered the best way to ensure credible findings,
and it involves the strategic comparison of data from different sources and multiple
methodologies (Merriam & Tisdell, 2016). A strong desire to have credible findings was one
motivation for utilizing a mixed methodology for this study, in which survey data, interview
data, and document analysis were woven together to address the research questions.
Interviewing until saturation was reached was another method used to increase the credibility of
findings, as continuing to seek answers until no new answers were emerging ensured a rich and
thorough data set.
Validity and Reliability
Validity and reliability are terms used to describe the quality of a quantitative study.
Validity is a reflection of how accurately a quantitative research instrument is measuring what it
intended to measure. Reliability can be broken down into descriptions of both internal and
external reliability. Internal reliability refers to the degree to which a study would yield the same
results if repeated, and external reliability describes the generalizability of the results to other
situations (Maxwell, 2013). Reliability can be considered a fruitless pursuit in qualitative
methodologies, where context, timing, and setting are crucial contributors to the research results
(Merriam & Tisdell, 2016). However, quantitative methodologies must strive to be both reliable
and valid.
There are three forms of validity for questions on a survey measure: content, criterion,
and construct validity (Salkind, 2017). Content validity was confirmed by sharing drafts of the
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

77
survey instrument with research experts who affirmed that the content of the conceptual
framework was well represented in the survey questions, and by conducting cognitive interviews
with peer researchers with early drafts of the instrument (Irwin & Stafford, 2016). These steps
also served to verify the construct validity, by checking that the questions were meaningful
measurements of the underlying psychological constructs, like self-efficacy and goal orientation
(Salkind, 2017). Beyond the validity of the individual questions, the validity of the study was
also dependent upon a meaningful response rate and respondents who completed the full survey.
The survey response rate of 85.07% in this study was significant, as it just surpassed the 85%
threshold necessary to make the sample responses automatically generalizable to the target
population (Pazzaglia et al., 2016). On the individual item level, the lowest response rate on any
of the 33 survey items was 98.83%, indicating that every item had a high enough density of
responses to be considered a valid representation of the population. To ensure valid responses
from students, the timing of the survey was mindfully scheduled to be when students weren’t
distracted by impending deadlines, major school events, or recent summative assessments that
could have left them tired or frustrated.
When considering reliability, the internal consistency reliability of the survey items was
of particular interest. Similar to construct validity, internal consistency reliability is a measure of
the alignment between a question and a singular dimension of interest (Salkind, 2017). In the
analysis of survey items, questions that aligned to the same theoretical construct of knowledge,
motivation, or organization were grouped together and the consistency of the responses was
investigated. It is important to note that while reliability and validity are closely related, that
reliability does not imply validity. A measure could be repeatable while consistently generating
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

78
incorrect data (Salkind, 2017). Social science research always has limitations due to the
complexity of studying human subjects.
Ethics
The driving ethical principle in research with human subjects is to do no harm to the
participants in a study (Glesne, 2011; Merriam & Tisdell, 2016; Rubin & Rubin, 2012).
Research participants must sign an informed consent document that: (1) describes both the
potential rewards and risks of involvement, (2) clarifies that participation is completely voluntary
and that subjects can quit the study at any time, and (3) describes the ways in which their
responses will be kept confidential (Krueger & Casey, 2009). Informed consent, alone, does not
dismantle the power dynamic between a researcher that is in a position of authority relative to the
subjects, and the subjects themselves (Glesne, 2011). Merriam and Tisdell (2016) argued that
beyond just the ethics of procedures, as prescribed by institutional review boards (IRB), there are
also layers of relational and situational ethics in qualitative research settings.
This mixed methods study began with a broad survey of students enrolled in regular
chemistry (n = 171) and then interviews of select female students (n = 12) based upon the survey
results. This required two separate informed consent processes, one for the survey measure and
a second for interview participation and audio recording. Because the subjects were minors,
thorough IRB review was required to ensure that the informed consent process involved both
students and their parents, and that the potential risk for these participants had truly been
minimized (Glesne, 2011; Rubin & Rubin, 2012). While the research incorporated mixed
methodology, qualitative interviewing of female students enrolled in regular chemistry courses at
the Hawaii School required the most careful oversight. A complete list of intended interview
questions, along with possible follow-up probe questions, was approved by IRB and the
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

79
researcher did not stray from the interview protocol during her interviews (Rubin & Rubin,
2012). In every informed consent document used throughout the study, it was made clear that
participation was voluntary and that students could extract themselves from the study at any
time, and this was reiterated at the time of survey administration and at the beginning of each
interview (Krueger & Casey, 2009). It was also made clear to students how the confidentiality of
the participants was maintained in the analysis of data and presentation of results by detailing the
processes for: removing student IDs from raw survey data during aggregate analysis of survey
data, using pseudonyms during interviews and on all transcribed interview data, and storing all
data in password protected files (Glesne, 2011).
The researcher’s relationship to the Hawaii School and to the research participants was
important in understanding the relational ethics of this study (Merriam & Tisdell, 2016). The
researcher was a senior administrator in the high school. The role of Assistant Principal for
Curriculum and Faculty put the researcher in a position of authority over the adults on campus,
but removed from the administration of student discipline as overseen by the role’s counterpart,
the Assistant Principal for Student Life and Athletics. Because the researcher was in a
supervisory role to all faculty in the high school, the choice to use the students as the research
subjects rather than the faculty avoided some relational complexity. However, students still
likely considered the researcher to be an authority figure. In addition, the researcher taught one
section (n = 22) of the 11 sections (n = 201) of the regular chemistry pilot course. For a teacher
to involve current students in a research study would have presented an ethical dilemma, where
students might have felt as though their participation in the study or their responses to interview
questions would impact their grade in the course (Merriam & Tisdell, 2016). For this study, that
dilemma was forgone through the decision not to include any of those 22 students as possible
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

80
interview subjects. This loss of 10% of possible qualitative research participants was a worthy
trade-off to avoid this ethical grey area.
The ethical considerations of a study extend beyond just the treatment of the test subjects
and include the honesty with which the researcher addresses her own biases (Merriam & Tisdell,
2016). The researcher entered in to this study with just over 13 years of science teaching
experience, where she had seen first-hand the gender differences in affect, confidence, and
efficacy of her students. In addition, the researcher had experience being an underrepresented
female in math-intensive STEM during her education, particularly when attending graduate
school in physical chemistry. Given this tension, interview questions were intentionally
constructed to not lead the participants to report on experiences in a way that resonated with the
researcher’s own background and feelings (Patton, 2002). The researcher’s previous
professional and personal experiences were, however, supported by theoretical and empirical
research, and all of these were used to construct the conceptual framework that hypothesized a
connection between mastery goals and self-efficacy (Maxwell, 2013). To ensure that the
research was ethical, other researchers were consulted in coding and interpreting the interview
data to avoid the pitfalls of confirmation bias. The researcher was also mindful to present the
research purpose to the study participants in a way that was clear but that did not ultimately
reveal the hypothesis and potentially influence responses (Glesne, 2011).
Limitations and Delimitations
A qualitative or mixed methods study does not have the luxury of a control group, a
random sample, and a true experiment that makes generalizability possible. As such, a careful
process of bounding the study and articulating limitations was necessary for a credible,
trustworthy, and ethical study (Merriam & Tisdell, 2016). The bounds of this study were visible
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

81
in the conceptual framework. The majority of the potential KMO influences discovered in the
literature review were represented, with the knowledge and motivation factors encapsulated
within the larger organizational context. Influences that would occur outside of the school
environment were not included in the conceptual framework or within the study, to narrow the
focus to what could be discovered about students’ experiences within the context of the mastery
learning initiative at the Hawaii School through their participation in the pilot chemistry course.
The questions that were asked on the survey and the interview connected directly to the
conceptual framework, creating coherence among all elements of this research process.
One delimitation of the interview protocol was to purposefully sample girls who have
been taught by a variety of chemistry instructors. This criterion helped to ensure that an
observed effect on self-efficacy was more likely from the new competency-based curriculum and
mastery assessment structure, rather than just from the effectiveness of a particular teacher.
Additional delimitations in this study included: the emergent qualitative design that incorporated
data analysis alongside data collection; the utilization of triangulation and saturation to increase
credibility; and the clarity of the researcher’s positionality within the study and within the
Hawaii School context (Bogdan & Biklen, 2007; Bowen, 2009; Merriam & Tisdell, 2016).
The limitations of this study arose mainly from the interviewer-respondent interactions in
the qualitative component of the study, based on the fact that both parties certainly brought
biases and predispositions that could affect both the in-person interactions and the data collected
(Merriam & Tisdell, 2016). The researcher was mindful to check her biases and create the
conditions of comfort, safety, and respect for the interview participants (Bogdan & Biklen,
2007). The researcher could not, however, control the respondent. It was impossible to know if
the student was giving honest responses or if she held a strong bias that swayed the way she
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

82
chose to answer questions (Weiss, 1994). It was also difficult to control how tired the
respondent was that day, if she happened to have a bad week, or if she was simply caught at a
moment of low efficacy or clarity. The knowledge of these limitations inspired the choices for
how to increase credibility and trustworthiness in this study. The combination of existing data
from early in the school year, survey data, and interviews gave both a time frame of analysis that
spanned nearly a school year and multiple points of data to triangulate with each interview. The
overarching goal of this research was to conduct a meaningful analysis of how the potential
KMO influences interact around the problem of gender inequities in self-efficacy and
participation in math-intensive STEM courses at the Hawaii School. The biggest limitation of
this study was that, however credible the results, they will not be generalizable beyond the
boundaries of the Hawaii School campus.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

83
CHAPTER 4
RESULTS AND FINDINGS
Introduction
A mixed methodology was utilized for this study, combining survey and interview data
with document analysis to answer the research questions. This chapter presents the results and
findings aligned to the study questions, beginning with a review of the organizational and
stakeholder goals and the degree to which the data indicated meaningful progress towards these
goals. Following the review of goals, the chapter is organized into knowledge-based,
motivational, and organizational assertions supported by numerical results and qualitative
findings. The chapter concludes with a summary of the validated KMO influences and the
discovered assets in the competency-based pilot chemistry course that will be used to generate
recommendations and an implementation and evaluation plan in Chapter 5.
The questions that guided data collection and which will frame this chapter were:
1. To what extent is the organization meeting its goals?
2. What are the knowledge, motivation, and organizational elements related to the
Hawaii School’s goal to achieve improved gender equity in self-efficacy in math-
intensive STEM courses aligned to the new competency-based curriculum and
mastery assessments in the high school by the end of the spring semester of 2019?
The third guiding question in this study will be addressed in Chapter 5:
3. What are the recommendations for organizational practice in the areas of knowledge,
motivation, and organizational resources that may be appropriate for solving the
problems of gender inequities in self-efficacy and gender gaps in participation in
math-intensive STEM courses at the Hawaii School?
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

84
Participating Stakeholders
Of the 201 students in the pilot chemistry course, 171 students (85.07% response rate)
elected to complete the study survey. Interviews were then conducted with 12 female survey
respondents, in an attempt to more deeply understand the factors that influence girls’ self-
efficacy and the other knowledge and motivation influences, in a mastery learning pilot course.
Table 6 compares the racial, ethnic, and gender breakdowns of the survey respondents and
interview subjects in comparison to the demography of the entire student body.

Table 6
Distribution of Study Participants by Race/Ethnicity and Gender
Student Body Data
from INDEX
(n = 1700)
Survey
Respondents
(n = 171)
Interview
Subjects (n = 12)
Race/Ethnicity
African American / Black 0.34% 0.00% 0.00%
Asian American 22.20% 45.03% 25.00%
White non-Latinx American 8.30% 18.71% 16.67%
Latinx 0.40% 1.75% 8.33%
Multiracial American 21.30% 19.88% 33.33%
Native American 0.06% 0.58% 0.00%
Pacific Islander American 9.40% 11.70% 16.67%
Unsure / Unreported 38.00% 2.34% 0.00%
Gender
Female 51.80% 45.03% 100.00%
Male 48.20% 53.22% 0.00%
Non-Binary / Prefer Not to
Answer
0.00% 1.75% 0.00%

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

85
The survey respondents included 91 males, 77 females, and three students identifying as
non-binary or prefer not to answer. From among the 77 female survey respondents, interview
subjects were purposefully selected from the 26 girls who had also consented to an audio-
recorded interview and for whom baseline self-efficacy data existed from an earlier measure. In
selecting the 12 interview subjects, the 26 girls were sorted into four categories of chemistry self-
efficacy change during the period of September to March: (1) students whose self-efficacy
decreased (n = 7); (2) students whose self-efficacy remained static (n = 6); (3) students whose
self-efficacy increased by less than 30% (n = 6); and (4) students whose self-efficacy increased
by 30% or more (n = 7). These categories were determined by calculating the percent change for
each subject between their average on three responses on the baseline self-efficacy measure and
their average on two comparable items on the survey administered in March. Three students
were selected from each category of self-efficacy change, striving for maximum variation of
student race/ethnicity, teacher, and grade level of entry into the Hawaii School. Responses to
additional survey items were also reviewed for outlying or noteworthy responses, and that data
was factored into selecting maximally diverse interview subjects. Table 7 introduces the 12
interview subjects by providing each girl’s pseudonym, race/ethnicity, and self-efficacy change
data.
In the analyses that follow, interview subjects are referred to by their pseudonym and
direct quotes from their interviews are used to support the assertions.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

86
Table 7
Pseudonyms, Race/Ethnicity, and Self-Efficacy Data for Interview Subjects
Pseudonym Race / Ethnicity Fall Self-Efficacy
Spring Self-
Efficacy % Change
“Ardeth” Pacific Islander American 3.67 3.00 -18.18%
“Hannah” Multiracial American 3.33 3.00 -10.00%
“Lois” White non-Latinx American 3.67 3.50 -4.55%
“Jordan” Asian American 3.00 3.00 0.00%
“Melissa” Latinx 3.00 3.00 0.00%
“Moana” Multiracial American 3.00 3.00 0.00%
“Emma” Asian American 3.67 4.00 9.09%
“Rose” Multiracial American 3.00 3.50 16.67%
“Kamalani” Pacific Islander American 3.33 4.00 16.67%
“Yams” Asian American 3.00 4.00 33.33%
“Jessica” Multiracial American 3.00 4.00 33.33%
“Margaret” White non-Latinx American 1.67 2.50 50.00%

Data Validity
This study utilized a mixed methodology to investigate the research questions and
explore the validity of the presumed KMO influences on the problem of practice. A census
survey was a meaningful initial measure to gather information about the full population’s
experience within the competency-based pilot course. The response rate of just over 85% was
significant, as the sample (n = 171), therefore, met the threshold for automatic generalizability to
the target population of 201 students in the pilot course (Pazzaglia et al., 2016). The survey data
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

87
was grouped by theoretical construct on the conceptual framework and the items were analyzed
both in aggregate and disaggregated by gender and by race/ethnicity to investigate any
noteworthy disparities.
The explanatory sequential methodology was employed both as a data gathering and data
analysis technique, as interview data was explored as a way to better understand survey results.
The decision to interview only girls not only created a detailed understanding of the evolution of
self-efficacy for girls in the pilot course, but the single-sex interview data also provided
particular insights into any gender disparities that emerged in the survey results. Alternatively,
for survey items on which no gender gaps existed, it was presumed that the commentary of the
interview subjects (n = 12) could be representative beyond just the female members of the larger
population. In both of those cases, survey data for the sample of female interview subjects
(n = 12) was compared to survey results for the larger population when making decisions to
determine if the interview subjects could accurately speak for all female survey respondents
(n = 77) or for the entire population (n = 171).
During data analysis, decisions had to be made regarding the thresholds for constructing
arguments. When making assertions in the sections that follow, 70% agreement on survey items
was the threshold for asserting agreement, however results above 90% were considered more
compelling evidence. Interview data used as evidence generally relied upon 50% alignment
among interview subjects, but code typicalities above 75% were considered more conclusive.
Research Question 1: To What Extent is the Organization Meeting its Goals?
The Hawaii School is working on the long-range performance goal of achieving greater
gender equality in participation in math-intensive STEM courses, particularly in the most
advanced courses in math, physics, computer science, and engineering, by the fall of 2021. This
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

88
long-range goal was supported by a short term goal of executing a competency-based pilot
course in chemistry during the 2018–2019 school year and evaluating its effectiveness at
shrinking presumed self-efficacy gender gaps. Gender inequity in self-efficacy is shown to be a
crucial contributing factor to participation gaps in math-intensive STEM courses in high school
(Bottia et al., 2015; Sadler et al., 2012). Due to the internal nature of motivational influences
like self-efficacy, the stakeholder group of focus for this study was the students themselves. The
goal for that group was that, by the spring semester of 2019, 100% of students who participated
in the competency-based STEM pilot courses would be able to self-assess their inquiry skills and
habits of mind, and create action plans to ensure their growth in these areas and support their
success in future STEM courses. Table 8 highlights the three goals for which progress was
evaluated in this study.

Table 8
Organizational and Stakeholder Goals Evaluated in this Study
Organizational Performance Goals
Long Range Goal: The goal of the Hawaii School is to achieve improved gender equality, as
defined by a gender representation that approaches the demography of the student body
(48.2% male and 51.8% female), in participation in math-intensive STEM courses in the
high school by the fall semester of 2021.
Supportive Shorter Term Goal: The shorter term goal of the Hawaii School is to evaluate
and, if necessary, improve gender equity in self-efficacy in math-intensive STEM pilot
courses aligned to the new competency-based curriculum and mastery assessments in the
high school by the end of the spring semester of 2019.
Student Stakeholder Goal
By the spring semester of 2019, 100% of students who participated in the competency-based
STEM pilot courses will be able to self-assess their inquiry skills and habits of mind, and
create action plans to ensure their growth in these areas and support their success in future
STEM courses.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

89
In the sections that follow, data from this study will be utilized to answer the first
research question guiding this study, “To what extent is the organization meeting its goals?”
Document analysis of student self-assessments from the pilot courses provides evidence towards
the student stakeholder goal. A comparison of existing institutional data on baseline self-
efficacy with survey data from this study, along with analysis of interviews with female
members of the pilot course, provides evidence towards the organizational short-term goal.
Finally, an analysis of trends in course enrollment data, which especially investigates STEM
course enrollment for the fall of 2019 for the students from the competency-based pilot course,
illuminates progress towards the long-range performance goal of the Hawaii School. The
findings and results from this study demonstrate notable progress on the stakeholder goal and
short-term organizational goal, with further room for improvement on the long range goal.
Results and Findings in Support of Stakeholder Goal
Students in the competency-based pilot course in chemistry were administered a third
quarter assessment by their teachers that asked them to reflect on their progress and set goals for
the final quarter. Document analysis from the sample of students who were present in class on
that day, provided the opportunity to evaluate 91.54% (n = 184 out of 201 students) of the
stakeholders on their progress towards the goal of self-assessing their inquiry skills and habits of
mind, and creating action plans to ensure their growth. The data provided evidence that students
have satisfactorily accomplished the stakeholder goal.
Students in the pilot chemistry course can self-assess STEM skills and habits of
mind and create targeted action plans. In the middle of the second semester of the
competency-based pilot course in chemistry, the students were asked to choose at least four of
the 14 course competencies on which to assess themselves and provide specific evidence of their
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

90
growing mastery. Students were also asked to set goals for building their skills in additional
competencies during the fourth quarter. Table 9 articulates each of the 14 course competencies
aligned to their category of relevant STEM skill or habit of mind. The table also shows the
percentage of students who provided evidence of their mastery of each competency and the
percentage of students who set goals and created action plans aligned to particular competencies.
Table 9 reveals that by the midpoint of the second semester of this year-long,
competency-based pilot course, there were four competencies for which over half the students
(n = 184) could provide evidence of mastery: persistence (69.02%), collaboration (61.96%),
problem-solving (55.43%), and scientific inquiry (50.54%). Meanwhile, the data for goal setting
indicated that the majority of students (67.39%) focused specifically on personal responsibility
and the competency that states “Students can engage in metacognitive reflection to determine
their true level of understanding and create plans for growth.” In addition, problem solving
(32.61%) and both communication competencies (20.11% and 20.65%) were also frequently
targeted in students’ fourth quarter goals.
In competency-based classrooms, the content of the discipline is meant to be the vehicle
by which students are given opportunities to practice and master skills and habits of mind. Of
the 14 course competencies, the three competencies in the areas of fundamental laws, nanoscale
explanations, and patterns and relationships connected most directly to the content of a chemistry
course. The fact that students chose to provide evidence of mastery towards content-oriented
competencies at a lower frequency than they did for competencies directly linked to skills and
habits of mind, is an indication that students experienced the course as predominantly about the
larger mindsets and skillsets of science.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

91
Table 9
Percentage of Students Demonstrating Mastery and Percentage of Students Setting Goals
Aligned to Particular Competencies on the Results from Student Self-Assessment
STEM Skill or
Habit of Mind Specific Course Competency
Students
Providing
Evidence of
Mastery
(n = 184)
Students
Aligning to
Future
Goals
(n = 184)
Adapt & Apply
Learning to
Relevant Issues
Students can contextualize and assign meaning to abstract
concepts and problems to enhance their understanding of the
real world applicability of the knowledge and skills of a
discipline.
9.78% 4.35%
Collaboration Students can enact open-minded listening and thinking skills
to engage empathetically with their peers during group work.
61.96% 15.22%
Communication Students can optimize graphical or visual displays of
information to both analyze data and communicate findings.
13.59% 20.11%
Communication Students can communicate clearly both in writing and orally. 21.74% 20.65%
Fundamental
Laws
Students can utilize fundamental laws of nature to describe,
explain, and predict a wide variety of physical phenomena.
15.76% 8.70%
Nanoscale
Explanations
Students can articulate the relationship between nanoscale
structure and macroscale behavior.
14.67% 4.89%
Patterns &
Relationships
Students can identify relationships and patterns and use them
to understand phenomena and make predictions.
25.54% 4.89%
Persistence Students can grapple with uncertainty and create action plans
to navigate setbacks.
69.02% 9.78%
Personal
Responsibility
Students can engage in metacognitive reflection to determine
their true level of understanding and create plans for growth.
26.09% 67.39%
Problem-Solving Students can organize their thought process clearly on paper
in order to both enhance and express their own
understanding.
55.43% 32.61%
Problem-Solving Students can apply their understanding of mathematics to
data analysis and problem solving.
26.63% 16.30%
Scientific
Inquiry
Students can develop their understanding through
experimentation and observation.
50.54% 3.26%
Scientific
Inquiry
Students can sort and prioritize messy data to form new
research questions.
21.20% 11.96%
Social
Responsibility
Students can accept responsibility for the care of their
community and design ways to improve the lives of others.
10.33% 5.43%

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

92
Goal-setting examples from document analysis. From among the 184 samples of
student work, three were selected to illustrate the targeted, personalized, self-reflective nature of
students’ goal-setting for the fourth quarter. One female student from the pilot course shared
these three reflections and goals:
I have a hard time with the math in chemistry, so I would set a goal of not asking my
table-mates what they got until I have a finalized number that I think is correct. When I
perform an experiment I want to hypothesize my answer before I carry out the
experiment instead of just doing the experiment. I want to get better at predicting
outcomes in ideal conditions vs. real conditions by thinking about both when we are in
the lab.
One male student also demonstrated his ability to set self-reflective, targeted goals for his own
growth as a STEM student when he wrote:
I need to review what I say on my packets to make sure that I am communicating my
answers in the most effective way (I have had problems with this). I need to make sure
that I am doing the best that I can in group work. I feel that I either let others lead and
feel kind of useless, or I take too much control over an experiment so I need to notice
those things when they happen and try to find a nice balance, which may include inviting
others to participate. My last goal is to participate more in class (at least once a day if I
can?) because I’m not completely aware of my participation, but I know sometimes I
hang back because I don’t like attention.
This final example of fourth quarter goals from a female student demonstrated her understanding
of the importance of academic risk taking in the pursuit of STEM inquiry:
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

93
In the fourth quarter, I really want to work on putting myself out there and giving my
ideas even when they could be wrong. I want to stop focusing on what other people will
think of me and take my learning into my own hands. I also want to try new things in
chemistry that will get me out of my comfort zone, like doing new experiments without
knowing where they will lead me.
Conclusion. A review of self-assessments from 184 students in the pilot course (91.54%
of population) indicated students’ strong abilities to self-assess their inquiry skills and habits of
mind, and create action plans to ensure their growth in these areas and support their success in
future STEM courses. This study found that the stakeholder goal was satisfactorily
accomplished.
Results and Findings in Support of Short-Term Organizational Goal
Baseline self-efficacy data existed for 194 students in the pilot course (n = 201) from a
self-efficacy survey that was administered by the Hawaii School’s institutional researcher at the
outset of the course. Survey data for this study was collected from 171 students in the course in
the spring, and it included two items that evaluated self-efficacy. Using student ID numbers to
align the two data sets, the institutional researcher was able to provide pre- and post- self-
efficacy data for 161 students (80.10%) in the pilot course. All interview subjects (n = 12) were
members of that sample of 161 students. Data from the interviews and both surveys were used to
assess the Hawaii School’s progress toward the organizational goal of evaluating and, if
necessary, improving gender equity in self-efficacy in math-intensive STEM pilot courses
aligned to the new competency-based curriculum and mastery assessments. The results and
findings indicated that there was a gender self-efficacy gap at the outset of the course which
narrowed over the duration of the course.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

94
Self-efficacy gender gaps narrowed in the pilot chemistry course. A comparison of
institutional data from the fall and survey results from the spring indicates that an existing gender
self-efficacy gap narrowed for the students in the pilot course for whom both this pre- and post-
data was available (n = 161). Interview findings support a correlation between participation in
the pilot course and the narrowing gender gap.
Existing institutional baseline data. A review of three questions from a self-efficacy
measure administered in the fall revealed gender disparities in self-efficacy, with male student
averages outranking female student averages on all three questions by between 0.18 and 0.29
points higher on a four-point scale. Table 10 shows the mean of the Likert scale responses on
each question by gender and the disparities between the average male and female responses.

Table 10
Mean of Likert Scale Responses to Three Self-Efficacy Questions by Gender and the Difference
Between Male and Female Means
Gender
Q1. When complicated
ideas are presented in
chemistry class, how
confident are you that
you can understand
them?
Q2. How confident
are you that you
can master all the
learning outcomes
in chemistry?
Q3. How confident are
you that you can do
the hardest work that
will be assigned in this
chemistry class?
Male (n = 81) 3.11 3.14 3.14
Female (n = 73) 2.93 2.85 2.86
Non-binary / Prefer not to
ans (n = 7)
2.57 2.43 2.86
Difference between male
and female
0.18 0.29 0.28

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

95
The breakdown by gender of the responses on these three questions can be found in
Figures 2, 3, and 4 that follow. Female respondents (n = 73) indicated disagreement in greater
amounts on all three questions, with 17.81% disagreement on Q1, 21.92% disagreement on Q2,
and 34.25% disagreement on Q3. In comparison, male respondents (n = 81) registered their
disagreement at frequencies of only 7.41% on Q1, 8.64% on Q2, and 11.11% on Q3. The self-
efficacy gap on Q3, “How confident are you that you can do the hardest work assigned in this
chemistry class?” was the most disparate with 23.14% more of the females surveyed disagreeing
with this statement than their male counterparts. The students identifying as non-binary or
preferring not to answer the question about gender (n = 7) disagreed with these statements at a
frequency of 42.86% on Q1, 28.57% on Q2, and 28.57% on Q3. Figures 2, 3, and 4 illustrate the
Likert scale responses by gender on each of these self-efficacy questions from the institutional
data set from the fall of 2018.

Figure 2. Likert scale responses by gender on Q1 from institutional self-efficacy data set (2018)
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

96

Figure 3. Likert scale responses by gender on Q2 from institutional self-efficacy data set (2018)

Figure 4. Likert scale responses by gender on Q3 from institutional self-efficacy data set (2018)

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

97
Post-intervention results. On the survey administered in the spring to students in the
pilot course as part of this study, results on two items related to self-efficacy showed increasing
alignment of male (n = 85) and female (n = 73) responses for the set of students for whom pre-
course data also existed (n = 161). The smaller disparities between both the means and the
overall percent agreement by gender illustrated a narrowing of the self-efficacy gap that was
evident in the data set from the beginning of the course. On an item that stated “I am confident
that I am able to perform the laboratory tasks required in this class,” two males (2.35%) and two
females (2.74%) disagreed, 52 males (61.18%) and 52 females (71.23%) agreed, and 31 males
(36.47%) and 19 females (26.03%) strongly agreed. On this item the overall agreement
frequency for male respondents and female respondents was nearly identical, at 97.65% and
97.26% respectively. Figure 5 illustrates these results.

Figure 5. Likert scale responses by gender to survey item 10 about laboratory self-efficacy
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

98
On a survey item that stated “I am confident I can perform the mathematical tasks
required in this class,” 2 females (2.74%) strongly disagreed, 13 males (15.29%) and 11 females
(15.07%) disagreed, 49 males (57.65%) and 43 females (58.90%) agreed, and 23 males (27.06%)
and 17 females (23.29%) strongly agreed. Again, overall agreement was much more strongly
aligned than in the fall institutional data set with 84.71% of male respondents and 82.19% of
female respondents registering agreement with this statement. Compared to laboratory self-
efficacy, however, both male and female respondents felt less confident in their execution of the
mathematical tasks required in the course. Figure 6 illustrates these results.

Figure 6. Likert scale responses by gender to survey item 11 about mathematical self-efficacy

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

99
It should be noted that while seven students identified as “non-binary” or “prefer not to
answer” on the existing data set from the fall, that number dropped to three students on the
spring survey. When the institutional data from the fall was aligned with the survey data for this
study using student ID numbers, it was evident that five students who had originally selected
“prefer not to answer” had switched their gender designation to “male” on the spring survey. In
addition, one student who identified as male in the fall switched gender designation to “non-
binary” in the spring. These changes explain how the gender distribution of the 161 participants
changed from 81 male, 73 female, and seven non-binary/prefer not to answer on the fall data set,
to 85 male, 73 female, and three non-binary/prefer not to answer on the spring survey.
Self-efficacy change for non-binary students. Due to the fluctuations in the population
of survey respondents who did not select a male or female gender designation, the self-efficacy
changes for this group were considered three ways: (1) comparing mean responses on both
surveys for the original seven students who identified as “non-binary” or “prefer not to answer”
in the fall; (2) comparing mean responses on both surveys for the three students who identified as
neither male nor female in the spring; and (3) comparing the mean responses for only the two
students whose “prefer not to answer” gender designation remained unchanged from spring to
fall. Figure 7 illustrates these three comparisons.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

100

Figure 7. Pre- and post- self-efficacy mean scores for students identifying as non-binary or
prefer not to answer

Depending upon how this data is grouped, self-efficacy could have improved, declined,
or remained static for students not selecting male or female as their gender on one or both of the
surveys. However, because four of the five students who switched their designation to male in
the spring appeared in the fall institutional data as “prefer not to answer” rather than “non-
binary,” it is possible that confidentiality rather than gender identity was their motivation for
selecting that category in the fall. It may be that the promising improvement from an average of
2.62 to 3.07 is the least likely trend to actually represent gender fluid students. Further study
would be necessary to determine how the self-efficacy of students who identify as non-binary
actually changed throughout the pilot course.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

101
Self-efficacy changes for male and female students. Comparing the institutional data
for males (n = 81) and females (n = 73) in the fall to the survey data from this study aligned to
students who identified as male (n = 85) and female (n = 73) in the spring, self-efficacy changes
were greater for the female students. While female respondents’ mean self-efficacy increased by
0.25 points, from 2.88 to 3.13, the surveyed males experienced an average growth of only 0.10
points, from 3.13 to 3.23. Although the males in this comparison still demonstrated higher
efficacy than the girls nearing the conclusion of the pilot course, the gap had narrowed to only
0.10 points. (It should be noted that if this comparison were undertaken with the post- self-
efficacy results from only the original 81 students identifying themselves as males, the mean
value for the males would drop to 3.22, narrowing the gap further to only 0.09 points.) Figure 8
displays the changes in mean self-efficacy values for male and female respondents on the
existing baseline data and on the post-intervention study survey.

Figure 8. Pre- and post- self-efficacy mean score comparisons for male and female respondents
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

102
While a correlation between participation in the pilot course and a narrowing self-
efficacy gender gap is evident, it is not possible to establish causation in the absence of a true
experiment with random sampling and a control group. However, qualitative data from
interviews with girls in the course (n = 12) demonstrates that many girls believe their
participation in this course is related to their growth in self-efficacy. Nine of the interview
subjects (75%) used language that indicated their perception of how the pilot course changed
their confidence in chemistry tasks and skills.
Interview findings. Kamalani indicated that “in the beginning of the year, I would have
told you, like, I’m not good at science. It’s not one of my strong subjects,” but she went on to
say, “now I feel, like, confident enough in myself that I think I can do that experiment [or] I can
do that stoichiometry problem.” Likewise, Jessica described coming into the course with
reservations when she said, “well all throughout middle school and freshman year I did not like
science at all, like, I thought I was just not a science person,” but she discussed that through her
experience this year “I’ve actually enjoyed learning about it, and I know that I can now.” Ardeth
answered one question in the interview about strategies for overcoming obstacles with the
phrase, “because now I feel confident,” and Margaret revealed that “I understand chemistry more
than I ever thought I would” and how she believed that the pilot course “kind of upped my
science identity, I guess, because I understand it.” Lois also revealed her surprise and delight at
having exceeded her own expectations in this course when she described how “at the beginning
of the year I definitely struggled,” and she celebrated “that persistence that I have used to get
from that point where I was in eighth grade all the way to where I am now, which is liking
chemistry — it’s kind of a beautiful thing.”
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

103
Four of the girls made reference in their interviews to the inquiry nature of the course and
the confidence they built through being consistently asked to figure things out on their own,
particularly in the lab. Hannah described how “now I just really enjoy science because it’s, like,
it has pushed me to think on my own” and she mentioned that “in Chem if someone were to ask
me to help them . . . I probably could help them.” Melissa recognized her growing independence
in experimental problem solving when she reflected, “so I thought that was really cool how I
figured that out. Because I didn’t think I would have,” and Moana connected the quantitative
nature of the experiments to an increase in her overall abilities in science as she described “I’ve
become more of a science thinker, too, because I enjoy math. And then math always has a tie in
to science.” Jordan summarized that through her experience in the pilot course “now I’m more
independent . . . [and] more curious” and that her confidence increased “because I know I can do
it myself now.”
Conclusion. The analysis of self-efficacy data collected prior to participation in the
competency-based curriculum indicated a gender gap in self-efficacy. The gap narrowed during
the year of participation in the pilot course, and interview data supports a correlation between
mastery learning and improved self-efficacy for girls. Notable progress was made on the short
term organizational goal; however, the self-efficacy gap, while narrowed, was not yet eliminated.
Results and Findings to Assess Progress Towards Long-Range Organizational Goal
Document analysis of enrollment data from the last four school years was performed to
evaluate any possible early progress towards the Hawaii School’s long-range goal of improving
gender equality, as defined by a gender representation that approaches the demography of the
student body (48.2% male and 51.8% female), in participation in math-intensive STEM courses
in the high school by the fall semester of 2021. The analysis revealed increased enrollment in
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

104
both physics and AP levels of biology and environmental science for girls who participated in
the pilot course. With a simultaneous decline in boys’ enrollment in physics, the gender equality
improved in overall physics enrollment. However, enrollment in honors and AP Physics showed
a slight decline in gender equality.
Enrollment trends reveal a narrowing gender gap in overall physics enrollment but
no improvement in equality of enrollment in honors or AP physics. Institutional enrollment
data for the last four school years was used to track STEM course enrollment for students in the
two years after they completed regular chemistry. The analysis was performed to evaluate the
early effectiveness of the pilot curriculum at creating improved gender equality in enrollment in
math-intensive STEM courses.
Chemistry students’ science course enrollment the fall after completion. Course
enrollment data for STEM courses in the 2017–2018 school year for 152 students (70 female, 82
male) who took regular chemistry in 2016–2017, was investigated alongside both the comparable
data from 167 students (75 female, 92 male) who took regular chemistry in 2017–2018 and the
projected enrollment for 185 (80 female, 105 male) students from this year’s competency-based,
pilot chemistry course. While there were 201 students in the pilot course, two students in the
course were seniors and graduated upon completion of chemistry, and for the remaining 199
students there is only a two year science graduation requirement. Continuing in science after
chemistry is not mandatory, however 92.96% of the non-graduated students from the pilot course
elected to continue in science the next year.
Upon completion of regular chemistry, students have the following options for science
course enrollment: regular physics; science elective courses, including engineering and various
levels of computer science; and the AP biological sciences, AP Biology or AP Environmental
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

105
Science (APES), provided that they were successful in their freshman biology course. Students
can enroll into honors or AP physics with special permission from their academic deans and
science faculty. Regular chemistry students are generally excluded from AP Chemistry, as
honors chemistry is a prerequisite for that course. Given these choices, the pursuit of any level
of physics or the AP biological sciences are considered the main pathways to pursue scientific
rigor after regular chemistry, while enrollment in physics is the only of those choices that aligns
to pursuing further math-intensive STEM. Figure 9 highlights the enrollment into physics and
AP biological sciences for the last three years of regular chemistry students in the fall
immediately following their completion of chemistry. The data points aligned to the year 2019
represent the students who participated in the competency-based pilot course.

Figure 9. Enrollment statistics by gender in physics and AP biological science courses in the fall
following participation in regular chemistry
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

106
Figure 9 illustrates that physics enrollment immediately following the completion of
chemistry declined for both genders from 2017 to 2018. For male students, 81.71% of male
chemistry students (n = 82) from 2016–2017 enrolled in physics in 2017, while only 65.22% of
male chemistry students from 2017–2018 (n = 92) enrolled in physics in 2018. The female
students experienced a less dramatic decline, with 60.00% of female chemistry students from
2016–2017 (n = 70) enrolling in physics in 2017, and 49.33% of female chemistry students from
2017–2018 (n = 75) enrolling in physics in 2018. Enrollment data from the pilot course students
shows a loss of another 4.27% of male students (n = 105) enrolling physics in the fall of 2019,
and a small, 4.42% increase in female students (n = 80) opting to take physics immediately
following their completion of chemistry. The gender gap between the percentage of chemistry
students electing to take physics immediately following chemistry narrowed over the three year
period, and female physics enrollment improved slightly with the pilot students.
Over the same three year period, enrollment in the AP level courses in biology and
environmental science immediately following the completion of regular chemistry has been on
the rise for female students, increasing from 30.00% in 2017 to a projected 47.50% in 2019.
Meanwhile, male students who just completed regular chemistry have oscillated between 21.90%
and 28.26% in their enrollment in these AP courses over the same three year period. A higher
percentage of female students each year are electing to challenge themselves with an AP science
course in the year immediately following chemistry. Female students from the pilot chemistry
course continued this upward trend.
Chemistry students’ science course enrollment one full year after chemistry. To fully
understand the continued science course trajectory for students who took regular chemistry, a
look at their course enrollment data a full school year after completing chemistry is also
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

107
necessary. Most students take chemistry in their sophomore year, and therefore have two more
years to pursue STEM coursework. While second year enrollment data will not be available for
students from the pilot course until Summer 2020, a look back at trends for the previous two
years gives an indication of the baseline. Figure 10 presents data about enrollment in the Fall of
2018 for students who participated in regular chemistry during the 2016–2017 school year and
enrollment for the upcoming Fall of 2019 for students who studied chemistry in 2017–2018.

Figure 10. Enrollment statistics by gender in physics and AP biological science courses one full
year after completing regular chemistry

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

108
Figure 10 shows that for the last two years, female students who took chemistry as
sophomores enrolled in physics at a steady frequency of 15.49% to 15.85% in their senior years.
Meanwhile, male students who took chemistry as sophomores have shown a slight decline in the
frequency with which they take physics as seniors, with 17.28% of the male students from the
2016–2017 chemistry course (n = 81) enrolling in physics in 2018 and 11.58% of male chemistry
students from 2017–2018 (n = 95) enrolled in physics in the fall of 2019.
Enrollment in AP Biology and AP Environmental Science in the senior year declined for
girls who took chemistry as sophomores, dropping from 77.46% in 2018 to 59.76% in 2019.
This decline during the senior year from the ’16–’17 chemistry students to the ’17–’18 chemistry
students, aligns with the large increase in enrollment in these courses for these girls as juniors, as
was seen in Figure 9. There was also a slight decline in male students enrolling in AP Biology
and AP Environmental Science as seniors after having taken chemistry as sophomores, dropping
from 66.67% in 2018 to 60.00% in 2019. In the summer of 2020 the pilot students’ science
course enrollments can be added to this analysis to investigate any impacts on trends.
The current complete enrollment data for these three groups of chemistry students is
summarized in Table 11. The table illustrates that in addition to a narrowing of the gender gap in
enrollment in physics with the pilot students, there was also more equality in enrollment into
computer science and engineering, although the numbers were low and ranged between 2.50%
and 4.76% of regular chemistry students selecting to take these courses.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

109
Table 11
Science Enrollment Data by Gender for the Two Years Following Completion of Regular
Chemistry
Physics
AP Bio
& APES
Comp
Sci Engineering
Science
Electives
’16–’17
chem
students
’17–’18
enrollment
Female 60.00% 30.00% 1.43% No data 14.29%
Male 81.71% 21.95% 2.44% No data 3.66%
’18–’19
enrollment
Female 15.49% 77.46% 1.41% 0.00% 38.03%
Male 17.28% 66.67% 13.58% 7.41% 41.98%
’17–’18
chem
students
’18–’19
enrollment
Female 49.33% 45.33% 0.00% No data 20.00%
Male 65.22% 28.26% 6.52% No data 15.22%
’19–’20
enrollment
Female 15.85% 59.76% 6.10% 1.22% 48.78%
Male 11.58% 60.00% 5.26% 6.32% 28.42%
’18–’19
PILOT
chem
students
’19–’20
enrollment
Female 53.75% 47.50% 3.75% 2.50% 17.50%
Male 60.95% 21.90% 4.76% 2.86% 17.14%

Chemistry students enrolling in honors or AP Physics. Despite a trend towards more
equal enrollment in math-intensive STEM courses for the pilot students and increased enrollment
in physics overall for female pilot students, the enrollment into honors or AP level physics
declined slightly for girls. With this drop, the existing gender gap widened from an 8.30%
difference between the percent of girls and boys enrolling into higher level physics straight out
of chemistry in 2018, to 10.89% more male students pursuing honors or AP physics out of the
pilot course. Figure 11 illustrates the honors and AP physics enrollment statistics for the three
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

110
groups of chemistry students over their two years after completing chemistry. It should be noted
that in 2017 separate courses existed for honors physics and AP Physics 1, but in 2018 the two
courses were merged together. Combining the courses seemed to correlate with a decline in
enrollment from regular chemistry students of both genders. The pilot students appear at the top
of the figure, with 6.25% of the girls and 17.14% of the boys from the pilot course enrolling in
honors or AP physics in the fall of 2019.

Figure 11. Enrollment statistics by gender of chemistry students taking honors or AP physics in
the two years following completion of regular chemistry

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

111
Conclusion. The analysis of three years of STEM enrollment data for courses that
students take after completion of regular chemistry showed a small increase in overall enrollment
in physics for female students after taking the pilot course. Coupled with a slight decline in male
enrollment in physics, the gender equality gap narrowed to only a 7.20% difference in
enrollment. However, trends in sign-ups for higher levels of physics for these students showed a
persistent and slightly worsened gender gap in enrollment. Early indicators show that the pilot
course may have been correlated with the decision to enroll immediately into another math-
intensive STEM course for a small number of girls, but it did not show promise at encouraging
enrollment into honors or AP levels of physics. A further investigation into all KMO influences
on the problems of self-efficacy and participation gaps in math-intensive STEM courses is
crucial in considering recommendations for making further progress on this long-range
organizational goal.
Research Question 2: Knowledge, Motivation, and Organizational Results and Findings
The next section of this chapter will address the second research question guiding this
study and investigate the relationship between presumed knowledge-based, motivational, and
organizational influences and the problems of gender gaps in self-efficacy and participation in
math-intensive STEM courses at the Hawaii School. While this is an evaluation study, the
context of the study participants being in a pilot course alters the nature of the evaluation.
Because the pilot curriculum was implemented with the hope to effect change on these gender
gaps, the evaluation is not just an attempt to validate root causes of the problem of practice but
also an investigation into the effectiveness of the pilot course at addressing presumed influences.
Not only will some pervasive KMO influences be validated in the sections that follow, but some
KMO assets will also be identified from the pilot curriculum that show promise at narrowing
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

112
gender gaps in self-efficacy and participation. Both validated influences and identified assets
will contribute in Chapter 5 to the generation of recommendations and an implementation plan
for the organization’s continuing work to achieve gender equality in the pursuit of math-
intensive STEM courses.
Table 12 summarizes the presumed influences that informed both the conceptual
framework for this study and the generation of survey items and interview questions.

Table 12
Assumed Knowledge, Motivation, and Organizational Influences Evaluated in this Study
Assumed Knowledge
Influences
Assumed Motivation
Influences
Assumed Organizational
Influences
Procedural Knowledge —
Students need to understand
the steps of inquiry inherent
in STEM disciplines.
Self-Efficacy — Students
need to believe they are
capable of the mathematical
and laboratory tasks required
in math-intensive STEM
courses.
Cultural Model — The
organization needs to sustain
a culture of gender equity
and inclusion in all STEM
courses.
Metacognitive Knowledge
— Students need to develop
science metacognition skills,
to include laboratory self-
regulation and conceptual
self-reflection.
Attributions — Students need
to feel their performance
depends upon their sustained
effort rather than believing
their STEM aptitude is fixed.
Cultural Model — The
organization needs to place a
high value on developing all
students’ competency in
STEM skills and habits of
mind.
Metacognitive Knowledge
— Students need to reflect
on their own beliefs about
how gender roles relate to
science.
Goal-Orientation — Students
need to engage in math-
intensive STEM courses with
the desire to master the skills
and habits of mind necessary
for authentic STEM inquiry.
Cultural Setting — The
organization needs to
develop and deliver STEM
curriculum that is focused on
mastery of skills over
content-based performance.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

113
In the sections that follow, assertions will be made aligned to each of the nine influences
outlined in Table 12. The assertions will be supported by a combination of quantitative and
qualitative data from the study survey, interviews, and document analysis. After a review of
data, each section will conclude with the validation of influences that persist despite the
intervention of the pilot curriculum and/or identification of KMO assets in the competency-based
course.
Knowledge Results and Findings
A number of items on the study survey and questions in the interview protocol were
designed to investigate students’ procedural and metacognitive knowledge. The procedural
knowledge of the steps to perform scientific inquiry, by enacting the requisite skills and habits of
mind of a scientist, is thought to be crucial in engaging students in science and deepening their
confidence and commitment in STEM fields. The metacognitive strengths of conceptual self-
reflection and laboratory self-regulation are thought to enhance students’ abilities to accomplish
goals and experience feelings of mastery in STEM classes. The metacognitive deficit of
attaching gender stereotypes to one’s understanding of who should be interested in or good at
STEM fields, is thought to inhibit girls’ self-efficacy, interest, and intention to pursue STEM.
The results and findings presented in this section investigate these presumed knowledge
influences in the context of the competency-based pilot course in chemistry.
Students in the pilot course demonstrated procedural knowledge about scientific
inquiry. Survey and interview data from this study indicated that pilot students had developed a
strong understanding of both the steps involved in scientific inquiry and the habits of mind
required to perform inquiry.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

114
Survey results. Surveyed students in the pilot chemistry course revealed their nearly
unanimous belief that they had developed a strong sense of the skills and habits of STEM
inquiry, with 95.3% of survey respondents (n = 170) agreeing or strongly agreeing to a survey
item that stated, “in this class I learned not just about science but how to do science.” More
specifically, the survey data revealed that students in the study (n = 171) understood that
engaging in scientific inquiry requires the habits of mind for persistence (98.83% agreement),
collaboration (96.49% agreement), and creativity (90.06% agreement). Study participants
(n = 171) also demonstrated an understanding of the inquiry process when they reported 86.56%
overall agreement with the statement “in scientific inquiry, asking questions is more important
than answering questions,” and 73.10% agreement with “I know the steps to follow in order to
interpret authentic, messy data sets.” Figure 12 shows the Likert scale distributions of responses
to these five related survey items.
Figure 12 highlights that the largest frequency of disagreement occurred on survey item
4, revealing that nearing the conclusion of the pilot course, 46 students (26.90%) were still
struggling to interpret authentic, messy data sets. While equipped with the mindsets for science
inquiry, the pilot students’ procedural knowledge of how to tackle real data still showed room for
growth.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

115

Figure 12. Distribution of Likert scale responses to procedural knowledge survey items 1, 2, 3, 4,
and 19 about the skills and habits of mind required for STEM inquiry

Interview findings. Interview data provides further evidence of the pilot course students’
understanding of STEM inquiry skills and their ability to articulate the appropriate habits of
mind for engaging in authentic research. In the interviews, 100% of interview subjects (n = 12)
could describe the steps of scientific inquiry and every girl also spoke about the requisite habits
of mind required to engage in STEM research.
Knowledge of the steps of inquiry. Jessica described her experience in the pilot course
labs as “we have to be able to find the answer to these actual real life questions without being
given a procedure [for] it.” Melissa also described a typical day in the lab as “we weren’t given
procedures, we had to figure it out on our own,” and Emma agreed that students in the pilot
course had to “make our own kind of procedures.” Hannah argued that, in fact, “knowing
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

116
science is actually being able to do it by yourself without any instructions.” Moana explained
that most days in the lab the “teacher runs over the experiment, what we’re doing, but, like, not
exactly what we’re doing, we get to do more of our own thinking of what we should do.”
Margaret described the structure of the inquiry-based pilot course as “you had a goal you had to
get to, so you had to figure out how to get there.”
In the context of these self-directed lab challenges, some interview subjects discussed
their growth in lab inquiry skills. Ardeth explained that “now I know, like, what chemistry is
about and how to do it” and Lois mentioned “that I’ve grown in reference to being able to make
up my own experiments.” Hannah explained that in preparation for experiments she
independently “started doing a lot of more outside research by myself.”
Beyond just experimental design, the interview subjects revealed the steps of inquiry to
include performing multiple trials and using data to answer questions and make predictions.
Yams emphasized the iterative nature of authentic scientific inquiry when she stated “we were
given the chance to go back into the lab and test out . . . try again or, like, alter our methods and
see why it [our original result] was off.” Kamalani clarified that in the inquiry process “we’re
tasked to, like, have a problem and we have to figure it out using critical thinking and
collaboration, all these different skills to figure out the answer to it,” and Jordan explained that
“actually doing science” means that in the lab students will ultimately be able to “predict, I
guess, what’s going to happen.”
Skills and habits of mind for inquiry. A particular set of skills and habits of mind is
required to engage meaningfully in the process of designing experiments, troubleshooting
experimental set-ups, iterating on data collection to improve accuracy and precision, and
utilizing data to answer questions and make predictions. The interview participants (n = 12)
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

117
identified persistence, collaboration, communication, open-mindedness, goal-setting, creativity,
problem-solving and critical thinking as crucial to success in scientific inquiry. All 12 interview
subjects (100%) discussed persistence, nine (75%) indicated collaboration, three (25%)
mentioned communication, two (16.67%) made reference to maintaining an open mind, two
(16.67%) added goal setting, and one mention (8.3%) was explicitly made of creativity, problem
solving, and critical thinking. Because of the relationship between persistence and attribution
theory, a robust analysis of interview data on persistence appears in the motivation section of this
chapter. The quotes selected present an overall snapshot of how the interview subjects described
the skills and habits of mind of a scientist.
When asked how they would describe this chemistry course to a new student, many
interview subjects responded by giving advice about being in the lab. Jordan began by
suggesting that students must be “open to try new things. I think you have to have an open
mind.” Kamalani also advised that it is important to “have an open mind” and Rose indicated
that lab work involves a person’s “creative side, which can easily factor into science.” Hannah
warned that “if you don’t communicate, if you don’t have persistence, then you’re not really
going to get anywhere.” Melissa suggested “I think it is really important for collaboration
because you can’t do anything on your own,” and Margaret agreed that “collaboration you need
because you can’t just have one perspective or else it doesn’t really work.” Emma also
encouraged collaboration, mentioning that “I mean, it’s kind of, like, important to collaborate
with your peers, especially if you don’t understand something.”
When asked how the pilot chemistry course provided opportunities to build and practice
skills, Moana described that in her work with her lab group “we would all work as a group and
communicate the general process that we would do, and do it.” Kamalani felt that “problem
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

118
solving and critical thinking are definitely, and like having the collaboration part, is a big part of
it.” Jessica described her reliance on “collaboration and communication” because she discovered
that “it’s really hard to figure something out with just like a piece of paper and the equipment all
by myself, like, I have to be able to talk to people a lot.” Ardeth described the goal-setting she
undertook when tackling lab challenges by saying “in order to do this you need to set goals . . .
and not only the scientific portion of the lab, but also to persistence per se,” and Yams
summarized the inquiry mindset as “if you have the qualities of, like, of a go-getter kind of
person, then you’ll be a good scientist.”
Validated influences and identified assets. An identified procedural knowledge asset of
students in the pilot course was that they had a solid understanding of the process of authentic
scientific inquiry and they could articulate the skills and habits of mind necessary to engage in
inquiry. A lingering influence validated in the survey data was that 26.90% of students surveyed
still did not feel that they knew the steps to follow to interpret authentic, messy data sets. From a
gender equity stand-point, there was no meaningful disparity by gender on this influence with
only 1.50% separating the frequency of agreement between males and females.
Students in the pilot course demonstrated metacognitive skills through conceptual
self-reflection and laboratory self-regulation. Survey and interview data, along with
document analysis of student self-assessments, demonstrated the strong metacognitive
capabilities of the students in the pilot course.
Survey results. On survey items targeting the students’ ability to self-reflect and discover
their own knowledge gaps, there was a high frequency of agreement. When asked if “self-
reflection is a crucial part of improving my science understanding,” 88.89% of survey
respondents (n = 171) agreed or strongly agreed. On a similar item asking students to rank
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

119
agreement with the statement “Once I can identify and explain what I do and do not understand
in this course, I am able to use that knowledge in new ways,” 95.32% of surveyed students
agreed or strongly agreed. Two additional survey items targeted the metacognitive skills
involved in self-regulating in the lab. When asked to rank agreement with “I can adjust my
process in the lab when my procedure is not producing useable data,” 96.49% of survey
respondents agreed or strongly agreed; and when asked if “I view set-backs in the lab as
additional data to interpret not as failures,” 81.29% of surveyed students agreed or strongly
agreed. Figure 13 highlights the breakdown of Likert scale responses on these four survey items.

Figure 13. Distribution of Likert scale responses to metacognitive knowledge survey items 5, 8,
9, and 25 about self-reflection and laboratory self-regulation

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

120
Just as the analysis of procedural knowledge revealed some students still struggling to
articulate and execute the steps to analyze authentic messy, data sets; the survey data from item
25 on metacognitive knowledge suggested that 32 of the students surveyed (18.71%) were still
struggling to frame set-backs in the lab as data rather than failures as they wrapped up the third
quarter of the course.
Document analysis. Document analysis from students’ fourth quarter goal-setting also
highlighted some students’ sense that they needed to continue to develop the skills to work with
real data and iterate in the lab, as 11.96% of students in the document analysis data set (n = 184)
set goals aligned to the competency that states “students can sort and prioritize messy data to
form new research questions.”
Similar to survey items 5 and 8, document analysis also confirmed students’ commitment
to self-reflection and strategy development, as 67.39% of student self-assessments (n = 184)
revealed goals specifically aligned to the competency that states “students can engage in
metacognitive reflection to determine their true level of understanding and create plans for
growth.” However, even for students who did not align their goals to this particular competency,
the very act of self-assessing one’s progress on competencies, providing evidence of mastery,
and setting targeted, specific goals for continued growth is a demonstration of metacognition.
With thorough self-assessment data from 184 students, out of the 201 students enrolled in the
pilot courses, there is evidence of metacognition from 91.54% of all students in competency-
based chemistry.
Interview findings. Interview data also exemplifies student self-reflection and highlights
their techniques for lab self-regulation. In conversations with the 12 interview subjects, 100% of
the girls revealed techniques for identifying what they do not understand and working to improve
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

121
their knowledge gaps, and 8 girls (66.67%) defined the specific strategies they use for self-
regulating as they overcome obstacles in the lab. These insights were elicited through responses
to questions about how they receive feedback, how they set goals, and what they do when they
are stuck or frustrated.
Conceptual self-reflection. Six of the interview subjects (50.00%) emphasized the value
of feedback in assisting them in accurate self-assessment and goal setting. Ardeth identified that
“[a] lot of feedback is good and that helps me see what I’m doing wrong,” and Hannah
mentioned “the first time I got feedback . . . that was really helpful because I was really lost
before that.” Kamalani also noted that feedback “really helps because sometimes I don’t know
where I am,” and Rose described how feedback is “really helpful because that way you know
how to improve and you’re just not blindly looking for a way to improve.” Margaret agreed that
feedback “more, like, just kind of show[s] where I am in the process, so I know exactly what I’m
doing wrong. And then it makes it easier to fix,” and Yams articulated her hopes for even more
frequent and robust feedback when she stated, “I feel like maybe a little bit more feedback
throughout the semester would be better.”
Eight girls (66.67%) shared specifically about strategies they cultivated to help them
learn or that they implemented when they were stuck or confused. Ardeth identified the
consistent practice of “reflect[ing] on what I did before to see where my mistakes were” as
crucial to her learning, and Hannah found it important to take “time to go home and think” as she
processed the activities from each class. Likewise, Yams shared that “another goal of mine is
probably just, like, going over the packets the day we don’t have class just to keep the chemistry
gears turning.” When confusion arose, Lois would “go to other classes, other teachers, just to
see if how they’re teaching the concept is different, to see if I latch on to that any better,” and
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

122
Rose explained how she took time to really probe her understanding when she was stuck when
she said, “I always try and make sure I’m really stumped and not just, like, being lazy.” Jordan
described “using different methods or, like, applying what I’ve accumulated over lessons in the
past,” and Emma would “just to try to recall the basic information” because she recognized that
“most of it is just ratios.” When Jessica was stuck, she would try reasoning out loud as she
found “by just explaining it to another person I would find my own answer,” and Lois agreed
that “if you can explain to somebody something, then you really know what you’re talking
about.”
Four girls (33.33%) identified the inquiry-based structure of the course as particularly
conducive to their deep learning. Jordan contrasted this course to a previous science course
where she would “just take notes the whole class” and she said she “did not feel like I was taking
anything in.” Kamalani also found “the structure of the class was, it’s just, like really fit for my
style of learning” and she described that with each lab experience “everything started flowing in
my head” and she would think “Oh, so that’s why we did this. And that’s why it’s going to help
me now.” Moana connected exploring concepts through experiments to deeper learning when
she stated “from doing it, you can actually comprehend it and visualize it easier in your brain,”
and Melissa agreed that with authentic inquiry you “not just remember, you understand how it
works, why it works, and you can use it later on.”
Laboratory self-regulation. Eight interview subjects (66.67%) also revealed their
specific techniques for self-regulating in the lab when obstacles arose. During labs Ardeth
recalled former lab experiences, stating, “I always looked over my lab notes from previous,” and
Lois agreed that it was helpful to keep “looking back at the stuff that we did in the past.” Jordan
also described “trying to use, like, past knowledge to figure things out yourselves,” and when
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

123
that didn’t work, to just start “playing around with the different equipment.” Rose described the
importance of discipline and focus in the lab, indicating that successful completion of lab
challenges required “working with your group members and making sure that everybody is, like,
on task.” After labs Ardeth would use reflection to plan ahead for future lab success when she
would “see I did this wrong,” and wonder “maybe next lab I can focus on this area more.”
Hannah also used thoughtful planning as a strategy, discovering with her group that “we thought
it was smarter to figure out our whole plan first, instead of just going along the way.”
In moments of high frustration, Kamalani would take a moment to self-assess and ask
herself, “OK, what am I doing? What do I need to do and what have I done already?,” and she
would just “take that step back and go at it again after.” Yams also knew the value in stepping
back, noting “When I’m stuck or frustrated, I kind of just, like, took a break. Just a deep breath.
Go eat a snack. And I just come back to the problem with a clear mind and then I just try again.”
Moana relied on trust to move through obstacles, asserting “you trust everybody in the group and
then you learn.”
Validated influences and identified assets. An identified metacognitive knowledge asset
of students in the pilot course was that they had developed robust conceptual self-reflection
techniques to support their learning, along with strong laboratory self-regulation strategies to
work through obstacles in the lab. A lingering influence validated in the survey data was that
18.71% of students surveyed still did not reframe set-backs in the lab as data rather than failure.
From a gender equity stand-point, there was a small disparity by gender on this influence with
5.59% more females than males struggling with this reframing.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

124
Students in the pilot course demonstrated gendered beliefs about STEM. Another
area of metacognition investigated in this study was students’ understanding of how they relate
their ideas of gender and STEM identities.
Survey results. Two questions on the survey were designed to reveal gendered beliefs
about science ability and performance. On a question which asked survey respondents to rank
agreement to the statement “there is a relationship between gender and science ability,” 7.79% of
females (n = 77) agreed, 8.89% of males (n = 90) agreed, and 0.00% of the non-binary/prefer not
to answer respondents (n = 3) agreed. One male participant declined to answer this question.
Figure 14 illustrates the Likert scale responses by gender to this survey item.

Figure 14. Likert scale responses by gender to metacognitive knowledge survey item 13 about
gendered STEM beliefs

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

125
A follow up question asked survey respondents to consider if “people of my gender are
better at science.” On this survey item, nine of the 77 female respondents (11.69%) agreed.
Among those nine girls, two of them had also agreed to a relationship between science ability
and gender on survey item 13, likely conveying their belief that girls are inherently better at
science. The other seven had disagreed to survey item 13, making their response to item 14
more challenging to interpret. The male responses (n = 89) to “people of my gender are better at
science,” included six boys (6.74%) who indicated agreement. Among those six boys, only one
had also agreed on the previous survey item, likely indicating only one boy conveying a belief
that boys are inherently better at science. Two male participants declined to answer this
question. As on survey item 13, there was 0.00% agreement on item 14 from respondents with
gender designations of non-binary/prefer not to answer (n = 3). Figure 15 illustrates the Likert
scale responses by gender to this survey item.

Figure 15. Likert scale responses by gender to metacognitive knowledge survey item 14 about
gendered STEM beliefs
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

126
As mentioned previously, forecasting backward from item 14 to item 13 revealed two
girls indicating their belief that girls are inherently better at science and one boy agreeing that
people of his gender are better at science. Meanwhile, forecasting forward from item 13 to 14
revealed that of the six females that agreed to a relationship between gender and STEM, four
disagreed that people of their gender were better at science. Meanwhile, of the eight boys who
linked gender to STEM ability on item 13, seven disagreed that people of their gender were
better at science on item 14. Table 13 displays these results.

Table 13
Relationship Between Responses to Survey Items 13 and 14
Gender
Agreed to a relationship
between gender and
science ability on survey
item 13.
Indicated belief that
boys are inherently
better at science on
survey item 14.
Indicated a belief that
girls are inherently
better at science on
survey item 14.
Male (n = 90) 8 1 7
Female (n = 77) 6 4 2

Table 13 clarifies that for both male and female students who agreed to a relationship
between gender and STEM ability, students of both genders were more likely to describe the
opposite gender as better at science. While survey data indicated only a very small number of
students displaying gendered beliefs, 8.89% of the male survey respondents and 7.79% of
females surveyed, interview data revealed more pervasive gendered beliefs about participation,
performance, and interest in STEM fields among female students.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

127
Interview findings. Of the girls interviewed (n = 12), Ardeth and Jordan represented two
of the four girls who indicated with their survey responses their belief that boys are inherently
better at science. Jordan expanded on these feelings more fully in the interview process,
however Ardeth made no indication that she held these beliefs during the interview. In addition,
Moana was one of the girls who disagreed on item 13 to an overall relationship between gender
and science, but who still agreed that people of her gender are better at science. Although only
those three girls indicated gendered beliefs in the survey, commentary in the interviews revealed
eight girls (66.67%) with ideas that align participation, performance, or interest in science to
gender.
Negative gendered beliefs. Emma mentioned in her interview her belief that “people say
that men are more successful than women in science.” She described these gendered beliefs
about STEM as an offshoot of a broader gender gap in society, stating “it’s a stereotype that men
are generally more successful than women. There’s also the thing that a man and woman in the
same position, men get paid more. So I mean generally science would also be a part of that.”
Rose also explained a sense that “as I’ve grown older I feel like it’s kind of unusual for a girl to
be a scientist.” She explained that “it kind of makes my scientific identity, like . . ., even though
I feel like I’m capable of doing science, it just seems like a strange thing to want to do,” and she
said about girls in general, “just for some reason we’re just not really taught to look towards
science and math for a career.” Hannah also mentioned “it’s kind of weird for a girl to be so into
science,” and Jessica described experiences in her science classes where “there’s a lot of times
where it’s unusual to be the girl that knows what’s going on.”
Some girls mentioned in their interviews a certain social stigma that accompanies being a
girl who excels at STEM. Jordan described how “the girls that tend to be, like, more into math
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

128
and science . . . they’re like put in this group. And they’re kind of pushed aside a little bit,” and
that this can make girls feel “kind of embarrassed about liking chem, math or science because
they don’t want to be associated with them.” Jessica added that “there’s a lot of girls, I feel, that
purposefully try to act like they don’t know what’s going on . . . I see it a lot in science.” Jordan
interpreted a certain reluctance in the way that girls engage in science, stating “I think maybe
girls are more timid, like they won’t just do things, while guys are like ‘let’s just try this or try
this’” and she shared “so I guess girls just like overthink.”
Positive gendered beliefs. In opposition to beliefs that relegate STEM interest and ability
to the male gender, a number of girls held gendered beliefs that elevate girls’ abilities above their
male counterparts. Hannah mentioned “I get the sense that boys, like, can’t be good at science or
something like that,” and Lois shared her irritation that “I’m with a group of three boys” in the
lab and she felt that “they do nothing.” Hannah also thought it is possible that “there’s more
pressure on girls to be really good at science,” and Moana recalled campus initiatives that were
STEM-oriented and female dominated. Examples included this year’s student initiated campus-
wide ban of single-use plastics, that Moana described as “the girl group [who] did the water
bottle and course thing,” and the environmental programming of the community service center
that she mentioned as “and then there’s also the girls that work in the center that do the beach
cleanups.” Agreeing with Moana about girls’ ability to dominate, at one point in her interview
Yams commented “I think they [girls] can . . . because they’re just as capable [as boys], if not
more.”
Science aspirations. Seven of the girls interviewed (Ardeth, Emma, Hannah, Kamalani,
Melissa, Moana, and Yams) shared their intent to pursue science as a major in college, many for
reasons aligned to career aspirations in medicine. Hannah, Moana, and Yams were among the
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

129
four girls whose gendered beliefs about STEM indicated their sense that girls are better at
science, although Hannah also mentioned that it still seemed strange for girls to be into science.
Emma was the only girl who intends to major in science who also openly discussed gender
stereotypes in STEM during her interview, and Ardeth indicated in her survey responses that she
felt boys are better at science than girls. Melissa and Kamalani did not reveal gendered beliefs
about STEM ability or interest in either the survey or interview data. The relationship between
gendered beliefs and intent to pursue STEM is not entirely clear, although the data suggests that
positive gendered beliefs or no gendered beliefs are more frequently correlated with intent for
future participation in STEM.
Validated influences and identified assets. The metacognitive deficit of holding
gendered beliefs about STEM ability appeared minimal from the survey data, with only 8.19% of
students agreeing to a relationship between gender and science ability. However, 66.67% of
interview subjects revealed gendered beliefs about STEM ability, participation, and performance
through commentary that indicated a presumed relationship between gender and STEM. Not all
of the gendered beliefs were negative, relegating superior STEM success and interest to the
opposite gender; nearly half of the interview commentary about gendered beliefs revealed a
sense of better skills and heightened interest in STEM among females. The majority of the girls
with these positive gendered beliefs intend to pursue STEM college majors.
Motivation Results and Findings
The survey instrument and research protocol were heavily weighted towards investigating
motivational influences on the problems of gender gaps in self-efficacy and participation in
math-intensive STEM. Self-efficacy is thought to be among the most powerful predictors of
both performance and continued participation in STEM fields. Unstable, controllable, internal
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

130
attributions are linked to a growth mindset, and are thought to foster persistence and resilience.
Mastery goal orientation is believed to decrease anxiety and improve intrinsic motivation
through the focus on individual growth and improvement. The results and findings presented in
this section investigate these presumed motivation influences in the context of the competency-
based pilot course in chemistry.
Physiological feedback and mastery experiences stand out as positive self-efficacy
influencers for girls in the pilot course. Earlier in this chapter self-efficacy data from this
study was analyzed in reference to the organizational short term goal of closing gender self-
efficacy gaps in the pilot course. In this section, the 12 interview subjects’ descriptions of how
they developed their STEM identities are analyzed aligned to Bandura’s (1986) theory of self-
efficacy.
Sample and population comparison. The interview subjects’ (n = 12) baseline self-
efficacy average was 3.11 on the three self-efficacy items in the institutional data set, placing
them 0.23 points higher than the baseline average of 2.88 for all females with both baseline and
study survey data (n = 73). Nearing the end of the course, the interview subjects (n = 12)
reported an average of 3.38 on the two self-efficacy items on the study survey, placing them 0.25
higher than the 3.13 average from the larger female population. Figure 16 illustrates the
differences between the population of females and the sample of interview subjects.
Despite having overall higher self-efficacy, the interview subjects experienced nearly
identical growth in self-efficacy, gaining 0.27 points on average as compared to the 0.25 point
increase for the larger female population. Data from the interviews, therefore, helps to
illuminate both the factors that influenced girls’ self-efficacy growth in the pilot course, as well
as characteristics of girls with relatively high STEM self-efficacy.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

131

Figure 16. Comparison of pre- and post- self-efficacy data for all females (n = 73) and female
interview subjects (n = 12)

Interview findings. Bandura’s (1986) theory of self-efficacy names four potential
influences on developing and sustaining high self-efficacy, including: (1) mastery experiences,
(2) vicarious experiences, (3) social persuasions, and (4) physiological feedback. Interview data
on self-efficacy development was coded according to these factors. In questions about the ways
in which they developed their science identities, nine interview subjects (75%) explicitly
described mastery experiences, seven girls (58.33%) discussed positive vicarious experiences,
six girls (50.00%) could explain the social persuasions that influenced their confidence in STEM,
and 11 interview subjects (91.67%) indicated a relationship between their physiological response
to the pilot course and their growth in self-efficacy.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

132
Mastery experiences. Nine interview subjects (n = 12) described feelings of mastery
aligned to either the mathematical or laboratory challenges presented in the pilot course.
Experiencing success on the mathematics of chemistry led Hannah to celebrate, “I did it by
myself, I actually got it and I was, like, actually really proud of myself,” and Lois revealed “hey,
I kind of know what I’m talking about, and I don’t know, feeling that smart is pretty cool.” Also
discussing the specific mathematical tasks in chemistry, Emma described how “it was pretty hard
in the beginning, so I thought it was an accomplishment to get it down,” and Jessica described
the satisfied feeling of “I can look at it and then I actually know how to do it.”
In regards to lab tasks, Jordan celebrated “being successful and also doing more labs by
myself, like it just kind of . . . see[s] you grow as a student more,” and Melissa was excited that
she was able to design her own experiment and she thought “that was really cool how I figured
that out. Because I didn’t think I would have.” Margaret described her feelings of success as “it
was like, wow, I can actually do this,” and Kamalani noted “now we understand that. So it’s
really satisfying.” Rose concluded “it just makes the whole class a lot more enjoyable if you
understand.”
Vicarious experiences. Seven of the girls interviewed (n = 12) indicated a relationship
between female models in science and their self-efficacy, and four of those girls spoke about the
impact of their mothers. Hannah described how her “mom, I think, studied science in college”
which “kind of had an impact on me wanting to go and learn science in college.” Lois explained
that “my mom went to UO for bio but then also went to Stanford” for her Master’s, so “my views
on science, just coming into the course, were always, like, OK well my mom did this, so I can do
it, too.” Kamalani revealed that her “mom’s a nurse and she loves science,” and Margaret
discussed how her “mom, she’s really good at science and math.” Margaret went on to say that
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

133
“I had a lot of strong female figures like my grandma and my mom. They’re both really strong,
empowering women. So I always kind of felt like I was able to do whatever I wanted.”
The remainder of the girls mentioned models in the form of other family members, peers,
teachers, and female scientists. Moana explained that she has “four sisters, older sisters and they
all love science too,” and Kamalani mentioned that “my cousin, actually, she’s working to
become a surgeon.” Jessica explained, “I went to an all girls school, that was 10 years of my
education. And so I never really doubted my abilities as a girl.” Melissa mentioned that “both of
my teachers in chemistry are actually women,” and Kamalani pointed out “we have a lot of
female teachers in the science department.” Melissa drew inspiration from professional female
scientists, noting that “seeing, like, other females doing good science work, that’s making a
difference in the world, I think that’s really inspiring.” Moana agreed that “just recently with the
black hole image, there are the two women scientists that helped work on the project. So that
was, like, very empowering as a woman.”
Social persuasions. Six of the interview subjects (n = 12) made reference to the social
persuasions that have impacted their sense of science self-efficacy. Two girls spoke about
specific encouragement from their teachers, and the other four described the supportive nudges
they receive from their parents. Kamalani explained how “my teacher would just, like, know my
potential and so she’d just push me. So, it really helped me as a learner and, like, to boost my
confidence.” Melissa mentioned how her math teacher “came up to me one day, and she said
‘hey, I think you’d do really well in this’. And ‘this’ was a STEM workshop over at another
school.” Melissa described how the interaction made her “feel like I was really doing something
good, and that I could maybe do something that involves science and math one day, and that
made me really proud.”
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

134
When describing supportive parents, Ardeth explained that her mom “likes that I like
science and that I am in this course, and she thinks it’s really cool,” and Yams said her parents
“encourage me to do whatever I want. But I think that they’re trying to push me more towards a
science path, just because I really like the medical field.” Margaret’s mother often encourages
her science ability by saying, “you have it in you somewhere you just gotta find it,” and Emma
believes “my parents being supportive, like, it allows me to think that men and women are
equally able to succeed in science.” Rose, on the other hand, has noticed that “I haven’t really
talked to my parents about, like, science in general.” From this absence of direct encouragement
to pursue science, Rose concluded “I guess it’s just the fact that it’s not brought up much is kind
of a message, it’s kind of unusual or not something that my parents would really see me as I
guess.”
Physiological feedback. Eleven interview subjects (n = 12) talked about their lower
stress or increased feelings of safety to take academic risks in the pilot course, connecting
positive physiological feedback to feelings of self-efficacy. Ardeth explained that the pilot
course was “less stressful than any regular class,” and Hannah explained that “it’s really a great
stress reliever to know that you don’t have grades.” Yams also expressed that narrative feedback
instead of numerical scores on assessments “does take a lot of pressure off ,” and she felt that
“it’s very low stress” and created an environment where “just to learn is fun.” Lois stated, “I’ve
come to appreciate this class more than I would have regular chemistry class, because I don’t
feel as stressed when I walk into the room.” Emma described her relief that there is “no pressure
to get a good grade,” and Melissa explained that without traditional testing “there was no
pressure of you have to remember [all of] this.” Moana agreed that “it helps our brain have a
calmer sense that we don’t need to worry about it being tests, and making our whole grade just
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

135
that test grade,” and Margaret explained that when it comes to mastering learning outcomes, “I
feel less stressed about them, like without grades, I feel like I have more time to understand
them.”
In addition to feelings of lower stress or pressure, some girls felt safer to take academic
risks in the inquiry-based pedagogy of the mastery pilot course. Rose expressed, “I feel a lot
more comfortable because I know that, like, everything I say is not going to be judged,” and she
explained that she had “teachers in the past who just, like, kind of shut you down if you had a
question. Just like that was not a smart question or you should have known the answer by now.”
Kamalani also felt that the way her teachers coached and supported the class through inquiry
“was, like, really comforting to me because like before I used to be like, Oh no, I can’t ask
questions, are they going to be like, she doesn’t know that?” Jessica also felt comfort in the
competency-based framework noting, “it does feel better to know that, like, the teacher is
actually paying attention to what you’ve been learning.”
Validated influences and identified assets. The interview data highlights the strengths of
the pilot course at creating opportunities for girls to experience feelings of mastery and positive
physiological feedback. Although a number of interview subjects also spoke about the
influences of modeling or social persuasion on their self-efficacy development, most of that
commentary connected to factors outside of the pilot course. Only vicarious experiences and
social persuasion connected to teachers or peers in the pilot course should be considered a
strength of the competency-based curriculum. Figure 17 highlights the typicality of each of
Bandura’s (1986) four factors in the interview data and separates those that align with the pilot
course experience from other sources.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

136

Figure 17. Number of interview subjects connecting self-efficacy to each Bandura (1986)
influencer and separation of the factors independent from or connected to the pilot course

The data suggests a correlation between the mastery experiences and physiological
feedback in the competency-based curriculum and growth in self-efficacy. Meanwhile, the
elevated self-efficacy of the interview subjects relative to their peers could be attributed to the
high frequency of vicarious experiences and social persuasions they were receiving from other
sources. As such, a validated influence on self-efficacy development could be that low levels of
vicarious experience and social persuasion were intentionally cultivated in the pilot curriculum.
Students in the pilot course attribute persistence to success in science. Survey and
interview data reveal students’ nearly unanimous success attributions linked to effort and
persistence.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

137
Survey results. On a group of Likert scale survey items linked to attributions, the
students surveyed demonstrated alignment to unstable, controllable, internal attributions for
success in science. When asked to rank agreement with the statement “persistence is important
for success in science,” 98.83% of survey respondents (n = 171) agreed or strongly agreed. In
addition, 95.91% of those surveyed agreed or strongly agreed that “confusion is temporary. I can
find ways to get unstuck,” and 91.81% of surveyed students reported that “I am in charge of my
success in this class.” Figure 18 represents the distribution of Likert scale responses on these
three attribution questions.

Figure 18. Distribution of Likert scale responses to motivation survey items 17, 18, and 19 about
success attributions

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

138
Figure 18 highlights that not only did the statement “persistence is important for success
in science” have the highest overall agreement, a much larger number (n = 104) of survey
respondents (n = 171) indicated strong agreement with this statement than with the other two
survey items.
Interview findings. Interview data also indicated attributions aligned to a growth
mindset, as 100% of students surveyed (n = 12) answered attribution questions with answers
about effort, persistence, and the incremental process of growth. Ardeth noted that “you’re not
going to get something right, everything, the first time. So you need to try again and try again.”
Emma discussed how “a lot of things aren’t especially easy to understand, especially at first . . .
so persistence is really important,” and Melissa suggested that in this pilot course “you’re
supposed to be gradually getting farther and farther.” Beyond just this course, Moana described
more broadly that the “resilience part of science is that you have to get into a stop, and you just
keep working on it.”
Many of the girls surveyed situated their commentary about persistence in the context of
the lab, where they felt they had opportunities to practice and cultivate persistence. As they
described encountering setbacks in the lab, Hannah mentioned that she thought “I can always do
another trial after this,” and Jordan described thinking “Oh, that’s not working. Let’s try
something else.” From her experience in the lab, Jessica concluded that “there’s nothing wrong
with being wrong the first time or the first few times, because then you know what not to do.
Then there’s only so many things left to try.” Likewise, Margaret mentioned “I can be wrong for
a long time and it’s not necessarily just wrong, because I understand what I’m doing and how to
do it.” Embracing the iterative nature of scientific inquiry and learning to persist in the lab were
mindful practices for some students, as Kamalani noted that when her group would hit a wall
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

139
“we just had to take that step back and go at it again after.” Lois also recognized that she had
opportunities to cultivate persistence through her participation in this pilot chemistry course,
stating “so I think that I’m a pretty persistent person, and this class has definitely grown my
persistence.”
In addition to discussing their persistence in overcoming obstacles in the lab, the
interview subjects made direct references to their feelings of control of their learning in the pilot
course. Rose mentioned that “I actually can improve what I have done in the past,” and Jordan
indicated “I feel, like, more in control, as opposed to my teacher being in control like in my other
classes.” Yams made clear that she believes effort outweighs ability when she asserted “I might
not be, like, the strongest in my abilities to do what is asked of me at first, but with a lot of
practice and asking questions to further my understanding then I can become exactly what I
want.”
Validated influences and identified assets. The combination of survey results and
interview findings support the assertion that students in the pilot chemistry course attribute their
success in science to persistence. However, despite interview data that was nearly devoid of
comments aligning to external, stable, or uncontrollable attributions for success, survey data
about poor performance on assessments elicited more mixed results. While 91.81% of students
surveyed agreed that they are in charge of their success, not as many students expressed the
belief that they are in charge of their failures on assessments.
Students in the pilot course hold conflicting attributions about their poor
performance on assessments. While the data about success beliefs indicated that the students
in this study assigned largely unstable, controllable, internal attributions; questions about failure
beliefs revealed more disparate attributions among the survey respondents.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

140
Survey results. On survey items about failure attributions, only 57.31% of students
surveyed (n = 171) agreed or strongly agreed with the statement “if I perform poorly on an
assessment, it is because I did not try hard enough to learn the material,” and 31.58% of those
surveyed still agreed that “if I perform poorly on an assessment, it is because it was too hard for
me.” Figure 19 illustrates the distribution of Likert scale responses on these two attribution
questions.

Figure 19. Distribution of Likert scale responses to motivation survey items 15 and 16 about
failure attributions

Sample and population comparison. Interview data lends some insights into this
phenomenon. However, to utilize the commentary of the interviews to explain these trends
toward more external, uncontrollable, and/or stable attributions, the interview subjects’ responses
to these questions first had to be checked against the average responses to ensure reasonable
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

141
alignment of this sample to the population. All survey respondents (n = 171) had overall 56.84%
agreement and female respondents (n = 77) had 53.25% agreement with the statement “if I
perform poorly on an assessment, it is because I did not try hard enough to learn the material,”
while interview subjects (n = 12) had somewhat higher agreement with this statement at 75.00%.
Similarly, while all survey respondents demonstrated overall 31.62% agreement and female
respondents registered 32.47% agreement with the statement “If I perform poorly on an
assessment, it is because it was too hard for me,” only 16.67% of interview subjects agreed with
this statement. Table 14 illustrates this comparison between all survey responses, female
responses, and interview subject responses.

Table 14
Comparison of Percent Agreement on Survey Items 15 and 16 for All Respondents, Female
Respondents, and Interview Subjects
15. If I perform poorly on
an assessment, it is because
I did not try hard enough to
learn the material.
16. If I perform poorly
on an assessment, it is
because it was too hard
for me.
All Survey Respondents (n = 171) 57.31% 31.58%
All Females (n = 77) 53.25% 32.47%
Interview Subjects (n = 12) 75.00% 16.67%

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

142
The interview subjects (n = 12) demonstrated overall greater agreement with question 15
and less agreement with question 16 than was typical for all other survey respondents, however
all groups showed greater agreement on question 15 than on question 16. Insights from the three
interview subjects who did disagree with question 15 (Jessica, Margaret, and Rose) and from the
two subjects who agreed with question 16 (Moana and Rose) were particularly considered to
clarify this attributional dissonance between effort and ability.
Interview findings. Many interview subjects contrasted their experiences in the pilot
course against other courses, particularly when describing grading and testing practices. Jordan
described “instead of it being based off of a grade and multiple tests and graded homework and,
like, graded assignments, it’s more based on your actual knowledge of the subject instead of
questions on a paper.” Moana, who agreed that her poor performance on an assessment is
because it is too hard, described the pilot course grading as “the grades not being based on the
tests, it’s really based on what we learned and what we can show that we know,” and Yams
explained that the grading is “geared towards something like your personal level not just, like,
one structured level.” All three of these comments reflect these students’ feelings that traditional
tests shift the locus of control from internal to external.
Some commentary from the interviews also indicated that feelings of controllability can
decrease in a testing situation. Hannah noted that “some tests don’t really reflect what you
actually learned, because like there’s some people that just aren’t good test takers, like they
know the material but you just can’t take a test,” and Kamalani mentioned “personally I’m not
the best, like, test taker.” Margaret, who disagreed that her poor performance on an assessment
is due to inadequate effort, described the pilot course grading as “measuring your depth of
understanding rather than, like, you got it wrong or you got it right” and “it’s more about your
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

143
understanding of the concept rather than like how well you perform.” Jessica, who also
disagreed with question 15, contrasted the alternative assessment methods in the pilot course
against traditional testing when she stated “they actually assess how much you’ve learned and
not just how much you memorize about the topic.”
Rose was the one interview subject who strongly disagreed with the statement “if I
perform poorly on an assessment, it is because I did not try hard enough to learn the material.”
Rose noted that in the pilot course “it’s more about the effort that you make to learn the material
than anything else.” She contrasted this to her experiences in other classes where “the grading
system has made you think that like if you’ve got a C on this then you’re actually not good at
science in life, even though you might be trying as hard as you can.” Rose seems to feel that she
has had experiences in the past where, despite putting forth great efforts, her results on
assessments did not highlight her effort. Rose was also one of the two interview subjects to
agree that a test might be too hard for her, however, her commentary in the interview seemed to
indicate that this was not about stability attributions. Rose reflected instead on the
uncontrollable nature of tests being administered to all students on the same day regardless of
their readiness, saying “I’m really hoping that, like, my teachers in the future will also
understand that not everybody learns at the same pace.”
Validated influences and identified assets. It is unclear if students understand the word
“assessment” to be inclusive of all the ways that their learning might be measured in a class. In
91.67% of the interviews the students talked about their experiences with tests, but only two
students (16.67%) utilized the word “assessment” to describe other methods of demonstrating
their understanding like reflective writing and lab practicums. It is possible that some students in
this study conflated the words test and assessment and interpreted survey questions 15 and 16 to
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

144
be about traditional tests, which were not utilized in the pilot course. The variable attributions
assigned to poor performance on assessments might be a representation of students’ perceptions
of their loss of choice and agency in traditional testing, where their progress is determined by
one common metric on a uniform time scale for all students.
The data does not make clear if students were actually struggling to align their failures
with effort as well as they were able to align their successes with effort in the pilot course, or if
there was a misunderstanding of the word “assessment” on the survey. In contrast to the variable
attributions revealed on survey items 15 and 16, which seemed to align to external,
uncontrollable attributions in traditional testing, 91.23% of pilot students surveyed (n = 171)
agreed or strongly agreed with a survey item that stated “my individual growth was rewarded in
this chemistry class.”
In terms of attributional equity, disaggregating the results from survey item 16 by
race/ethnicity reveals notable gaps, with Multiracial, Asian American, and Pacific Islander
students being more likely to agree that “if I perform poorly on an assessment, it is because it
was too hard for me.” Figure 20 displays the percent agreement to survey item 16 by racial or
ethnic identity.
Whether or not this attributional gap was present for the alternative assessments in the
pilot course, the evidence that students of color are more often experiencing the fixed mindset of
thinking their ability is preventing performance, validates this influence for this population of
students.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

145

Figure 20. Percent agreement to attributional survey item 16 disaggregated by race/ethnicity

Students in the pilot course excelled at mastery goal setting, but continued to also set
performance goals. Survey and interview data revealed that while students in the pilot course
were demonstrating high alignment to a mastery goal orientation, they were concurrently setting
performance goals.
Mastery goal survey responses. A number of survey items were designed to elicit
students’ goal orientation. Responses to four survey items relating to a mastery goal orientation
were reviewed to look for patterns, and data was disaggregated by female (n = 77) and male
respondents (n = 91). On survey item 21, 89.61% of female respondents and 91.21% of male
respondents agreed or strongly agreed with the statement, “As I approach class each day, my
goal is to gain a deeper understanding of science.” Item 22, which stated “My goals in this class
tie directly to the learning outcomes/competencies” revealed lower overall agreement and greater
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

146
gender differences, with agreement frequencies of 66.23% for females surveyed and 73.63% for
male respondents. When asked to rank agreement with item 23, “my goals are about self-
improvement,” 94.81% of female respondents and 92.31% of males surveyed agreed or strongly
agreed; and even stronger alignment was demonstrated to the statement “leaving the Hawaii
School with excellent skills (like inquiry, persistence, communication, and personal
responsibility) is very important to future success,” on item 33 with 100% of females and
96.70% of males surveyed agreeing or strongly agreeing. Figure 21 illustrates the Likert scale
responses by gender on each of these four survey items.

Figure 21. Distribution of Likert scale responses by gender to motivation survey items 21, 22,
23, and 33 about mastery goal orientation

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

147
Survey data makes clear that students in the pilot course were setting goals to deepen
their understanding and improve their skill sets, the hallmarks of a mastery goal orientation.
Responses to item 22, with 29.24% disagreement, reveal that there is room for growth in the
students’ abilities to link their mastery goal setting to the actual course competencies.
Mastery goal interview findings. Interview data also indicates a strong orientation
towards mastery goal setting, as 100% of interview subjects (n = 12) described their mastery
goals in the pilot course in response to the question “describe how you set goals for yourself in
this class.”
Moana explained that she “set goals by always working on what I need to improve on,”
and Rose agreed that “seeing what needs to be improved helped me set a goal as, like, what I can
do better.” Jordan described that in the pilot course “you have to actually improve your
understanding,” which Lois explained as, “so if I don’t understand it, then my goal can be to
understand more.” Kamalani described her goal setting as, “I was trying to figure out what I was
doing wrong and how I could fix it.” Ardeth argued that, “I definitely think that it’s important to
have goals in science” and “you can never really master anything but you’re, like, you’re always
learning.” Yams described her goals as “keep on asking questions and keep learning.”
A few interview subjects differentiated their mastery goals in the pilot course from their
performance goals in other contexts. Jessica described her goals as, “I’m not as focused on
finding the right answer as I am actually understanding the concept and how I can apply it to
other situations,” and Margaret explained “it’s more about your understanding of the concept
rather than like how well you perform.” Melissa also contrasted her mastery goals with
performance goals when she stated, “but I think I set my goals instead of getting 100% on this
test, just understanding something.”
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

148
As in the survey data, the interview data was not rich with examples of girls aligning
their goals to particular competencies, as only two girls made direct reference to the learning
outcomes. Emma stated, “I mean, I have to make sure I know exactly what each outcome is
asking for so that I can actually master it,” and Hannah explained that she used the outcomes
more for reflection than for goal setting when she described how she “can align with it [the
learning outcomes] later on when I need to figure out which one I need to work on.”
Despite compelling evidence that students in the pilot course were demonstrating a
mastery goal orientation, survey and interview data revealed that they continued to also set
performance goals.
Performance goal survey responses. Three survey items were designed to reveal a
performance goal orientation. Item 20, which stated “as I approach class each day, my goal is to
get a good grade in science,” produced 89.47% agreement from female respondents (n = 76) and
84.62% agreement from male respondents (n = 91). Agreement frequencies were also high on
item 32, “good grades are very important to future success,” with 85.71% of females surveyed
(n = 77) and 75.82% of male respondents (n = 91) agreeing or strongly agreeing with that
statement. Low levels of agreement and greater gender disparities appeared on item 24, “my
goals are about outperforming others.” While only 9.09% of female respondents (n = 77) agreed
with this statement, 23.08% of males surveyed (n = 91) agreed or strongly agreed. Figure 22
illustrates the Likert scale responses to these three items by gender.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

149

Figure 22. Distribution of Likert scale responses by gender to motivation survey items 20, 24,
and 32 about performance goal orientation

Survey data provided evidence that students in the pilot course were still setting goals
that were performance oriented, specifically goals aligned to grades. While performance goals
are typically linked with competition, comparing survey data from items 20 and 24 does not
seem to indicate that the students in the pilot course see grade-centered goals as synonymous
with goals about hoping to outperform others.
Performance goal interview findings. Interview data also provided evidence of students
setting goals about grades, with no indication that the girls were hoping to compete with and beat
their peers. Six of the interview subjects (50.00%) made comments that align to a performance
goal orientation, in addition to the commentary they provided indicating their mastery goal-
setting.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

150
Three girls specifically named a goal related to a grade outcome in response to the
question “describe how you set goals in this class.” Emma said “So I guess my main goal, that’s
to get an A in the class,” and Melissa agreed “Well obviously everyone wants that A, that’s like
my clear goal.” Margaret even went so far as to explain “if I get the good grade, then I get into
the good college, and then I get a good job. So it’s like that, it’s all those things.” Two girls did
not articulate their goal to get a good grade explicitly, but did mention a conversation with their
teacher that was about grades. Hannah mentioned that in a conversation with her teacher that
“she just told me I need to complete four more [competencies] to get a good grade,” and Ardeth
expressed her appreciation for “feedback from our teachers and being able to converse with them
about what I can do to better my grade in the class.”
Two interview subjects also shared their perception of the value of the extrinsic
motivation grades can provide. Emma explained that having grades “pushes me to, like, try to
understand everything better and put a lot of effort into the labs and stuff.” Yams did not feel
that she, personally, needed the extrinsic motivator, but she remarked that the course does
demand intrinsic motivation and “if you don’t want to learn then you’re probably not going to
get much out of it, because there’s no grade, you are probably gonna just chillax.”
Validated influences and identified assets. A powerful motivational asset of the
competency-based curriculum is the mastery goal setting demonstrated by the students, with a
focus on self-improvement in skills and understanding. A validated influence remains in the
ability of students to set targeted goals that align to the actual course competencies, as 29.24% of
students report this lack of alignment. A more pervasive validated influence was students’
continued performance goal-setting, as demonstrated through a focus on grades, with 86.55% of
students still naming good grades as a goal in the course and 80.12% of student asserting that
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

151
good grades are very important to future success. Gender inequity existed for both of these
sentiments, with the larger difference represented in 9.89% more female students agreeing to the
importance of grades for future success.
Organization Results and Findings
Some survey items and interview questions were designed to elicit the pilot students’
understanding of the existing cultural model of gender inclusion at the Hawaii School and their
interpretations of the cultural setting of a competency-based curriculum and mastery assessment
practices. To achieve the long-term organizational goal of improved gender equality in math-
intensive STEM courses, there must be a shared cultural model throughout the organization for
the value of gender equity and inclusion in STEM. A cultural setting of mastery learning has
been identified by the Hawaii School as a significant tool for improving equity in educational
experiences and outcomes. As such, students’ understanding of their experiences in the pilot
course can serve to deepen the organization’s understanding of this cultural setting and perhaps
inspire a more commonly shared cultural model among all stakeholder groups of the value of
competency-based education.
The organization’s cultural model of gender inclusion in STEM is evident to the
pilot course students. Survey and interview data support the assertion that pilot students
perceive a cultural value system of gender equity and inclusion in STEM.
Survey results. Three survey items were intended to reveal students’ perceptions of the
cultural model of gender inclusion in STEM at the Hawaii School. All three items elicited nearly
unanimous agreement with little variation between genders. Overall agreement to the statement
“the Hawaii School culture promotes equal participation in STEM by male and female students,”
was 95.32% among all survey respondents (n = 171). Questions specific to gender produced
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

152
100% agreement among all survey respondents (n = 171) to the statement “it is socially
acceptable to be a boy who loves science at the Hawaii School,” and 98.25% agreement
(n = 170) to the converse, “ it is socially acceptable to be a girl who loves science at the Hawaii
School.” Survey responses on these three items indicated that students in the pilot course
perceived an equitable and inclusive cultural value system surrounding gender and participation
in STEM.
Interview findings. Seven interview subjects (58.33%) made specific mention of their
perception of a thriving cultural model of gender equity and inclusion in STEM at the Hawaii
School. Ardeth explained that, “I feel that it is a very diverse school. And a lot of our students
are really supportive towards each other.” Rose described the Hawaii School as “super accepting
of whatever you want to do,” and Lois agreed that “it’s a pretty, like, accepting place.” Emma
expressed that “the Hawaii School believes both genders have, like, the same amount of
capability,” and Melissa agreed “Yeah, I feel like there’s no real barrier between like, ‘oh you’re
a girl, you can’t do that’ — like anywhere in any aspect.” Kamalani shared “I think, like, that
science here for girls is just like anybody else. There’s a lot of encouragement and no
discrimination.” Ardeth also talked about encouragement when she mentioned “all the adults on
campus really support all the students,” and Margaret agreed that “from the adults, I think they
all encourage science very strong.”
Earlier in this chapter the analysis of metacognitive knowledge influencers indicated that
a number of the girls interviewed held gendered beliefs about participation, performance, and
interest in STEM. Students’ overall perception of the existing cultural model of inclusion does
not seem to indicate that these gendered beliefs are being largely initiated or perpetuated at the
Hawaii School. However, Jordan did describe her feelings of how “the girls that tend to be, like,
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

153
more into math and science . . . they’re like put in this group. And they’re kind of pushed aside a
little bit,” indicating that interest in science could come at the cost of some social exclusion. In
addition, Jessica described “there was a time in biology freshman year, where there were a few
guys that definitely tried to make me feel like I couldn’t do it because I was a girl.” Jessica had
just transferred to the Hawaii School from an all girls school. She shared that, “my dad told me
that that’s what happens at co-ed schools.” Jessica’s experience indicates that despite an
overarching sense of inclusion in the organization, some gendered discouragement of STEM
pursuit could exist at the peer to peer level.
Validated influences and identified assets. An organizational asset of the Hawaii School
was reflected in the pilot students’ strong belief that a cultural model of gender equity and
inclusion drove encouragement for students to pursue STEM. However a potential validated
influence arose from some interview data that indicated social exclusion or social
discouragement from peers towards girls with interest in STEM.
Student anxiety was low and sense of belonging was high in the cultural setting of a
competency-based and mastery-assessed STEM class. Students’ sense of belonging and
reduced anxiety could be considered markers of success for the pilot curriculum, given that the
population in regular chemistry consists of students who have not demonstrated success in
science in the past or who tend to see themselves not as “science people.” Survey and interview
data support the assertion that the cultural setting of a competency-based course was correlated
with low anxiety and a sense of belonging.
Survey results. Survey data indicated that the majority of students in the pilot course do
not feel anxious in science class. Surveyed students reported disagreement with the statement “I
feel anxious in science class” at frequencies of 75.32% for female respondents (n = 77), 79.12%
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

154
for male respondents (n = 91), and 66.67% for students identifying as “non-binary” or “prefer
not to answer.” Figure 23 illustrates the Likert scale responses to this item, disaggregated by
gender.

Figure 23. Likert scale responses by gender to survey item 12 about science anxiety

In addition, 76.61% of survey respondents (n = 171) agreed that “when I am in science
class, I feel as if I belong.” The gender disparity on this survey item was only 0.71%, however,
when the results were disaggregated by race and ethnicity it was found that for Asian American
(n = 77), Multiracial (n = 34), and Pacific Islander (n = 20) students, the sense of belonging was
above average at 80.26%, 82.35%, and 85.00% respectively. Figure 24 illustrates the full set of
percent agreement data disaggregated by survey respondents’ self-identified race or ethnicity.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

155

Figure 24. Percent agreement statistics by race and ethnicity on survey item 26 about feelings of
belonging

Survey data suggests that both gender and racial/ethnic equity exists in the competency-
based pilot course in regards to low anxiety and high feelings of belonging, and in fact, this
cultural setting could be more inclusive for students of color.
Interview findings. Earlier in this chapter, interview data was reviewed in alignment to
Bandura’s (1986) four factors that influence self-efficacy. In that section it was discussed that
91.67% of interview subjects explicitly identified a sense of lower stress, decreased anxiety, or
increased empowerment in the pilot course, aligning to Bandura’s (1986) influence of
physiological feedback.
Validated influences and identified assets. A cultural setting asset of the pilot course is
the students’ reports of low anxiety and a strong sense of belonging in their science course. The
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

156
possibilities for educational equity, with students of color reporting higher than average sense of
belonging, is another promising asset. A validated persisting influence is that 22.81% of
students do still feel anxious and 23.39% of students do not yet feel as if they belong in science
class.
Girls connected the cultural setting of a competency-based curriculum and mastery
assessment to personalized, authentic learning. Interview findings reveal that girls perceive
deeper, authentic, personalized learning opportunities in the cultural setting of a competency-
based pilot course.
Interview findings. Interview data indicated that 100% of the girls interviewed (n = 12)
saw advantages to the cultural setting cultivated by a competency-based curriculum and mastery
assessment practices. Specifically, the interview subjects contrasted the impacts of alternative
assessments and competency-aligned, narrative feedback against earning numerical scores on
traditional assessments. Themes of personalization, authenticity, and deep learning
predominated their commentary.
Ardeth indicated that the feedback practices in the pilot course made it “how it is more in
the real world,” noting how, “You don’t get As if you turn in, like, a grant or something.”
Kamalani described it as the “teacher will tell you what you’re doing good and what you need to
work on,” which she appreciated because she feels, “personally I’m not the best, like, test taker.”
Margaret also explained that “we get feedback . . . she’ll say, like, this looks good or you’re
almost there” and “rather than just being, like, wrong for a whole semester I could be half right
and getting there, and it gives me more opportunity to actually understand the concept.” Moana
agreed that “it’s more based off of your learning.”
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

157
Many of the interview subjects felt that a focus on traditional tests or grading structures
actually inhibits their deep learning. Lois described that in the pilot course “it’s more like I’m
learning instead of cramming, which I think is really important,” and Jessica agreed that in this
course “they actually assess how much you’ve learned and not just how much you memorize
about the topic.” Hannah explained, “it pushes me to actually learn the material instead of just,
like, knowing it,” and Emma claimed that in the pilot course the teachers “get us to focus on
learning instead of actually, like, just doing stuff to get the grade.” Ardeth described how in labs
“that you work towards, like, competencies instead of a regular grade,” and Jordan felt “you’re
growing more as a student instead of just trying to get perfect test scores.” Melissa discussed
how when preparing for traditional tests in other courses she “was remembering it just for the
test, because I’d forget it after,” but she noted that in the pilot course, “since it is just, like, not
focused on having to remember, I feel like I’m absorbing more of the information that I’m
learning.”
Two girls described specifically how their approach to learning has changed in a cultural
setting without traditional tests. Rose revealed that when there is a test coming in other classes,
her questions relate to “is this gonna be on the test?,” and she went on to explain “then if it is,
then you know you have to learn it and if not, you, like, just throw the information away.” Rose
felt that in the pilot course “we’ve learned . . . a lot of things that I probably wouldn’t have
picked out if it weren’t going to be on the test [in a traditional course].” Likewise, Yams
remarked, “I’m able to ask more questions that I probably wouldn’t care to ask, because without
focusing on the grade it actually lets me appreciate what this class is about and actually care
about the content and want to learn it.” She went on to remark, “I’m not worried about the
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

158
grade, I just want to understand what’s going on; but with math [class] I’m just like — formula,
formula, formula.”
Validated influences and identified assets. A cultural setting asset of the pilot course is
students’ strongly shared perception of deeper, more authentic, personalized learning. But
despite the unanimous sentiment among the female interview subjects that their learning is
deepened by a greater focus on feedback and growth and less focus on grades, survey
respondents aligned to an item comparing the competency-based method to traditional grading
with less unanimity. On a question that stated “compared to traditional grading, the competency-
based method gives me a better sense of what I do and don’t understand,” survey respondents
(n = 171) indicated overall 68.42% agreement. This gap presents a validated cultural setting
influence yet to be addressed.
Summary
This chapter utilized quantitative results and qualitative findings to provide answers to
the first two study questions, reporting first on the progress towards organizational and
stakeholder goals and then validating KMO influences or identifying KMO assets in the pilot
course. At the opening of this chapter it was clarified that when making assertions from the data,
70% agreement on survey items was the threshold for asserting agreement, and interview data
used as evidence generally relied upon code typicalities of at least 50% among interview
subjects. Similarly, the degree to which an influence was validated depended upon the overall
percent agreement on the survey and/or the typicality in the interview data. Given these metrics,
influences were validated when data indicated more than 70% of students surveyed or 50% of
students interviewed were experiencing that KMO challenge. Influences impacting between 10–
70% of survey respondents or 10–50% of interview subjects were considered partially validated,
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

159
and those influences affecting fewer than 10% of students from either sample would have been
classified as not validated. There were no influences in the “not validated” category in this
study. Table 15 reiterates the nine influences explored in this study and highlights whether each
was validated (V) or partially validated (PV).

Table 15
Degree of Validation of Assumed Influences (V = validated, PV = partially validated)
Assumed Knowledge
Influences
V
or
PV
Assumed Motivation
Influences
V
or
PV
Assumed Organizational
Influences
V
or
PV
Procedural Knowledge
— Students need to
understand the steps of
inquiry inherent in
STEM disciplines.
PV Self-Efficacy — Students
need to believe they are
capable of the
mathematical and
laboratory tasks required
in math-intensive STEM
courses.
PV Cultural Model — The
organization needs to
sustain a culture of
gender equity and
inclusion in all STEM
courses.
PV
Metacognitive
Knowledge — Students
need to develop science
metacognition skills, to
include laboratory self-
regulation and
conceptual self-
reflection.
PV Attributions — Students
need to feel their
performance depends
upon their sustained effort
rather than believing their
STEM aptitude is fixed.
PV Cultural Model — The
organization needs to
place a high value on
developing all students’
competency in STEM
skills and habits of mind.
PV
Metacognitive
Knowledge — Students
need to reflect on their
own beliefs about how
gender roles relate to
science.
V Goal-Orientation —
Students need to engage
in math-intensive STEM
courses with the desire to
master the skills and
habits of mind necessary
for authentic STEM
inquiry.
V Cultural Setting — The
organization needs to
develop and deliver
STEM curriculum that is
focused on mastery of
skills over content-based
performance.
V

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

160
Most of the influences in Table 15 were directly connected to the assertions made
throughout this chapter. However, the final two organizational influences were addressed a bit
more subversively by the data from student stakeholders. The partial validation of the cultural
model influence relating to the value of developing students’ competencies derived from the data
showing that 31.58% of survey respondents still did not see how competency-based feedback
improved their ability to understand their own learning. As for the cultural setting influence
asserting that the organization needs to develop and deliver STEM curriculum that is focused on
mastery of skills over content-based performance, it was challenging to directly gather the
student perspective on this influence, as the student stakeholder group has no agency in making
this cultural setting more prevalent in the organization. This influence is instead validated by the
researcher’s institutional knowledge that competency-based practices are only being utilized in a
small handful of pilot courses. However, the urgency to address this cultural setting gap is
strengthened by survey data indicating that student anxiety is lower and their sense of belonging
is higher in the competency-based pilot course and 100% of interview subjects sharing their
perceptions of greater personalization, depth, and authenticity of learning in a competency-based
model.
Table 16 summarizes the key assertions and supporting data articulated in Chapter 4,
categorized as either KMO identified assets or KMO validated influences. The third research
question guiding this study asks for knowledge-based, motivational, and organizational
recommendations to address the problem of practice. The results and findings from this chapter
will guide the selection of principles and the creation of context-specific recommendations that
can be incorporated into an implementation and evaluation plan in Chapter 5. This plan will be
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

161
informed both by the persistent influences within the pilot course and the strengths of the
program identified by student data.

Table 16
Summary of Results and Findings, Reported as Identified Assets and Validated Influences
KMO Identified Assets KMO Validated Influences
Procedural Knowledge
Pilot students can articulate and enact the
skills and mindsets necessary to conduct
STEM inquiry.
26.90% of pilot students still do not feel
comfortable interpreting authentic, messy data
sets.
Metacognitive Knowledge
Pilot students have robust strategies for
conducting accurate conceptual self-
reflection and setting targeted goals for
improvement.
18.71% of pilot students could still improve lab
self-regulation, particularly reframing lab
obstacles as data (slightly more common issue
among female students).
Pilot students have developed a number of
techniques for overcoming obstacles in the
lab.
66.67% of interviewed girls hold gendered
beliefs about ability, interest, and participation in
STEM that relegate their gender to lower status.
Only 8.19% of surveyed students identified
any relationship between STEM ability and
gender.

Motivation — Self-Efficacy
Mastery experiences and physiological
feedback from the pilot course seemed to be
most profound influencers on girls’ self-
efficacy.
It appeared that low levels of intentional
modeling & social persuasions for girls existed
in the pilot curriculum.
96.49% of survey respondents demonstrate
laboratory self-efficacy.
16.96% of survey respondents do not
demonstrate mathematical self-efficacy.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

162
Table 16, continued
KMO Identified Assets KMO Validated Influences
Motivation — Attributions
98.83% of pilot students attribute success in
science to persistence.
Pilot students have mixed attributions for failure
on (traditional?) assessments. 31.5% attribute
their poor performance to assessments being too
hard and 57.3% attribute poor performance to
insufficient effort.
91.81% of pilot students felt in charge of
their success in the pilot course.
It is likely that pilot students do not have a robust
understanding of alternative summative
assessments beyond traditional tests.
Pilot students of color are more likely to attribute
poor performance on assessment to ability.
Motivation — Goal Orientation
Students in the pilot course excelled at
mastery goal setting, looking to deepen their
skills and understanding and focusing on
self-improvement.
29.24% of pilot students are not yet linking their
mastery goal setting to the actual course
competencies.
Pilot students continued to also set performance
goals, with 86.55% of students still identifying
the achievement of good grades as a goal.
Girls surveyed are more likely to believe that
good grades are important for future success.
Cultural Models — Gender Equity and Inclusion
95.32% of pilot students agreed that the
Hawaii School promotes equal gender
participation in STEM.
Some interview data indicated social exclusion
or social discouragement from peers for girls
with interest in STEM.
Cultural Settings — Competency-Based Curriculum and Mastery Assessment
77.19% of students do not feel anxious in
the pilot course and 76.61% of students
report a sense of belonging (and students of
color reported above average sense of
belonging).
22.81% of students do feel anxious and 23.39%
of students do not feel as if they belong in the
pilot course.
100% of girls interviewed described greater
personalization, depth, and authenticity of
learning in a competency-based model.
31.58% of survey respondents do not agree that
the competency-based model helps them
understand their own learning better than
traditional grading.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

163
Conclusion
The conceptual framework that guided this study (see Figure 1 in Chapter 2) highlighted
the relationships, as determined by a review of the literature, between self-efficacy and other
theories of knowledge and motivation. One specific set of connections linked both attributions
and goal orientation to metacognition, which in turn was linked to self-efficacy. Evidence from
this study supports the conclusion that students in the pilot course held mastery goal orientations,
attributed their success to internal, controllable, unstable conditions, and were able to self-reflect
and self-regulate to support their goals in pursuing scientific inquiry. In turn, self-efficacy
increased for all students in the course, with more dramatic increases for the females in the study.
A link between procedural knowledge and self-efficacy also appeared on the conceptual
framework, and evidence from this study highlighted the pilot students’ strong procedural
knowledge of the skills and habits of mind necessary to perform authentic inquiry.
The conceptual framework situated all knowledge and motivation influences within the
organization’s cultural settings and cultural models. The data from this study was situated within
the cultural setting of a competency-based pilot course that utilized mastery assessment
practices. While the results and findings highlighted both KMO identified assets and validated
influences which will frame the recommendations in Chapter 5, the overall data from this
evaluation study aligns quite well with the body of research that informed the conceptual
framework. The correlations between improved self-efficacy for girls in math-intensive STEM
and other knowledge, motivation, and organizational factors are evident.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

164
CHAPTER 5
SOLUTIONS AND INTEGRATED IMPLEMENTATION AND EVALUATION PLAN
In Chapter 4 progress towards organizational and stakeholder goals was evaluated in the
context of a competency-based pilot course, and a number of persistent knowledge-based,
motivational, and organizational influences on the problem of practice were identified. In
Chapter 5 these results and findings serve as the targets that guide the selection of evidence-
based solutions and recommendations, specific to the categories of validated and/or partially
validated knowledge, motivation, and organization challenges. This chapter will then utilize the
New World Kirkpatrick Model to generate an integrated implementation and evaluation plan for
the recommended solutions (Kirkpatrick & Kirkpatrick, 2016). In order to frame this chapter in
the context of the organizational mission, performance goals, stakeholders, and research
questions that guided the study, this chapter begins by revisiting pertinent sections from
Chapter 1.
Organizational Context and Mission
The Hawaii School (a pseudonym) is a large, independent, K-12, day school in the state
of Hawaii, with a mission to cultivate within each student the capacity to collaborate,
communicate, create, think critically, empathize, embrace challenge, engage with a global
perspective, and honor self and place. The academic mission has a specific additional focus on
developing students’ skills and habits of mind for engagement in authentic inquiry. The Hawaii
School has a well-regarded high school, described as a thought leader and an innovation
incubator in the independent school world.
The high school is comprised of 1700 students (48.2% male and 51.8% female) and 168
faculty members. A Principal, two Assistant Principals, and a team of eight Student Deans lead
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

165
the high school, and the entire K-12 school is led by a senior administrative team of 17 people
including the school President, Vice Presidents, division directors, and team leaders. Table 17
shows self-reported diversity data for the 2018–2019 school year, which indicates the racial and
ethnic demography of the students, faculty, and school leadership (indexgroups.org, 2018).

Table 17
Hawaii School Diversity Data from INDEX

African
American
/ Black Latinx
Asian
American
Native
American
Multiracial
American
Pacific
Islander
American
White Non-
Latinx
American
Unsure /
Unreported
Students 0.34% 0.4% 22.2% 0.06% 21.3% 9.4% 8.3% 38.0%
Faculty 0.3% 2.2% 26.2% 23.8% 1.5% 46.0%
High School
Leaders
18.2% 9.1% 72.7%
Senior Leader
Team
41.2% 23.5% 35.3%

Organizational Goal
Although the Hawaii School enjoys a high status both locally and nationally and its
students achieve impressive college admissions outcomes, the school is not immune to the
national trend of underrepresentation of girls in upper-level, math-intensive STEM courses
(College Board, 2014). In the 2017–2018 school year, overall enrollment in all AP courses in
calculus, computer science, chemistry and physics was 47.6% female and 52.4% male (n = 519).
Larger gender disparities existed in the highest levels of calculus and physics, with 64% male
students and 36% female students in AP Calculus BC (n = 44), and 96% male students with only
4% females in AP Physics C (n = 26). In addition, AP Computer Science was 76% male
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

166
students and 24% female students (n = 34) and a course in Engineering projects was 73% male
and 27% female (n = 11). Given the confirmed participation gap, the organizational goal of the
Hawaii School is to achieve improved gender equality, as defined by a gender representation that
approaches the demography of the student body (48.2% male and 51.8% female), in participation
in math-intensive STEM courses in the high school by the fall semester of 2021.
Currently, the Hawaii School is a leader in its commitment to the work of the Mastery
Transcript Consortium (MTC), a consortium of over 300 independent and public schools that are
working together to disrupt the traditional transcript and create curriculum, pedagogy, and
assessments that promote mastery over performance goals. Along with other MTC member
schools, the Hawaii School is actively working to reshape its curriculum to highlight desired
skills and competencies over factual knowledge in all content areas. These changes are
redefining what it means to be “good at STEM,” which could have a positive impact on girls’
self-efficacy, and ultimately their participation, in math-intensive STEM courses (Dembo &
Eaton, 2000; Flowers III & Banda, 2016; Leaper et al., 2012). Given the research in support of
mastery learning as a tool to enhance self-efficacy and the context of the Hawaii School as a
leading school in the MTC, a shorter term organizational goal was to evaluate and, if necessary,
improve gender equity in self-efficacy in math-intensive STEM pilot courses aligned to the new
competency-based curriculum and mastery assessments in the high school by the end of the
spring semester of 2019.
Description of Stakeholder Groups
The key stakeholder groups involved in meeting the goals of improved gender equality in
participation in math-intensive STEM courses and greater gender equity in self-efficacy, include
the Hawaii School administration, faculty, and students. The school administration will continue
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

167
to be responsible for connecting the faculty with research on gender-based best practices in
STEM education and providing time and guidance for faculty to innovate their pedagogy to be
more inclusive as they make the shift towards mastery curriculum. The STEM faculty members
will keep working to create curriculum, pedagogy, and assessments that are more gender
inclusive, with a focus on techniques that will enhance girls’ STEM self-efficacy. The students
will need to keep living the school’s mission by developing the skills and habits of mind to
perform STEM inquiry, and engaging in self-reflection and targeted goal setting to promote their
continued success in STEM courses.
Goal of the Stakeholder Group for the Study
While the joint efforts of all stakeholder groups will be required to achieve greater gender
equality in participation in math-intensive STEM courses, the STEM self-efficacy of the students
is thought to be a crucial determining factor in reaching this goal. Therefore, the stakeholder
group of focus for this study was the Hawaii School students. The school’s mission is a call to
students to develop their skills in inquiry and adopt the habits of mind for persistence and
resourcefulness in their pursuit of rigorous studies. Self-efficacy, or the belief in one’s own
ability, determines the resilience a student will have in the face of setbacks, and it is a predictor
of the interest a student will have in a subject and the persistence she will demonstrate in pursuit
of her learning goals (Bandura, 1986; Peters, 2013; Sadler et al., 2012). Focusing on students as
the stakeholder group for this study allowed for an evaluation of the factors that influence the
development of girls’ self-efficacy in math-intensive STEM subjects. Table 18 articulates the
organizational goal and cascading stakeholder goals in the context of the organization’s mission.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

168

Table 18
Organizational Mission, Global Goal and Stakeholder Performance Goals
Organizational Mission
The mission of the Hawaii School is to cultivate within each student the capacity to collaborate,
communicate, create, think critically, empathize, embrace challenge, engage with a global
perspective, and honor self and place. The academic mission has a specific additional focus on
developing students’ skills and habits of mind for engagement in authentic inquiry.
Organizational Performance Goals
Long Range Goal: The goal of the Hawaii School is to achieve improved gender equality, as
defined by a gender representation that approaches the demography of the student body (48.2%
male and 51.8% female), in participation in math-intensive STEM courses in the high school by
the fall semester of 2021.
Supportive Shorter Term Goal: The shorter term goal of the Hawaii School is to evaluate and, if
necessary, improve gender equity in self-efficacy in math-intensive STEM pilot courses aligned
to the new competency-based curriculum and mastery assessments in the high school by the end
of the spring semester of 2019.
Stakeholder 1 Goal Stakeholder 2 Goal Stakeholder 3 Goal
School Administration:
By the fall semester of 2019,
the school administration will
have provided 100% of high
school STEM faculty members
with professional development
in the area of gender-inclusive
STEM best practices for
pedagogy and assessment, with
a focus on techniques for
enhancing students’ self-
efficacy.
STEM Faculty Members:
By the spring semester of
2021, the Hawaii School
STEM faculty members will
have critically assessed 100%
of their curriculum for gender
inclusivity and will have
responded to all areas for
improvement with meaningful
changes in pedagogy and
assessment, with a focus on
techniques for enhancing
students’ self-efficacy.
Students:
By the spring semester of
2019, 100% of students who
participated in the
competency-based STEM
pilot courses will be able to
self-assess their inquiry
skills and habits of mind,
and create action plans to
ensure their growth in these
areas and support their
success in future STEM
courses.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

169
Purpose of the Project and Questions
The purpose of this project was to evaluate the degree to which the Hawaii School is
meeting its goals of gender equality in math-intensive STEM course enrollment and gender
equity in self-efficacy in competency-based pilot courses. The analysis focused on the
knowledge, motivation, and organizational elements related to achieving the organizational
goals. While a complete performance evaluation would have focused on all stakeholders, for
practical purposes the stakeholder focused on in this analysis was the Hawaii School students.
As such, the questions that guided this study were the following:
1. To what extent is the organization meeting its goals?
2. What are the knowledge, motivation, and organizational elements related to the
Hawaii School’s goal to achieve improved gender equity in self-efficacy in math-
intensive STEM courses aligned to the new competency-based curriculum and
mastery assessments in the high school by the end of the spring semester of 2019?
3. What are the recommendations for organizational practice in the areas of knowledge,
motivation, and organizational resources that may be appropriate for solving the
problems of gender inequities in self-efficacy and gender gaps in participation in
math-intensive STEM courses at the Hawaii School?
Introduction and Overview
The content of Chapter 4 provided insights into the first two questions guiding this study.
The purpose of Chapter 5 is now to answer the third and final research question, by
recommending solutions for the validated knowledge-based, motivational, and organizational
influences and creating an integrated implementation and evaluation plan for enacting those
solutions. In the sections that follow, each set of validated influences is aligned with principles
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

170
from the literature to generate context-specific recommendations. These recommendations then
inform the development of a program to be implemented as a research-based solution to the
problem of practice. In this study, the articulated program is the next round of competency-
based pilot STEM courses at the Hawaii School, which will incorporate data and findings from
this study to improve effectiveness and value for students.
After aligning influences with recommendations, later sections in this chapter articulate
the development of an integrated implementation and evaluation plan using the New World
Kirkpatrick Model (Kirkpatrick & Kirkpatrick, 2016). While the Kirkpatrick model for training
evaluation has always emphasized four levels of intended outcomes for training events, including
reactions, learning, behavior, and results, the New World Model approaches design using an
integrated approach and beginning with Level 4 Results (Kirkpatrick & Kirkpatrick, 2006;
Kirkpatrick & Kirkpatrick, 2016). In this chapter the planning for the program begins with Level
4, by articulating the leading indicators of successful accomplishment of the organizational and
stakeholder goals. Level 3 planning follows, in which critical stakeholder behaviors for
accomplishing goals are identified, along with the factors that would drive the development of
those behaviors. In Level 2 learning goals are articulated which, along with the context-specific
recommendations from the first half of this chapter, inform the design of the program.
Simultaneous to backwards designing the program, which would be considered the
implementation component of the integrated implementation and evaluation plan, metrics and
timelines for evaluation at all four levels are articulated. The integrated plan concludes with a
presentation of sample survey items and proposals for data visualization for various key
stakeholders. The planning is performed with the end in mind and evaluation is integrated into
implementation, both during planning and executing the program.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

171
Recommendations for Practice to Address KMO Influences
Knowledge Recommendations
Introduction. The data from this study validated procedural and metacognitive
knowledge influences on the problem of practice. Despite strengths in executing the steps of
STEM inquiry, students demonstrated procedural knowledge gaps relating to the processes of
prioritizing, interpreting, and analyzing authentic, messy data sets. In addition, students revealed
self-regulation gaps in the lab through their struggles to reframe lab setbacks as data rather than
failures. In addition, girls’ gendered beliefs about STEM participation and performance were
revealed in the interview data. It is predicted that providing training and education on these
influences could help to narrow the gender gap in self-efficacy in math-intensive STEM subjects
and the related participation gaps in high school and beyond. Given this study’s focus on self-
efficacy, there is a high density of recommendations aligned to metacognitive knowledge
influences. The conceptual framework for this study highlighted the relationship in the literature
between metacognition and self-efficacy, connecting a robust set of self-regulation and self-
assessment strategies to improved STEM performance, a stronger sense of STEM identity, and
ultimately an increased STEM self-efficacy (Bandura, 1977, 1986, 1998; Dembo & Eaton, 2000;
Flowers III & Banda, 2016; Goeden et al., 2015; Lundeberg & Moch, 1995; Mathabathe &
Potgieter, 2017). Table 19 articulates the specific approaches to training and education predicted
to impact this problem of practice, and it references the theoretical principles in support of these
recommendations.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

172
Table 19
Summary of Knowledge Influences and Recommendations
Assumed
Knowledge
Influence
Validated
as a Gap?
Yes, High
Probability
or No
(V, HP, N)
Priority?
Yes, No
(Y, N) Principle and Citation
Context-Specific
Recommendation
Students need to
understand the
steps of inquiry
inherent in
STEM
disciplines. (P)
V Y An inquiry-based
curriculum creates an
environment where the
behaviors of STEM
inquiry can be practiced
(Tuckman, 2009).
Performance feedback
given during learning
enhances information
processing (Mayer, 2011).
Training — provide
classroom instruction that
presents information about
the skills and mindsets of
scientific inquiry,
followed by repeated
practice in authentic
inquiry experiences with
frequent formative
feedback.
Students need to
develop science
metacognition
skills — to
include
laboratory self-
regulation and
conceptual self-
reflection. (M)
V Y A classroom that provides
instruction, modeling and
opportunities for
metacognitive reflection
could increase the
occurrence and transfer of
that behavior (APA, 2015;
Baker, 2006; Denler et al.,
2006; Tuckman, 2009).
Education — provide
targeted learning
opportunities that both
demonstrate and ask
students to reflect on
strategies that they can use
to:
• overcome obstacles in
the lab
• identify what they do
not understand
• correct misconceptions
Students need to
reflect on their
own beliefs
about how
gender roles
relate to science.
(M)
V Y Students need to be
provided the opportunity
and structure to create
meaning by connecting
their knowledge to their
interests and beliefs
(Schraw & McCrudden,
2006).
Education — provide
opportunities for reflection
and discussion about
gender identity and gender
roles and their relationship
to science.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

173
Increasing students’ procedural knowledge of the steps of inquiry in STEM. The
results and findings from this study indicate that 26.9% of students lack in procedural knowledge
about how to conduct STEM inquiry, particularly the steps involved in interpreting authentic,
messy data sets. Behavioral and information processing theories inform a recommended solution
to this procedural knowledge gap. Tuckman (2009) asserted that environment shapes behavior
and that immediate feedback creates reinforcement of desirable behaviors. Information
processing theory also highlights the effectiveness of frequent performance feedback during
learning (Mayer, 2011). These theories would suggest that constructing an environment using
inquiry-based curriculum and formative assessment strategies could enhance students’ ability to
engage in the steps of STEM inquiry. Thus, the recommendation is to provide training to
students in the form of classroom instruction that presents information about the skills and
mindsets necessary to conduct scientific inquiry, followed by repeated practice in authentic
inquiry experiences with frequent formative feedback.
Clark and Estes (2008) indicated that training is the appropriate tool for closing a
procedural knowledge gap when learners would benefit from the combination of demonstration,
guided practice, and formative feedback. The researchers go on to assert that good training
begins with a clear articulation of what learners will be able to do when the training is complete
(Clark & Estes, 2008). A foundational construct in competency-based education (CBE) aligns
with this principle, stating that deeper and more enduring learning will take place if the desired
skills and habits of mind are made transparent for students (Sturgis & Casey, 2018). In CBE, the
learning outcomes are explicit and measurable, and formative feedback is used to assist learners
in setting goals for continuous improvement (Giammatteo & Obaya, 2018; Sturgis & Casey,
2018). For girls in math-intensive STEM, it has been found that focusing on the competencies
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

174
necessary to perform authentic, meaningful scientific inquiry not only deepens their learning but
activates their engagement and increases their future participation in STEM (Fredricks et al.,
2018; Kang & Keinonen, 2017; Wang, 2013). Therefore, training in the classroom that focuses
on guided practice of the skills and mindsets of STEM inquiry with frequent formative feedback
is a well-aligned recommendation for closing this procedural knowledge gap and addressing the
larger problem of practice.
Developing students’ self-regulation and self-reflection strategies. The data from this
study indicates that students lack key skills and strategies in laboratory self-regulation and
conceptual self-reflection, as 18.7% of students still struggle to reframe setbacks in the lab as
data and 29.2% of students are not self-assessing nor setting goals aligned to course
competencies. A variety of learning theories offer possibilities for addressing these
metacognitive knowledge gaps, including behavioral, information processing, and social
cognition theories. Research from Rueda (2011) asserted that self-regulation and metacognitive
thinking skills can be taught; and Tuckman (2009) emphasized the importance of identifying and
designing for the behavioral objectives for learning. Baker (2006) suggested that metacognition
can be modeled by teachers who talk through strategy selection and emphasize the processes of
strength identification and targeted goal setting. Additionally, it is considered important to
provide students with opportunities for self-evaluation and the skills to optimize and adjust
learning strategies (Denler et al., 2006). This body of research indicates that a classroom that
provides instruction, modeling and opportunities for metacognitive reflection could increase the
occurrence and transfer of that behavior. The recommendation then is to provide education to
students through targeted learning opportunities that both demonstrate and ask students to reflect
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

175
on strategies that they can use to: (1) overcome obstacles in the lab, (2) identify what they do not
understand, and (3) correct misconceptions.
Clark and Estes (2008) clarified that while training is appropriate for teaching and
practicing the execution of a particular technique, education is the more appropriate tool to
prepare learners to solve future, novel challenges. Education, therefore, is a good fit for closing
metacognitive knowledge gaps. Rueda (2011) noted that “a major goal of education is to
produce self-regulated learners who have acquired expertise and can transfer their knowledge
and skills to real world problems” (p. 17). Self-regulation and self-evaluation skills are
necessary for adaptive learning, resilience, and autonomy (Giammatteo & Obaya, 2018; Rueda,
2011). In fact, Rueda (2011) defined metacognitive knowledge as the source of strategic
problem-solving behaviors. Therefore, it is likely that self-regulation and self-reflection will be
enhanced with an educational intervention in the classroom that focuses on modeling and
guiding metacognitive skill development.
Addressing gendered beliefs about STEM. While only 8.2% of survey respondents
indicated a relationship between gender and science ability, the interview findings from this
study reveal that 66.7% of girls interviewed held implicit and unaddressed gendered beliefs
about who is good at STEM, who likes STEM, and who should pursue STEM. Information
processing theory informs a potential solution to this metacognitive gap. Schraw and
McCrudden (2006) emphasized that meaning-making comes from the intersection of interests,
beliefs, and learning events. Students need to be provided the opportunity and structure to
confront their understanding of the relationship between gender and STEM performance and
participation by connecting their knowledge to their interests and beliefs. The recommendation,
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

176
therefore, is to provide education that structures opportunities for reflection and discussion about
gender identity and gender roles and their relationship to math-intensive STEM.
For members of underrepresented groups in STEM fields, there is a risk of developing a
weak science identity and exhibiting behaviors that are driven by stereotype threat (Flowers III &
Banda, 2016; Leaper et al., 2012; Lee et al., 2015). Rueda (2011) included the ability to read
context and conditions as part of his definition of metacognition, which implies that girls’ deep
understanding of both their own personal science identity and the larger context of women’s
participation in STEM are crucial metacognitive gaps to close. Leaper et al. (2012) discovered a
positive correlation between girls’ choice to participate in STEM fields and their ability to apply
feminist and gender-egalitarian beliefs in their strategic thinking. Meanwhile, Flowers III and
Banda (2016) argued that the cultivation of a positive science identity is crucial to closing gender
gaps in STEM self-efficacy and participation. Therefore, educational opportunities to explore
gendered beliefs and implicit biases in the realm of women in math-intensive STEM are likely to
deepen girls’ metacognitive sense of their own decision-making about their participation and
persistence in these fields.
Motivation Recommendations
Introduction. Data analysis for this study revealed motivation influences aligned to the
theories of self-efficacy, attributions, and goal orientation. Pintrich (2003) predicted the
presence of these three influencers, as he asserts that competence beliefs, control beliefs, and
goals are among the keys to motivating students. Motivation involves the processes of initiating,
sustaining, and ensuring the quality of goal-oriented activity (Rueda, 2011). These three
motivational indexes are also referred to as active choice, persistence, and mental effort (Clark &
Estes, 2008). In the context of K-12 education, recommendations for addressing student
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

177
motivation issues exist mostly in the design of curriculum, pedagogy, and assessment; and in the
creation of nurturing, supportive classroom environments. Table 20 highlights the theoretical
principles aligned to addressing the three motivation influences and suggests context-specific
recommendations. While all three influences have been found to affect motivation in this study,
a significant amount of research concludes that STEM self-efficacy is, in fact, the single most
important influence on a girl’s choice to pursue STEM fields and her persistence during the
hardships she encounters during that pursuit (Bandura, 1977, 1986, 1998; DiBenedetto &
Bembenutty, 2013; Leaper et al., 2012; Pajares, 2006; Peters, 2013; Uitto, 2014; Zeldin et al.,
2008; Zeldin & Pajares, 2000).

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

178
Table 20
Summary of Motivation Influences and Recommendations
Assumed
Motivation
Influence
Validated
as a Gap?
Yes, High
Probability,
No
(V, HP, N)
Priority?
Yes, No
(Y, N) Principle and Citation
Context-Specific
Recommendation
Self-Efficacy —
Students need to be
confident in
executing the
mathematical and
laboratory tasks
required in math-
intensive STEM
courses.
V Y Feedback and modeling
increases self-efficacy
(Pajares, 2006).
Learning and motivation are
enhanced when learners have
positive expectancies for
success (Pajares, 2006).
Provide students a classroom
environment in which they:
• receive frequent, specific
feedback which includes
procedural advice
• celebrate their successes on
close, concrete goals
• experience a positive
emotional environment where
it is made clear that all students
are capable of success
• follow successful, similar
models to encourage
expectation of success
Attributions —
Students need to
feel their
performance
depends upon their
sustained effort
rather than
believing their
STEM aptitude is
fixed.
V Y Learning and motivation are
enhanced when individuals
attribute success or failures to
effort rather than ability
(Anderman & Anderman,
2006).
Provide feedback that stresses
the process of learning,
including the importance of
effort, strategies, and potential
self-control of learning
(Anderman & Anderman,
2006).
Provide attributional retraining
through structured feedback in
which success or failure is
attributed to effort; and ensure that
feedback is specific to identifying
and supporting growth in the skills
and habits of mind on which
individual students need to work
in order to achieve better learning
outcomes.
Goal-Orientation
— Students need
to engage in math-
intensive STEM
courses with the
desire to master
new skills and
enact their
curiosity,
resourcefulness,
persistence, and
resilience.
V Y Focusing on mastery,
individual improvement,
learning, and progress
promotes positive motivation
(Yough & Anderman, 2006).
Goals motivate and direct
students (Pintrich, 2003).
Provide students with curriculum,
pedagogy, and assessment
structured around mastery,
learning, effort, and progress; and
emphasize mastery goal-setting
with assessment centered on
measuring individual growth
towards meeting those goals.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

179
Increasing students’ self-efficacy. The results and findings of this study indicate that,
prior to completing the competency-based pilot chemistry course, 34.2% of female students
(n = 73) lacked confidence in their ability to “do the hardest work that will be assigned in this
chemistry class.” A recommendation rooted in self-efficacy theory was selected to address this
confidence gap in the design of the pilot course. Pajares (2006) asserted that learning and
motivation are enhanced when learners have positive expectancies for success, and he goes on to
suggest that feedback and modeling are tools for increasing self-efficacy. These findings imply
that self-efficacy will be bolstered, particularly for underrepresented groups, within a curriculum
that uses specific feedback to create feelings of success on incremental goals and an environment
of safety, support, and effective peer modeling. Nearing completion of the course, only 2.7% of
the female students surveyed reported low confidence in their ability to “perform the laboratory
tasks required in this class,” while 17.8% of the girls in this study still lacked confidence in their
ability to “perform the mathematical tasks required in this class.” The recommendation moving
forward is to continue to foster a classroom environment in which students: (1) receive frequent,
specific feedback which includes procedural advice, particularly about executing the
mathematical tasks required in the course; (2) celebrate their successes on close, concrete goals;
(3) experience a positive emotional environment where it is made clear that all students are
capable of success; and (4) follow successful, similar models to encourage expectation of
success.
It is found that students with low self-efficacy tend to avoid difficult tasks and give up
more quickly, creating a link between self-efficacy and participation in a discipline (Bandura,
1998). In addition, Pajares and Miller (1994) determined that self-efficacy is the most powerful
predictor of performance over other theories of learning and motivation. With enduring gender
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

180
gaps in both participation and performance in math-intensive STEM fields, many researchers
have looked closely at gender disparities in self-efficacy as a root cause for these gaps (Kitts,
2009; Pajares & Miller, 1994; Uitto, 2014; Zeldin et al., 2008; Zeldin & Pajares, 2000).
Bandura’s 1986 articulation of his theory of self-efficacy stated four key influences on self-
efficacy development: (1) mastery experiences, (2) vicarious experiences, (3) social persuasions,
and (4) physiological states. To cultivate a strong sense of self-efficacy, students must
experience success on a task, see similar models experiencing success, receive meaningful
encouragement, and experience limited negative emotions during the learning process.
Therefore, there is an excellent alignment between Bandura’s (1986, 1998) work on self-efficacy
and the recommendation to foster a classroom environment in which students: (1) receive
frequent, specific feedback, (2) celebrate their successes on close, concrete goals, (3) experience
a positive emotional environment where it is made clear that all students are capable of success,
and (4) follow successful, similar models to encourage expectation of success.
Providing attributional retraining. The results from the survey given near the
conclusion of the pilot chemistry course indicate that 98.8% of survey respondents (n = 171)
agree that “persistence is important for success in science” and 91.8% of those surveyed agree
that “I am in charge of my success in this class.” However, there were still only 57.3% of
respondents agreeing that “if I perform poorly on an assessment, it is because I did not try hard
enough to learn the material” and 31.5% of those surveyed agreeing that “if I perform poorly on
an assessment, it is because it is too hard for me.” A recommendation rooted in attribution
theory has been selected to address this imbalance in students’ beliefs about the stability of their
STEM performance and the controllability of their success. It has been found that learning and
motivation are enhanced when individuals attribute success or failures to effort rather than
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

181
ability; and attributional retraining can occur with targeted feedback that stresses the process of
learning, including the importance of effort, strategies, and potential self-control of learning
(Anderman & Anderman, 2006). The recommendation, therefore, is to ensure that all feedback
is specific to identifying and supporting growth in the skills and habits of mind on which
individual students need to work in order to achieve better learning outcomes, and to be mindful
that all feedback and modeling attribute success or failure to effort.
Rueda (2011) described attributions as part of the human tendency towards sense-
making. If a student attributes success to her efforts, she will be more likely to persist when a
task is challenging (Rueda, 2011). Effort-oriented attributions are considered internal, unstable,
and controllable (Anderman & Anderman, 2006; Hochanadel & Finamore, 2015). Alternatively,
if a student were to believe that her performance was dependent upon innate intelligence, her
attributions would be similarly internal, but instead stable and uncontrollable. Such a fixed
mindset about set-backs experienced while studying STEM runs counter to the nature of
scientific inquiry (Kang & Keinonen, 2017; Lundeberg & Moch, 1995). Bauer (2005) described
the scientific habits of mind as a set of shared values, attitudes, and skills that makes possible a
person’s continued engagement in the process of inquiry. The literature, therefore, supports the
recommendation that all feedback ties success to effort and is specific to identifying and
supporting growth in the skills and habits of mind on which individual students need to work in
order to achieve better learning outcomes.
Orienting goals towards mastery and learning. The data from the survey given near
the conclusion of the pilot chemistry course shows that while 89.5% of survey respondents
(n = 171) are setting goals to gain a deeper understanding of science, 86.5% of those surveyed
also indicate that they are setting goals related to getting a good grade in science. In addition,
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

182
while 93.6% of students surveyed assert that their goals are about self-improvement, there were
still 17.0% of respondents who claimed that their goals were about outperforming others. A
recommendation rooted in goal-orientation theory is suggested to address this problem of
competing goal orientations. Pintrich (2003) asserted that goals motivate and direct students,
and Yough and Anderman (2006) specified that positive motivation comes from goals that focus
on mastery, individual improvement, learning, and progress. Structural changes to the way
courses are designed and assessed could promote changes in students’ goal orientation. The
recommendations, therefore, are to provide students with curriculum, pedagogy, and assessment
structured around mastery, learning, effort, and progress; and to emphasize mastery goal-setting
with assessment centered on measuring individual growth towards meeting those goals.
Mastery goals relate to improvement and are benchmarked against one’s own previous
performance, while performance goals are driven by competition and comparing oneself to
others (Yough & Anderman, 2006). Mastery goals are set with specific tasks in mind, and goal
achievement is measured by tracking an individual’s evolving performance on those tasks. Task-
specific mastery goals are correlated to higher academic achievement; and an orientation towards
setting goals to improve on one’s own previous achievement is found to be highly effective at
maintaining engagement and motivation in learning environments (Lewis, 2018; Martin & Elliot,
2016). There is also thought to be a connection between the attributes of a growth mindset and a
mastery goal orientation, as mastery-oriented students are said to be more adaptive in the face of
setbacks (Dweck, 1986; Lau & Roeser, 2008; Lewis, 2018). Pintrich (2003) suggested that an
orientation towards mastery will inspire motivation for authentic, deep learning in the classroom.
The literature on goal orientation, therefore, supports the recommendations to emphasize, model,
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

183
and teach mastery goal-setting and to provide students with curriculum, pedagogy, and
assessment structured around mastery, learning, effort, and progress.
Organization Recommendations
Introduction. The data from this study verified organizational influences on both the
problem of gender self-efficacy gaps in math-intensive STEM courses in high school and the
related problem of participation gaps in these courses. Table 21 highlights the cultural models
and cultural settings that could potentially be adjusted through context-specific recommendations
in order to effect change on these problems of practice. Strategies for addressing organizational
influences include: confronting the current culture by raising questions and engaging in dialogue
about equity and inclusion, adapting research-based best practices to be culturally relevant for
the organization, investigating and adjusting policies and practices, and engaging in small-scale
pilots of change initiatives that utilize the plan-do-study-act (PDSA) cycle (Bensimon, 2005;
Chavez, Duran, Baker, Avila, & Wallerstein, 2008; Clark & Estes, 2008; Elmore, 2002). Table
21 illustrates how these theoretical principles could be applied to address cultural models that
don’t fully value a competency-based curriculum or gender equality in participation in math
intensive STEM courses. The table also describes how cultural settings of both organizational
policies and procedures and classroom practices could be adjusted to address the gender gaps in
self-efficacy and participation.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

184
Table 21
Summary of Organization Influences and Recommendations
Assumed
Organization
Influence
Validated
as a Gap?
Yes, High
Probability,
No
(V, HP, N)
Priority?
Yes, No
(Y, N) Principle and Citation
Context-Specific
Recommendation
Cultural Model —
The organization
needs to sustain a
culture of gender
equity and inclusion
in all STEM courses.
V Y Effective leaders are aware
of biases and prejudices that
occur in the organization at
the individual and structural
levels (Bensimon, 2005;
Chavez et al., 2008).
Effective leaders address
institutional policies and
practices that create barriers
for equity (Bensimon,
2005).
Define broadly the
constructs of diversity,
equity, and access as they
currently exist in STEM
classrooms.
Consider how current
practices (particularly the
course enrollment process)
either promote or inhibit
equity, diversity, and
inclusion.
Cultural Model —
The organization
needs to place a high
value on developing
all students’
competency in
STEM skills and
habits of mind.
V Y Effective change begins by
addressing motivation
influencers; it ensures the
group knows why it needs
to change (Clark & Estes,
2008).
Effective change efforts use
evidence-based solutions
and adapt them, where
necessary, to the
organization’s culture
(Clark & Estes, 2008).
Create more space for two-
way dialogue about
competency-based teaching
and learning, where
evidence of best practices
can be shared and concerns
can be aired.
Create a team to articulate
how any evidence-based
change effort should be
adapted for the organization.
Cultural Setting —
The organization
needs to develop and
deliver STEM
curriculum that is
focused on mastery
of skills over
content-based
performance.
V Y Effective change efforts will
test (and if needed, modify)
a change or innovation on a
small scale before
implementing it widely.
(PDSA) (Clark & Estes,
2008).
Accountability is increased
when individual roles and
expectations are aligned
with organizational goals
and mission (Elmore, 2002).
Pilot competency-based
STEM courses and collect
data to measure
effectiveness.
Develop strategies to align
individual practices with
organizational goals.
Support accountable
autonomy.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

185
Verifying that a cultural model of gender equity and inclusion guides student
experiences in STEM courses and enrollment processes. While interview respondents
(n = 171) agreed with the statement “the Hawaii School culture promotes equal participation in
STEM by male and female students” at a frequency of 95.32%, gender inequalities in enrollment
in math-intensive STEM courses persist. At the conclusion of the pilot course, 60.95% of the
male students and 53.75% of the female students enrolled in physics for the following school
year. Greater disparities existed in enrollment into honors or AP levels. Of the male students
from the pilot chemistry course who enrolled in physics (n = 64), 28.12% selected honors or AP
level physics, while only 11.63% of female pilot chemistry students who enrolled in physics
(n = 43) elected to take a course with rigor beyond the basic “principles of physics” course. A
recommendation rooted in diversity theory has been selected to close this organizational gap.
Ensuring equity and inclusion requires effective leaders, who are aware of biases and prejudices
that occur in the organization at the individual and structural levels (Bensimon, 2005; Chavez et
al., 2008). Bensimon (2005) asserted that it is crucial for leadership to address institutional
policies and practices that create barriers for equity. An evaluation of how students are formally
and informally guided in their course selections will be a necessary step to address this gap. The
recommendation, therefore, is to investigate and define broadly the constructs of diversity,
equity, and access as they currently exist in STEM classrooms; and to consider how current
practices, particularly the formal course enrollment process, either promote or inhibit equity,
diversity, and inclusion.
Ensuring positive experiences in science and math for high school girls is crucial in
addressing gender gaps in participation in STEM at both the high school and college levels
(Bottia et al., 2015; Uitto, 2014). A mindful focus on establishing a shared organizational value
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

186
system of inclusivity in the classroom experience is one important lever in building gender
equity in self-efficacy, participation, and performance in high school STEM. Research indicates
that gender inclusive classrooms will need to utilize an individualized approach, matching
curriculum, pedagogy, and assessment decisions with the values and strengths of the diverse
students in the class (Aragón et al., 2017; Burkam et al., 1997; Chetcuti, 2008; Darby, 2005;
Demetriou & Wilson, 2009; Kanny et al., 2014; Ramsey et al., 2013; Schuster & Martiny, 2017).
In addition to attending to the institutional value system that cultivates girls’ experiences within
classrooms, institutions must also critically evaluate procedures and policies for gender equity
and inclusion. Coleman and Stevenson (2013) asserted that for independent schools to really
becomes diverse communities they must ensure alignment between their mission, practices, and
policies. Lumby (2009) acknowledged that this includes the hard work of accepting,
confronting, and replacing the past with a new inclusive present. The literature on diversity,
equity, and inclusion supports the recommendations to investigate and define broadly the
constructs of diversity, equity, and access as they currently exist in STEM classrooms; and to
consider how current practices, particularly the formal course enrollment process, either promote
or inhibit equity, diversity, and inclusion.
Investigating and aligning the organization’s cultural model around the value of
competency-based education. Results and findings from this study revealed that 77.19% of
survey respondents from the pilot course (n = 171) disagreed with the statement “I feel anxious
in science class,” and 91.67% of interview subjects (n = 12) explicitly named the cultural setting
of a competency-based classroom and mastery assessments as a cause of reduced stress and
lower anxiety for them in science. In addition, 76.61% of survey respondents (n = 171) agreed
that “when I am in science class, I feel as if I belong.” When the results of that survey item are
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

187
disaggregated by race and ethnicity it is found that for Asian American (n = 77), Multiracial
(n = 34), and Pacific Islander (n = 20) students, the sense of belonging is above average at
80.26%, 82.35%, and 85.00% respectively. Decreasing student anxiety and ensuring equity and
inclusion are foundational reasons why the Hawaii School is pursuing mastery learning. Given
the promising results from this early pilot, a recommendation rooted in organizational change
theory has been selected. Clark and Estes (2008) advised that effective change begins by
addressing motivation influencers to ensure the group knows why it needs to change, and then
implementing evidence-based solutions that are adapted to the organization’s culture. At this
time it is unclear if the faculty stakeholder group shares a common cultural model about the
value of competency-based education. The recommendations, therefore, are to create more
space for two-way dialogue about competency-based teaching and learning, where evidence of
best practices can be shared and concerns can be aired; and to establish a team to articulate how
any evidence-based change effort should be adapted for the organization.
The research in support of competency-based, or mastery, learning leans heavily on the
comparison of the affective impacts of orienting students’ goals towards mastery over
performance. Performance goals, which emphasize comparing one’s performance and abilities
to others, are linked to high levels of student anxiety (Simon et al., 2015). Because mastery
goals are focused on individual growth, they are found to create a healthier mindset that can lead
to higher self-efficacy, positive affect, increased interest, and self-regulation strategies that
increase perseverance (Senko & Tropiano, 2016; Simon et al., 2015). A growing body of
research demonstrates that mastery curriculum, pedagogy, and assessment have a positive effect
on the affect, self-efficacy, and engagement of girls in math-intensive STEM (DiBenedetto &
Bembenutty, 2013; Fredricks et al., 2018; Goeden et al., 2015; Kang & Keinonen, 2017;
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

188
Lundeberg & Moch, 1995; Senko & Tropiano, 2016; Simon et al., 2015; Vincent-Ruz & Schunn,
2017). Given the evidence-based research in support of the competency-based model of
education, Clark and Estes (2008) would advise that aligning buy-in and understanding among
all stakeholders must precede any full-scale implementation efforts. Alongside helping to
familiarize the community with the evidence, it is advised to create space for members of the
organization to evaluate and adjust recommended best practices to fit the organization’s culture
(Clark & Estes, 2008). Organizational change theory supports the recommendations to create
more space for two-way dialogue about competency-based teaching and learning, where
evidence of best practices can be shared and concerns can be aired; and to establish a team to
articulate how any evidence-based change effort should be adapted for the organization.
Piloting competency-based STEM courses to achieve a cultural setting that focuses
on students’ mastery of crucial skills. In this study, 100% of interview subjects (n = 12)
described their learning experience in the pilot course as deep, authentic, and personalized in
comparison to traditional performance-oriented classes. However, only 68.42% of survey
respondents (n = 171) agreed with the statement, “compared to traditional grading, the
competency-based method gives me a better sense of what I do and don’t understand.” A pair of
recommendations rooted in organizational change and accountability theories has been selected
to address this organizational gap. Clark and Estes (2008) asserted that effective change efforts
will test (and if needed, modify) a change or innovation on a small scale before implementing it
widely using the Plan Do Study Act (PDSA) model. Elmore (2002) added that accountability is
increased when individual roles and expectations are aligned with organizational goals and
mission. Because the Hawaii School is engaged in a period of change and innovation, a
continuation of the piloting process will likely be crucial to the organizational change. The
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

189
recommendation, therefore, is to pilot additional competency-based STEM courses that build on
the learning of previous pilots and collect data to measure their effectiveness, and to support
accountable autonomy by developing strategies to align individual practices of pilot teachers
with organizational goals.
If executed well, competency-based education can create authentic partnerships between
faculty and students, allowing student learning to advance to higher levels of application and
transfer (Curry & Docherty, 2017). These strong relationships with teachers and peers are often
more important for girls than boys in math-intensive STEM classes, as a sense of connection and
collaboration are particularly motivating for underrepresented groups (Fredricks et al., 2018).
Pedagogies that create opportunities to relate and collaborate can lead to gains in content
knowledge, critical thinking, and intellectual risk-taking, particularly for students that are
currently underrepresented in math-intensive STEM fields (Goeden et al., 2015; Lundeberg &
Moch, 1995). In addition, assessment structures that focus on the individual are believed not
only to personalize and deepen learning, but to improve equity and inclusion (Chetcuti, 2008;
DiBenedetto & Bembenutty, 2013). An organization attempting to reach these aspirational goals
for teaching and learning will have to undertake a deliberate adjustment of current practices.
Clark and Estes (2008) advised a period of iteration and refinement, in which small-scale pilots
are executed and evaluated to inform future pilots. For the PDSA model to function well, the
organization must commit to heightened accountability for practitioners to align their work with
the organizational goals (Elmore, 2002). Organizational change and accountability theories
support the recommendations to pilot additional competency-based STEM courses that build on
the learning of previous pilots and collect data to measure their effectiveness, and to support
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

190
accountable autonomy by developing strategies to align individual practices of pilot teachers
with organizational goals.
Integrated Implementation and Evaluation Plan
Implementation and Evaluation Framework
The model that guided the design of this implementation and evaluation plan is the New
World Kirkpatrick Model (Kirkpatrick & Kirkpatrick, 2016), based on the original Kirkpatrick
Four Level Model of Evaluation (Kirkpatrick & Kirkpatrick, 2006). In the New World Model,
the four levels of reactions, learning, behavior, and results are considered in reverse so that
training and education can be backwards designed to ensure accountability for intended
outcomes. In the sections that follow, Level 4 is considered first as the desired internal and
external outcomes are defined. Level 4 outcomes are informed by the organization’s mission
and goals, and in this study the outcomes align to the stakeholder group of the students, in
pursuit of their gender equality in participation in math-intensive STEM courses. Level 3
follows with defining the critical behaviors that must be cultivated in the student stakeholder
group, identifying the necessary drivers to correct or adjust students’ knowledge and motivation
influences, and articulating the necessary organizational support structures to enact the drivers.
In Level 2 the learning goals are articulated in support of the desired critical behaviors, and a
program is planned to both target and evaluate declarative and procedural knowledge, attitude,
confidence, and commitment. In Level 1 a plan is developed for measuring students’
engagement, satisfaction, and their perception of the relevance of the program. The inverted use
of the Kirkpatrick and Kirkpatrick (2006) levels is meant to increase the likelihood that enduring
learning, behavior change, and organizational results are the ultimate outcomes of education or
training interventions.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

191
Organizational Purpose, Need and Expectations
The Hawaii School’s mission is to cultivate within each student the capacity to
collaborate, communicate, create, think critically, empathize, embrace challenge, engage with a
global perspective, and honor self and place. The academic mission has a specific additional
focus on developing students’ skills and habits of mind for engagement in authentic inquiry. The
Hawaii School is working on the long-range performance goal of achieving greater gender
equality in participation in math-intensive STEM courses, particularly in the most advanced
courses in math, physics, computer science, and engineering, by the fall of 2021. This long-
range goal was supported by a short term goal of executing a competency-based pilot course in
chemistry during the 2018–2019 school year and evaluating its effectiveness at shrinking
presumed self-efficacy gender gaps and affecting change on enrollment trends. Due to the
internal nature of motivational influences like self-efficacy, the stakeholder group of focus for
this study was the students themselves. The goal for that group was, by the spring semester of
2019, 100% of students who participated in the competency-based STEM pilot courses would be
able to self-assess their inquiry skills and habits of mind, and create action plans to ensure their
growth in these areas and support their success in future STEM courses.
A focus on gender equality in enrollment in any discipline is supported by the school’s
mission, as a well-rounded education in all fields of study is necessary to cultivate the school’s
desired competencies in all graduates. Achieving the school’s mission relies upon equal access
to educational opportunities and experiences for all students, and the area with a current
identified gender participation gap is the math-intensive STEM courses.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

192
Level 4: Results and Leading Indicators
Table 22 shows the proposed Level 4: Results and Leading Indicators organized into
external and internal outcomes and the metrics and methods that could be used to evaluate them.
The outcomes are the lead indicators of continual, successful attainment of the long-range goal to
achieve improved gender equality in enrollment into math-intensive STEM courses, and the
related stakeholder goals to improve the metacognition and self-efficacy of female students.
Internal indicators are likely to occur if critical behaviors of female students can be cultivated by
a unified and skilled faculty stakeholder group. External indicators should follow upon
successful attainment of internal outcomes.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

193
Table 22
Outcomes, Metrics, and Methods for External and Internal Outcomes
Outcome Metric(s) Method(s)
External Outcomes
Female Hawaii School students
graduate from college with a math-
intensive STEM degree
Number of female alumni
granted math-intensive STEM
degrees
Annual alumni survey to gather
degree attainment statistics
Female Hawaii School students
pursue careers in math-intensive
STEM
Number of female alumni
pursuing math-intensive STEM
careers
Annual alumni survey to gather
career pursuits and job titles
Families pursue enrollment in the
Hawaii School because of a
reputation for fostering girls’
participation in STEM
Number of families enrolling
due to reputation for fostering
girls’ participation in STEM
Admissions enrollment survey
or interviews
Presence in the media highlighting
the STEM accomplishments of
Hawaii School students, focusing
on those of females
Frequency and quality of
reporting about girls’ STEM
education at the Hawaii School
Communications department
outputs (communications to
media)
Internship opportunities and
connections built for female
students in math-intensive STEM
Number of STEM-oriented
community partners and the
number of girls involved in
math-intensive STEM
internships
Public service &
entrepreneurship center statistics
on community partnerships and
girls’ involvement
Internal Outcomes
Gender equality in enrollment in
math-intensive STEM courses,
particularly in highest levels
The percentage of females
enrolled in specific math-
intensive STEM courses
Enrollment statistics
Gender equity in self-efficacy in
math-intensive STEM fields at the
Hawaii School
The score on a self-efficacy
measure
Self-efficacy measures
completed and compared
annually (and pre- and post- in
individual classes)
Greater involvement by female
students in engineering and
computer science projects and
enrichment opportunities
The number of girls involved in
these activities
Attendance records and rosters
Girls pursue physics after chemistry The number of girls pursuing
physics
Enrollment statistics
Alignment of teaching practices to
gender-inclusive best practices
The number of science faculty
adjusting practice based upon
gender-inclusive best practices
Review of syllabi and lesson
plans, classroom observations

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

194
Level 3: Behavior
Critical behaviors. The stakeholder group of focus in this study was the Hawaii School
students. The first critical behavior identified is that students must cultivate the skill sets and
mindsets of scientists and engage in authentic STEM inquiry. The second critical behavior is
that students set goals in STEM courses aligned to their personal growth on skills and habits of
mind, rather than on competitive, performance outcomes. The third critical behavior is that
students sustain and nurture their self-efficacy through a focus on competency-aligned mastery
experiences. Table 23 specifies the metrics, methods, and timing for the evaluation of each of
these critical behaviors.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

195

Table 23
Critical Behaviors, Metrics, Methods, and Timing for Evaluation
Critical
Behavior Metric(s) Method(s) Timing
1. Cultivate the
skill sets and
mindsets of a
scientist and
engage in
authentic STEM
inquiry.
a. The number of
trials of an
experiment a
student is willing to
conduct in pursuit
of an answer to a
research question.
a1. The teacher shall
observe and monitor the
quantity of trials a lab
group performs.
a2. Students will be asked
to self-report on number of
trials performed and their
reasoning for the quantity.
For the first quarter, teachers
will monitor this metric for
every research question.
If successful, students will
begin self-reporting, with
monthly accountability
checks by the teacher.
b. The numerical
precision and
accuracy results of
the experimental
outcomes.
Students will be asked to
calculate precision and
accuracy in experiments for
teacher to check and
monitor.
For the first quarter, teachers
will double check all
calculations.
In subsequent quarters,
teachers will intermittently
spot-check students’ work.
c. The degree of
alignment of the
experimental
design to the
research question.
c1. The teacher shall give
feedback as to the ability of
the procedure to produce
data that could answer the
question.
c2. Students will be asked
to self-reflect on alignment.
For the first quarter, teachers
will monitor this metric for
every research question.
If successful, students will
begin self-reporting, with
monthly accountability
checks by the teacher.
d. The frequency of
proposed follow-up
experiments to
explore curiosities
about new
questions that arose
in the lab.
d1. The teacher shall
observe and monitor the
quantity of follow-up
experiments designed, and
assess the depth of curiosity
demonstrated.
d2. Students will be asked
to self-report on follow-up
experiments and self-assess
curiosity.
For the first quarter, teachers
will monitor this metric for
every research question.
If successful, students will
begin self-reporting, with
monthly accountability
checks by the teacher.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

196

Table 23, continued
Critical
Behavior Metric(s) Method(s) Timing
2. Set goals in
STEM courses
aligned to
personal growth
on skills and
habits of mind,
rather than on
performance
outcomes.
a. The frequency of
setting mastery-
oriented goals vs.
performance-
oriented goals.
Students will write goals
that are reviewed by the
teacher for mastery or
performance orientation.
Every mid-quarter.
b. The degree of
alignment of goals
to competencies.
Students will write goals
that are reviewed by the
teacher for alignment to
competencies.
Every mid-quarter.
With feedback, students
begin to assess alignment
for themselves.
c. The degree to
which goals are
SMART (specific,
measurable,
attainable, relevant,
and timely).
Students will write goals
that are reviewed by the
teacher to determine how
SMART they are.
Every mid-quarter.
With feedback, students
begin to assess SMART-
ness for themselves.
3. Sustain and
nurture self-
efficacy through
a focus on
competency-
aligned mastery
experiences.
a. Scores on self-
efficacy measure.
Use pre-existing self-
efficacy measure to
quantify self-efficacy.
Administered three times in
STEM courses: pre-, at the
semester, and post-
b. accuracy of self-
assessment of skills
and habits of mind.
Students will present work
as evidence of mastery of
competencies, teachers will
provide feedback on
accuracy of self-assessment
and validity of evidence.
2x per month.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

197
Required drivers. In order to influence the achievement of the desired outcomes,
students will need supportive classroom environments with teachers that have the knowledge and
skills to cultivate the critical behaviors described in Table 23. As such, teachers and their
curricular, pedagogical, and assessment choices will be the drivers for developing students’
critical behaviors. Kirkpatrick and Kirkpatrick (2016) categorized drivers as either reinforcing,
encouraging, rewarding, or monitoring. Many of the knowledge-based recommendations align
with the category of reinforcing, as training and education solutions are incorporated in
classrooms. Motivational recommendations fall primarily into the encouraging category, as
classroom practices support students in initiating and sustaining goal-oriented behaviors.
Occasionally motivation solutions involve an incentive, which places those drivers in the
rewarding category as students are celebrated for their successes. Finally, monitoring is most
often an organizational level solution, laying the foundation for accountability measures and data
driven decision making. Table 24 identifies and categorizes the required drivers identified in this
study, outlines the time interval for enacting each strategy, and demonstrates the alignment of
each driver to particular critical behaviors.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

198
Table 24
Required Drivers to Support Critical Behaviors
Method(s) Timing
Critical
Behaviors
Supported
1, 2, 3, etc.
Reinforcing
Training in the classroom about the required skills and habits of mind for
scientific inquiry
Weekly 1, 2
Repeated practice with authentic inquiry Weekly 1, 2, 3
Education that includes both instruction on and targeted opportunities for
students to self-reflect about concepts or self-regulate in the lab
Weekly 1, 2, 3
Education that confronts students’ gendered beliefs about STEM
participation, interest, and performance
Quarterly 3
Emphasis on and modeling of mastery goal-setting Bi-weekly 2
Encouraging
Frequent, specific feedback which is aligned to competencies and
includes procedural advice
Weekly 1, 2, 3
A positive emotional environment where it is made clear that all students
are capable of success
Daily 2, 3
Access to similar, successful models Monthly 1, 3
Attributional retraining, particularly around poor performance on
assessments
Weekly 2, 3
Curriculum, pedagogy, and assessment designed around mastery,
learning, effort, and progress
Daily 1, 2, 3
Rewarding
Celebration of successes on close, concrete goals Weekly 2, 3
Assessments that measure and reward individual progress and growth Weekly 2, 3
Monitoring
Engagement in PDSA cycle with pilot courses, ensuring that data is
collected and effectiveness is determined
Quarterly 1, 2, 3

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

199
Organizational support. The aforementioned critical student behaviors delineated in
Table 23 and the required drivers described in Table 24 are predicated on the implementation of
recommendations at the organizational level. In order to address the gender gaps in self-efficacy
and participation for the student stakeholder group, the cultural models, cultural settings,
policies, and practices of the Hawaii School must be investigated and mindfully realigned in
support of the organizational goals. The cultural model of gender equity and inclusion must be
defined broadly in relation to the constructs of diversity, equity, and access as they currently
exist in STEM classrooms. A shared belief system that competency-based education is a
powerful tool for educational equity will also be important in order to engage faculty members in
coherently enacting the drivers. Establishing this shared value will require two-way dialogue
about competency-based teaching and learning where best practices can be shared and concerns
can be aired, and intentional work to articulate how evidence-based best practices should be
adapted for the organization’s culture. Informed by shared cultural models, cultural settings will
have to be modified through a series of small-scale pilots which align individual practices with
organizational goals, enable accountable autonomy, and are continuously reviewed using the
PDSA cycle. Finally, an evaluation and elimination of any systemic barriers to goal
achievement, particularly those due to current enrollment policies and practices, will be crucial
in ensuring that internal outcomes are authentically attainable through the development of
students’ critical behaviors.
Level 2: Learning
Learning goals. The following program learning goals target the Level 3 critical
behaviors to support stakeholder of focus behavior change en route to achieving internal leading
indicators at Level 4 for ultimate alignment to the stakeholder and organizational goals.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

200
Following the implementation of the recommended solutions, particularly as facilitated in
competency-based, pilot STEM courses, the student stakeholders will be able to:
1. Summarize the steps of authentic STEM inquiry. (D)
2. Carry out the steps of STEM inquiry including: question generation, experimental
design, data analysis, and iteration to achieve more conclusive, precise, and accurate
results. (P)
3. Integrate individualized feedback into goal-setting. (P)
4. Utilize habits of mind of scientists in the lab including: persistence, curiosity,
resourcefulness, and collaboration. (M)
5. Reflect on personal growth in knowledge and skills to accurately self-assess and set
goals. (M)
6. Indicate confidence that they can execute the steps of authentic STEM inquiry.
(confidence)
7. Indicate confidence that they can create action plans for continuing growth on course
competencies. (confidence)
8. Value the personalization and opportunities for deep learning in a competency-based
curriculum. (value)
9. Value the shift in focus away from competitive performance pressure in mastery
assessment practices. (value)
10. Commit to the candid self-reflection and self-assessment necessary to set meaningful
and well-aligned SMART goals. (commitment)
Program. The competency-based pilot course in chemistry that was featured in this
study was designed to integrate guided-inquiry and project-based pedagogies into a competency-
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

201
based curriculum with mastery assessment practices. The course, therefore, required the
chemistry teaching team to reimagine pedagogy, curriculum, and assessment, as part of the
Hawaii School’s larger movement toward mastery learning. Over the next period of years, small
scale pilots will continue to run in specific sub-departments and in the form of new
transdisciplinary courses designed to create unique opportunities for students to pursue their
growth in competencies. Each iteration of the pilot course program will be methodically
informed by evaluation of the pilot courses from the previous year. As such, the program
described in this section represents suggestions for the next iteration of pilot courses,
incorporating the KMO needs analysis from this dissertation study.
The recommended program consists of evidence-based suggestions for how to design and
improve the next round of year-long, competency-based, STEM pilot courses. STEM courses
meet 39 times each semester for 90 minutes each class period, making a total of 58.5 hours of
class time each semester and total of 117 hours for the year. The 117 hours is spread over 9.5
months of classroom time that focuses heavily on training in procedural knowledge and
education in metacognition. The identified assets from the chemistry pilot courses should be
mindfully incorporated into the next round of pilot courses, while correcting for the validated
influences with the context-specific recommendations highlighted in Tables 19, 20, and 21
earlier in this chapter.
The pilot courses should continue to focus on classroom instruction that presents
information about the skills and mindsets for STEM inquiry, followed by repeated practice in
authentic inquiry experiences with frequent formative feedback. Specific additional training
should be added in the form of modeling the interpretation of authentic, messy data sets; and
students should be provided additional, scaffolded opportunities to practice this skill with real-
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

202
time feedback. To generate actual messy data for students, the courses must continue to
emphasize open-ended questions for which students must design their own experiments and draw
their own conclusions. Students need additional support, however, in developing laboratory self-
regulation strategies, specifically for reframing failed experiments as simply more data to
interpret. Incorporating more demonstrations of how this reframing is possible and allowing
students to articulate strategies for their own reframing and reflect on their effectiveness will be
important additions to future pilot courses.
To ensure that students can build and sustain their self-efficacy in the pilot courses, a
stronger emphasis on modeling and social persuasions should be incorporated. The chemistry
pilot course featured in this dissertation demonstrated frequent opportunities for students to
celebrate successes on close, concrete goals and to experience assessment practices that reduced
anxiety and addressed negative physiological feedback. There was not compelling evidence,
however, that the pilot chemistry course was strong in social persuasions and vicarious
experiences. In future pilots, connecting students with successful, similar models to encourage
an expectation for success and ensuring that it is clearly stated and reaffirmed that all students
are capable of success will be important components of fostering self-efficacy. It is also
recommended to openly address students’ gendered beliefs about STEM participation, interest,
and performance by providing opportunities for reflection and discussion about gender identity,
gender roles, and their relationship to science.
Future pilot courses should continue to provide students with curricula, pedagogies, and
assessments that are structured around mastery, learning, effort, and progress. However,
students need to be trained how to set specific goals aligned to course competencies and teachers
need to give more frequent and timely feedback on students’ progress. The next generation of
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

203
pilot courses also needs to create opportunities for attributional retraining aligned to competency-
based assessments, particularly when students are failing to demonstrate mastery. Feedback
from teachers that consistently identifies specific growth areas in skills and habits of mind is
thought to be capable of shifting students’ attributions toward effort rather than fixed ability.
In order to develop the desired critical behaviors articulated in Table 23, the next iteration
of competency-based, pilot STEM courses must benefit from the evaluation of influences
performed in this study. To increase students’ procedural and metacognitive knowledge,
enhance their confidence and commitment, and instill in them the value of developing
competencies, some of the 117 hours of contact time with students need to be repurposed to
enact the recommended solutions.
Evaluation of the components of learning. To determine the effectiveness of the
program, checks for students’ acquisition of declarative and procedural knowledge must be
incorporated throughout instruction. Beyond knowledge, assessment of confidence will be
crucial to ensure that this motivational influence is not inhibiting students’ learning. Attitude
and commitment are also important to methodically evaluate as the competency-based
framework requires a lot of self-direction and ownership from students. If there are gaps in
students’ perceptions of the value of this model and a lack of commitment on their part to setting
challenging goals and candidly reflecting on progress, the cultivation of critical behaviors is less
likely. Table 25 highlights the methods and timing for evaluating these knowledge-based and
motivational components of learning.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

204
Table 25
Evaluation of the Components of Learning for the Program
Method(s) or Activity(ies) Timing
Declarative Knowledge — “I know it.”
Knowledge checks through formative
quizzes
Early in each project to ensure foundational facts
and concepts are internalized
Knowledge checks in real time through
“pair, think, share” and other
individual/group activities
Periodically throughout instruction, documented
through field notes
Pre- test and post- tests Beginning, middle, and end of course
Procedural Skills — “I can do it right now.”
Observations of students’ application of
STEM inquiry skills and habits of mind
Periodically throughout instruction, documented
through field notes
Lab reports that reflect the inquiry process End of each project
Scenario questions on pre- test and post-
test
Beginning, middle, and end of course
Attitude — “I believe this is worthwhile.”
Discussions with students about value,
rationale, and issues
Ongoing informally during the course, targeted
focus groups at the half-way point of each semester
Pre- and post- test assessment Beginning, middle, and end of course
Confidence — “I think I can do it on the job.”
Likert scaled survey items related to
confidence
Ongoing to monitor progress during each project
(beginning, middle, and end of each project)
Discussions with students while executing
tasks
Ongoing, informal — recorded in field notes
Pre- and post- test assessment Beginning, middle, and end of course
Commitment — “I will do it on the job.”
Goal setting — quality of individual
action plans
Ongoing to monitor progress during each project
(beginning, middle, and end of each project)
Observations by instructor during class Ongoing, informal — recorded in field notes
Likert scaled survey items related to
commitment
Ongoing to monitor progress during each project
(beginning, middle, and end of each project)

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

205
Level 1: Reaction
Level 1 evaluation seeks to measure reactions to the program in the categories of
engagement, relevance, and customer satisfaction. Table 26 articulates the methods or tools for
evaluating these reactions and indicates the frequency and timing of each evaluation.

Table 26
Components to Measure Reactions to the Program
Method(s) or Tool(s) Timing
Engagement
Completion of assignments Ongoing during the course
Quality and frequency of goal setting Ongoing during the course, particularly at the
conclusion of each major project
Observations by instructor during class that indicate
not just compliance, but actual cognitive engagement
Ongoing during the course
Observations of out-of-class behaviors that indicate
engagement like setting up teacher meetings or
participating in study groups or STEM enrichment
activities
Ongoing during the course
Course evaluation At the conclusion of each semester
Relevance
Student survey At the half-way point of each semester
Discussions with students Ongoing informally during the course, targeted
focus groups at the half-way point of each
semester
Course evaluation At the conclusion of each semester
Customer Satisfaction
Student survey At the half-way point of each semester
Discussions with students Ongoing informally during the course, targeted
focus groups at the half-way point of each
semester
Course evaluation At the conclusion of each semester

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

206
Evaluation Tools
Immediately following the program implementation. In Tables 25 and 26 course
evaluations and specific survey items were listed as examples of ways to measure Level 1 and
Level 2 outcomes immediately following program implementation. In this evaluation plan,
Level 1 questions were designed to be reflective of just the post-course reactions, measuring
students’ perceptions of their engagement while learning, their satisfaction with the experience,
and the relevance of what they learned using a four-point, forced choice, Likert scale.
Level 2 evaluations incorporate measures of declarative and procedural knowledge,
commitment, confidence, and attitude. The Level 2 rating items in this evaluation plan were
designed to include both post-course assessments and pre-course reflections using a five-point
scale. These items strive to measure both the effectiveness of the program at achieving the
intended learning goals, while also assessing students’ perceptions of their opportunities for
growth in knowledge, confidence, commitment, and attitude. Appendix D provides examples of
Level 1 and Level 2 rating items, such as those that might be used on a course evaluation at the
conclusion of each semester of a pilot course.
Delayed for a period after the program implementation. Kirkpatrick and Kirkpatrick
(2016) advised additional post-program evaluation at a time after the required drivers have been
activated and program participants have actually had opportunities to enact the knowledge and
skills they acquired in the program. The recommended timeframe will vary from organization to
organization, depending upon how long it takes to activate both the drivers and the stakeholders’
critical behaviors (Kirkpatrick & Kirkpatrick, 2016).
The drivers articulated in this implementation and evaluation plan describe actions to be
taken by teachers in pilot STEM courses to reinforce, encourage, and reward students as they
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

207
work on developing critical behaviors. Given the 9.5 month duration of the program and
students’ ongoing acquisition of learning goals and continuous development of their critical
behaviors, it would be supportive of the program to evaluate progress on Level 3 drivers mid-
way through the course. This would provide the opportunity for teachers to adjust their
implementation of drivers in order to achieve better outcomes for students. Appendix E
showcases Level 3 rating items for drivers using a four-point, forced choice Likert scale, for a
survey to be administered mid-way through a pilot course.
While drivers are more pertinent to measure mid-stream in this evaluation plan, Level 3
critical behaviors are better measured at some time frame after completion of a pilot course to
indicate the tenacity and transfer of students’ critical behaviors developed during a pilot course.
In addition, Level 4 outcomes describe the more long-range internal and external indicators of
success aligned to the accomplishment of the stakeholder and organizational goals. As such,
Level 4 rating items would be most appropriate at least three months after completion of a pilot
course. Kirkpatrick and Kirkpatrick (2016) encouraged also revisiting Level 1 metrics of
relevance and customer satisfaction in a delayed survey, and they suggest a check for retention of
Level 2 knowledge and skills-based learning goals. Appendix F highlights sample Level 1, 2, 3,
and 4 rating items that could be used for a delayed survey administered to students three months
after their completion of a pilot course. The rating items include a combination of open-ended
questions, indications of degree of application of behaviors on a five-point scale, and check box
questions.
Data Analysis and Reporting
The implementation and evaluation plan presented in this section will generate a
significant amount of data. Kirkpatrick and Kirkpatrick (2016) offered advice for how to avoid
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

208
the common pitfalls in data collection and analysis including: (1) spending too much time and
energy on Level 1 and Level 2 feedback; (2) asking questions that don’t generate useable data;
(3) making presentations of data analysis too complicated; and (4) simply not using the data that
has been collected. The plan for data analysis and reporting has been designed to mindfully
avoid these pitfalls.
The plan for data analysis and reporting includes creating two dashboards for use by
teachers and administrators. Given the experimental nature of piloting curricular change and the
necessity to continuously collect and analyze data as part of the PDSA cycle, faculty members
could benefit from a visual display of results from the measures for Level 1, Level 2, and Level 3
drivers administered at the conclusion of each semester. Teachers may even find that they would
prefer to collect Level 3 driver data more frequently, if they sense that students need for them to
adjust their practice to optimize the learning environment. Table 27 highlights the information
that would be featured on the teacher dashboard and Appendix G illustrates a visual of how this
display might appear at a glance. The guiding concept would be for all results to display in
aggregate and for survey responses to be kept confidential with no identifying information
collected from the students. Students would be encouraged to share honestly so that survey
results could reflect actual class trends; and the measures would be considered flexible so that
teachers could add questions if they determined that there is data that is failing to be captured.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

209
Table 27
Faculty Dashboard with Kirkpatrick Level 1, Level 2, and Level 3 Drivers Evaluation Data
New
World
Model Data included on the dashboard Visual presentation of the data
Level 1 —
Reactions
Four-point Likert scale data from
survey items evaluating engagement,
relevance, and customer satisfaction
Pie charts that illustrate the percentage
of respondents that strongly disagree,
disagree, agree, and strongly agree for
each metric
Level 2 —
Learning
Pre- and post- responses on five-
point scale for items aligned to
declarative and procedural
knowledge, commitment,
confidence, and attitude
Bar charts that illustrate the means of
pre- and post- responses on each metric
Level 3 —
Drivers
Four-point Likert scale data from
survey items evaluating students’
perceptions of their teachers’
effectiveness at reinforcing,
encouraging, and rewarding their
learning
Pie charts that illustrate the percentage
of respondents that strongly disagree,
disagree, agree, and strongly agree
from the most recent administration of
each metric
In addition, mean responses to each
metric compared to means from
previous administrations in a column
chart to show trends in effectiveness of
teacher practices over time

To serve the broader systemic work of administrators, a second dashboard will be
necessary that would reflect the results from the delayed survey alongside student STEM
enrollment trends and faculty participation statistics in competency-based pilot courses. This
dashboard would allow for monitoring of the Level 3 behaviors, Level 4 results, and institutional
data on the defined internal outcomes articulated earlier in this chapter in Table 22. The goal of
the administrator dashboard would be to monitor if the program is working to accomplish the
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

210
organizational goal and close the gender self-efficacy and participation gaps in math-intensive
STEM courses at the Hawaii School. Table 28 delineates the data reporting choices for the
administrator dashboard and Appendix H illustrates a visual of how this might look at a glance.

Table 28
Administration and Faculty Dashboard with Kirkpatrick Levels 3 and 4 Delayed Evaluation
Data and Additional Institutional Data
Source Data included on the dashboard Visual presentation of the data
Delayed
Evaluation —
Level 3 Critical
Behaviors
Five-point scale data for degree
of application of critical
behaviors, along with
identified preventative factors
Column chart comparing means on
each behavior by gender, stacked bar
chart of preventative factors
disaggregated by gender
Delayed
Evaluation —
Level 4 Results
Frequency data for agreement
to lasting positive outcomes of
participation in pilot course
Stacked bar chart of number of
students agreeing to each outcome,
disaggregated by gender
Enrollment
Statistics
STEM enrollment data for
students in the years following
their completion of a
competency-based pilot STEM
course
Scatterplots of annual enrollment
trends in math-intensive STEM
courses, disaggregated by gender
Faculty Teaching
Assignments
Annual statistics of STEM
faculty involvement in
competency-based pilot
courses
Table of data to include
subdepartment, number of faculty in
each subdepartment, and number of
faculty involved in pilot courses each
year

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

211
The ultimate purpose of collecting, visualizing, and analyzing data is to monitor progress
toward the achievement of stakeholder and organizational goals. The administrator dashboard in
Appendix H can illustrate potential closure of the gender gap in enrollment in math-intensive
STEM courses over time, aligned to the long-term organizational goal. Both the teacher and
administrator dashboards (see Appendices G and H) display results aligned to the short-term
organizational goal of addressing gender self-efficacy gaps. The teacher dashboard visualizes
the aggregate result of changes in Level 2 confidence throughout the course, while the
administrator dashboard illustrates the enduring Level 3 critical behaviors and Level 4 indicators
relating to self-efficacy, disaggregated by gender. Finally, progress on the stakeholder goal is
evident in questions relating to students’ abilities to set competency-aligned goals for
improvement. The teacher dashboard (Appendix G) shows both students’ perceptions of their
ability to set these goals and their feedback on the effectiveness of drivers that would promote
their growth in this area. The administrator dashboard (Appendix H) illustrates students’
perceptions of the enduring positive outcomes from a competency-based pilot course including:
strategies for improving understanding, valuing feedback from teachers, and setting goals that
are targeted and specific. In addition to data from the evaluations designed in this chapter,
progress toward the stakeholder goal should continue to be monitored using samples of student
self-assessments as was modeled in Chapter 4 of this dissertation.
Summary
In the preceding sections, the New World Kirkpatrick Model was employed to design an
integrated implementation and evaluation plan to enact recommended solutions to the problems
of gender inequities in self-efficacy and participation in math-intensive STEM courses at the
Hawaii School (Kirkpatrick & Kirkpatrick, 2016). A program was designed using backwards
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

212
planning, starting from the Level 4 leading indicators of goal attainment and Level 3 stakeholder
critical behaviors. The suggested program offered guidance on how to adjust and improve the
next round of competency-based pilot courses in STEM, making it a natural fit for the current
organizational change initiative at the Hawaii School. Successful implementation requires a
suggested program that is culturally aligned to create maximum buy-in. Successful execution of
the next round of pilot courses in STEM could lead to more consistent acquisition of Level 2
learning goals by all students, driven by teaching practices and teacher behaviors that were
articulated by the Level 3 drivers.
Kirkpatrick and Kirkpatrick (2016) recommended not waiting until the end of
implementation to see if a program has worked by changing behaviors and creating desired
results in an organization, but rather to gather data continuously to see if the program appears to
be working. The evaluation plan articulated in the preceding sections highlights the iterative
design for frequent data collection. The teacher dashboard highlighted in Table 27 and
Appendix G can provide teachers with real-time data to adjust their practices mid-program. The
administrator dashboard highlighted in Table 28 and Appendix H provides data each year that
will support the large scale development of the next iteration of pilot courses, improving upon
the previous year using data and reflection as the PDSA cycle demands.
Frequent and repeated data collection allows the opportunity to explore the following
three questions about outcomes at any of the four Kirkpatrick levels: (1) Does this outcome meet
expectations?; (2) if so, why?; and (3) and if not, why not? (Kirkpatrick & Kirkpatrick, 2016). In
seeking answers to the last two questions, it is likely that new evaluation instruments will have to
be created that involve additional stakeholder groups. The evaluation tools designed in this
chapter were constructed for use by the student stakeholder group. However, as data is reviewed
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

213
to determine if students are meeting expectations, measures of students’ Level 3 critical
behaviors and Level 2 learning from the vantage point of teachers will add valuable data to help
in ascertaining why the students are or are not meeting expectations.
Kirkpatrick and Kirkpatrick (2016) define return on expectations (ROE) as the ultimate
indicator of the value of a training program. The New World Kirkpatrick Model is meant to
ensure this value by clearly articulating expectations from the outset and making clear that
successful attainment of the leading indicators is the measure of success for the implemented
program. Change initiatives that end with leaders wondering how to prove if they were
successful, leave room for disagreement about both the success and value of an initiative.
Change management supported by an integrated implementation and evaluation plan defines
value at the outset, creating both greater buy-in and a greater likelihood of success. The Hawaii
School’s goal to improve gender equality in enrollment in math-intensive STEM courses by
2021 has a greater likelihood to be met with the support of the planning process articulated in
this chapter.
Strengths and Weaknesses of the Approach
In this study two methodological approaches were merged in investigating and
recommending solutions for the problem of practice. The Clark and Estes (2008) gap analytic
framework informed the organization of literature in Chapter 2, the design of the survey
instrument and interview protocol in Chapter 3, the analysis of data in Chapter 4, and the
investigation of research-aligned recommended solutions in Chapter 5. In this chapter, the New
World Kirkpatrick Model was also introduced as the methodological framework for converting
recommendations into a plan for implementing and evaluating a training program.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

214
All methodological and theoretical approaches have strengths and weaknesses. The
strengths of using the Clark and Estes (2008) KMO gap analysis technique in this study were
largely due to the incredible alignment of this framework with a school context. The theories of
learning and motivation that lend insight into K and M influences are the same theories that drive
the design of curriculum, pedagogy, and assessment in high schools. In the independent school
world, the O component of KMO is particularly relevant as each institution is accountable only
to its own mission, creating unique and complex cultural models and cultural settings. Using the
Clark and Estes (2008) gap analytic framework to study a problem of practice in the context of
an independent school was an excellent approach for root cause analysis.
The New World Kirkpatrick Model was well aligned as a method to design a program to
address K and M influences for the student stakeholder group. Because K and M influences
were significant in this study, the backwards planning towards a program to address these root
causes was very effective. In a school setting, however, students occupy a unique position that is
simultaneously internal stakeholder and customer. As such, adjusting the knowledge and
motivation influences of this stakeholder group may not lead to changes in the cultural models of
the stakeholder groups empowered to enact organizational change. Faculty buy-in to the
competency-based initiative will be crucial for effective implementation of the recommended
solutions laid out in this chapter. As the Hawaii School implements the program articulated in
this chapter as a means to achieve organizational and stakeholder goals, administrators should
also direct their attention towards evidence-based strategies for overcoming organizational
obstacles to change such as: resistance, lack of engagement, silence, conflict, poor
communication, inadequate supervision, and prohibitive resources, policies, or procedures
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

215
(Agócs, 1997; Alper, Tjosvold, & Law, 2000; Berger, 2014; Harvey, 1988; Johnson, 1992;
Miller & Lee, 2014; Morrison & Milliken, 2000; Ross, Pinder, & Coles-White, 2015).
Limitations and Delimitations
Potential limitations and delimitations of this study were first considered in the design of
the methodology, as can be seen at the conclusion of Chapter 3. With data analysis and
recommendations for solutions now complete, a broader discussion of limitations and
delimitations of the results and implementation plan can be presented.
The use of a mixed methodology to enable data triangulation was successful as a tool to
strengthen assertions from qualitative data or deepen understanding of quantitative results. In
Chapter 4, the alignment of survey results, interview findings, and document analysis afforded
greater confidence in the validity of quantitative data, the credibility of qualitative data, and the
decisions to validate influences. Another delimitation of the mixed methodology arose when
disaggregated survey data revealed gender inequities, as the choice to interview only female
subjects provided opportunities to interrogate the interview data for possible sources of those
disparities. Alternatively, the choice not to interview male subjects prevented the opportunity to
use interviews to deepen understanding of the male perspective, making this choice a limitation
in data analysis.
An additional limitation in the data analysis resulted from the inability to investigate the
root causes of interview non-participation. While 77 females completed the survey, only 29
consented to be interviewed. The 12 girls selected for interviews were a subset of a self-
selecting interview population, and no part of the interview process investigated why they chose
to participate. Another limitation identified during data analysis was the inability to confirm
students’ shared understanding of the vocabulary used on survey items. In the analysis of
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

216
attributions, for example, it was unclear if the mixed survey results arose due to a
misinterpretation of the word “assessment” as a synonym for traditional test.
One delimitation in the development of the integrated implementation and evaluation
plan was the choice to situate the recommended program within the current Hawaii School
curricular change initiative. By making a plan in support of a current curricular initiative, the
plan is more likely to be enacted and, in turn, the curricular initiative is more likely to succeed.
Another delimitation of the recommended program is that its implementation is the responsibility
of the researcher. Given her position overseeing all curricular initiatives, creating a plan that she
can be responsible for enacting ensures that the recommendations from this chapter will be
utilized by the organization.
A limitation of the implementation plan is that the key drivers are to be executed by a
variety of faculty members, and despite accountability measures it could be difficult to ensure
equal emphasis on those drivers by all faculty members. In addition to possible inconsistent
implementation, it is likely that the largest limitation to the recommended solutions from this
chapter is the still unexplored cultural model alignment, within and across stakeholder groups,
towards the inherent value of competency-based education.
Future Research
This study sought to explore the knowledge-based, motivational, and organizational root
causes for the underrepresentation of girls in math-intensive STEM courses at the Hawaii
School, with a special focus on investigating gender inequities in self-efficacy. This focus was
selected because the body of research on gender gaps in participation identified low self-efficacy
as the most significant influencer on girls’ choices not to participate. The self-efficacy changes
observed in this study, in the cultural setting of a competency-based STEM pilot course, inspire
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

217
the need for an opportunity to explore causality through a methodology that includes random
selection and a control group. It would serve more than just the Hawaii School to discover if
competency-based education could be a proven solution for addressing gender gaps in self-
efficacy in high school STEM.
The relationships between student self-efficacy, teacher gender, and the impacts of
teachers’ implicit biases could also warrant future investigation, as inclusive classrooms must be
built by the educators who implement curricular or pedagogical change. No change initiative
will succeed without ensuring the elimination of bias and the ability of all faculty to create
inclusive experiences for their students.
Outside of the gender self-efficacy results of this study, promising student outcomes
aligned to procedural knowledge, metacognition, attributions, and goal orientation were evident.
Absent pre- and post- data it was not possible to report on students’ growth in these areas, only
on their status at the conclusion of the pilot course. Future studies that incorporate pre- and post-
assessments could better reflect students’ growth and, similar to the self-efficacy proposal above,
a true experimental methodology could investigate causation. The body of research on the
positive effects of competency-based curriculum and mastery assessments is still small,
indicating the need for further research in this area.
To serve the Hawaii School, a future study is needed to explore a variety of stakeholders’
responses to the curricular change initiative to convert to a competency-based curriculum and
align to the values of the Mastery Transcript Consortium. A KMO investigation of both the
causes for resistance and the inspiration for buy-in could serve school leaders well in managing
this organizational change.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

218
Conclusion
This study explored the student experience in a competency-based STEM pilot course,
using Clark and Estes’ (2008) KMO framework. A deeper understanding of the students’
knowledge and motivation at the conclusion of the course provided the opportunity to generate
recommendations for the next iteration of pilot courses at the Hawaii School. The cultural
setting of a competency-based curriculum and mastery assessment practices framed this study
and defined the pilot course experience. The Hawaii School’s rationale for exploring
competency-based education, or mastery learning, includes five aspirations: (1) to build the
emotional resilience of students and take actions to lessen student stress and anxiety; (2) to move
beyond valuing only behavioral engagement, or compliance, and increase students’ cognitive
engagement in their high school coursework; (3) to ensure equity and inclusion for the diverse
student population by creating individualized pathways towards mastery; (4) to enhance intrinsic
motivation by giving students more autonomy and agency in designing their educational journey;
and (5) to focus the curriculum on the acquisition of not just content knowledge, but also the
skills and habits of mind for success in life beyond the Hawaii School.
The organizational goals that inspired this study, goals to achieve gender equality in
enrollment in math-intensive STEM courses and improve gender equity in self-efficacy in these
subjects, are aligned to the equity and inclusion aspiration listed above. At the conclusion of this
evaluation study, it is clear that gender self-efficacy gaps closed for students enrolled in the
competency-based pilot course in chemistry at the Hawaii School. However, the broad study of
all KMO factors for all students in the pilot course provided evidence that the students in the
pilot course were realizing aspirations beyond improved gender equity in self-efficacy. Data
revealed that students were experiencing lower stress and anxiety, demonstrating increased
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

219
interest, reporting feelings of both greater agency and a sense of belonging, and developing the
skills and habits of mind of scientists. As future iterations of the pilot program use data to
improve the execution of competency-based education at the Hawaii School, the school hopes to
realize its bold aspirations for all students.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

220
REFERENCES
Agócs, C. (1997). Institutionalized resistance to organizational change: Denial, inaction, and
repression. Journal of Business Ethics, 16, 917–931. Retrieved from
https://www.jstor.org/stable/25072959
Alkin, M. C. (2011). Section R: How are quantitative data analyzed? In M. C. Alkin & A. T. Vo
(Eds.), Evaluation essentials: From A to Z (pp. 166–179). New York, NY: The Guilford
Press.
Alper, S., Tjosvold, D., & Law, K. S. (2000). Conflict management, efficacy, and performance in
organizational teams. Personnel Psychology, 53(3), 625–642. doi:10.1111/j.1744-
6570.2000.tb00216.x
American Psychological Association, Coalition for Psychology in Schools and Education.
(2015). Top 20 principles from psychology for preK–12 teaching and learning.
Washington, D.C.: Author. Retrieved from http://www.apa.org/ed/schools/cpse/top-
twenty-principles.pdf
Anderman, E., & Anderman, L. (2006). Attributions. Retrieved from
http://www.education.com/reference/article/attribution-theory/
Aragón, O. R., Dovidio, J. F., Graham, M. J. (2017). Colorblind and multicultural ideologies are
associated with faculty adoption of inclusive teaching practices. Journal of Diversity in
Higher Education, 10(3), 201–215. doi:10.1037/dhe0000026
Baker, L. (2006). Developmental differences in metacognition: Implications for metacognitively
oriented reading instruction. In S. E. Israel, C. Collins Block, K. L. Bauserman, & K.
Kinnucan-Welsch (Eds.), Metacognition in literacy learning (pp. 83–102). New York,
NY: Routledge. doi:10.4324/9781410613301
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

221
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change.
Psychological Review, 84, 191–215. doi:10.1037/0033-295X.84.2.191
Bandura, A. (1986). The explanatory and predictive scope of self-efficacy theory. Journal of
Social and Clinical Psychology, 4(3), 359–373. doi:10.1521/jscp.1986.4.3.359
Bandura, A. (1998). Personal and collective efficacy in human adaptation and change. In J. G.
Adair, D. Belanger, & K. L. Dions (Eds.), Advances in psychological science (Vol. 1),
Personal, social, and cultural aspects (pp. 51–71). Hove, UK: Psychology Press.
Bandura, A. (2005). The evolution of social cognitive theory. In K. G. Smith & M. A. Hitt
(Eds.), Great minds in management (pp. 9–35). Oxford, UK: Oxford University Press.
Bauer, C. F. (2005). Beyond “student attitude”: Chemistry self-concept inventory for assessment
of the affective component of student learning. Journal of Chemical Education, 82(12),
1864. doi:10.1021/ed082p1864
Bensimon, E. M. (2005). Closing the achievement gap in higher education: An organizational
learning perspective. New Directions for Higher Education, 2005(131), 99–111.
doi:10.1002/he.190
Berger, B. (2014). Read my lips: Leaders, supervisors, and culture are the foundations of
strategic employee communications. Research Journal of the Institute for Public
Relations, 1(1), 1–17. Retrieved from http://www.instituteforpr.org/wp-
content/uploads/BergerFinalWES.pdf
Bogdan, R. C., & Biklen, S. K. (2007). Qualitative research for education: An introduction to
theories and methods. Boston, MA: Allen and Bacon.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

222
Bottia, M. C., Stearns, E., Mickelson, R. A., Moller, S., & Parker, A. D. (2015). The
relationships among high school STEM learning experiences and students’ intent to
declare and declaration of a STEM major in college. Teachers College Record, 117(3),
1–46. Retrieved from https://eric.ed.gov/?id=EJ1056740
Bowen, G. A. (2009). Document analysis as a research method. Qualitative Research Journal,
9(2), 27–40. doi:10.3316/QRJ0902027
Britner, S. L. (2008). Motivation in high school science students: A comparison of gender
differences in life, physical, and earth science classes. Journal of Research in Science
Teaching, 45(8), 955–970. doi:10.1002/tea.20249
Burkam, D. T., Lee, V. E., & Smerdon, B. A. (1997). Gender and science learning early in high
school: Subject matter and laboratory experiences. American Educational Research
Journal, 34(2), 297–331. Retrieved from https://www.jstor.org/stable/1163360
Chavez, V., Duran, B., Baker, Q. E., Avila, M. M., & Wallerstein, N. (2008). The dance of race
and privilege in CBPR. In M. Winkler & N. Wallerstein (Eds.), Community-based
participatory research for health: From process to outcomes (pp. 91–105). San
Francisco, CA: Jossey-Bass/John Wiley & Sons.
Chetcuti, D. (2008). Identifying a gender‐inclusive pedagogy from Maltese teachers’ personal
practical knowledge. International Journal of Science Education, 31(1), 81–99.
doi:10.1080/09500690701647996
Clark, R. E., & Estes, F. (2008). Turning research into results: A guide to selecting the right
performance solutions. Charlotte, NC: Information Age.
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

223
Coleman, S., & Stevenson, H. C. (2013). The racial stress of membership: Development of the
faculty inventory of racialized experiences in schools. Psychology in the Schools, 50(6),
548–566. doi:10.1002/pits.21693
College Board. (2014). The 10th annual AP report to the nation: Subject supplement [Data file].
New York, NY: Author. Retrieved from
https://research.collegeboard.org/programs/ap/data/nation/2014
Corbin, J., & Strauss, A. (2008). Techniques and procedures for developing grounded theory
(3rd ed.). Thousand Oaks, CA: Sage. doi:10.4135/9781452230153.n4
Creswell, J. W. (2014). Research design: Qualitative, quantitative, and mixed methods
approaches (4th ed.). Thousand Oaks, CA: Sage.
Curry, L., & Docherty, M. (2017). Implementing competency-based education. Collected Essays
on Learning and Teaching, 10, 61–73. Retrieved from https://eric.ed.gov/?id=EJ1147189
Daly, E. (2006). Behaviorism. Retrieved from
http://www.education.com/reference/article/behaviorism/
Darby, L. (2005). Science students’ perceptions of engaging pedagogy. Research in Science
Education, 35(4), 425–445. doi:10.1007/s11165-005-4488-4
Deemer, E. D. (2015). Women’s science major satisfaction: Regulatory focus and the critical
mass hypothesis. Journal of Career Development, 42(1), 60–71.
doi:10.1177/0894845314543190
Dembo, M. H., & Eaton, M. J. (2000). Self-regulation of academic learning in middle-level
schools. The Elementary School Journal, 100(5), 473–490. doi:10.1086/499651
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

224
Demetriou, H., & Wilson, E. (2009). Synthesising affect and cognition in teaching and learning.
Social Psychology of Education: An International Journal, 12(2), 213–232.
doi:10.1007/s11218-008-9075-x
Denler, H., Wolters, C., & Benzon, M. (2006). Social cognitive theory. Retrieved from
http://www.education.com/reference/article/social-cognitive-theory/
DiBenedetto, M. K., & Bembenutty, H. (2013). Within the pipeline: Self-regulated learning, self-
efficacy, and socialization among college students in science courses. Learning and
Individual Differences, 23, 218–224. doi:10.1016/j.lindif.2012.09.015
Duckworth, A., & Quinn, P. D. (2009). Development and validation of the Short Grit Scale
(Grit-S). Journal of Personality Assessment, 91(2), 166–174. Retrieved from
doi:10.1080/00223890802634290
Dweck, C. S. (1986). Motivational processes affecting learning. American Psychologist, 41(10),
1040–1048. doi:10.1037/0003-066X.41.10.1040
Elmore, R. F. (2002). Bridging the gap between standards and achievement: The imperative for
professional development in education. Washington, D.C.: Albert Shanker Institute.
Retrieved from http://www.shankerinstitute.org/sites/shanker/files/Bridging_Gap.pdf
Elster, D. (2014). First-year students’ priorities and choices in STEM studies — IRIS findings
from Germany and Austria. Science Education International, 25(1), 52–59. Retrieved
from https://eric.ed.gov/?id=EJ1022899
Espinoza, O. (2007). Solving the equity-equality conceptual dilemma: A new model for analysis
of the educational process. Educational Research, 49(4), 343–363.
doi:10.1080/00131880701717198
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

225
Fink, A. (2013). How to conduct surveys: A step-by-step guide (5th ed.). Thousand Oaks, CA:
Sage.
Flavell, J. H. (1977). Cognitive development. Englewood Cliffs, NJ: Prentice-Hall.
Flowers III, A. M., & Banda, R. (2016). Cultivating science identity through sources of self-
efficacy. Journal for Multicultural Education, 10(3), 405–417. doi:10.1108/JME-01-
2016-0014
Fredricks, J. A., Hofkens, T., Wang, M., Mortenson, E., & Scott, P. (2018). Supporting girls’ and
boys’ engagement in math and science learning: A mixed methods study. Journal of
Research in Science Teaching, 55(2), 271–298. doi:10.1002/tea.21419
Gallimore, R., & Goldenberg, C. (2001). Analyzing cultural models and settings to connect
minority achievement and school improvement research. Educational Psychologist,
36(1), 45–56. doi:10.1207/S15326985EP3601_5
Gehlbach, H. (2015). User guide: Social-emotional learning. Boston, MA: Panorama Education.
Retrieved from http://www.panoramaed.com/social-emotional-learning
Giammatteo, L., & Obaya, A. V. (2018). Assessing chemistry laboratory skills through a
competency-based approach in high school chemistry course. Science Education
International, 29(2), 103–109. Retrieved from https://eric.ed.gov/?id=EJ1184773
Glesne, C. (2011). Becoming qualitative researchers: An introduction (4th ed.). Boston, MA:
Pearson.
Goeden, T. J., Kurtz, M. J., Quitadamo, I. J., & Thomas, C. (2015). Community-based inquiry in
allied health biochemistry promotes equity by improving critical thinking for women and
showing promise for increasing content gains for ethnic minority students. Journal of
Chemical Education, 92(5), 788–796. doi:10.1021/ed400893f
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

226
Harvey, J. (1988, Summer). The Abilene paradox: The management of agreement.
Organizational Dynamics, 17–43. Retrieved from
https://dppa.wildapricot.org/Resources/AbileneParadoxJerryHarvey.pdf
Hedges, L. V., & Nowell, A. (1995). Sex differences in mental test scores, variability, and
numbers of high-scoring individuals. Science, 269, 41–45. doi:10.1126/science.7604277
Hochanadel, A., & Finamore, D. (2015). Fixed and growth mindset in education and how grit
helps students persist in the face of adversity. Journal of International Education
Research, 11(1), 47–50. Retrieved from https://eric.ed.gov/?id=EJ1051129
Hong, H., & Lin-Siegler, X. (2012). How learning about scientists’ struggles influences students’
interest and learning in physics. Journal of Educational Psychology, 104(2), 469–484.
doi:10.1037/a0026224
Irwin, C. W., & Stafford E. T. (2016). Survey methods for educators: Collaborative survey
development (part 1 of 3) (REL 2016-163). Washington, D.C.: U.S. Department of
Education, Institute of Education Sciences, National Center for Education Evaluation and
Regional Assistance, Regional Educational Laboratory Northeast & Islands.
Johnson, B. (1992). Polarity management: Identifying and managing unsolvable problems.
Amherst, MA: HRD Press.
Johnson, R. B., & Christensen, L. B. (2015). Educational research: Quantitative, qualitative,
and mixed approaches (5th ed.). Thousand Oaks, CA: Sage.
Kang, J., & Keinonen, T. (2017). The effect of inquiry-based learning experiences on
adolescents’ science-related career aspiration in the Finnish context. International
Journal of Science Education, 39(12), 1669–1689. doi:10.1080/09500693.2017.1350790
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

227
Kanny, M. A., Sax, L. J., & Riggers-Piehl, T. A. (2014). Investigating forty years of STEM
research: How explanations for the gender gap have evolved over time. Journal of
Women and Minorities in Science and Engineering, 20(2), 127–148.
doi:10.1615/JWomenMinorScienEng.2014007246
Kekelis, L., Ryoo, J. J., & McLeod, E. (2017). Making and mentors: What it takes to make them
better together. Afterschool Matters, 26, 8–17. Retrieved from
https://eric.ed.gov/?id=EJ1160841
Kirkpatrick, D., & Kirkpatrick, J. (2006). Evaluating training programs: The four levels. San
Francisco, CA: Berrett-Koehler.
Kirkpatrick, J. D., & Kirkpatrick, W. K. (2016). Kirkpatrick’s four levels of training evaluation.
Alexandria, VA: ATD Press.
Kitts, K. (2009). The paradox of middle and high school students’ attitudes towards science
versus their attitudes about science as a career. Journal of Geoscience Education, 57(2),
159–164. doi:10.5408/1.3544253
Krathwohl, D. R. (2002). A revision of Bloom’s Taxonomy: An overview. Theory Into Practice,
41(4), 212–218. doi:10.1207/s15430421tip4104_2
Krueger, R. A., & Casey, M. A. (2009). Focus groups: A practical guide for applied research
(4th ed.). Thousand Oaks, CA: Sage.
Lau, S., & Roeser, R. W. (2008). Cognitive abilities and motivational processes in science
achievement and engagement: A person-centered analysis. Learning and Individual
Differences, 18(4), 497–504. doi:10.1016/j.lindif.2007.11.002
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

228
Leaper, C., Farkas, T., & Brown, C. S. (2012). Adolescent girls’ experiences and gender-related
beliefs in relation to their motivation in math/science and english. Journal of Youth and
Adolescence, 41(3), 268–282. doi:10.1007/s10964-011-9693-z
Lee, J. K., Alston, A. T., & Kahn, K. B. (2015). Identity threat in the classroom: Review of
women’s motivational experiences in the sciences. Translational Issues in Psychological
Science, 1(4), 321–330. doi:10.1037/tps0000050
Lewis, S. E. (2018). Goal orientations of general chemistry students via the achievement goal
framework. Chemistry Education Research and Practice, 19(1), 199–212.
doi:10.1039/C7RP00148G
Li, L. (2002). Gender equity in science — who cares? Journal of Chemical Education, 79(4),
418–419. doi:10.1021/ed079p418
Lourenço, O. (2012). Piaget and Vygotsky: Many resemblances, and a crucial difference. New
Ideas in Psychology, 30, 281–295. doi:10.1016/j.newideapsych.2011.12.006
Lumby, J. (2009). Performativity and identity: Mechanisms of exclusion. Journal of Education
Policy, 24(3), 353–369. doi:10.1080/02680930802669284
Lundeberg, M. A., & Moch, S. D. (1995). Influence of social interaction on cognition:
Connected learning in science. The Journal of Higher Education, 66(3), 312–335.
doi:10.1080/00221546.1995.11774782
Martin, A. J., & Elliot, A. J. (2016). The role of personal best (PB) and dichotomous
achievement goals in students’ academic motivation and engagement: A longitudinal
investigation. Educational Psychology, 36(7), 1282–1299.
doi:10.1080/01443410.2015.1093606
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

229
Mastery Transcript Consortium. (n.d.). A broken tool. Sawyer, MI: Author. Retrieved from
http://mastery.org/a-broken-tool/
Mathabathe, K. C., & Potgieter, M. (2017). Manifestations of metacognitive activity during the
collaborative planning of chemistry practical investigations. International Journal of
Science Education, 39(11), 1465–1484. doi:10.1080/09500693.2017.1336808
Maxwell, J. A. (2013). Qualitative research design: An interactive approach (3rd ed.). Thousand
Oaks, CA: Sage.
Mayer, R. E. (2011). Applying the science of learning. Boston, MA: Pearson Education.
McEwan, E. K., & McEwan, P. J. (2003). Making sense of research. Thousand Oaks, CA: Sage.
Merriam, S. B., & Tisdell, E. J. (2016). Qualitative research: A guide to design and
implementation (4th ed.). San Francisco, CA: Jossey-Bass.
Miller, L. J., & Lee, J. S. (2014). Policy barriers to school improvement: What’s real and what’s
imagined? Seattle, WA: Center on Reinventing Public Education. Retrieved from
https://eric.ed.gov/?id=ED545230
Morrison, E. W., & Milliken, F. J. (2000). Organizational silence: A barrier to change and
development in a pluralistic world. Academy of Management Review, 25(4), 706–725.
doi:10.5465/amr.2000.3707697
National Science Foundation. (2002). NSF’s program for gender equity in science, technology,
engineering and mathematics: A brief retrospective 1993-2001 (NSF Publication No. 02-
107). Alexandria, VA: Author. Retrieved from
https://www.nsf.gov/pubs/2002/nsf02107/nsf02107.pdf
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

230
National Science Foundation. (2016). Bachelor’s degrees awarded, by sex and field: 2004-14.
Alexandria, VA: Author. Retrieved from
https://www.nsf.gov/statistics/2017/nsf17310/static/data/tab5-1.pdf
Pajares, F. (2006). Self-efficacy theory. Retrieved from
http://www.education.com/reference/article/self-efficacy-theory/
Pajares, F., & Miller, D. M. (1994). Role of self-efficacy and self-concept beliefs in
mathematical problem solving: A path analysis. Journal of Educational Psychology,
86(2), 193–203. doi:10.1037/0022-0663.86.2.193
Patterson, J. V., & Johnson, A. T. (2017). High school girls’ negotiation of perceived self-
efficacy and science course trajectories. Journal of Research in Education, 27(1), 79–
113. Retrieved from https://eric.ed.gov/?id=EJ1142363
Patton, M. Q. (2002). Qualitative research and evaluation methods (3rd ed.). Thousand Oaks,
CA: Sage.
Patton, M. Q. (2015). Qualitative research and evaluation methods (4th ed.). Thousand Oaks,
CA: Sage.
Pazzaglia, A. M., Stafford, E. T., & Rodriguez, S. M. (2016). Survey methods for educators:
Selecting samples and administering surveys (part 2 of 3) (REL 2016-160). Washington,
D.C.: U.S. Department of Education, Institute of Education Sciences, National Center for
Education Evaluation and Regional Assistance, Regional Educational Laboratory
Northeast & Islands.
Peters, M. L. (2013). Examining the relationships among classroom climate, self-efficacy, and
achievement in undergraduate mathematics: A multi-level analysis. International Journal
of Science and Mathematics Education, 11(2), 459–480. doi:10.1007/s10763-012-9347-y
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

231
Pintrich, P. R. (2003). A motivational science perspective on the role of student motivation in
learning and teaching contexts. Journal of Educational Psychology, 95, 667–686.
doi:10.1037/0022-0663.95.4.667
Ramsey, L. R., Betz, D. E., & Sekaquaptewa, D. (2013). The effects of an academic environment
intervention on science identification among women in STEM. Social Psychology of
Education: An International Journal, 16(3), 377–397. doi:10.1007/s11218-013-9218-6
Reilly, D., Neumann, D. L., & Andrews, G. (2015). Sex differences in mathematics and science
achievement: A meta-analysis of national assessment of educational progress
assessments. Journal of Educational Psychology, 107(3), 645–662.
doi:10.1037/edu0000012
Renshaw, C. E. (2014). Design and assessment of a skills-based geoscience curriculum. Journal
of Geoscience Education, 62(4), 668–678. Retrieved from
https://files.eric.ed.gov/fulltext/EJ1164163.pdf
Ross, D., Pinder, G., & Coles-White, D. (2015). The impact of structural barriers and facilitators
on early childhood literacy programs in elementary charter schools. Equity & Excellence
in Education, 48(1), 118–136. doi:10.1080/10665684.2015.991550
Rubin, H. J., & Rubin, I. S. (2012). Qualitative interviewing: The art of hearing data (3rd ed.).
Thousand Oaks, CA: Sage.
Rueda, R. (2011). The 3 dimensions of improving student performance. New York, NY:
Teachers College Press.
Sadler, P. M., Sonnert, G., Hazari, Z., & Tai, R. (2012). Stability and volatility of STEM career
interest in high school: A gender study. Science Education, 96(3), 411–427.
doi:10.1002/sce.21007
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

232
Salkind, N. J. (2017). Statistics for people who (think they) hate statistics: Using Microsoft Excel
2016 (4th ed.). Thousand Oaks, CA: Sage.
Schraw, G., & McCrudden, M. (2006). Information processing theory. Retrieved from
http://www.education.com/reference/article/information-processing-theory/
Schuster, C., & Martiny, S. E. (2017). Not feeling good in STEM: Effects of stereotype
activation and anticipated affect on women’s career aspirations. Sex Roles: A Journal of
Research, 76(1-2), 40–55. doi:10.1007/s11199-016-0665-3
Schwarzer, R., & Jerusalem, M. (1995). Generalized self-efficacy scale. In J. Weinman, S.
Wright, & M. Johnston (Eds.), Measures in health psychology: A user’s portfolio. Causal
and control beliefs (pp. 35–37). Windsor, UK: NFER-NELSON.
Senko, C., & Tropiano, K. L. (2016). Comparing three models of achievement goals: Goal
orientations, goal standards, and goal complexes. Journal of Educational Psychology,
108(8), 1178–1192. doi:10.1037/edu0000114
Simon, R. A., Aulls, M. W., Dedic, H., Hubbard, K., & Hall, N. (2015). Exploring student
persistence in STEM programs: A motivational model. Canadian Journal of Education,
38(1), 1–27. Retrieved from http://journals.sfu.ca/cje/index.php/cje-rce/article/view/1729
Stafford, M. E. (2006). Testing and measurement. Thousand Oaks, CA: Sage.
Strauss, A., Schatzman, L., Bucher, R., & Sabshin, M. (1981). Psychiatric ideologies and
institutions (2nd ed.). New York, NY: Wiley.
Sturgis, C., & Casey, K. (2018). Designing for equity: Leveraging competency-based education
to ensure all students succeed (CompetencyWorks Final Paper). Vienna, VA: iNACOL.
Retrieved from https://www.inacol.org/wp-content/uploads/2018/03/CompetencyWorks-
DesigningForEquity.pdf
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

233
Tuckman, B. (2009). Operant conditioning. Retrieved from
http://www.education.com/reference/article/operant-conditioning/
Uitto, A. (2014). Interest, attitudes and self-efficacy beliefs explaining upper-secondary school
students’ orientation towards biology-related careers. International Journal of Science
and Mathematics Education, 12(6), 1425–1444. doi:10.1007/s10763-014-9516-2
Vincent-Ruz, P., & Schunn, C. (2017). The increasingly important role of science competency
beliefs for science learning in girls. Journal of Research in Science Teaching, 54(6), 790–
822. doi:10.1002/tea.21387
Wang, M.-T., & Degol, J. L. (2016). Gender gap in science, technology, engineering, and
mathematics (STEM): Current knowledge, implications for practice, policy, and future
directions. Educational Psychology Review, 29(1), 119–140. doi:10.1007/s10648-015-
9355-x
Wang, X. (2013). Why students choose STEM majors: Motivation, high school learning, and
postsecondary context of support. American Educational Research Journal, 50(5), 1081–
1121. doi:10.3102/0002831213488622
Weiss, R. S. (1994). Learning from strangers: The art and method of qualitative interview
studies. New York, NY: The Free Press.
Yough, M., & Anderman, E. (2006). Goal orientation theory. Retrieved from
http://www.education.com/reference/article/goal-orientation-theory/
Zeldin, A. L., Britner, S. L., & Pajares, F. (2008). A comparative study of the self-efficacy
beliefs of successful men and women in mathematics, science, and technology careers.
Journal of Research in Science Teaching, 45(9), 1036–1058. doi:10.1002/tea.20195
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

234
Zeldin, A. L., & 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. doi:10.3102/00028312037001215

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

235
APPENDIX A
EXISTING DATA
The Hawaii School’s Institutional Researcher (IR) distributed this questionnaire to all students in
regular chemistry in September 2018. The intent was to collect data on chemistry self-efficacy
early in the school year. As the Hawaii School works to convert its entire curriculum to a
competency-based structure and to utilize mastery assessment techniques throughout the high
school, the IR will continually collect data on pilot courses to enable data-driven decision
making as innovation progresses. This questionnaire was adapted from two sources: (1) the self-
efficacy questions from the 2015 Panorama Social-Emotional Learning Measure developed by
Dr. Hunter Gehlbach (2015) in his role as Director of Research at Panorama Education, and (2)
Schwarzer and Jerusalem’s (1995) Generalized Self-Efficacy Scale as found on pages 35–37 in
Weinman, Wright, and Johnston’s book of measures in health psychology.

The following questionnaire was distributed by the five chemistry teachers during class time
using Survey Monkey, with results that went directly back to the IR.

Chemistry Self-Efficacy Questionnaire
“Hello! We would like to know about your experiences in Chemistry classes. Please take a
couple of minutes to fill out this questionnaire. Your responses will help to make improvements
to the curriculum.”

1. Fill in “Student ID #”

2. When complicated ideas are presented in chemistry class, how confident are you that you can
understand them?
○ Not all confident
○ Barely confident
○ Somewhat confident
○ Very confident

3. How confident are you that you can master all the learning outcomes in chemistry?
○ Not all confident
○ Barely confident
○ Somewhat confident
○ Very confident

4. How confident are you that you can do the hardest work that will be assigned in this chemistry
class?
○ Not all confident
○ Barely confident
○ Somewhat confident
○ Very confident

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

236
5. How confident are you that you will remember what you learned in your chemistry class?
○ Not all confident
○ Barely confident
○ Somewhat confident
○ Very confident

6. If you invest the necessary effort, to what extent are you able to answer most chemistry
questions?
○ Not all able
○ Barely able
○ Somewhat able
○ Very able

7. When unforeseen situations in the lab occur, how resourceful are you in handling it?
○ Not all resourceful
○ Barely resourceful
○ Somewhat resourceful
○ Very resourceful

8. When confronted with a challenging problem in chemistry, to what extent are you able to
propose several possible solutions?
○ Not all able
○ Barely able
○ Somewhat able
○ Very able

9. To what extent are you able to handle whatever comes your way in a science class?
○ Not all able
○ Barely able
○ Somewhat able
○ Very able

10. What is your graduation year?
○ 2019
○ 2020
○ 2021
○ 2022

11. What is your gender?
○ Female
○ Male
○ Non-Binary
○ Prefer not to answer

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

237
12. At what grade did you enter the Hawaii School?
○ Kindergarten
○ Grades 1–3
○ Grades 4–5
○ Grades 6–8
○ Grade 9
○ Grades 10–12

13. Fill in “What science course did you take last year?”

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

238
APPENDIX B
SURVEY ITEMS
The survey was prepared by aligning the KMO influences with closed, forced choice, 4-point
Likert scale items (Fink, 2013). The potential influences are listed below, followed by a table
that outlines the survey items, lists the possible responses, and maps each item to a particular
KMO construct.

Knowledge Influences

Procedural: Students need to understand the steps of inquiry inherent in STEM disciplines.

Metacognitive: Students need to develop skills in laboratory self-regulation and conceptual self-
reflection.

Motivation Influences

Self-Efficacy: Students need to have confidence in their ability to perform the mathematical and
laboratory tasks required in math-intensive STEM courses.

Attributions: Students need to feel their performance depends upon their sustained effort rather
than believing their STEM aptitude is fixed.

Goal Orientation: Students need to engage in math-intensive STEM courses with the desire to
master new skills and enact their curiosity, resourcefulness, persistence, and resilience.

Organizational Influences

Cultural Model: The organization needs to place a high value on developing all students’
competency in STEM skills and habits of mind.

Cultural Model: The organization needs to sustain a culture of gender equity and inclusion for
students in all STEM classrooms.

Cultural Setting: The organization needs to develop and deliver STEM curriculum that is
focused on mastery of skills over content-based performance.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

239
KMO
Construct Survey Item (question and response)
K-P Collaborating with others is one important step in the process of scientific
inquiry.
(strongly disagree, disagree, agree, strongly agree)
K-P In scientific inquiry, asking questions is more important than answering
questions.
(strongly disagree, disagree, agree, strongly agree)
K-P Personal creativity is an important consideration in performing scientific inquiry.
(strongly disagree, disagree, agree, strongly agree)
K-P I know the steps to follow in order to interpret authentic, messy data sets
(strongly disagree, disagree, agree, strongly agree)
K-M Self-reflection is a crucial part of improving my science understanding.
(strongly disagree, disagree, agree, strongly agree)
K-M Compared to traditional grading, the 1–4 grading scale gives me a better sense of
what I do and don’t understand.
(strongly disagree, disagree, agree, strongly agree)
K-M I am able to create projects to focus on learning the outcomes on which I need to
improve.
(strongly disagree, disagree, agree, strongly agree)
K-M Once I can identify and explain what I do and do not understand in this course, I
am able to use that knowledge in new ways.
(strongly disagree, disagree, agree, strongly agree)
K-M I can adjust my process in the lab when my procedure is not producing useable
data.
(strongly disagree, disagree, agree, strongly agree)
M-SE I am confident that I am able to perform the laboratory tasks required in this
class.
(strongly disagree, disagree, agree, strongly agree)
M-SE I am confident I can perform the mathematical tasks required in this class.
(strongly disagree, disagree, agree, strongly agree)
M-SE I feel anxious in science class.
(strongly disagree, disagree, agree, strongly agree)
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

240
KMO
Construct Survey Item (question and response)
M-SE There is a relationship between gender and science ability.
(strongly disagree, disagree, agree, strongly agree)
M-SE People of my gender are better at science.
(strongly disagree, disagree, agree, strongly agree)
M-A If I perform poorly on an assessment, it is because I did not try hard enough to
learn the material.
(strongly disagree, disagree, agree, strongly agree)
M-A If I perform poorly on an assessment, it is because it was too hard for me.
(strongly disagree, disagree, agree, strongly agree)
M-A I am in charge of my success in this class.
(strongly disagree, disagree, agree, strongly agree)
M-A Confusion is temporary. I can find ways to get unstuck.
(strongly disagree, disagree, agree, strongly agree)
M-A Persistence is important for success in science.
(strongly disagree, disagree, agree, strongly agree)
M-GO As I approach class each day, my goal is to get a good grade in science.
(strongly disagree, disagree, agree, strongly agree)
M-GO As I approach class each day, my goal is to gain a deeper understanding of
science.
(strongly disagree, disagree, agree, strongly agree)
M-GO My goals in this class tie directly to the learning outcomes.
(strongly disagree, disagree, agree, strongly agree)
M-GO My goals are about self-improvement.
(strongly disagree, disagree, agree, strongly agree)
M-GO My goals are about outperforming others.
(strongly disagree, disagree, agree, strongly agree)
M-GO I view set-backs in the lab as additional data to interpret not as failures.
(strongly disagree, disagree, agree, strongly agree)
O-CM When I am in science class, I feel as if I belong.
(strongly disagree, disagree, agree, strongly agree)
ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

241
KMO
Construct Survey Item (question and response)
O-CS In this class I learned not just about science but how to do science.
(strongly disagree, disagree, agree, strongly agree)
O-CM The Hawaii School culture promotes equal participation in STEM by male and
female students.
(strongly disagree, disagree, agree, strongly agree)
O-CM It is socially acceptable to be a boy who loves science at the Hawaii School.
(strongly disagree, disagree, agree, strongly agree)
O-CM It is socially acceptable to be a girl who loves science at the Hawaii School.
(strongly disagree, disagree, agree, strongly agree)
O-CS My individual growth was rewarded in this class.
(strongly disagree, disagree, agree, strongly agree)
O-CM Good grades are very important to future success.
(strongly disagree, disagree, agree, strongly agree)
O-CM Leaving the Hawaii School with excellent skills (like inquiry, persistence,
communication, and personal responsibility) is very important to future success.
(strongly disagree, disagree, agree, strongly agree)
K-P = Knowledge-Procedural, K-M = Knowledge-Metacognitive, M-SE = Motivation-Self-
Efficacy, M-A = Motivation-Attributions, M-GO = Motivation-Goal Orientation,
O-CM = Cultural Models, O-CS = Cultural Settings

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

242
APPENDIX C
INTERVIEW PROTOCOL
The interview began with a scripted introduction, in which the researcher welcomed the
participant and reminded her of the structures put in place to ensure confidentiality, protection of
data, and voluntary participation (Merriam & Tisdell, 2016). She then thanked the subject for
her participation and revealed a little bit about why this project mattered to the researcher,
establishing a sense of rapport (Bogdan & Biklen, 2007; Merriam & Tisdell, 2016). The
interview began with the following opening:

“Thank you for agreeing to participate in this interview today. I wanted to begin by reminding
you of the research protocols that are being used to ensure your confidentiality and comfort.
Throughout the analysis of the interview data I will use a pseudonym in place of your real name.
I will be recording today’s interview, and as we chat today I will also avoid using your real
name. All recorded data and the transcriptions of these interviews will be kept in password
protected files on my computer. Your participation in this study is completely voluntary, and as
such you may choose to skip any questions today or to withdraw from the interview at any time.
Do you have any questions about these procedures?

I am so grateful that you were willing to spend part of your very busy day to sit and talk with me
about your experiences as a young woman in science. As a woman in science, myself, this topic
is very near and dear to my heart. I am hoping that through my dissertation research I can
develop a deeper understanding of how this school’s culture and classroom practices shape a
girl’s sense of herself as a scientist. I ask that you be very honest with me today. Please know
that there are no right answers to my questions, I am simply looking for an authentic glimpse into
your experiences. We will go ahead and begin with a few questions about the structure of your
chemistry course this year . . .”

Following this opening the research began to ask the scripted interview questions. Twelve
questions are listed below. Written in red font immediately following each question is the
identification of question type using Patton’s (2015) six types and/or Strauss et al.’s (1981) four
types (as cited in Merriam & Tisdell, 2016). In italicized red font following the question type is
a brief goal for each question, tying the interview question directly to the study’s presumed
KMO influences. In bulleted lists after each of the 12 questions are possible probes, written by
imagining that the student interview subjects might need further prodding to reveal rich
responses. Between each of the twelve questions is proposed transitional language that was used
to help the flow of the interview, creating a conversational feel to the interaction despite the
interview being highly structured (Merriam & Tisdell, 2016).

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

243
1. Suppose I was a new student in this class, how would you describe to me the way that
chemistry is taught in this course?
Knowledge (Patton, 2015) & Hypothetical (Strauss et al., 1981) — goal is to reveal student
understanding of the cultural setting (O influencer) of inquiry pedagogy and competency-based
course structure.
Possible Probes
• What could I expect to do each day in class?
• What could I expect each “cycle” to look like?
• Can you describe a specific lab experience to me?

Transition: “Great! Thank you for sharing some details about the day to day class structure.
Next we will talk a little bit about the grading structure in chemistry this year . . .”

2. How would you describe the way that your work is graded?
Knowledge (Patton, 2015) — goal is to reveal student understanding of cultural setting (O
influencer) of mastery assessment practices.
Possible Probes
• Can you share a specific example of something on which you were graded and how that
grade was assigned?
• Can you explain what “learning outcomes” are?

Transition: “As is true with every new year in school, I imagine that some of these structures and
practices were new and different for you . . .”

3. Can you describe any ways in which your approach to science learning has been different in
this class than in your previous science courses in middle school and high school?
Experience & Behavior (Patton, 2015) — goal is to reveal student’s metacognitive sense of
improved laboratory self-regulation or conceptual self-reflection (K influencer).
Possible Probes
• Describe your experience with a particular lab. How did you decide what to do in the
lab?
• How did you know when you had mastered a particular learning outcome?

Transition: “I would like to talk now specifically about goal setting . . .”

4. How did you set goals for yourself in this class?
Experience & Behavior (Patton, 2015) — goal is to understand if student was setting
performance or mastery goals (M influencer).
Possible Probes
• Describe ways in which you were able to align your goals with learning outcomes.
• Were you driven by the number (1–4) or were you driven by the description of the
learning outcome? Explain.

Transition: “As you know, one of the learning outcomes in this course was persistence . . .”

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

244
5. How was your persistence related to your learning in this course?
Opinion & Value (Patton, 2015) — goal is to understand to what students attribute their success,
do they have a growth or fixed mindset (M influencer).
Possible Probes
• Did you feel as if you were in control of your ability to master the material?
• What strategies did you use when you were stuck or frustrated?

Transition: “We have been talking very specifically about this year’s chemistry course for the
first five questions. Now I want you to think more broadly about yourself as a learner . . .”

6. A science identity describes the degree to which a person thinks of herself as a person capable
of doing science. Describe your sense of your science identity.
Feeling (Patton, 2015) — goal is to establish a sense of the girl’s science self-efficacy (M
influencer).
Possible Probes
• Tell me more about how you developed this sense.
• Could you paint a more detailed picture for me?

Transition: “I am curious if this sense changed at all over the course of this year, so I’d like to
ask . . .”

7. What impact has taking this chemistry course had on your science identity?
Opinion & Value (Patton, 2015) — goal is to create a moment of self-reflection that is related to
self-efficacy changes over time (M influencer).
Possible Probes
• Tell me a story of a moment in this class when you felt like you could really do science
(perhaps in a way you never felt before).
• Tell me a story of a moment in this class when you wondered if science was too hard
(perhaps in a way you never felt before).

Transition: “Of course, your science identity is not just developed at school . . .”

8. What messages do you hear at home that influence the way you feel about yourself as a girl in
science?
Feeling (Patton, 2015) — goal is to explore one of Bandura’s (1986) factors that influence self-
efficacy development outside of school (M influencer).
Possible Probes
• Could you paint a more detailed picture for me?
• What words of encouragement or discouragement do your family members use when
talking to you about taking science courses (or your future in science)?
• How does that make you feel about your science ability?

Transition: “Context seems important in this conversation, too. Can you tell me . . .”

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

245
9. What is the “Hawaii School” culture around girls in science?
Opinion & Value (Patton, 2015) — goal is to get a sense of how the girl would describe the
current cultural model (O influencer) of gender equity and inclusion in STEM.
Possible Probes
• Could you paint a more detailed picture for me?
• Are girls encouraged and included in science the same as boys? Explain.
• In your opinion, is this culture dictated by faculty or by students? Explain.
• How does that make you feel about your science ability?

Transition: “Getting back to thinking about your work in chemistry this year . . .”

10. Describe what you believe is the correlation between the “AIMS” (persistence,
communication, personal & social responsibility, empathy & compassion, scientific inquiry, and
adapting & applying learning to relevant topics) and doing science?
Opinion & Value (Patton, 2015) — goal is to reveal if students can describe the procedural
knowledge (K influencer) of how to succeed in doing actual science by focusing on habits of
mind.
Possible Probes
• What is the difference between knowing science and being able to do science?
• When did your abilities in the AIMS matter in the lab? (especially AIMS besides
scientific inquiry)

Transition: “You are only half done with high school, and you have decisions to make about how
you spend the rest of your time here . . .”

11. Describe your ideal experience for your next two years of science coursework at the “Hawaii
School”?
Ideal position (Strauss et al., 1981) — the goal is to reveal the cultural settings and/or models (O
influencer) that a girl would describe as ideal.
Possible Probes
• What courses do you plan to take?
• How about in college?
• In what ways do you believe this course shaped these hopes?

Transition: “That wraps up my formal questions about your experiences in chemistry this year
and your reflections about how you view your place as a girl in science. Before we go, I would
like to ask one last thing . . .”

12. What should I have asked you that I didn’t think to ask about your experiences as a girl in
STEM?
Open-ended final question — the goal is to capture thoughts and experiences of the participant
that I did not think to ask about.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

246
APPENDIX D
SAMPLE SURVEY ITEMS MEASURING KIRKPATRICK LEVELS 1 AND 2
1. This course consistently held my interest. (Level 1 Engagement)
Strongly Disagree, Disagree, Agree, Strongly Agree
2. I was constantly learning and growing in this course. (Level 1 Engagement)
Strongly Disagree, Disagree, Agree, Strongly Agree
3. The competencies that were the focus of this course will have relevance in my life beyond
this course. (Level 1 Relevance)
Strongly Disagree, Disagree, Agree, Strongly Agree
4. I enjoyed the inquiry-based approach to learning science. (Level 1 Customer Satisfaction)
Strongly Disagree, Disagree, Agree, Strongly Agree
Questions 5–10. Use the five-point scale articulated below to respond to the prompts. Each
question asks you to consider the way you would have responded before participating in this
course compared to how you respond now, at the conclusion of the course.

1 2 3 4 5
not at all barely somewhat quite a bit or
quite well
certainly a lot or
extremely well

Before this course:
1 2 3 4 5
5. I am committed to applying the
skills and habits of mind I
developed in this course to future
science classes. (Level 2
Commitment)
After this course:
1 2 3 4 5

Before this course:
1 2 3 4 5
6. I can summarize the necessary
skills and habits of mind to
engage in the steps of authentic
scientific inquiry. (Level 2
Declarative Knowledge)
After this course:
1 2 3 4 5

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

247
Before this course:
1 2 3 4 5
7. I can execute the steps of
scientific inquiry, including
question generation, experimental
design, and data analysis. (Level
2 Procedural Knowledge)
After this course:
1 2 3 4 5

Before this course:
1 2 3 4 5
8. I feel confident that I can set
specific, targeted goals for my
growth aligned to course
competencies. (Level 2
Confidence)
After this course:
1 2 3 4 5

Before this course:
1 2 3 4 5
9. I feel confident that I can
master the competencies in a
science class. (Level 2
Confidence)
After this course:
1 2 3 4 5

Before this course:
1 2 3 4 5
10. I see value in the competency-
based model of education. (Level
2 Attitude)
After this course:
1 2 3 4 5

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

248
APPENDIX E
SAMPLE SURVEY ITEMS MEASURING KIRKPATRICK LEVEL 3 DRIVERS
It is recommended that Level 3 drivers be assessed at the midpoint of the course, as monitoring
the drivers at the midpoint (approximately 4 months after the beginning of the course) could help
faculty members to adjust their practice. Sample Level 3 rating items for a student survey are
shown below.

1. I have received adequate training in the classroom about the required skills and habits of
mind for scientific inquiry. (Level 3 Required Driver — Reinforcing)
Strongly Disagree, Disagree, Agree, Strongly Agree
2. I have opportunities for repeated practice with authentic inquiry. (Level 3 Required Driver
— Reinforcing)
Strongly Disagree, Disagree, Agree, Strongly Agree
3. I receive frequent, specific feedback from my teacher which is clearly aligned to
competencies and includes procedural advice for how to improve. (Level 3 Required Driver
— Encouraging)
Strongly Disagree, Disagree, Agree, Strongly Agree
4. The pilot course classroom fosters a positive emotional environment where it is made clear
that all students are capable of success. (Level 3 Required Driver — Encouraging)
Strongly Disagree, Disagree, Agree, Strongly Agree
5. My teacher helps me to acknowledge and celebrate my successes on small, concrete goals.
(Level 3 Required Driver — Rewarding)
Strongly Disagree, Disagree, Agree, Strongly Agree
6. The assessments in this pilot course measure and reward my individual progress and growth.
(Level 3 Required Driver — Rewarding)
Strongly Disagree, Disagree, Agree, Strongly Agree
7. What could your teacher do to better support your learning in this competency-based pilot
course?

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

249
APPENDIX F
SAMPLE BLENDED EVALUATION ITEMS MEASURING KIRKPATRICK
LEVELS 1, 2, 3, AND 4
It is recommended to revisit Level 1 relevance and satisfaction and Level 2 knowledge and skills
in a delayed survey to pilot course participants. In addition, Level 3 critical behaviors and Level
4 indicators and results should also be assessed in this measure administered at least 3 months
after completion of the pilot course. Sample items are shown below.

Open-Ended Questions for Revisiting Level 1 and Level 2
1. What projects and/or competencies from the pilot STEM course continue to feel relevant to
you now? (Level 1 Relevance)
2. Knowing what you know now, what would you change about the competency-based pilot
course? (Level 1 Customer Satisfaction)
3. Scenario Question: You are asked to create a series of experiments to investigate the
chemical composition of Bufferin (an over the counter drug that is a combination of aspirin
and antacid). Explain the steps you would take to design the experiments. Discuss what
hurdles you could anticipate in the lab during data collection and the strategies you would
use to overcome them. (Level 2 Procedural Knowledge)

Five-Point Scale Questions for Evaluating Level 3 Critical Behaviors
For questions 4–6 below, identify the degree to which you have continued to practice the
behaviors that were cultivated in your competency-based STEM pilot course. (Level 3 Critical
Behaviors)

1 — Little or no application
2 — Mild degree of application
3 — Moderate degree of application
4 — Strong degree of application
5 — Very strong degree of application and desire to help others do the same

4. I cultivate the skill sets and mindsets of a scientist and engage in
authentic STEM inquiry.
1 2 3 4 5
5. I set goals in STEM courses aligned to my personal growth on skills and
habits of mind, rather than on performance outcomes.
1 2 3 4 5
6. I reflect upon my growth in skills and competencies in order to build my
confidence in STEM.
1 2 3 4 5

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

250
8. If you circled 3 or below on any of the questions above, please indicate the reason(s) why
you are not continuing to practice and apply these behaviors in your next STEM courses.

£ my next class is not competency-based and therefore these behaviors don’t apply.
£ my next class in not competency-based, however these do still seem relevant but I
don’t have support to incorporate them.
£ my next class is also competency-based, but I don’t understand exactly how to apply
these behaviors in this new context.
£ my next class is also competency-based, but I don’t have the confidence to apply what
I learned.
£ Other, please specify _______________________________

Level 4 Indicators and Results Sample Metrics

9. I have noticed the following continued positive outcomes from my participation in the pilot
course. Check all that apply.

£ I have more self-confidence in science class
£ I contribute more to group processes in the lab
£ I am comfortable with setbacks in the lab
£ I don’t expect everything to work the first time I try
£ I have strategies for improving my understanding when I am stuck
£ I can identify the relevance of learning science and conducting inquiry
£ I produce higher quality work with a better depth of science understanding
£ I value feedback from my teachers use it to set goals
£ I am better at setting specific, targeted goals for my improvement
£ I am less anxious about grades
£ Other positive outcomes, please specify __________________________
£ None of the above — I don’t feel any continued positive outcomes.

10. To what degree do you feel that the competency-based framework might create more gender
equality in math-intensive STEM course enrollments? Have you changed your mind about
pursuing upper level courses in physics, math, computer science, or engineering at the
Hawaii School? Explain your thinking.

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

251
APPENDIX G
SAMPLE TEACHER DASHBOARD USING LEVELS 1, 2, AND 3 DRIVER
EVALUATION DATA
As described in Table 27, this dashboard displays Level 1 results about students’ perceptions of
engagement, satisfaction, and relevance as pie charts. Feedback on Level 3 drivers is also
displayed as pie charts and labeled as “Feedback on Teaching Practices” on the dashboard. In
the upper right corner is a comparison of means for reinforcing, encouraging, and rewarding
drivers as measured at different points in a course. Level 2 learning pre- and post-data is in the
bottom left corner and labeled as “Knowledge & Motivation Changes.”

ADDRESSING GENDER INEQUITIES IN SELF-EFFICACY

252
APPENDIX H
SAMPLE ADMINISTRATOR DASHBOARD USING LEVEL 3 BEHAVIORS,
LEVEL 4 RESULTS AND INSTITUTIONAL DATA
As described in Table 28, the dashboard on the following page displays the Level 3 behaviors
and Level 4 results data from the delayed survey. Students’ perceptions of their degree of
application of critical behaviors and an analysis of identified preventative factors are on the top
left of the dashboard. On the top right is frequency data on the positive outcomes from the
program reported by students. The bottom right show annual STEM enrollment trends for
students who took competency-based pilot courses. It should be noted that all student graphics
on this dashboard are disaggregated by gender to align to the organizational goals relating to
gender equity and inclusion. The bottom left of the dashboard allows administrators to track
trends in faculty involvement in STEM competency-based pilot courses by subdepartment.

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Improving pilot training by learning about learning: an innovation study

PDF

Gender inequity and leadership in the large state militia: an innovation study

PDF

Quality literacy instruction in juvenile court schools: an evaluation study

PDF

Application of professional learning outcomes into the classroom: an evaluation study

PDF

Impact of mindfulness on early education teacher well-being: an evaluation study

PDF

The racially responsive facilitator: an evaluation study

PDF

Raising women leaders of Christian higher education: an innovation study

PDF

Spiritual persistence of high school alumni: an evaluation model

PDF

Civic learning program policy compliance by a state department of higher education: an evaluation study

PDF

The Bridge Program and underrepresented Latino students: an evaluation study

PDF

An evaluation of project based learning implementation in STEM

PDF

Online graduate-level student learning and engagement: developing critical competencies for future leadership roles: an evaluation study

PDF

Teacher role in reducing the achievement gap: an evaluation study

PDF

The knowledge, motivation, and organization influences affecting the frequency of empathetic teaching practice used in the classroom: an evaluation study

PDF

Evaluation of thesis completion in a graduate blended learning program

PDF

Implementing problem-based learning to develop student supply chain skills

PDF

STEM industries apprenticeships: organization influences, skill gaps, and challenges facing the 21st-century workforce: an evaluation study

PDF

Technology integration and self-efficacy of in-service secondary teachers in an international school

PDF

Collaborative instructional practice for student achievement: an evaluation study

PDF

One to one tablet integration in the mathematics classroom: an evaluation study of an international school in China

Asset Metadata

Creator Mingarelli, Sally Diane (author)

Core Title Using mastery learning to address gender inequities in the self-efficacy of high school students in math-intensive STEM subjects: an evaluation study

School Rossier School of Education

Degree Doctor of Education

Degree Program Organizational Change and Leadership (On Line)

Publication Date 09/23/2019

Defense Date 08/16/2019

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

Tag competency-based learning,gender inequity,mastery learning,math-intensive STEM,OAI-PMH Harvest,participation gaps,self-efficacy

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Robles, Darline (committee chair), Datta, Monique (committee member), Tiwana, Ravneet (committee member)

Creator Email mingarel@usc.edu,smingarelli@punahou.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-220181

Unique identifier UC11674768

Identifier etd-Mingarelli-7833.pdf (filename),usctheses-c89-220181 (legacy record id)

Legacy Identifier etd-Mingarelli-7833.pdf

Dmrecord 220181

Document Type Dissertation

Rights Mingarelli, Sally Diane

Type texts

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

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA