University of Southern California Dissertations and Theses

Self-efficacy beliefs and intentions to persist of Native Hawaiian and non-Hawaiian science, technology, engineering, and mathematics majors

(USC Thesis Other)

Self-efficacy beliefs and intentions to persist of Native Hawaiian and non-Hawaiian science, technology, engineering, and mathematics majors

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Running head: SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

1

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST OF NATIVE HAWAIIAN
AND NON-HAWAIIAN SCIENCE, TECHNOLOGY, ENGINEERING, AND
MATHEMATICS MAJORS

by

Joshua Kananimauloa Kaakua

_____________________________________________________________________________

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 2014

Copyright 2014 Joshua Kananimauloa Kaakua

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

2
DEDICATION
This dissertation is dedicated to my wife Laura H.E. Kaakua and father Joseph W.L. Kaakua.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

3
ACKNOWLEDGEMENTS
I wish to acknowledge and thank my mentor and advisor, Dr. Darnell Cole, whose
guidance and example reminded me that I could do it when I thought otherwise. Thank you to
my committee member Dr. Melora Sundt for teaching me to attend to my writing and to the one
percent. Mahalo nui e Dr. Maenette Benham no kou ‘ike hohonu a me kou kākoʻo. To my
colleagues, mentors, and friends at the University of Southern California, University of Hawaii,
and NHSEMP, thank you. I have been blessed.
I wish to acknowledge my first teachers and heroes, Mom and Dad, who prepared me to
seek happiness and meaningful work. I wish to acknowledge my wife and best friend Laura for
being my biggest supporter and source of encouragement along the long road through my
doctoral program. Thank you for the chance to pursue my dreams. Lastly, to my favorite
dissertation distractions Pono and Melia, Dad’s pau and loves you.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

4
TABLE OF CONTENTS
Dedication 2
Acknowledgements 3
List of Tables 5
List of Figures 7
Abstract 8
Chapter 1: Introduction 9
Chapter 2: Literature Review 22
Chapter 3: Methodology 49
Chapter 4: Results and Analysis 68
Chapter 5: Discussion 124
References 149
Appendices 164
Appendix A: UH Human Studies Program Approval 164
Appendix B: Informed Consent and Survey Instrument 165
Appendix C: University of Hawaii at Manoa STEM Majors 177
Appendix D: Descriptive Statistics by Major, Level, and Pre-Institution Status 178
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

5
LIST OF TABLES
Table 1. University of Hawaiʻi at Manoa Peer Group 53
Table 2. Ethnic Background of UHM Undergraduates, 2012 54
Table 3. Enrollment of UHM Undergraduates by STEM College, Spring 2014 55
Table 4. Input and Background Variables Collected 57
Table 5. Environment and Outcome Variables Collected 62
Table 6. Participant Gender (N = 638) 70
Table 7. Sample Gender (N = 638) by College 71
Table 8. Participant Ethnicity (N = 638) 72
Table 9. Sample and Population by STEM College 73
Table 10. Sample and Population by Highest and Lowest Enrolled STEM Academic 75
Majors
Table 11. Native Hawaiian Status Count and Frequency by Input Variable 77
Table 12. Descriptive Statistics for GPA Items 79
Table 13. Frequency of Program Participation (N = 638) 80
Table 14. Descriptive Statistics for SES Items 81
Table 15. SES Characteristics by Native Hawaiian Status 82
Table 16. Factor Analysis Component Matrix for SES 83
Table 17. Descriptive Statistics for Family Support Items 85
Table 18. Descriptive Statistics for Peer Interaction Items 86
Table 19. Factor Analysis Component Matrix for Peer Interaction 87
Table 20. Descriptive Statistics for Faculty Interaction Items 88
Table 21. Factor Analysis Component Matrix for Faculty Interaction 89
Table 22. Descriptive Statistics for Faculty Support Items (N = 566) 90
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

6
Table 23. Factor Analysis Pattern Matrix for Faculty Support 92
Table 24. Descriptive Statistics for Sense of Belonging Items 93
Table 25. Factor Analysis Pattern Matrix for Sense of Belonging 95
Table 26. Descriptive Statistics for Satisfaction Items 96
Table 27. Factor Analysis Component Matrix for Satisfaction 98
Table 28. Descriptive Statistics for STEM Self-Efficacy Items and Intent to Persist 99
Table 29. Factor Analysis Component Matrix for STEM Self-Efficacy 100
Table 30. Summary of Composite Variables and Reliability Analysis 102
Table 31. Significant t-test Results by Native Hawaiian Status 104
Table 32. Independent Variables Used for Self-Efficacy Sequential Multiple Correlation 107
Table 33. Sequential Multiple Regression Predicting Self-Efficacy 109
Table 34. Regression Coefficients Predicting STEM Self-Efficacy 111
Table 35. Sequential Logistic Regression Predicting Intent to Persist 116
Table 36. ANOVA of Select Characteristics by Native Hawaiian Status 118
Table 37. Robust Tests of Equality of Means of Select Characteristics by Native 118
Hawaiian Status
Table 38. MANOVA of Selected Characteristics by Hawaiian and Non-Hawaiian 120
Groups (N = 345)
Table 39. Significant Predictors and Native Hawaiian Differences 122
Table 40. Sample and Population by STEM Academic Major 178
Table 41. Frequency Counts by Educational Level and Pre-Institution Status (N = 638) 180

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

7
LIST OF FIGURES
Figure 1. Conceptual Input – Environment – Outcomes (I-E-O) model for STEM 34
self-efficacy and intent to persist
Figure 2. Ho’okahua conceptual framework 137
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

8
ABSTRACT
This study applies the framework of Social Cognitive Career Theory and Astin’s (1999) Inputs –
Environment – Outcomes model to investigate the personal input and environmental factors
associated with self-efficacy beliefs and intentions to persist in Science, Technology,
Engineering, and Mathematics (STEM) and to examine the differences of these factors and
outcomes between Native Hawaiian and non-Hawaiian students. Conducted at a large, public
tier-one research institution in Hawaii, this cross-sectional study gathered survey data from 638
undergraduate STEM majors and analyzed data through factor analysis, regression, and
MANOVA techniques. The findings indicate that sense of belonging to major, past performance,
and family support explained STEM self-efficacy. Self-efficacy, in turn, predicted intent to
complete a STEM degree at the institution. This study also found higher levels of peer
interaction, program involvement, family support, and intentions to persist for Native Hawaiians
relative to non-Hawaiians. A Ho’okahua or foundation building framework is presented based
on self-efficacy, sense of belonging, and involvement to guide educational practice and theory.
The implication for practice is that academic communities at the department or discipline level,
especially for underclassmen and Native Hawaiians, are important to improve degree completion
in STEM. The findings provide direction for Native Hawaiian education research to further
investigate socio-cultural aspects of learning and Native Hawaiian congruence in STEM.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

9
CHAPTER 1
INTRODUCTION
Student persistence is one of the most studied topics in the higher education literature.
Institutions, educators, and researchers want to better understand the nature of student
persistence (or departure), student self-beliefs about their capabilities and motivations to
complete a degree, and the influences and capabilities of the institution to support students to
graduation. Leading theorists such as Astin (1977, 1993, 1999), Tinto (1975), Kuh (1993), Bean
(1980) and Pascarella and Terenzini (1991) have postulated different models and frameworks to
identify, study, and understand the complex mix of factors associated with college success.
U.S. educational leaders, researchers, and policy makers have become increasingly
interested in understanding persistence issues to improve success in Science, Technology,
Engineering, and Mathematics (STEM) as well as retention issues pertaining to underrepresented
ethnic minorities. The focus on the STEM educational pipeline is due to the importance of
STEM as a national driver of technological advancement, economic prosperity, and global
competitiveness (National Research Council [NRC], 2010). Initiatives to develop models of
effective programs aimed to increase participation of minority STEM students have gained
traction since the 1970s (Landis, 1985). Educators at the University of Hawaii are interested in
determining how the research informs smart retention strategies for Native Hawaiians, the
indigenous peoples of Hawaii. While the broader context for the motivation of this study is to
address the underperforming U.S. STEM educational pipeline, the specific context is the goal of
improving college outcomes for Native Hawaiian STEM majors at the University of Hawaii at
Manoa. The aim of this study was to investigate two outcomes of interest — self-efficacy beliefs
and intent to persist — for Native Hawaiian and non-Hawaiian STEM majors.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

10
This chapter first presents the need for an improved science and engineering workforce
and educational pipeline to address local, national, and global demands. The case is made that
improving the ways in which higher education involves, prepares, and successfully graduates
underrepresented students is a critical piece to meeting the nation’s science and engineering
demands (NRC, 2010; Museus & Liverman, 2010). Using Astin’s theory of student involvement
(1977, 1993, 1999), social cognitive theory (Bandura, 1977, 1997), and social cognitive career
theory (Lent, Brown, & Hackett, 1994; Lent, 2013) as a framework, this study highlights what is
known and what is not known related to improving STEM persistence for underrepresented
minorities in general, and for Native Hawaiian STEM students in particular. Finally, this chapter
introduces the purpose, parameters, and research questions driving this quantitative study.
Background of the Problem
The United States has, in many ways, led the globe in economic prosperity rooted in its
science and engineering enterprises. Science, Technology, Engineering, and Mathematics have
been well documented as providing a foundation for the nation’s competitiveness (NRC, 2007,
2010). Economic studies have shown over half of the growth in United States’ gross national
product (GDP) in recent decades has been attributable to direct or indirect results of
advancements in science and technology (Boskin & Lau, 1992). National Science Board (NSB)
(2010) indicators reveal that while only 4% of the nation’s workforce are scientists or engineers,
this group disproportionately creates jobs for the other 96%. Advancements in knowledge,
leading to technology and innovation, have been a primary driver for the creation of jobs in the
twentieth century and are expected to be the source of competitive edge in the future economy.
The nation’s strategic leadership in science and technology, however, is diminishing
relative to global competition. Consider that the World Economic Forum (2010) ranked the
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

11
United States 48
th
in quality of mathematics and science education, the United States ranks 27
th

among developed nations in the proportion of college students receiving undergraduate degrees
in science or engineering (Organization for Economic Cooperation and Development, 2009), and
that 49% of United States adults do not know how long it takes for the Earth to revolve around
the Sun (NSB, 2010). The inability of the U.S. educational system to keep pace with global
competitors to produce a citizenship literate in STEM and a STEM workforce that is well-trained
and well-educated has been called “the quiet crisis” (Friedman, 2006; Jackson, 2004). The
United States faces an increasing threat to economic prosperity, security, and strategic leadership
in science and technology in the wake of a flattening world.
Recent national reports echo the daunting outlook of U.S. prosperity rooted in America’s
comparative edge in innovation. The National Academies’ Rising Above the Gathering Storm
(NRC, 2007) detailed a national call to action to address America’s eroding leadership in science
and technology. The follow-up report Rising Above the Gathering Storm, Revisited: Rapidly
Approaching Category 5 (NRC, 2010) updated that in the five years since the original report, the
nation’s outlook had worsened. President Obama (2011) characterized the problem of global
competition as an opportunity in science and engineering as “our generation’s Sputnik moment”.
Education, government, and industry leaders are challenged to identify the opportunities to
improve STEM education and capacity.
Improving the nation’s outlook is associated with improving the U.S. human capital in
science and engineering. The percentage of U.S. students pursuing first (undergraduate) degrees
in engineering (6%) is the second lowest among developed countries (NSB, 2010). In
comparison, over one-third of undergraduate students in China are enrolled in engineering study
(NSB, 2010). For students in the United States that do enter in science and engineering, less than
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

12
half complete their science or engineering degree within five years. Underrepresented minorities
drop out of science and engineering programs at a higher rate than other groups (NSB, 2010).
This translates to a smaller domestic talent pool to enter advanced degrees and the science and
engineering workforce.
Major segments of the domestic U.S. population, particularly ethnic minority groups and
women, are significantly underrepresented in STEM. African Americans, Hispanics, Native
Americans, Native Hawaiians, females, and persons with disabilities account for a significant
portion of the population and workforce, but are disproportionately found in science and
engineering classrooms, research laboratories, and the corporate environment (Jackson, 2004).
Achievement gaps exist between minority and non-minority students with regard to pre-college
preparation, college access, and STEM degree completion (Huang, Taddese, Walter, & Peng,
(2000). Although only half of all U.S. engineering majors graduate in engineering (National
Science Foundation, 2011), the completion rate for minority students in engineering is even
lower (Hurtado, Newman, Tran, & Chang, 2010; NRC, 2007). This implies that the higher
education pipeline accelerates minority underrepresentation in the STEM workforce, rather than
reverses it.
Underrepresented ethnic minorities are those whose group composition (college
enrollment) in education is below that of their composition in the general population (NRC,
2010). African Americans, Hispanics, Native Americans, Alaskan Natives, Native Hawaiians
and Pacific Islanders are largely underrepresented in STEM higher education and in the STEM
workforce (Duderstadt, 2008; Leggon & Pearson, 2009; NSB, 2010). Asians, while a minority
group in the U.S. population, are typically overrepresented in science and engineering fields.
Pacific Islanders, including indigenous peoples to Samoa, Guam, Micronesia, and Polynesia, are
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

13
considered an underrepresented group although they are commonly aggregated with Asians in
national data sets on STEM fields.
While ethnic minorities are largely underrepresented in science and engineering they are
also the most rapidly growing segment of the U.S. population (Committee on Science,
Engineering, and Public Policy [CSEPP], 2011). Demographic trends project a steep increase in
minority populations, especially amongst the college-going 18–24 year old age group, such that
by 2050 almost half of the U.S. population will be non-White. Diversifying the domestic STEM
workforce by increasing participation from all populations is a key element to address the quiet
crisis.
STEM researchers, industry professionals, and policy makers believe if U.S. educational
institutions improved the recruitment, retention and success rates of minority students in STEM,
then the country would be better equipped to innovate, compete, and problem solve (NRC, 2007,
2010; Duderstadt, 2008). Slaughter (2008) writes of the New American Dilemma, which is
marked by the post-Sputnik generation of white male engineers and technology workers retiring
in record numbers. Slaughter (2008) states, “Given the demographic changes in the U.S.
population, we cannot–and should not–expect white males to replace them. The solution to
America’s competitiveness problem lies in bringing young underrepresented minorities into
STEM careers in dramatically increased numbers” (p. 4). While there are many different
pathways and solutions to “fixing” the STEM education problem, researchers recognize that
“there are issues that are specific to underrepresented minorities, in general and in STEM,
focused on preparation, access, and motivation, financial aid, academic support, and social
integration” (CSEPP, 2011, p. 5).
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

14
In light of these concerns, several questions loom: How will the science and engineering
professions draw more students from an increasingly diverse population? What factors most
influence student success of underrepresented minorities in STEM and how can higher education
utilize this information? What role will Hawaii educators and Native Hawaiian students play in
addressing the increasing the quality, quantity, and diversity of tomorrow’s STEM workforce?
Statement of the Problem
The National Academies highlight three challenges in their strategic plans to increase
participation of minorities in science and engineering for America’s competitiveness (NRC,
2007, 2010). First, the sources for the future science and engineering workforce are uncertain
and changing increasingly relying on non-US citizens and fluctuating from a predominantly
white male workforce. Second, demographics show that the groups most underrepresented in
STEM are also the fastest growing in the domestic population, including the school-age
population from which the workforce may draw future talent. Finally, the strategy recognizes
that diversity is an asset in to enhance innovation, sustainability, and health of the nation. The
challenges associated with improving a national STEM educational pipeline can be investigated
on a local level. This section discusses Native Hawaiians in STEM, Native Hawaiians at the
University of Hawaii, and the effects of Native Hawaiian underrepresentation.
Native Hawaiians in STEM
Native Hawaiians, the indigenous peoples of Hawaii, have traditionally held a familial
relationship with their environment. Because of this inseparable connection, Native Hawaiians
were keenly aware and therefore experts in the STEM fields of Hawaii (University of Hawai’i
Hawaiian Studies Task Force [UoHHSTF], 1986; Kame’eleihiwa, 1992). For example, the NH
people established a well-functioning resource management system, the ahupuaʻa, which was
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

15
able to sustain a population of over one million NHs prior to western contact (Stannard, 1988).
Hawaii’s natural laboratory and geographic advantages have contributed to it being one of the
leading locations in the world for the study of ocean sciences, astronomy, and geology.
Therefore, in today’s higher education system in Hawaii, it would be expected that these STEM
areas would be highly populated with Native Hawaiian students. However, in the University of
Hawaii at Manoa (UHM), the flagship campus of the state’s higher education system, this is not
the case. Of the 2,520 Native Hawaiian students enrolled at UHM, there are only 30 students in
Marine Biology, 9 students in Geology, and 1 student in Astronomy (University of Hawaii
Institutional Research Office [UoHIRO], 2012). NHs are not found pursuing culturally-important
STEM disciplines at the University in high numbers.
Like other indigenous U.S. citizens, many Native Hawaiians have not experienced
success in the STEM educational pipeline. High school, pre-college preparation, and college
success outcomes are lower for NHs than their non-Hawaiian counterparts. NH children lag
behind statewide averages by approximately 10 percentile points in reading and math and the
achievement gap widens as students progress to higher grades (Kamehameha Schools Press
[KSP], 2005). NHs have lower transition rates between middle and high school, are retained in
high school more often, and are less likely than non-Hawaiians to graduate, to enroll in college,
or to complete a bachelor’s degree in the expected timeframe (Benham, 2006; Hokoana, 2010).
At UH Manoa, NH students are the least likely of Hawaii’s major ethnic groups to graduate
within six years and are most likely to be working full-time while attending school (22.3%
versus 17.8% statewide) (Hokoana, 2010; KSP, 2005). These findings support Benham’s (2006)
assessment that Native Hawaiians are not fairing well in their own homeland.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

16
In postsecondary education, Native Hawaiians continue to be underrepresented,
particularly in the STEM fields. Native Hawaiians comprise 23.1% of the state of Hawaii
population but comprise only 12.8% of the student body and 3.8% faculty at the University of
Hawaii at Manoa (UoHIRO, 2010). In the STEM disciplines, Native Hawaiian undergraduate
and graduate student enrollments in each of the STEM colleges/schools are well below parity
with the State population (23.1%): 13.2% in Tropical Agriculture, 11.9% in Engineering, 8.8%
in Natural Sciences, 6.5% in Medicine, and 4.4% in Ocean and Earth Sciences (UoHIRO, 2010).
More so, the Native Hawaiian population is expected to double in size from 2000 to 2050 (Hsu
& Nielson, 2010). Overall Native Hawaiian enrollment numbers increased by 21% at the
University of Hawaii at Manoa between 2004 to 2008, but the graduation rate with a four year
degree has not significantly changed, remaining between nine and twelve percent in the past
three decades (Matsumoto, 2010).
Finally, NHs are significantly underemployed in the STEM workforce. The combined
working population of NHs, Pacific Islanders, and ‘Other Race’ (grouped by U.S. Census due to
small sample size) represents 4.6% of the total U.S. workforce, but only 1.4% of STEM
occupations (Landivar, 2013). This makes NHs and Pacific Islanders the most underrepresented
ethnic group in the nation in STEM employment (factor of 3.3), more so than Hispanic (2.3),
African American (1.7), and American Indian and Alaska Native (1.5) groups.
University of Hawaii and Native Hawaiians
The University of Hawaii bears a unique responsibility to improve the educational
success, such as STEM persistence and graduation, of Native Hawaiians. The mission of the
University the Board of Regents affirm:
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

17
as the only provider of public higher education in Hawaii, the University embraces its
unique responsibilities to the indigenous people of Hawaii and to Hawaii’s indigenous
language and culture. To fulfill this responsibility, the University ensures active support
for the participation of Native Hawaiians at the University. (University of Hawaii Board
of Regents [UoHBR], 2012, 4-2)
Native Hawaiian educational attainment is a performance measure guiding the UH System in
their 2008-2015 strategic plan to “position the University of Hawaii as one of the world’s
foremost indigenous-serving universities by supporting the access and success of Native
Hawaiians” (University of Hawaii Office of the Vice President for Academic Planning & Policy
[UoHVPAPP], 2008, p. 2). The associated performance goal is to increase degree attainment of
Native Hawaiians at UH by 6-9% per year.
Although there has not always been commitment by leadership or successful educational
outcomes for Native Hawaiians at the University of Hawaii at Manoa (UoHHSTF, 1986; KSP,
2005), the institution has more recently affirmed in its UH Manoa 2011-2015 Strategic Plan the
mission:
dedicated not only to academic and research excellence but also to serving with aloha the
local, national, and internal communities that surround us. Taking as its historic trust the
Native Hawaiian values embedded in the concepts of kuleana, ‘ohana, and ahupua’a that
serve to remind us of our responsibilities to family, community and the environment.
(University of Hawaii at Manoa, 2011)
The UHM faculty senate Strategic Plan Implementation Committee unanimously approved the
2012-2013 Native Hawaiian Scholarship initiative based, in part, on the 2012 UHM Native
Hawaiian Task Force report (University of Hawai’i at Manoa Native Hawaiian Advancement
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

18
Task Force [UoHMNHATF], 2012). Improved Native Hawaiian participation is a strategic
initiative of the institution.
Effects of NH Underrepresentation
Effects of Native Hawaiian underrepresentation in STEM are two-fold. First, Native
Hawaiians continue to be underemployed in STEM professional careers translating to a loss in
talent nationally and diminishing benefits for individuals and their families. Advancement in the
areas of STEM is associated with economic benefits and has been directly correlated to a higher
living standard and improved quality of life (Burke & Mattis, 2007). NH families have the
lowest mean family income and NH individuals have the highest percentage of individuals living
below the poverty threshold compared with all other major ethnic groups in the state (KSP,
2005). The studies are clear that a college degree is economically and socially beneficial
(Adelmann, 1999; Choy, 2001; Hokoana, 2010; Day & Newburger, 2002). A focus on STEM
college and career pathways is warranted for NHs to address socioeconomic disadvantage, job
stability, and future economic well-being as well as for the nation to address the most
underrepresented population group in the national STEM workforce.
Second, because Native Hawaiians are underemployed in the STEM professions, they
have an underrepresented voice and less impact on policy and practices that affect communities
and environments. For example, forecasts of scarce freshwater supply and watershed recharge,
damage to fisheries, reefs, and ocean ecosystems, and congested infrastructure development to
accommodate the swelling population magnify global issues locally. Challenges faced by Native
Hawaiians, Hawaii, and the United States including energy dependency, climate change, and
environmental protection requires STEM solutions from diverse perspectives. If the University
of Hawaii is successful in increasing STEM participation (capabilities, motivation, and
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

19
completion) among the growing number of Native Hawaiians, then the local and global
workforce will strengthen.
Purpose of the Study
The purpose of this study is to investigate the factors related to involvement, self-
efficacy, and intent to persist for Native Hawaiian and non-Native Hawaiian college students in
STEM fields. Astin’s involvement and I-E-O framework (1975, 1993, 1999) was utilized as a
relevant model to understand undergraduate college inputs, environments, and outcomes. A
conceptual model based on the frameworks of social cognitive theory (Bandura, 1977, 1997) and
Social Cognitive Career Theory (Lent, Brown, & Hackett, 1994, 2000) were used to investigate
the influences of eight input characteristics: ethnicity, gender, socioeconomic status, financial
ability, pre-college academic performance, STEM college/major, academic level, and incoming
student status; and ten environmental factors: family support, peer interaction, faculty
interaction, faculty support, participation in a minority STEM program, college academic
performance, satisfaction, and sense of belonging to major, to school, and to campus community;
on two outcomes: STEM self-efficacy and intent to persist in STEM major.
Self-efficacy in STEM was measured to assess undergraduate STEM major’s beliefs
about their own capabilities to complete their Bachelor’s degree in STEM. Intent to persist was
measured to assess STEM major’s beliefs about their intentions or commitment to complete their
Bachelor’s degree in STEM. These cognitive self-beliefs describe the level of confidence in what
one can do and what one will do. Research has shown these outcome measures to be strong
predictors of actual persistence and degree completion (Cabrera, Castaneda, Nora, & Hengstler,
1992; Pascarella & Chapman, 1983; Hausmann, Schofield, & Woods, 2007; Lent et al., 1994;
Lent, 2013).
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

20
To contribute to the literature regarding STEM persistence and self-efficacy for
underrepresented college students, this study focused on three research questions:
1. What are the personal input and environmental factors associated with STEM self-
efficacy beliefs of undergraduate STEM students?
2. What are the personal input and environmental factors associated with intent to
persist in STEM of undergraduate STEM students?
3. How do these factors and outcomes differ, if at all, amongst Native Hawaiian and
non-Hawaiian students?
This cross-sectional, single-institution study consisted of administration of a web-based
survey to all undergraduate STEM majors at the University of Hawaii at Manoa (N=3,592)
including a subset of Native Hawaiian STEM majors. A 17.7% response rate netted a sample
size of n=638. A quantitative approach was taken to answer the three research questions by
examining the relationships between input, environment, and outcome variables. Analyses
involving descriptive statistics, factor analysis, regressions, and analysis of variance were
performed to address this study’s three research questions.
Importance of the Study
For the United States to best address the “quiet crisis” of global competition, each state
and educational institution must improve its STEM education pipeline and pathways. Engaging
groups historically underrepresented in the STEM fields can significantly increase the domestic
talent pool in science and engineering (Duderstadt, 2008). If the intellectual talent inherent in
ethnic minority groups, which will soon constitute a new majority of the domestic population,
are identified, nurtured, and encouraged, the projected gap of scientists and engineers can be
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

21
filled (Jackson, 2004). There is potential for the University of Hawaii to improve the STEM
education pipeline by addressing STEM participation of the indigenous peoples on Hawaii.
This study will address a gap in the extant research by investigating college outcomes for
Native Hawaiians in the STEM fields. While some studies exist on the college-going
experiences of Native Hawaiian students in general and for Native Hawaiian students in
community college vocational programs, few, if any, focus on behavior or motivation in STEM
fields in particular. The literature analyzes interventions and promising programs to improve
minority representation at the undergraduate, graduate, and faculty level and this study will
extend that analysis to minority STEM programs at the University of Hawaii at Manoa.
Understanding the factors that influence student persistence and self-efficacy in STEM, for
Native Hawaiian and non-Hawaiian students at a single-institution, can be used to develop
potential improvements and recommendations at the local and national level.
Organization of the Dissertation
Chapter 1 of this dissertation presents the backround of the study, the statement of the
problem, the purpose of the study, the research questions, and a brief description of the research
methodology. Chapter 2 presents a review of the literature highlighting the conceptual
framework guiding the study and a discussion of the variables selected for investigation. Chapter
3 presents the methodology including the research design and a description of the setting,
sample, instrumentation, and data collection procedures. Chapter 4 presents the analysis and
findings of this quantitative study. Chapter 5 summarizes the findings, addresses implications of
the study, and provides recommendations and conclusions. Finally, references and an appendix
conclude this dissertation.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

22
CHAPTER 2
LITERATURE REVIEW
This chapter describes the focus areas of this study related to involvement, self-efficacy,
and intent to persist for Native Hawaiian and non-Native Hawaiian college students in STEM
fields. First, Astin’s involvement and I-E-O framework (1975, 1993, 1999) will be presented as a
relevant model to understand undergraduate college inputs, environments, and outcomes. Second,
social cognitive theory and self-efficacy theory will be presented as they inform Lent et al.’s
(1994, 2000) Social Cognitive Career Theory (SCCT). Empirical support for SCCT is presented
related to its limitations and utility in studying the influences of background and environmental
factors as they contribute to self-efficacy and intent to persist in STEM. A conceptual model
based on the presented literature is detailed to provide a guide for this study. The research
questions focus on eight input characteristics: gender, ethnicity, socioeconomic status, financial
ability, pre-college academic performance, STEM College/major, educational level, and
incoming student status; ten environmental factors: family support, involvement in a minority
STEM program, peer interaction, faculty interaction, faculty support, college academic
performance, sense of belonging to school, major, and campus community, and satisfaction; and
their association with two outcomes: STEM self-efficacy and intent to persist to STEM degree
attainment. Theoretical and empirical support for the selection and utility of these variables will
be presented. Finally, STEM retention studies regarding involvement, self-efficacy, and
intention to persist of underrepresented minorities in STEM and Native Hawaiian students in
particular will follow. This chapter will provide readers with adequate knowledge to comprehend
the nature of this study’s three investigative questions:
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

23
1. What are the personal input and environmental factors associated with STEM self-
efficacy beliefs of undergraduate STEM students?
2. What are the personal input and environmental factors associated with intent to
persist in STEM of undergraduate STEM students?
3. How do these factors and outcomes differ, if at all, amongst Native Hawaiian and
non-Hawaiian students?
Theory of Student Involvement
Involvement theory builds on and provides insight to the conversation of higher
education student development literature. Beyond providing a clear characterization of what
supports student persistence, Astin’s longitudinal studies (1993, 1999) provide a breadth of data
on student inputs, environmental factors, and their correlation with student outcomes and effects.
Other theorists such as Kuh (1993), Tinto (1993), and Tierney (2004) have contributed to the
conversation with regard to the understanding of the effects of the college environment and the
role of the institution in the area of student engagement, sense of belonging, and departure.
Astin’s theory of student involvement emerged from his 1975 longitudinal study of
college dropouts seeking to identify significant factors affecting student college persistence. The
key finding was that virtually all effects contributing to college persistence suggested increased
student involvement, while all effects contributing to student departure suggested a lack of
involvement (Astin, 1999). In What Matters in College (1993), Astin reports highly consistent
results with his involvement theory of student retention (1975, 1999). Significant positive
associated variables with persistence suggest high involvement with other students, with faculty,
and with academic work. More so, significant negative correlates with persistence included
working off campus, commuting, reading for pleasure, and other involvements that take time and
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

24
energy away from the academic experience (Astin, 1993). The analysis of a variety of input
characteristics, environmental factors, and effects led to a theory of involvement. The theory of
student involvement is better understood in terms of the nature of input characteristics and their
correlated effects.
Astin (1999) defines “student involvement” as the “amount of physical and psychological
energy that the student devotes to the academic experience” (p. 519). Involvement is related to
the concepts of “effort,” “time on task,” and “vigilance” (p. 518). Astin (1999) gives additional
postulates of student involvement theory: involvement occurs across a continuum such that a
student can manifest different degrees of involvement in different objects at different times;
involvement has both quantitative and qualitative features; student learning and personal
development is directly proportional to the quality and quantity of student involvement; and the
effectiveness of educational policy or practice is directly related to the capacity of the policy or
practice to increase student involvement. These postulates suggest that the means for faculty,
educators, and institutions intending to increase student learning and development are to focus on
the motivation, behavior, and involvement of the student. The charge for educators is to develop
environments that elicit sufficient student effort and investment of time, energy, and active
participation by the student. Thus, Astin’s theory suggests that the most important institutional
resource is the student’s time.
Inputs-Environment-Outcomes Model
A key element to Astin’s methodology is the I-E-O or Inputs, Environment, Outcomes
model. Pascarella and Terenzini (2005) calls Astin’s I-E-O model “one of the first and most
durable and influential college impact models” (p. 53). In order to study college affects, the
conceptual/methodological guide views college outcomes as functions of inputs (such as pre-
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

25
college academic experiences, demographic characteristics, socioeconomic status), environment
(such as the college experiences students encounter in college), and outcomes (such as student
characteristics, knowledge, beliefs, values, and behaviors). Astin’s extensive variables and
measures were analyzed to find patterns, correlations, and findings that gave empirical credibility
to the I-E-O and student involvement theory.
Student involvement and the I-E-O framework provide a model for investigating college
student development. However, Astin is more concerned with defining and identifying
involvement in the behavioral sense (how a student behaves) rather than in the cognitive or
motivational sense (how a student thinks or feels). In order to investigate the attitudinal,
affective, and cognitive beliefs of college students, such as self-efficacy and intent to persist in
STEM, involvement theory by itself is insufficient.
This cross-sectional study, further described in Chapter 3, does not intend to directly
investigate STEM persistence or STEM graduation behavioral measures, but instead to focus on
predictive cognitive measures of STEM self-efficacy beliefs and intent to persist. It is not only
important to identify what a student does but why a student does (or does not). Social cognitive
theory and social cognitive career theory expand on the role of input and environmental factors
as they influence cognitive beliefs and student interpretations of their environment and
experiences.
Social Cognitive Theory
Bandura’s (1986) social cognitive theory of human behavior offers a key framework
related to the study of motivation and in mediating academic persistence. The beliefs
individual’s hold about their abilities and the outcome of their efforts strongly influence their
actions and behaviors (Parajes, 1996; Bandura, 1977, 1986). Built on the concept of reciprocal
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

26
determinism, social cognitive theory offers the view that (a) personal factors (e.g. thoughts and
beliefs), (b) behavior, and (c) environment mutually influence each other (Parajes, 1996).
Personal beliefs, including those stemming from self-reflection and self-evaluation, are
influenced by and can influence the individual’s environment and behavior.
The key factor for human agency, Bandura (1997) argues, is self-efficacy. Bandura
(1977, 1997) formally defined self-efficacy as personal judgments of one’s capabilities to
organize and execute courses of action to attain desired goals. Self-efficacy is a performance-
based measure of perceived capability related to specific tasks in a given domain. Self-efficacy
beliefs influence three behaviors: the individual’s goal choice, the effort enacted to reach those
goals, and the persistence when difficulties arise (Bandura, 1997; Pajares & Urdan, 2006;
Rittmayer & Beier, 2009). The construct of self-efficacy, first introduced extensively in Social
Cognitive Theory (Bandura, 1977), is the focus of this study as applied to choice and persistence
in STEM.
Self-Efficacy Theory
Belief in one’s capabilities to perform a specific task is referred to as self-efficacy.
Bandura (1977, 1997) describes three characteristics of self-efficacy: level, generality, and
strength. These, respectively, pertain to dependence of the difficulty of the task, the
transferability of self-efficacy beliefs — such as from writing to algebra, and the amount of one’s
certainty about performing a certain task. Self-efficacy is focused on performance capabilities
as opposed to personal qualities such as physical or psychological characteristics (Zimmerman,
2000). Judgments are made about one’s confidence in accomplishing a task, not about who they
are or how they feel about themselves in general.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

27
With regard to content, perceptions of efficacy depend on mastery criterion of
performance as opposed to how well one expects to do in comparison with others. Self-efficacy
beliefs may differ across domains (a student has high levels of self-efficacy in mathematics but
not in biology) and may differ across tasks (a student has high self-efficacy in conducting
engineering research but low levels of self-efficacy in completing an engineering design project).
In addition, measures of self-efficacy are context specific, and are sensitive to changes in the
performance context. For example, a student’s self-efficacy for getting an A in an exam can
differ if administered in an uncomfortable, noisy environment, if taken after three other exams,
or if taken under optimal conditions. Finally, self-efficacy beliefs specifically refer to capabilities
of future performance.
Related Constructs
Self-efficacy differs from the related constructs self-concept, self-esteem, and ability.
Efficacy beliefs refer to contextual and task-specific capabilities such as, “I believe I can score a
B or better on my next chemistry assignment.” It is not meaningful to say someone has high (or
low) self-efficacy in general. It is appropriate to say one has high math self-efficacy or low self-
efficacy in writing a term paper. By definition self-efficacy beliefs are goal specific. Self-
efficacy beliefs to complete a STEM degree were the focus of this study.
In contrast, self-concept describes self-perceptions that are more general than self-
efficacy. Whereas self-efficacy focuses on beliefs about capabilities of future performance of a
specific task, self-concept includes affective and evaluative components of the broader domain.
For example, “I am good at chemistry” or “I hate chemistry” are beliefs describing self-concept.
An individual can have a high self-concept in engineering, but may have low self-efficacy to
complete a specific engineering project due to reasons affecting their capabilities (can’t
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

28
understand the professor, team members are not reliable contributors, insufficient budget to
accomplish design, etc.). However, self-concept is often positively related to self-efficacy and
both motivational constructs develop in similar ways through reinforcements and evaluations of
others, self-judgments about past experiences, and interpretations of their environmental
conditions (Schunk & Pajares, 2002; Rittmayer & Beier, 2009).
Beliefs about general abilities in a domain (self-concept) and task-specific capabilities of
achievement (self-efficacy) are closely related to, and partly based on self-esteem. Self-esteem
relates to global, evaluative feelings of self-worth. Research has shown self-efficacy to be a
stronger predictor of task-specific performance than self–concept and self-esteem (Pajares &
Urdan, 2006; Zimmerman, 2000).
Much of the discussion thus far, has centered on the notion of self-perceived ability to
accomplish a task, what then of actual ability, its relation to self-efficacy, and their collective
results on academic performance? Certainly self-efficacy alone cannot enable someone to
accomplish a task without some level of ability. However, given two persons of equal ability,
unequal levels of self-efficacy will lead to unequal performance behaviors and goal attainment
(Pajares & Urdan, 2006). For example, Betz and Hackett (1981) found that among women and
men of equal prior academic performance (grade point average) in science, engineering, and
mathematics fields, men tended to estimate their capabilities higher (higher self-efficacy) than
that of women (lower self-efficacy) in future coursework. This difference, in part, led to uneven
subsequent performance and decision to persist or leave the science/engineering field. It is noted
in this example that self-efficacy is derived from socio-cultural factors including gender
differences in one’s interpretations of their experiences and environmental cues. Higher self-
efficacy allows individuals to better organize, manage, and make the most of their talents. Those
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

29
who doubt their capabilities may be less likely to persist or expend additional effort on a goal or
they may chose to avoid the goal altogether.
Although research has shown the benefits of high self-efficacy, too much self-efficacy
(relative to actual abilities) is not a good thing for future motivation and performance.
Overconfidence in self-capabilities in relation to actual abilities may lead to under-preparation,
underperformance, and ill-fitted choice goals that lead to failure and future discouragement.
Similarly, when self-efficacy levels are too low in relation to actual abilities (under-confidence),
performance may be negatively affected by diminished effort and persistence in the face of
setbacks, anxiety and unnecessary physiological detriment, lower goals, and avoidance of
realistic challenges (Bandura, 1986; Lent, 2013). Both types of misjudgments of actual ability
can hamper skill development, motivation, and performance (Zimmerman, 2000). Lent (2013)
discusses the benefit of slight overconfidence or congruence with actual abilities which
encourages motivation for pursuing task challenges, promotes proximal learning development,
and promotes future and ongoing performance.
Development of Self-Efficacy
Bandura (1986, 1997) theorizes four sources of influence on self-efficacy: mastery
experiences (from past accomplishments), vicarious learning experiences, social persuasion, and
physiological reaction/affective states. Mastery experiences refer to prior performance outcomes
and one’s interpretation of their successes, challenges, and abilities. Prior experience affords the
individual a better understanding of their capabilities to succeed in the future. Successful prior
performance will likely lead an individual to feel confident in their capabilities to succeed in a
like task in the future, however, poor prior performance is likely to cast doubt on the individual’s
self-perceived ability to do well on the next, like task.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

30
Vicarious learning experiences refer to observations of similar others succeeding or
failing at particular activities. This is another strong source of influence for determining one’s
own self-efficacy, especially if one has little direct experience to estimate one’s own capabilities.
Role models or persons perceived as similar to the subject are best influences for the subjects’
self-efficacy for like tasks. If a model is viewed as having much higher talent or abilities as the
observer, than the relevance may be discounted by the observer.
Social and verbal persuasion refers to others feedback, support, and influence. In
academic settings, feedback and judgments from faculty/teachers, counselors, and peers can
enhance and erode self-efficacy. Verbal persuasion, however, may have less impact on the
learner than observer or direct experiences because outcomes are described and not witnessed
and thus relies heavily on the credibility or influence of the other (faculty, parents, etc.). Verbal
persuasion is strongest when tied to mastery experience such as feedback relative to a specific
previous task performance (Pintrich, 2003; Betz & Schifano, 2000).
Finally, physiological reaction refers to how one interprets their emotional and physical
states to determine their self-efficacy beliefs. Fatigue, nervousness, “butterflies”, stress, and fear
of failure are all physiological reactions that can affect self-efficacy and the increased anxiety on
subsequent performance.
Social Cognitive Career Theory
Social Cognitive Career Theory (SCCT) (Lent et al., 1994; Lent, 2013) is an application
and extension of Bandura’s (1986) social cognitive theory. It shares the core elements of social
cognitive theory that emphasizes the role of triadic reciprocity between people, their behavior,
and their environment. Concurrent with social cognitive theory, personal agency or self-
direction also plays a central role in mediating behavior, although the interchange with
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

31
environmental supports, barriers, and other factors can strengthen, weaken, or even override
personal agency (Lent, 2013). Individuals possess self-regulatory skills to organize, reflect, and
regulate their own behavior and make alterations to their environment and personal factors. This,
in turn, leads to changes in their subsequent behavior. Individuals are seen as contributors to
their life circumstances rather then as by products of their life circumstances.
SCCT is a relatively recent framework to better understand educational and career
development behavior. The theory attempts to describe the interactions between background
characteristics (e.g. gender, ethnicity, predisposition) and the environment with cognitive-person
factors (e.g. self-efficacy, outcome expectations, and goals). The four academic and career
development outcomes modeled by social cognitive career theory are career and academic
interest, choice, performance, and most recently satisfaction/well-being. SCCT is useful in this
study to explore the relationships between self-efficacy and intent to persist, while also exploring
the complex ways in which social cognitive factors, environment and personal influences
describe persistence, performance and development in STEM.
Career development and academic choice and success, Lent et al. (1994) argue, are quite
similar although they typically appear in different literatures. SCCT offers a segmental model of
career behavior focusing primarily on issues of career interest, preparation, choice, and entry.
This dovetails (in late adolescence and early adulthood) with academic development. Causal
models and mechanisms appear in both career and academic development study. The dynamic
nature of the influential factors within persons, environments, contextual supports and contextual
barriers are relevant. For example, investigating the factors related to choice, performance, and
persistence for undergraduate students considering a STEM major closely mirrors that of
individuals considering a STEM career.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

32
Empirical Support for Theoretical Models
There is empirical support for the use of SCCT to explain how personal inputs, social
cognitive variables, and college environment aid in our understanding of academic behavior,
motivation, and outcomes. Meta-analytic methods provide a quantitative way to integrate,
compare, and contrast the results from multiple, independent studies. This is advantageous in
reviewing a model because larger data sets may be able to increase statistical power and better
estimate true effect size. Meta-analytic reviews also allow researchers to investigate
inconsistencies and variation between studies. Caution must be taken into account, however, for
selection of poorly designed (methodologically unsound) studies, sources of bias, and
combination of summary measures.
Several meta-analysis of research focusing on young adults have directly tested a number
of SCCT’s hypotheses. Rottinghaus, Larson, and Borgen (2003) empirically synthesized and
evaluated 60 independent samples (N=39,154) finding a strong overall relationship (r = .59)
between self-efficacy and career interests. Regarding choice hypothesis, Sheu, Lent, Brown,
Miller, Hennessey, and Duffy (2010) found in their meta-analysis of SCCT studies that self-
efficacy, interests, and outcome expectations strongly predicted choice goals. Multon, Brown,
and Lent (1991) conducted an efficacy-performance meta-analysis from 36 studies yielding a
total of 38 samples (N=4,998) of subjects. Studies included a mix of experimental (18) and
correlational (13) design and included subjects from elementary to college level. Multon et al.
(1991) found support for the hypothesized relationships of self-efficacy to academic performance
and persistence. The relation of self-efficacy to performance varied by students’ prior academic
achievement with stronger relations found among low-achieving students. College and high
school student samples also evidenced stronger effect sizes for efficacy-performance relationship
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

33
than did elementary students. In summary, research findings suggest that self-efficacy beliefs are
generally related to academic behaviors in the ways that support Bandura’s (1977, 2001) social
cognitive theory and its extension to social cognitive career theory.
Betz and Hackett (1981) were one of the first to apply SCCT to investigate differences
among women and men in undergraduate career choice interest and persistence in engineering.
They noted that gender role socialization tend to provide experiences that limit self-efficacy in
nontraditional career and academic domains such as Science, Technology, Engineering and
Mathematics. Interest, consideration, and academic persistence in nontraditional choice options
were found to be lower for women than men. Subsequent to the Betz and Hackett (1981) work,
many other studies have applied SCCT and investigated socio-cognitive variables on diverse
populations. Researchers have focused on the application of SCCT on student populations
taking into account gender, race/ethnicity, culture, socioeconomic status, age, and disability.
Conceptual Model
Background, socio-cultural, college environment, and cognitive factors on the diverse
populations in particular academic domains (such as STEM) have received growing attention
from researchers. Few studies focus on the combined effects of the variables to Native Hawaiian
students in general and none exist to the knowledge of the researcher focusing on Native
Hawaiian STEM students in particular. The intent of this study is to examine the differences
between Native Hawaiian and non-Hawaiian STEM students on various contextual and
environmental factors and their impact on STEM self-efficacy and intention to persist.
This section describes the conceptual model used in this study based on the Inputs-
Environments-Outcomes framework and SCCT. The selected input and environmental factors
are presented based on relevant research as well as a review of the dependent variables. It was
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

34
hypothesized that all variables will directly or indirectly mediate self-efficacy and intent to
persist, but the level and nature of influences are yet to be determined. Furthermore, differences
between Native Hawaiian and non-Hawaiian STEM students were explored.

Figure 1. Conceptual Input – Environment – Outcomes (I-E-O) model for STEM self-efficacy
and intent to persist. Adapted from Astin (1993).

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

35
Outcome Variables
The two dependent variables of interest in this study were STEM self-efficacy and intent
to persist. Self-efficacy in STEM was measured to assess undergraduate STEM major’s beliefs
about their own capabilities to complete their Bachelor’s degree in STEM. Intent to persist was
measured to assess STEM major’s beliefs about their intentions or commitment to complete their
Bachelor’s degree in STEM.
Self-Efficacy
In this study, self-efficacy in STEM will pertain to student’s self-beliefs about their own
capabilities to complete their intended STEM degree at their current institution. Self-efficacy has
garnered increased attention from researchers due to its influence on choice goals, task
performance, and motivation discussed prior in this chapter. STEM self-efficacy is modeled as a
dependent variable based on the influences of the other personal input and environmental factors
outlined in this study’s conceptual model.
Intent to Persist
Although research has shown self-efficacy to be strongly directly related to task
motivation and interest, capabilities do not directly measure intent or interest. For example one
can have a high self-efficacy in washing dishes (high confidence in their capability to complete
the task) but make no intention of washing the dishes. Similarly, one can feel confident and
capable of flying a kite, but have little or no interest in kite flying. These examples highlight that
self-efficacy beliefs are statements about what one can do and not what one will do. Bandura
(1997) argues that individuals tend to avoid tasks that they do not feel capable of successfully
completing, but the opposite may not be true. For these reasons, self-efficacy is often examined
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

36
in concert with outcome expectations, interests, and choice goals. This study investigated self-
efficacy for STEM degree completion along with intent to persist in STEM.
The use of intent to persist as an outcome in research is substantiated by prior studies
showing a strong association between intentions to persist and actual persistence (Bean, 1980;
Cabrera et al., 1992; Pascarella & Chapman, 1983, Hausmann et al., 2007). Tinto (1975, 1987,
1993) theorized students’ integration into their social and academic environment were critical to
student persistence. Commitments to finishing college (goal commitment) and to completing a
degree at the college in which they are enrolled (institutional commitment) were also determined
to be importance predictors of student persistence. Although research has shown intent to persist
tends to overestimate actual persistence behavior, intent not to persist is an excellent indicator of
student attrition.
It is important to note that this study defines intent to persist as intent to continue in a
STEM major at their current institution to degree completion (goal commitment + institutional
commitment). Persistence is often measured in other studies as year-to-year continuation,
whereas this study is focused on graduation and degree attainment. Other definitions of
persistence not applicable to this study are concerned with major or domain persistence, in which
an science student is labeled as a persister in science after transferring to one or more institutions
and maintain their focus on completing their science degree. Conversely, some studies of
persistence are concerned with institutional persistence regardless of degree whereby
institutional transfers, leavers, and stop-outs are labeled as non-persisters. Intent to persist in this
study refers to the student’s self-reported intention to complete a STEM degree at the current
institution. This definition of persisters includes students that may switch between STEM majors
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

37
at the institution but does not include STEM students that change majors from a STEM degree to
a non-STEM degree.
Environment Factors
This study examined ten environment factors: family support, participation in a minority
STEM program, peer interaction, faculty interaction, faculty support, sense of belonging to
major, belonging to institution, belonging to campus community, and satisfaction. The ten
environmental factors will be treated as independent variables. Although there is disagreement
to the level and nature of influence, the extant literature has shown these factors to influence self-
efficacy belief and intent to persist.
Family Support
Parents, family members, and caregivers provide experiences, role modeling, and
persuasion that differentially influences individual’s self-efficacy and intentions. Home
influences can help children interact effectively with their environment, stimulate curiosity, gain
mastery experiences, and positively affect self-efficacy (Bandura, 1997). Lack of family support
has been shown to be a barrier to success in STEM, whereas ongoing encouragement from
parents positively influenced persistence (Sandler, 1999; Swail & Perna, 2002). Family dynamics
can positively and negatively influence self-efficacy and intention to persist.
Participation in Minority STEM Program
This factor identifies the involvement and potential influence of intervention activities at
the institution to promote the academic development of Native Hawaiian and other minority
students in the STEM disciplines. Minority STEM programs of interest in this study are the: (1)
Native Hawaiian Science & Engineering Mentorship Program (NHSEMP), (2) Louis Stokes
Alliance for Minority Participation (LSAMP) Scholars Program, (3) the Center for Microbial
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

38
Oceanography (CMORE) Scholars Program; (4) Na Pua No’eau Center for Gifted and Talented
Native Hawaiian Children (NPN); (5) Hui Manawa Kupono Native Hawaiian Scholarship
Program; (6) Minority Access to Research Careers; (7) Undergraduate Research at Mentoring
(URM) in the Biological Sciences; (8) Pacific Internship Programs for Exploring Science
(PIPES); (9) UH Manoa Honors Program; (10) Kua’ana Native Hawaiian Student Services
Program; and the (11) UH Manoa Honors Program.
The extant literature analyzes interventions and promising programs at the undergraduate,
graduate, and faculty level that address increasing the success of minority students in STEM.
Jackson (2004) identified the Meyerhoff Scholars Program (MSP) at University of Maryland,
Baltimore County (UMBC) as an exemplary program targeted to produce African American and
other minority students to complete STEM BS degrees and continue to earn STEM doctorates.
The MSP is largely research-based, utilizes residential intensive peer study groups, and focuses
on the need to build a strong sense of community. MSP students achieved higher GPAs,
graduated in STEM majors at higher rates, and gained acceptance to graduate schools at higher
rates than current and historical samples. More so, faculty involvement and institutional
commitment (the intervention will not disappear if external funding for MSP ends) are key
components cited by evaluators and researchers.
Successful programs including MSP and LSAMP agree with Astin’s (1993) theory of
student involvement. Expectations are very high and holistic support systems including strong
faculty mentoring and financial assistance are present. For Louis Stokes Alliance for Minority
Participation (Clewell, de Cohen, Tsui, & Deterding, 2006; Leggon & Pearson, 2009) student
participants, the four strategies most often cited (among national programs) are student research
(82%), summer bridge academic preparation (67%), mentoring (60%), and stipends (48%).
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

39
Similar to MSP the program stresses an integrated approach to provide financial, academic,
social, and professional support. Factors negatively affecting degree completion or continuation
include lack of quality mentoring, poor academic preparation, and poor connection with campus
or department culture. Although this study does not aim to directly investigate the activities and
outcomes of indivudal minority STEM programs at the research setting, it is predicted that
student participation has an effect on STEM self-efficacy and intent to persist.
Peer Interaction
The research evidence indicates that peer relationship and interaction is linked with
identity development and social adjustment for college students. In a study of National
Longitudinal Survey of Freshmen data (N=3,924), Fischer (2007) found a positive relationship
between relatedness to peers and college persistence for on-campus peers, and a negative
relationship off-campus peers. Bandura (1997) argues that self-efficacy development is, in part,
dependent on an individual’s social relations within their environment and culture. This study
will focus on peer interaction with respect to its association with intent to persist and STEM self-
efficacy.
Faculty Interaction and Support
Research has shown that student-faculty interaction is a strong predictor of student
learning and persistence for all students, including minority students (Kuh & Hu, 2001;
Pascarella & Terenzini, 1991, 2005; Tinto, 1993). In a study on 7,603 college students, Cole
(2007) investigated student-faculty interaction as both an intermediate outcome variable from
interracial interactions and an environmental variable to inform intellectual self-concept. In
addition, Cole and Espinoza (2009) looked at support from faculty in terms of encouragement,
feedback, and help in areas such as professional goals, intellectual challenge and stimulation, and
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

40
respect. Faculty support was found to have positive influence on academic performance for
URM STEM students and that the quality of faculty interaction was important than the quantity
of faculty interaction (Cole & Espinoza, 2009). Through descriptive analysis, factor analysis,
and regression analysis, this study investigated the effects of faculty interaction on self-efficacy
and intent to persist as they apply to Hawaiian and non-Hawaiian STEM students.
College GPA
Cumulative college grade point average is measured in this study to represent academic
achievement. While other indicators of academic achievement exist such as degree aspirations,
degree attainment, degree attainment with honors, knowledge or skill based assessment,
demonstration, and performance (Astin, 1993), grades and grade point average were found to
predict completion of bachelor’s degree even after controlling for other factors (Pascarella &
Terenzini, 1991, 2005). In a meta-analytic review of 109 published studies including 279
correlations with cumulative GPA, Robbins et al. (2004) found the largest true score correlations
of .50 between self-efficacy and cumulative GPA, and .45 between high school GPA and
cumulative college GPA.
Sense of Belonging
Sense of belonging refers to the psychological sense that one is a valued member of the
whole and was investigated in areas of belonging to major, belonging to institution, and
belonging to the campus community. Due to its impact on student persistence and other
academic outcomes, sense of belonging has been a closely examined construct. Tinto (1975,
1987, 1993) theorized students’ integration into their social and academic environment as being
critical to student persistence along with institutional commitment. Bean (1985) identified
institutional fit, the extent in which students felt they “fit in” at the university, as a key element
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

41
to successful socialization and student persistence. In a study focusing on sense of belonging as a
predictor of intent to persist on a sample of first year African American and Caucasian students
(N=365), Hausmann et al. (2007) found sense of belonging to be positively related to intent to
persist and other environmental factors such as faculty interaction, peer support, and parental
support. Sense of belonging was not related to race, gender, financial difficulty, or academic
integration. Hausmann et al. (2007) also found sense of belonging to decrease with time over the
students first year possibly due to the initial high expectations and excitement new college
students bring with them to college. Commitment to the institution and intentions to persist also
declined with time (Haussmann et al., 2007). This cross-sectional study assessed senses of
belonging across educational levels to evaluate their relation to the research outcome variables.
Sense of belonging has additional importance for Native Hawaiians in STEM in terms of
interpretations of acculturation and enculturation. Native Hawaiian individuals, like other
groups, vary greatly in their sense of cultural identity, pride, cultural practice, and ability to
manage potential differences and similarities in their sense of academic, cultural, and home
community. Makuakane-Drechsel and Hagedorn (2000) postulated but did not explore the
influence of Native Hawaiian culture, values, and traditions on Hawaiian student persistence.
Matsumoto (2010) found that Hawaiian sense of belonging, gender, Hawaiian blood quantum,
and residential boarding as a high school student were insignificant factors among Hawaiians in
degree completion. Hagedorn and Tibbetts (2003) found high family responsibilities, high job
responsibilities, cultural obligations, and starting out at a community college were negative
factors to retention and completion, which relate to the inquiry of Native Hawaiian student
cultural identity.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

42
Native Hawaiian cultural identity, integration, and development and their associated
impacts on STEM degree completion are interesting questions left unanswered. Some past
studies and reports (Alu Like, 1988) have used anecdotal rather than empirical evidence to show
the relationship between knowledge of and involvement in Hawaiian cultural traditions and
college persistence. Further qualitative studies would add to the findings. Matsumoto (2010)
also cautions that thinking of the institutional culture and Native Hawaiian culture as being in
opposition (i.e. that one needs to subsume one’s culture to integrate into the institution’s culture)
as a faulty assumption. Program administrators, educators, and stakeholders in Native Hawaiian
education can incorporate cultural factors within strategies to support the student holistically.
Similarly, Pacific Policy Research Center (PPRC, 2010) reviewed studies finding greater levels
of learning for Native American students who attended institutions with a deep commitment to
diversity. University of Hawaii campuses, departments, and programs seek to build an
environment and culture that addresses Native Hawaiian student needs that reinforces their level
of satisfaction and success (UoHVPAPP, 2002) but do not exactly know how to do so. Assessing
sense of belonging among Native Hawaiian STEM students is an initial step in understanding
this process.
Satisfaction
Research has shown student educational satisfaction to be a mediating concept positively
associated with academic performance, persistence, and student development (Astin, 1993;
Pascarella & Terenzini, 1991; Tinto, 1975). Astin (1993) found students reported highest levels
of satisfaction with their major courses, opportunities to interact with faculty, opportunities to get
involved in extracurricular activities, and overall college experience. Lowest levels of
satisfaction were found to be related to university life policies and regulations and with student
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

43
support services such as academic advising, financial aid, and housing (Astin, 1993). The
environmental variable satisfaction was based on self-reported survey items focusing on student
perceptions of quality of instruction, interactions with peers, satisfaction with STEM major and
advising, and satisfaction with overall college environment.
Input Factors
This study examined eight background personal inputs and academic classification
variables: gender, ethnicity, social economic status, financial ability, pre-college (high school)
grade point average, STEM college, academic level, and incoming student status. These
background variables were treated as independent variables as they influence two dependent
variables: STEM self-efficacy and intent to persist in STEM.
Gender
Gender is a key personal input in this study of student development in STEM. Byars and
Hackett (1998) reviewed the literature on academic development of women through the lens of
SCCT describing the process through which gender as well as ethnicity, socioeconomic, and
cultural factors influence social cognitive and academic outcomes. Racism, sexism, and gendered
role orientation can influence environmental factors of social support, learning experiences, and
perceived and encounters barriers (Lent et al., 1994; Flores & O’Brien, 2002). Research on the
underrepresentation of females in certain disciplines such as Engineering, Computer Sciences,
Physics, and Mathematics have shown gender to be an important variable in understanding
college outcomes (Cole & Espinoza, 2008; Jackson, 2004; Leggon & Pearson, 2009).
Ethnicity
This study was more concerned with the socio-cultural constructs of ethnicity and gender
that influence contextual, cultural, and cognitive effects, as opposed to the view of race and sex
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

44
as categorical physical or biological factors. Ethnicity was measured based on University of
Hawaii categories, with Native Hawaiian reported measure taking priority in instances of mixed
Native Hawaiian measures. Ethnicity was a self-report measure assessing which of one or
combination of 25 ethnic categories students self-identified.
National reports show that college outcomes vary significantly by ethnicity group (NSB,
2010). Research has shown Native Hawaiians perceive different barriers to college success than
Non-Hawaiians (Hokoana, 2010; Hagedorn & Tibbetts, 2003). A central question to this study
was to explore any differences of input, environment, or outcome variables between Native
Hawaiians and non-Hawaiians.
Socioeconomic Status and Financial Ability
Socioeconomic status in this study was viewed as a personal input because it indirectly
influences socio-cognitive variables (e.g. self-efficacy and intent to persist) due to its impact on
choice behaviors, past experiences, and academic achievement, particularly with minority
populations. For example, since many racial ethnic minority students have been found to be from
low-income households, have parents with lower levels of education, and may have less
exposure to high quality schools and learning experiences, they are at risk for lower academic
achievement, entry, or completion of higher education (Sirin, 2005). Thus, socioeconomic status
as an input can affect student’s college access and choice not only directly, but also indirectly
through the lack of social and cultural capital and the creation of poor signals sent to post-
secondary education (Perna, 2000; PPRC, 2010). Social strata in which a person belongs
continues to be a critical factor in the academic development of minority and non-minority
students (Sirin, 2005).
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

45
Although financial-aid will not be directly measured in this study, it is noted that self-
assessment of financial ability can provide some insight to potential barriers to college
persistence. Given the lower socioeconomic status of Native Hawaiians in Hawaii (Benham,
2006; Hsu & Nielson, 2010; Makuakane-Drechsel & Hagedorn, 2000), it was expected that
college financial assistance was a strong influence on college outcomes. Kumashiro (2006)
elaborates that “Native Hawaiians and other Pacific Islanders are living in poverty at almost one-
and-a-half times the national average” (p. 131). Hagedorn and Tibbetts (2003) found that receipt
of financial assistance “clearly overshadowed other important variables such as high school
grades and family responsibilities” (p. 15) as the leading factor increasing the likelihood that
Native Hawaiian students will complete college. Financial support was also cited as the most
significant factor contributing to recruitment and retention of Pacific Islander students (Ah Sam
& Robinson, 1998) at the University of Hawaii at Manoa. This study investigated the
association of SES and financial ability with STEM self-efficacy and intent to persist as well as
explored significant difference between Native Hawaiian and non-Hawaiian groups.
High School Grade Point Average
High school grade point average was the variable selected to evaluate pre-college
academic performance and preparation. Grade point average in high school (academic
preparation) and cumulative college grade point average were strong indicators of success
(degree completion) found in the literature for Native Hawaiians and other URM students.
Researchers further note that while cognitive factors (including performance on standardized
tests) are strong predictors of success, models that also incorporate non-cognitive factors (such as
motivation, leadership, and self-efficacy) provide a more promising tool to identify students who
may leave STEM or have potential in STEM (Besterfield-Sacre, Altman, & Shuman, 1997;
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

46
Zhang, Anderson, Ohland, Carter, & Thorndyke, 2002). Academic factors commonly found in
successful URM STEM programs include academic engagement such as learning communities
and involvement in research.
STEM College and Major
This variable referred to the particular STEM College or school to which a student was
currently enrolled. In this single-institution study, there were four STEM colleges or schools that
housed undergraduate academic STEM majors: the College of Engineering (Engineering),
College of Natural Sciences (Natural Sciences), College of Tropical Agriculture and Human
Resources (Agriculture), School of Ocean and Earth Science and Technology (Ocean). A full
list of STEM academic majors is presented in Appendix C.
The STEM college variable could be viewed as an input factor and an environmental
factor. Upon enrollment into the institution of study, undergraduate students self-select and are
admitted to their academic college/school of choice. An exception, however, is engineering in
which self-selected students are granted admission based on criteria (high school Trigonometry,
Chemistry, Physics, SAT-Quantitative, and high school GPA) in addition to institutional
admission criterion. Students can also elect to change majors at any time during their academic
journey thereby switching between departments, Non-STEM or STEM colleges, and/or
institutions. STEM college was viewed as an input variable because it signifies a prior choice
goal and choice action (major in particular academic domain).
Although STEM colleges are structured primarily due to organizational reasons, they
certainly represent like academic domains based on STEM content areas. STEM colleges have
their own academic leadership, faculty, curriculum requirements, student learning objectives,
community of students, and culture(s). STEM college was hypothesized to influence other
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

47
independent variables such as sense of belonging, faculty interaction, satisfaction, minority
STEM program participation, and dependent variables self-efficacy and intent to persist.
STEM college and major were treated as academic classification input variables because
of its proximal influence on the ongoing student experience (environment). In the model, they
are conceptual inputted after background or personal input variables such as gender, ethnicity
and SES that are pre-college factors. The influence of academic classification variables, STEM
College, educational level and incoming student status, were analyzed after background inputs
and before college environment variables such as peer and faculty interaction.
Educational Level and Incoming Student Status
Educational level refers to class standing in major such as freshmen, sophomore, junior,
and senior. This factor is related to but different from total college credits completed or
semesters in college because it is focused on degree completion in selected major. A student
may have completed 100 credits and eight semesters of college but only have sophomore level
class standing as a Biology major based on their degree curriculum. Higher academic level
denotes persistence towards degree completion.
Educational level was of particular interest in this cross-sectional study. Self-efficacy
and intent to persist may change as students progress along their curriculum and have more
experiences at the college. Changes over time can also be inferred by looking at the progression
from freshmen to senior levels.
A related academic status variable is incoming student status. This variable refers to the
student’s pre-institution experience directly before entrance into the University such as first-time
clean freshmen, transfer from another four-year institution, transfer from a community college,
and non-traditional or returning student. Research indicates that perceptions and experiences of
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

48
students vary by incoming student status (Makuakane-Drechsel & Hagedorn, 2000; Kuh, 1993;
Pascarella & Terrenzini, 1991, 2005). In this study, incoming student status was treated as an
independent variable in predicting STEM self-efficacy and intent to persist.
Summary
The literature provides some empirical and theoretical foundation to guide this
investigation of Native Hawaiian and Non-Hawaiian STEM majors. A conceptual model based
on Astin’s I-E-O framework, social cognitive theory, and social cognitive career theory was
utilized to understand the complex influences of background and environmental factors on
STEM self-efficacy and intent to persist in STEM. Limitations and voids in the literature
warrant further research in self-beliefs and influential persistence factors for undergraduate
Hawaiian and non-Hawaiian STEM undergraduate students. Few studies focus on the combined
effects of the variables to Native Hawaiian students in general and none exist focusing on Native
Hawaiian STEM students in particular. The intent of this study was to contribute to the literature
in this regard.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

49
CHAPTER 3
METHODOLOGY
The aim of this quantitative study was to extend the literature on Native Hawaiians by
investigating factors associated with positive college outcomes. Using social cognitive career
theory (Lent et al., 1994), this study explored the influence of input and environment factors on
two college outcomes of interest: STEM self-efficacy and intent to persist. In addition, Native
Hawaiian and non-Hawaiian groups were investigated for significant difference. The research
questions for this study were:
1. What are the personal input and environmental factors associated with STEM self-
efficacy beliefs of undergraduate STEM students?
2. What are the personal input and environmental factors associated with intent to
persist in STEM of undergraduate STEM students?
3. How do these factors and outcomes differ, if at all, amongst Native Hawaiian and
non-Hawaiian students?
Chapter 3 provides an overview of the research design, institutional context, population, and
sample. Then, a description of the instrumentation and measures for all variables in the study is
presented. A discussion of the data collection and quantitative analysis procedures concludes the
chapter.
Research Design
This section describes the plans and procedures selected to best address the research
questions given the scope, researcher’s experience, and audiences for the study. The use of a
quantitative approach, cross-sectional design, and self-reported survey research are discussed.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

50
Quantitative Approach
To address the research questions, this study utilized a quantitative, non-experimental,
survey design. A quantitative approach is often chosen as a means for isolating and measuring
variables, investigating their relationships, and testing objective theories (Creswell, 2008).
Consistent with prior studies investigating SCCT, this study gathered input, environmental, and
outcome variable data to be analyzed with quantitative methods to seek trends and correlations
among variables.
In addition, a quantitative approach was chosen (as opposed to qualitative or mixed-
method approach) to address the three research questions to thereby inform a broader objective
to improve the college outcomes of Native Hawaiians at the University of Hawaii. In terms of
generalizability and grounds to inform policy or guidelines, it was important for the study to
gather and analyze data from as many individuals as possible. The quantitative survey approach
provided the researcher with efficient means to invite participation from the entire population of
UH Manoa STEM majors.
Lastly, the quantitative approach was chosen in efforts to limit bias and perceptions of
undue bias in the study. Given that the researcher works in Native Hawaiian STEM education at
the institution, guidelines of a quantitative, objective approach were utilized to reduce systematic
error and increase validity of results.
Cross-Sectional Design
A cross-sectional, single-administration survey was chosen to allow for the collection of
current, self-reported data, in a single-point in time. Experimental and longitudinal designs,
which are better suited for investigating claims of cause and effect or impacts of interventions,
were not as appropriate for this study (Creswell, 2008). This study examined the relationships
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

51
between variables and between groups. Cross-sectional designs are important and applicable in
this study because they describe things as they are so that people can plan for change (Fink,
2013). More so, the measures of interest to this study, such as self-efficacy beliefs and sense of
belonging, relate to dynamic, social psychological processes that change. A cross-sectional
single-administration survey allowed the study to capture data on variables at the same point in
time to allow for meaningful investigation of association and patterns.
Self-Reported Survey Research
The purpose of survey research is to collect information directly from people to describe,
compare, or explain their knowledge, feelings, values, and behavior (Fink, 2013). Survey
research can also be used to generalize from a sample to a population so that inferences can be
made about some characteristic, attitude, or behavior of the population of study (Babbie, 1990;
Creswell, 2008). This strategy of inquiry was the preferred type of data collection given this
study’s research questions, scope, and nature of what is being measured.
A self-administered, questionnaire survey instrument (presented in Appendix B) was
utilized to collect self-reported input, environmental, and outcome variables including measures
assessing beliefs and motivation such as satisfaction and intent to persist. These measures are
not available in existing or historical data sources and are best derived from the subjects
themselves. Although there are some concerns with truthfulness or accuracy of self-reported
data, it is the only source for many variables of interest to this study.
Setting, Context, and Environment
The specific context of the study was the motivation to improve college outcomes for
Native Hawaiian students at the University of Hawaii. This single-institution study was focused
at the University of Hawaii at Manoa (UHM), the flagship campus of the state’s sole public
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

52
University of Hawaii System, where most of the Baccalaureate degrees in STEM are conferred.
UHM is classified by the Carnegie Foundation as “very high research activity,” ranked in the top
30 public universities in federal research funding for engineering and science and 49
th
overall by
the National Science Foundation. UHM is one of only 32 institutions nationwide to hold the
distinction of being a land-, sea-, and space-grant research institution.
UHM can be compared to “peer institutions” of like characteristics such as urban city
location (Honolulu), sector (Public, 4-year and above), basic Carnegie classification (research
universities, very high research activity), and size by enrollment (20,400). The University of
Hawaii at Manoa peer group listing, shown in Table 1 (UoHIRO, 2009), was determined by
National Center for Higher Education Management Systems (NCHEMS) Information Service
using variables including finance, degrees awarded, faculty, institutional characteristics,
professional judgment, and other student and research data from the national Integrated
Postsecondary Education Data System (IPEDS).
Institutional characteristics including undergraduate ethnic diversity, commitment to
Native Hawaiians, STEM colleges and majors, and this study’s sample are described in the
following sections.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

53
Table 1
University of Hawaiʻi at Manoa Peer Group
Institution City State
Colorado State University Fort Collins CO
Iowa State University Ames IA
Louisiana State University and Agricultural and Mechanical College Baton Rouge LA
Oregon State University Corvallis OR
The University of Utah Salt Lake City UT
University of California – Davis Davis CA
University of Georgia Athens GA
University of Kentucky Lexington KY
University of Missouri – Columbia Columbia MO
University of North Carolina at Chapel Hill Chapel Hill NC
University of Tennessee – Knoxville Knoxville TN
University of Virginia Charlottesville VA
University of Hawaii at Manoa Honolulu HI

Undergraduate Ethnic Diversity
The University of Hawaii at Manoa in many ways reflects the State of Hawaii’s ethnic
population with 71.2% of the undergraduate enrollment comprised of in-state students. Table 2
shows the Race/Ethnic Background of all undergraduate enrollments at UH Manoa in 2012
(UoHIRO, 2012). It is noted that UHM is a non-majority population, with similar proportions
from Mixed, Caucasian, Native Hawaiian, Japanese, and Other Asian race/ethnicity groups.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

54
Relative to national minority-serving institutions, UHM enrolls a low percentage of African
American (1.6%), Hispanic (2.0%), and Native American or Alaskan Native (0.3%) students but
a large percentage of Native Hawaiian and Pacific Islander students (17.4%).

Table 2
Ethnic Background of UHM Undergraduates, 2012
Ethnicity n %
Asian (Chinese, Filipino, Japanese, Korean, Other) 5,914 40.3
White 3,069 20.9
Native Hawaiian or Pacific Islander 2,556 17.4
Mixed 2,089 14.2
International 436 3.0
Hispanic 300 2.0
Black or African American 233 1.6
American Indian or Alaskan Native 46 0.3
Unknown 22 0.2
Total 14,665 100.0

Commitment to Native Hawaiians
Additional context for this study is that improved Native Hawaiian participation is a
stated strategic initiative of the institution. Native Hawaiian educational attainment is a
performance measure guiding the UH System in their 2008-2015 strategic plan to “position the
University of Hawaii as one of the world’s foremost indigenous-serving universities by
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

55
supporting the access and success of Native Hawaiians” (UoHVPAPP, 2008, p. 2). Increasing
degree attainment of Native Hawaiians (in STEM and non-STEM fields), is a performance goal
of the campus and ten-member University system.
STEM Colleges and Majors
Roughly one-third of the undergraduate study body at UHM major in STEM. UH Manoa
offers 93 Bachelor’s programs including 35 STEM majors. Undergraduate STEM majors fall
into academic departments within one of four STEM colleges shown in Table 3. Regarding
transferability of this study’s findings to other settings, it should be noted that the size (total
STEM undergraduate enrollment) of STEM programs at UH Manoa is moderate compared to
many other land-grant, public universities.

Table 3
Enrollment of UHM Undergraduates by STEM College, Spring 2014
College/School All Native Hawaiian
Natural Sciences 2170 226
Engineering 916 115
Tropical Agriculture 310 58
Ocean & Earth Sciences 117 12
Total 3,513 411

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

56
Sample
The population in this study was the undergraduate STEM majors at the University of
Hawaii at Manoa. An inclusive subset of this population that is of particular interest in this study
was Native Hawaiians in STEM. As shown in Table 3, for the Spring 2014, 411 Native
Hawaiians represent 12% of the overall STEM population.
The sample in this study consisted of the 638 respondents the survey. All enrolled UHM
undergraduate STEM majors were invited to participate in the study by completing an online,
self-reported survey. Of the 953 students who started the online questionnaire, 241 did not
submit their survey, 71 opted out of final participation, and 4 responses were excluded from the
sample because they did not meet the study criterion (3 had graduated and were not current
undergraduate STEM majors and one respondent had transferred to the business school, a non-
STEM major). The total number of participants was 638 for a study response rate of 17.7%.
Chapter 4 tests for differences between the sample and available information about the
population to inform the generalizability of the findings.
Instrumentation
This study utilized a 63-item, closed-ended questionnaire to collect data on the 18 input,
environment, and outcome variables (presented in Appendix B). No single existing instrument
tool was found to cover the range of variables sought to answer this study’s research questions,
necessitating the motivation to merge and modify multiple scales. Many of the subscales utilized
were from two source instruments in particular, the Academic Pathways of People Learning
Engineering Survey (APPLES) (Sheppard et al., 2010) and the Longitudinal Assessment of
Engineering Self-Efficacy (LAESE) (Marra & Bogue, 2006). The instrument was constructed by
a combination of a subscales of tested and published instruments focusing on motivation, self-
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

57
efficacy, and persistence of science or engineering students and a researcher-developed
demographics survey. A discussion of all variables and their measurement items are presented.
Input Variables
Table 4 summarizes the background variables to be collected in the demographics section
of the survey developed by the researcher.

Table 4
Input and Background Variables Collected
Variable Items Source
Gender 1 demographic
Ethnicity 2 demographic
Socioeconomic status 3 APPLES
Financial ability 1 APPLES
High School GPA 1 demographic
College 1 demographic
Major 1 demographic
Educational level 1 demographic
Incoming status 2 demographic

Gender was collected with a single dichotomous item on the demographics survey.
Ethnicity was collected with 2 items on the survey. Ethnicity response categories
followed the UH System application for admissions allowing students to select one or more
categories of African American, American Indian or Alaskan Native, Caucasian, Chinese,
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

58
Filipino, Guamanian/Chamorro, Hawaiian, Hispanic, Indian, Japanese, Korean, Laotian,
Micronesian, Other Asian, Other Pacific Islander, Samoan, Thai, Tongan, and Vietnamese.
Consistent with the UH System application form, a second survey item queried, “were any of
your ancestors Hawaiian?” As is consistent with UH Manoa race and ethnicity reporting
procedures, student respondents with more than one ethnicity were derived into multiple
ethnicity categories. For data analysis, ethnicity responses were hierarchically clustered into
categories as follows: (1) Native Hawaiian (any); (2) Mixed Asian for multiple ethnicity reports
that fall under the larger Asian grouping only (example: Japanese, Chinese, and Korean);
(3) Mixed Pacific Islander for multiple ethnicity reports that fall under the larger Pacific Islander
grouping only (example: Samoan and Tongan); (4) Mixed Race for multiple ethnicity reports
that fall across several larger groupings (example: Japanese and White); (5) Reported ethnicity
for single ethnicity reports; (6) Hispanic for self-indicated ethnicity report; and (7) Unknown for
unreported ethnicity.
Socioeconomic status was measured using 3 items from the APPLES instrument
querying: (1) Highest level of education mother completed; (2) Highest level of education father
completed; and (3) Self-perceived family income (low, lower middle, middle, upper-middle, or
high income). A fourth close-ended survey item taken from the APPLES instrument queried “Do
you have any concerns about your ability to finance your college education?” In SES component
analysis, the fourth item was found to load onto a separate construct, used in subsequent analysis
as “financial ability.”
Self-reported high school grade point average was reported by letter grade treated as a
continuous scalar variable from 1.3 to 4.0. Two categorical input variables were reported for
College and Major. Educational level was a single-item variable assessed self-reported
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

59
University class standing as: (1) Freshmen, (2) Sophomore, (3) Junior, (4) Senior, or (5) 5
th
year
Senior. Two additional items related to academic status queried “when you first entered this
institution, were you: (first-time, returning, transfer student from a two-year college, and transfer
student from a four-year college)” and “are you a full-time/part-time student?” and “where were
you immediately before starting at this institution used in analysis as “Pre-institution status.”
Environment Variables
This study investigated ten environmental factors: family support, program participation,
peer interaction, faculty interaction, faculty support, college GPA, sense of belonging to school,
major, and campus community, and satisfaction. The survey items and scales from extant
literature used to measure these independent variables are discussed.
Family support was measured with two-items taken from Cabrera et al. (1992) querying,
“My family approves of my attending this university” and “my family encourages me to continue
attending this university.” Reliability analysis, further described in Chapter 4, found good
internal consistency (α = .838) for this composite measure with respondent data.
Regarding program participation, one survey item prompted students to check any
programs that they have participated in from a list eleven minority STEM programs at UH
Manoa known to the researcher. Programs of interest include (1) C-MORE Scholars Program,
(2) Hui Manawa Kupono Native Hawaiian Scholarship Program, (3) Kua’ana Native Hawaiian
Student Services, (4) Minority Access to Research Careers, (5) Na Pua No’eau, (6) NHSEMP,
(7) Pacific Internship Programs for Exploring Science, (8) Undergraduate Research and
Mentoring in the Biological Sciences, (9) UH Manoa Honors Program, (10) Society for the
Advancement of Chicanos and Native Americans in Science, (11) American Indian Science &
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

60
Engineering Society. An open-ended response was available for respondents to write in
participation in other programs.
Interaction with peers was queried on frequency (not at all, occasionally, frequently) of
three items: studied with other students, tutored another college student, and worked on a group
project. Relationship with peers will be queried on frequency (not at all, occasionally, frequently)
of four items: “I worked cooperatively with other students on course assignments”; “I discussed
ideas with my classmates (individuals or groups)”; “I got feedback on my work or ideas from my
classmates”; and “I interacted with classmates outside of class.” The subscale of peer interaction
is used with permission from the Engineering – National Student Survey of Engagement (E-
NSSE) (Cady, Fortenberry, Drewery, & Bjorkland, 2009). In a pilot study of the E-NSSE
instrument, 261 students completed the test twice. Cady et al. (2009) found significant test-retest
Pearson’s coefficients on the individual items and α = 0.918 for relationships with peers. In this
study, reliability analysis of the 7-item peer interaction measure found good internal consistency
(α = .837).
Capturing faculty interactions, three items from the APPLES survey will be used to query
frequency (not at all, occasionally, frequently) of interaction with faculty and/or instructors
during class, during office hours, and outside of class or office hours. Inter-item reliability
measures for the faculty interaction construct for administration of APPLES to 4,266 students
were acceptable (Cronbach’s alpha = 0.74). In the current study, internal consistency was found
to be moderate (α=.609).
Cole and Ahmadi (2010) generated a factor “general faculty support” from 9 items (α =
0.90) including encouragement for graduate school, opportunity to work on research project,
advice about educational program, respect, emotional support/development, letter of
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

61
recommendation, intellectual challenge/stimulation, opportunity to discuss coursework outside of
class, and help in achieving professional goals (p. 130). In the current study, good internal
consistency was found for the composite 9-item measure (α = .826).
Based on scales of prior research on the construct, the instrument contained a total of 9-
items to measure sense of belonging to major, to the institution, and to the campus community.
Smith, Wilson, Jones, Plett, Bates, and Veilleux (2012) studied sense of belonging for more than
900 engineering and science undergraduates in four different settings to assess sense of
belonging to major and sense of belonging to the university as an institution. Analysis of the
study demonstrated strong internal reliability (Cronbach’s alphas ranged from 0.80 to 088).
Sense of belonging to major consisted of 3 items including “I feel accepted in my major,” “I feel
comfortable in my major,” “I feel that I am a part of my major.” Sense of belonging to the
university as an institution included three items: “I feel like I really belong at this school,” “I
really enjoy going to school here,” and “I wish I had gone to another school instead of this one
(reverse scoring).” Hurtado and Carter (1997) measured sense of belonging to campus subscale
(3 items, α = 0.94) measured “I see myself as a part of the campus community,” “I feel that I am
a member of the campus community,” and “I feel a sense of belonging to the campus
community.” For this study, reliability analysis found strong internal consistency with
Cronbach’s α values to school, major, and campus community of .731, .885, and .952
respectively.
The environmental variable satisfaction was based on 6 survey items of student’s ratings
of their satisfaction with quality of instruction, amount of contact with faculty, interaction with
peers, academic advising and student support, STEM major, and overall quality of their
collegiate experience so far. These items are taken with permission from the Academic Pathways
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

62
of People Learning Engineering Survey (APPLES). In this study, the 6-item scale demonstrated
good internal consistency (α=.842). College grade point average was measured by self-report to
letter grade ranging from D (1.3) to A (4.0).
Table 5 presents the ten environmental independent variables and two outcome variables
highlighting source instrument and internal consistency found in this study.

Table 5
Environment and Outcome Variables Collected
Variable Items Source Cronbach’s α
Family support 2 Cabrera et al. (1992) .838
Program participation 1 Researcher -
Peer Interaction 7 Cady et al. (2009), (α = .92) .837
Faculty Interaction 3 APPLES, (α =.70) .609
Faculty Support 9 Cole and Ahmadi (2010), α = .90 .826
College GPA 1 Researcher -
Belonging to Major 3 Smith et al. (2012), .80< α < .88 .885
Belonging to School 3 Smith et al. (2012), .80< α < .88 .731
Belonging to Campus 3 Hurtado & Carter (1997), α =.94 .952
Satisfaction 6 APPLES .842
Intent to Persist 1 APPLES -
STEM Self-Efficacy 8 LAESE, α = .82 .914

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

63
Outcome Variables
The purpose of this study was to investigate the associations between the input and
environmental measures on two outcomes of interest – intent to persist and STEM self-efficacy.
These two outcomes were chosen because of their influence on motivation, choice goals, and
actual persistence (Lent, 2013; Cabrera et al., 1992). Intent to persist was a single-item measure
queried students to rate their agreement on a 4-point scale (1 = strongly disagree and 4 = strongly
agree) to the prompt “I intend to complete a STEM degree at UH Manoa.”
Eight items from the Longitudinal Assessment of Engineering Self-Efficacy (LAESE)
engineering self-efficacy subscale (11-item scale, α = 0.82) to assess self-efficacy beliefs related
to completing a STEM degree at the institution. Wording of the LAESE items was modified
from ‘Engineering’ to ‘STEM major’ to include the science, technology, and mathematics
domains of interest in this study. For example, the item “I can succeed in an engineering
curriculum” was modified to “I can succeed in my STEM major curriculum.” The STEM self-
efficacy variable was comprised of eight items rating their agreement on a 4-point scale
(1=strongly disagree and 4 = strongly agree) to the following prompts:
1. I can succeed in my STEM major curriculum;
2. I can succeed in my STEM major curriculum while not having to give up
participation in my outside interests (e.g. extra curricular activities, family, sports);
3. I can complete the math requirements for my STEM major;
4. I can complete the science requirements for my STEM major;
5. I can excel in my current STEM major during the current academic year;
6. I can persist in my STEM major during the next year;
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

64
7. I can complete my STEM major at this institution;
8. I feel confident in my ability to complete a STEM degree at UH Manoa.
The survey instrument consisted of 63 items and gathered data found to load onto 21
factors. The median time for survey completion was 8 minutes.
Validity and Reliability Measures
Three procedures were utilized to enhance the ability of survey instrument to elicit valid
responses. First, survey items were directly taken or modified from existing published tools that
have been tested for reliability and validity. Then, aspects of the survey including directions that
accompany the survey and the ordering and wording of questions were critiqued by the
dissertation faculty chair. Expert review is a commonly accepted method for validating survey
instruments (American Educational Research Association [AERA], 1999). Lastly, the web-based
survey was tested with group of like individuals (graduate students and recent alumni) not in the
intended study sample. Pilot testing is a highly regarded step to improve validity and quality of
data collected (Suskie, 1996). Pilot participants (n=25) were instructed to interpret and complete
the survey from the perspective of an undergraduate student, then provide feedback on the clarity
of survey items and response options to the researcher. Modifications to survey language,
response options, layout, and elimination of three items were made based on the pilot results and
feedback.
Data Collection
This study utilized Qualtrics web-based software to create and administer the online
survey instrument. Web-based data collection was the desired form of data collection in this
study for many reasons. First, computer access, computer literacy, and online access was found
to be high among the target population (University of Hawai’i Office of the Vice President for
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

65
Student Affairs [UoHVPSA], 1999) supporting the samples’ candidacy survey completion.
Second, the web-based option allowed for rapid turnaround, and economy of design (Fink, 2013)
enabling a large sample size to be taken. Collection of data from the total population was
desirable because it would potentially increase the utility of the data and was a feasible option for
this study.
This study applied for and received Human Studies Program approval as exempt
(presented in Appendix A) from the University of Hawaii Human Studies Program and the USC
University Park Institutional Review Board. Email addresses for enrolled Spring 2014 STEM
undergraduates were granted to the researcher by the College of Engineering, Natural Sciences,
CTAHR, and SOEST for the purposes of this study. Students with more than one major were
classified by primary major.
Data collection from participants took place in February 2014, coinciding with week 4
through week 7 of the Spring semester. Collecting data from Spring enrollment would allow
first-year college students at least a semester to interpret their thoughts related to their college
environment and might reduce the number of first semester STEM leavers from the sample. An
initial email on February 6 was sent to all subjects (n=3,592) introducing the voluntary study and
inviting them to participate in the web-based survey via a unique link to Qualtrics. A total of
three reminder emails was sent to participants that did not complete the survey every six days.
The study utilized unique links within the Qualtrics Mailer to prevent multiple responses from
the same link. The intent was to collect data only from specific individuals (STEM
undergraduates directly emailed).
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

66
Data Analysis
Data analyses, conducted using SPSS Statistics v.21 software, consisted of descriptive
statistics, factor analysis, reliabilities, regression analysis, and analysis of variance techniques.
First, the data file of raw survey responses was prepared using Excel and SPSS software
and SPSS. Data were screened for errors and scores out of range. Some data used for composite
measures and reverse-scored data were recoded to provide for more meaningful and usable
variables for analysis. New variables were derived from raw data such as mixed ethnicity as well
as dichotomous variables such as intent to persist and program participation.
Preliminary analysis including frequencies and descriptives checked for mistakes due to
recoding or data entry. Tests for normal behavior and investigation of outliers were conducted to
ensure clean data. Principal factor analysis was conducted to confirm clustering effects leading to
higher order factors as predicted in the scales used from prior research. It is noted reliability of a
scale can vary depending on the sample (Pallant, 2013). Reliability tests investigated inter-item
consistency by inter-item factor correlations as well as overall correlation. Descriptive statistics
were utilized to organize, summarize, and describe the experiences of the sample in aggregate for
each individual or composite variable.
Subsequent analysis involved logistic and multiple regression to explore relationships
among input and environment variables with outcome variables STEM self-efficacy and intent to
persist, respectively. Sequential (or hierarchical) regressions were chosen (as opposed to
standard or step-wise) in order for the study to enter independent (input and environmental)
variables based on the theoretical grounds outlined in Lent’s (2013) Social Cognitive Career
Theory and Astin’s (1999) Inputs – Environment – Outcomes model and to interpret to
contribution of independent variables after prior variables were controlled.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

67
Analysis of variance (ANOVA) and multiple analysis of variance (MANOVA) were
chosen to explore differences between the Native Hawaiian and non-Hawaiian groups. These
were the ideal techniques to investigate whether the mean differences between groups were
likely to have occurred by chance.
Summary
This chapter described the methodology of this quantitative study including an overview
of the research design, approach, and context. Details of the instrumentation were provided with
discussion of the validity and reliability of measures used to assess each variable. Finally, the
data collection and data analysis were presented. Chapter 4 presents the analysis and results of
the study as they relate to the three research questions.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

68
CHAPTER 4
RESULTS AND ANALYSIS
The purpose of this study is to examine the factors associated with STEM self-efficacy
and intent to persist for Native Hawaiian and non-Hawaiian students. The goal is to better
understand the dynamics that correlate with student’s beliefs about their own capabilities and
commitment to complete a Bachelor’s degree in STEM. Data was collected on a range of
input/background, environment, and outcome measures from undergraduate STEM majors at the
University of Hawaii at Manoa to answer the following research questions:
1. What are the personal input and environmental factors associated with STEM self-
efficacy beliefs of undergraduate STEM students?
2. What are the personal input and environmental factors associated with intent to
persist in STEM of undergraduate STEM students?
3. How do these factors and outcomes differ, if at all, amongst Native Hawaiian and
non-Hawaiian students?
Chapter 4 will present the data analysis for the key variables and findings of the study as they
relate to the three research questions.
The results and analysis are organized into two sections. The first section of this chapter
describes the sample, data on variables collected, and construction of composite variables.
Descriptive statistics are presented to also display similarities and differences between non-
Hawaiian and Hawaiian groups. Results of factor analysis and reliability tests are presented on
multi-item variables to be used in subsequent analysis. The second part of the chapter presents
the analysis applied to answer the research questions. First, the results of sequential multiple
regression and sequential logistic regression models used to explore relationships between the
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

69
input and environmental variables to outcome variables STEM self-efficacy and intent to persist
are discussed. Then, the results from a series of analysis of variance and multivariate analysis of
variance used to explore the similarities and differences between non-Hawaiian and Hawaiian
group are presented. A summary of the major results concludes the chapter.
Participant Characteristics
The entire population of UHM undergraduate STEM majors was invited to participate in
the study (N=3592). Of the 953 students who started the online questionnaire, 241 did not
submit their survey, 71 opted out of final participation, and 4 responses were excluded from the
sample because they did not meet the study criterion (3 had graduated and were not current
undergraduate STEM majors and one respondent had transferred to the business school, a non-
STEM major). The total number of participants was 638 (N=638) for a study response rate of
17.7%. Sample characteristics are presented by gender, ethnicity, STEM college, and academic
major and are evaluated via a chi-square goodness of fit to test for significant difference from
existing demographic data on all UH Manoa STEM undergraduates. Environmental and
outcome variables and construction of multi-item variables were utilized.
Gender
Table 6 displays frequency and percentage of students in the sample by gender as well as
reference percentages derived from the total population of undergraduate STEM students at UH
Manoa (N=3592). The majority of the respondents in the sample were female (54.9%) versus
male (43.9%). A chi-square test for goodness of fit indicates that the sample was significantly
different from the population χ
2
(2, n=638) = 67.76, p = 0.000) noting that the reference STEM
undergraduate population has a higher percentage of males (59.4%) than found in the sample and
lower percentage of females (40.1%) than found in the sample.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

70
Table 6
Participant Gender (N = 638)
Gender Frequency % Reference %
Female 350 54.9 40.1
Male 280 43.9 59.4
Prefer not to answer/No Data 8 1.2 0.5
Total 638 100.0 100.0

Gender representation of the all undergraduate STEM majors varies by College with the
College of Engineering enrolling 80.2% male, College of Natural Sciences enrolling 53.6%
male, and the College of Tropical Agriculture and Human Resources (CTAHR) enrolling 57.7%
female. School of Ocean and Earth Sciences and Technology (SOEST) has a near balanced
gender enrollment of 58 men and 57 women. Table 7 presents gender characteristics of the
sample by College. The sample represents a greater percentage of females in all colleges than
expected from population parameters. Therefore, it should be noted that the results of this study
reflect a larger voice from females (+15% or 1.4x) when generalizing to the UHM population.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

71
Table 7
Sample Gender (N = 638) by College
Male Female Unreported
College Frequency % Frequency % Frequency %
Natural Sciences 101 33.3 198 65.3 4 1.3
Engineering 150 70.8 59 27.8 3 1.4
Tropical Agriculture 18 21.4 65 77.4 1 1.2
Ocean & Earth Sciences 7 23.3 23 76.7 0 0
Prefer not to answer 4 44.4 5 55.6 0 0
Total 280 43.9 350 54.9 8 1.2

Ethnicity
Table 8 displays frequency and percentage of students in the sample by ethnic group as
well as reference percentages of each ethnic group relative to the entire population of
undergraduate STEM students (N=3592). A chi-square goodness of fit indicates significant
difference in the ethnic composition of the sample as compared with Spring 2014 total STEM
enrollment ethnicity data, χ
2
(21, n = 637) = 14270.76, p = 0.000). Hawaiian respondents
(n=109) represent 17.1% of the sample, an overrepresentation of Hawaiians compared to the
reference percentage. Other groups overrepresented in the sample include Unreported and
Chinese. Groups underrepresented in the sample include White, Japanese, and Korean.
Overrepresentation and a sufficient subsample of Native Hawaiians are desirable for this study to
conduct statistically powerful analysis and investigation of the Native Hawaiian subgroup.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

72
However, it should be noted that the results of this study reflect a larger voice from Native
Hawaiians (1.5X) when generalizing to the UHM population.

Table 8
Participant Ethnicity (N = 638)
Ethnicity Frequency % Reference %
White 112 17.6 21.7
Native Hawaiian 109 17.1 11.7
Mixed (2 or more ethnicities) 96 15.0 15.4
Mixed Asian 72 11.3 10.9
Filipino 67 10.5 10.7
Chinese 64 10.0 8.1
Japanese 51 8.0 11.1
Unreported 17 2.7 .2
Korean 12 1.9 3.1
Hispanic 8 1.3 1.7
Vietnamese 5 .8 1.7
Black or African American 5 .8 .9
Samoan 4 .6 .3
Micronesian 3 .5 .2
Other 3 .5 0
Laotian 2 .3 .1
Other Asian 2 .3 .6
Tongan 2 .3 .1
American Indian or Alaskan Native 1 .2 .4
Asian Indian 1 .2 .1
Chamorro/Guamanian 1 .2 .3
Other Pacific Islander 1 .2 .2
Thai 0 0 .1
Mixed Pacific Islander 0 0 .1
Total 638 100.0 100.0

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

73
STEM College and Academic Major
Table 9 presents the frequency and percentage of students by college for both the sample
and population of all undergraduate STEM students. The sample was found to be significantly
different than the population with an underrepresentation of Natural Science students and a
larger than expected percentage of students from SOEST, CTAHR, and Engineering. Response
rates from the four different STEM Colleges were 16.0% from the College of Natural Sciences,
18.4% from the College of Engineering, 19.7% from the College of Tropical Agriculture and
Human Resources (CTAHR), and a high of 26% from the School of Ocean and Earth Sciences
and Technology (SOEST).

Table 9
Sample and Population by STEM College
Sample (N=638) Population (N=3592)
College Frequency % Frequency %
Natural Sciences 303 47.5 1896 52.8
Engineering 212 33.2 1154 32.1
Tropical Agriculture 84 13.2 427 11.9
Ocean & Earth Sciences 30 4.7 115 3.2
Prefer not to answer 9 1.4 0 0.0
Total 638 100.0 3592 100.0
Note. χ
2
(3, n = 629) = 9.016, p = 0.029).

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

74
This study invited participation from all enrolled students in any one of 34 undergraduate
STEM majors offered at UH Manoa. The complete frequency and percentage of student by
academic major for both the sample and population is presented in Appendix D: Table 40. Table
10 presents a reduced set of data for the highest enrolled and lowest enrolled majors. It is noted
that Biology is the highest enrolled academic major representing 20.6% of the total STEM
undergraduates followed by Mechanical Engineering (8.6%), Civil & Environmental
Engineering (8.3%), Marine Biology (8.0%), and Electrical Engineering (6.1%). Similarly, the
highest number of sample respondents came from 5 highest enrolled academic majors. There
were less than 10 respondents in the sample for 12 of the smaller academic majors and no
respondents from Geology, Environmental Studies/Interdisciplinary Studies, or Pre-Physical
Therapy in the sample.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

75
Table 10
Sample and Population by Highest and Lowest Enrolled STEM Academic Majors
Sample (N=638) Population (N=3592)
Major Frequency % Frequency %
Biology 111 17.4 741 20.6
Civil & Environmental Engineering 71 11.1 298 8.3
Mechanical Engineering 69 10.8 309 8.6
Electrical Engineering 43 4.7 219 6.1
Marine Biology 35 5.5 286 8.0
Computer Engineering 9 1.4 76 2.1
Other 9 1.4 0 0
Biological Engineering 6 .9 41 1.1
Meteorology 6 .9 23 .6
PEPS 5 .8 21 .6
Geology & Geophysics 5 .8 35 1.0
Plant & Environmental Biotechnology 4 .6 18 .5
TPSS 4 .6 42 1.2
Pre-Medicine 4 .6 0 0
Botany 3 .5 25 .7
Ethnobotany 3 .5 20 .6
Molecular Biosciences & Biotechnology 2 .3 0 0
Geology 0 0 8 .2
Prefer not to answer 2 .3 0 0.0
Total 638 100.0 3592 100.0
Note. Students enrolled in one or more academic majors (double majors) were classified into
their primary major for this study. GES = Global Environmental Sciences; ICS = Information &
Computer Sciences; NREM = Natural Resources & Environmental Management; PEPS = Plant
and Environmental Protection Sciences; TPSS = Tropical Plant and Soil Sciences.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

76
Academic Level and Pre-UH Manoa Status
In addition to the gender, ethnicity, college, and major, the research model tested
variation of self-efficacy and intentions to persist by academic level and pre-institution status.
Frequency counts of the sample by academic level and pre-UH Manoa status are presented in the
Appendix D: Table 41. Combined, Seniors and 5
th
year seniors represent the largest group
(34.6%), followed by Juniors (30.4%), Sophomores (17.7%), and Freshmen (14.6%). The
majority of the respondents (61.9%) defined themselves as first-time college students when they
first entered UH Manoa. Regarding incoming students status, 22.1% of respondents classified
themselves as transfer students from any 2-year college (although many students wrote in
comments regarding their UH community college), 9.7% as transfers from a 4-year college, and
5.6% as returning or non-traditional students. The majority of the sample (61%) was classified as
first-time clean freshmen enrolling at UHM from high school.
Input Characteristics by Native Hawaiian Status
Data for all variables are also presented by Hawaiian and non-Hawaiian groups. This
will inform the analysis, results, and potentially recommendations. For example, the study may
find some key differences that allow Native Hawaiian serving programs to better customize their
approach, whereas similarities should be noted such that interventions to affect Native Hawaiians
will affect all students. Table 11 presents counts and percentages of Native Hawaiian status by
gender, college, educational level, and incoming student status (first-time clean freshmen,
transfer student, etc.). Not shown in Table 11 are percentages within variable, however, data
revealed highest percentages of NH respondents within College of Engineering (22.0%),
followed by CTAHR (21.7%), SOEST (16.7%), and Natural Sciences (12.7%).

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

77
Table 11
Native Hawaiian Status Count and Frequency by Input Variable
Non-HW HW
Variable n % n %
Gender (N=627)
Female* 301 58.2 48 43.6
Male* 216 41.8 62 56.4
College (N=622)
Engineering* 163 31.7 46 43.0
Natural Sciences* 262 50.9 38 35.5
Tropical Agriculture 65 12.6 18 16.8
SOEST 25 4.9 5 4.7
Level (N=615)
Freshmen 82 16.1 11 10.3
Sophomore 93 18.3 2 18.7
Junior 162 31.9 31 29.0
Senior 106 20.9 26 24.3
5
th
year Senior 65 12.8 19 17.8
Incoming student status (N=628)
First time college student 331 63.9 62 56.4
Community college transfer 107 20.7 32 29.1
4-year University transfer 50 9.7 11 10.0
Returning or non-traditional 30 5.8 5 4.5
Note. * column proportions differ from each other at the .05 level.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

78
A Chi-square test for independence (with Yates Continuity Correction for gender)
indicated a significant association between Native Hawaiian status and gender, χ
2
(1, n=627) =
7.237, p = 0.007, phi = .112, and for college, χ
2
(3, n=622) = 8.902, p = 0.031, phi = .120. In
other words, there were more Hawaiian males than Hawaiian females represented in the sample
compared to more females than males in the non-Hawaiian group. Similarly, Hawaiian
respondents were found distributed in higher levels in Engineering than non-Hawaiians, and in
lower levels in Natural Sciences than non-Hawaiians. No significant association was found for
educational level, χ
2
(4, n=615) = 4.33, p = 0.363, phi = .084, or for incoming status, χ
2
(3,
n=628) = 4.004, p = 0.261, phi = .080. It is noted that a higher percentage of Native Hawaiians
as compared to Non-Hawaiians in the sample came from two-year community colleges.
High School and College GPA
Of the respondents 613 provided a cumulative high school GPA by category (A-, B+, B,
etc.) and 25 (4.9%) did not provide a GPA. A value of 1.30 was assigned to one student who
indicated a GPA of “D+ or lower (less than 1.4).” High school GPA of A- (3.5 – 3.8) was the
most common response, followed by A/A+, and B+ categories. As is sometimes seen in research
with self-reported GPA, this variable exhibited negative skew. Of the respondents, 618 provided
a College GPA by category and 20 (3.1%) did not. Table 12 shows descriptive statistics for GPA
items.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

79
Table 12
Descriptive Statistics for GPA Items
GPA Item
Total
M (SD)
Non-HW
M (SD)
HW
M (SD)
HS GPA* (N = 613) 3.55 (.46) 3.58 (.43) 3.44 (.58)
College GPA (N = 618) 3.18 (.52) 3.20 (.53) 3.10 (.53)
Note. * Non-HW and HW groups significantly differ at the p <0.05 level.

Independent t-tests were conducted to explore if GPA variables varied by Native
Hawaiian status (non-Hawaiian vs. Hawaiian group). There was no significant difference found
in College GPA t(612)=1.708, p = .088, two-tailed. There was, however, significant difference
found in HS GPA t(609) = 2.309, p = .023, two-tailed. Inspection of means show Non-HW
students reported a higher GPA than HW students. The magnitude of the differences in the
means (mean difference = .14, 95% CI: .02 to .26) was very small (eta
2
= .009).
Participation in Program
The majority of the sample (66%) did not indicate participation in any academic
preparation program, 21.2% participated in one program, 6.3% in two programs, and 6.6% or 52
students in 3 to 5 programs. Participation in the UHM Honors Program was identified by 105
respondents (16.5%) and 82 students (12.9%) indicated participation in the Native Hawaiian
Science & Engineering Mentorship Program. Table 13 displays descriptive data for program
participation by count and percentage. Thirty-one students (4.9%) indicated participation in a
program not listed. Other programs identified by respondents included ACE learning
communities (3), ASME (3), Marine Option Program (3), IKE (2), IEEE (2), Undergraduate
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

80
Research Opportunities Program (2), HNEI research, Hawaii Space Grant Consortium
fellowship, College Opportunities Program, McNair Scholars Program, EPSCoR Hawaii
Program, JABSOM DMEAP, TASSO, Trio, and INBRE.

Table 13
Frequency of Program Participation (N = 638)
Program n %
None or prefer not to answer 421 66.0
UHM Honors Program 105 16.5
NHSEMP 82 12.9
AISES 42 6.6
Other 31 4.9
Kuaana Native Hawaiian Student Services 29 4.5
SACNAS 22 3.4
Na Pua No’eau 22 3.4
Undergraduate Research and Mentoring (URM) in Biological Sciences 13 2.0
C-MORE Scholars Program 8 1.3
Hui Manawa Kupono Native Hawaiian Scholarship Program 7 1.1
MARC 5 .8
PIPES 4 .6
Note. NHSEMP = Native Hawaiian Science & Engineering Mentorship Program; AISES =
American Indian Science & Engineering Society; SACNAS = Society for Advancement of
Chicanos and Native Americans in Science; MARC = Minority Access to Research Careers;
PIPES = Pacific Internship Program for Exploring Science.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

81
Native Hawaiians participated in at least one program (67.9%) at a higher rate than non-
Hawaiians (26.9%). This was found to be a significant difference, χ
2
(1, n=637) = 67.76, p <
0.001.
Level of Parental Education, Self-Reported Family Income, and Financial Ability
Descriptive statistics for SES variables and financial ability are presented in Table 14.
Level of parental education was recoded for analysis for less than high school (.14), graduated
from high school (.29), some college (.43), Associates degree (.57), Bachelor’s degree (.71),
Master’s degree (.86), or professional or doctoral degree (1.0). Self-reported family income was
recoded as 0 for low income, .25 for lower-middle income, .50 for middle income, .75 for upper-
middle income, and 1.0 for high income. Financial concern was reverse scored such that higher
measures denote higher financial ability to allow for easier interpretation with the SES measures.

Table 14
Descriptive Statistics for SES Items
SES Item
Total
M (SD)
Non-HW
M (SD)
HW
M (SD)
Mother’s education (N=610) .560 (.229) .563 (.232) .546 (.213)
Father’s education (N=592) .540 (.243) .542 (.247) .527 (.222)
Perceived family income (N = 607) .433 (.233) .433 (.234) .416 (.233)
Financial ability (N=623) 2.51 (.878) 2.547 (.864) 2.373 (.937)
Composite SES (N=595) .479 (.186) .482 (.189) .466 (.174)

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

82
No significant differences between non-HW and HW groups were found for perceived
family income, level of parental education, the calculated SES scale (a composite of parental
income and self-reported income), or financial ability. It is noted, however, that the non-HW
group exhibited higher means relative to the HW group on all measured items of SES. As
presented in Table 14, both non-HW and non-HW groups have some degree of financial concern
with means of 2.55 and 2.37 respectively falling between 3 - “Some, I probably will have
sufficient funds” and 2 – “Major, I have funds but will graduate with significant debt.” 10.7% of
Non-HW and 20.2% of HW students rate their concern, 1 – “Extreme – Not sure if I will have
sufficient funds to complete college.” Table 15 presents further descriptive analysis of SES data
by NH status.

Table 15
SES Characteristics by Native Hawaiian Status
Variable non-HW HW
Mother’s education, % reporting mother with Bachelor’s or higher: 46.4 40.4
Father’s education, % reporting father with Bachelor’s or higher: 41.1 34.9
Financial ability, % who report extreme concern 10.7 20.2

SES Factor Analysis and Reliability Analysis
The grouping and reliability of the SES scale to be used in subsequent analysis was tested
using SPSS version 22. Factor analysis was conducted on the three-items related to SES and 1-
item financial concern as a technique to investigate any grouping or ‘clumps’ among the set of
variables. A potential four-item SES scale was subjected to principal component analysis (PCA)
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

83
using oblimin rotation. The Kaiser-Meyer-Olkin value of 0.615 met the recommended minimum
value of .6 (Kaiser, 1970; Kaiser & Rice, 1974; Pallant, 2013), the correlation matrix found
many coefficients of .3 and above, and the Bartlett’s Test of Sphericity (Bartlett, 1954) reached
statistical significance. Given that the sample size (n=638) and ratio of participants to items
exceeded minimum recommendations in the literature (Nunnally, 1978; Tabachnick & Fidell,
2007), the researcher found the SES data suitable for factor analysis.
Two components were found with eigenvalues exceeding 1, explaining 47.5% and 25.5%
of the variance respectively. Table 16 displays the component matrix for the 2-component
solution showing strong unrotated loadings of 3 items for component 1 and a single item for
component 2. A component correlation value of .248 suggests a weak correlation between the
two components.

Table 16
Factor Analysis Component Matrix for SES
Component
Variables 1 2
Highest level of education completed by Mother .728 -.481
Highest level of education completed by Father .766 -.339
Self-reported family income .730 .316
Concerns about ability to finance college .501 .757

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

84
The factor analysis did not support the inclusion of the item financial ability into the SES
scale given that it was loading onto a separate (r=.248) component from the 3-item SES scale.
Thus, the study treated financial ability as a separate, single-item measure.
A reliability analysis of the 3-item SES scale found low to moderate internal consistency
with a Cronbach’s alpha = .649. A slightly higher alpha (.661) could be achieved if the self-
reported family income item was deleted. Both respondent self-report and highest completed
level of parent’s education are considered acceptable proxies for SES of college students
(Entwisle & Astone, 1994; Walpole, 2003). In this study, however, inter-item correlation
between self-reported family income to mother’s education and father’s education was .325 and
.323 respectively. Some researchers (Briggs & Cheek, 1986; Clark & Watson, 1995) advocate
for an optimal level of homogeneity when the mean inter-item correlation falls in the .2 to .4
range if the items target diverse measures defining a latent construct rather than emerge from it.
Donaldson, Lichtenstein, and Sheppard (2008) utilize a combined approach for approximating
SES giving equal weight to respondent judgment and the traditional literature-grounded parent’s
education levels. This study retained the 3-item SES scale consistent with the method of prior
studies (Cady et al., 2009; Donaldson et al., 2008) that believe a more reliable estimate of SES is
obtained using measures based on research (parental education) and measures based on self-
perception (self-reported income). Socio-economic status was determined using a composite
measure ranging from 0 to 1.0.
Family Support
Respondents reported high levels of family support with mean response rates between 3
(agree) and 4 (strongly agree) to items ‘my family approves of my attending this university’ (M
=3.45, SD = .57) and ‘my family encourages me to continue attending this university’ (M = 3.38,
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

85
SD = .65). A reliability analysis showed good internal consistency (Cronbach’s α = .838) for the
2-item composite measure. Inter-item correlation was 0.727 with N=622 following list wise
deletion.

Table 17
Descriptive Statistics for Family Support Items
Family Support Item
Total
M (SD)
Non-HW
M (SD)
HW
M (SD)
Approves of my attending this university** 3.45 (.57) 3.43 (.58) 3.63 (.51)
Encourages me to continue attending* 3.38 (.65) 3.36 (.67) 3.51 (.52)
Composite Family Support ** 3.42 (.57) 3.34 (.58) 3.58 (.47)
Note. Non-HW and HW groups significantly differ at: * p <.05, **p <0.01 level.

A significant difference of means was found in scores for Family Support (a composite
measure of items my family approves of my attending this university and my family encourages
me to continue attending this university) between non-HW (M=3.398, SD = .580) and HW
(M=3.576, SD=.474; t(616)=-3.362, p = .001, two-tailed. The magnitude of differences in the
means (mean difference = -.18, 95% CI: -.282 to -.073) was a very small effect (eta squared =
.018). It is noted that HW group exhibited higher mean scores on family approval to attend UHM
and family encouragement to continue attending UHM.
Peer Interaction
Table 18 displays descriptive data related to peer interaction/support for responses to the
prompt “how often do you do the following activities?” ranging from 1-not at all, 2-occasionally,
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

86
and 3-frequently. Inspection of means show that HW group reported higher scores (more
frequent peer interaction) than non-HW group on all seven items. Independent t-tests were
conducted and found significant differences among groups for five items. The magnitudes of the
differences in the means, however, were found to be very small with eta
2
values ranging from
.006 to .009.

Table 18
Descriptive Statistics for Peer Interaction Items
Peer Interaction Item
Total
M (SD)
Non-HW
M (SD)
HW
M (SD)
Discussed ideas 2.46 (.60) 2.44 (.60) 2.50 (.59)
Interacted outside of class* 2.44 (.62) 2.42 (.63) 2.55 (.58)
Worked cooperatively on assignments* 2.38 (.61) 2.35 (.61) 2.48 (.62)
Studied with other students* 2.30 (.68) 2.26 (.68) 2.42 (.66)
Got feedback on my work and ideas* 2.21 (.65) 2.18 (.65) 2.33 (.67)
Worked on a group project* 2.16 (.66) 2.14 (.67) 2.27 (.65)
Tutored another college student 1.67 (.71) 1.64 (.71) 1.72 (.71)
Note. * Significant at the p <0.05 level.

A factor analysis (PCA method) was applied to the peer interaction data with the intent to
confirm that the items were grouped appropriately. A Kaiser-Meyer-Olkin measure of .866 and
statistical significance reached of Bartlett’s test of sphericity showed the data was suitable for
factorability. The component matrix is shown in Table 19. A single component was found with
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

87
an eigenvalue of 3.672 explaining 52.5% of the variance. The results of the factor analysis
confirmed the utility of all 7-items for a single peer interaction scale.

Table 19
Factor Analysis Component Matrix for Peer Interaction
Variables Component #1
Discussed ideas with classmates .844
Worked cooperatively with other students on course assignments .804
Got feedback on my work and ideas from classmates .769
Interacted with classmates outside of class .747
Studied with other students .728
Worked on a group project .638
Tutored another college student .477

A reliability analysis of the 7-item peer interaction scale was conducted. A list-wise
deletion based on all variables in the procedure excluded 11 missing cases for an N of 627. A
Cronbach’s alpha of 0.837 for the 7-item scale demonstrated good internal consistency. The
Cronbach’s alpha would increase to 0.851 if item “tutored another student” were deleted,
however, this study chose to keep the item in the overall scale to be consistent with the findings
from the literature (Cady et al., 2009). A composite measure for Peer Interaction ranging from 7
to 21 consisted of a sum of the seven items (M=15.63, SD=3.226).

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

88
Faculty Interaction
Three items on the survey asked students to rate the frequency of interaction with
instructors on 3-point likert scale ranging from not at all, occasionally, and frequently as well as
the option “I prefer not to answer.” Students reported higher than “occasional” interaction
during class, less during office hours, and least outside of class and office hours. Students of the
HW group were slightly more likely than non-HW group to interact with faculty during class and
during office hours, but less likely to interact with faculty outside of class and office hours based
on mean scores. However, analysis showed no statistical significance between groups of faculty
interaction.

Table 20
Descriptive Statistics for Faculty Interaction Items
Faculty Interaction Item
All
M (SD)
Non-HW
M (SD)
HW
M (SD)
During class 2.09 (.62) 2.08 (.64) 2.11 (.57)
During office hours 1.71 (.60) 1.70 (.60) 1.73 (.64)
Outside of class or office hours 1.59 (.64) 1.60 (.64) 1.59 (.61)

A direct oblimin principal components analysis (PCA) was conducted to check the
underlying nature of the three faculty interaction items. The Kaiser-Meyer-Olkin value was .631
and Bartlett’s test of sphericity reached statistical significance supporting the factorability of the
items. A single component was found with an eigenvalue exceeding 1, explaining 56.4% of the
variance. Table 21 revealed the three items loading onto a single component.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

89
Table 21
Factor Analysis Component Matrix for Faculty Interaction
Variables Component 1
During class .708
During office hours .791
Outside of class or office hours .752

According to APPLES, the frequency of interaction with instructors scale demonstrated
good internal consistency with a Cronbach’s α of 0.74. In the current study, the Cronbach’s
alpha coefficient was .609. Ideally, a Cronbach’s alpha greater than .7 is desired (DeVellis,
2012). Alpha values are sensitive to the number of items on a scale and with short scales (less
than 10 items) it is common to find low Cronbach’s alpha values. In short scales, a mean inter-
item correlation for the items may be a more appropriate measure with an optimal range of inter-
item correlation of .2 to .4 (Pallant, 2013; Briggs & Cheek, 1986). The inter-item correlation
means for this study were, .284, .345 and .400.
Although reliability analysis showed moderate internal consistency (α = .609), no
increase to Cronbach’s α values would be achieved if any of the three items were deleted. The 3-
items were retained giving a composite measure of faculty interaction ranging from 3 to 9,
consistent with the use of the faculty interaction scale in prior studies (Sheppard et al., 2010).
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

90
Faculty Support
Table 22 displays the means and standard deviations for responses to the prompt “how
often do you receive the following from your instructors?” ranging from 1-not at all, 2-
occassionally, and 3-frequently.

Table 22
Descriptive Statistics for Faculty Support Items (N = 566)
Faculty Support Item
All
M (SD)
non-HW
M (SD)
HW
M (SD)
Respect 2.55 (.58) 2.56 (.57) 2.56 (.60)
Intellectual challenge and stimulation 2.44 (.65) 2.45 (.65) 2.47 (.62)
Discuss coursework outside of class 2.25 (.73) 2.25 (.73) 2.29 (.72)
Help in achieving professional goals 1.84 (.71) 1.84 (.72) 1.81 (.67)
Emotional support/development 1.77 (.72) 1.81 (.71) 1.70 (.75)
Advice about educational program 1.66 (.68) 1.67 (.69) 1.61 (.62)
Opportunity to work on a research project 1.60 (.67) 1.62 (.68) 1.56 (.68)
Encouragement for graduate school 1.60 (.67) 1.60 (.68) 1.65 (.66)
Letter of recommendation* 1.59 (.67) 1.55 (.66) 1.71 (.71)
Note. * p < .05.

Students of the HW group were slightly more likely (by inspection of means) than non-
HW group to report higher levels faculty support based on frequency of intellectual challenge
and stimulation, opportunity to discuss coursework outside of class, encouragement for graduate
school, and letter of recommendation. Students of the HW group were less likely to report higher
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

91
levels of faculty support based on frequency of help in achieving professional goals, emotional
support/development, advice about educational program, and opportunity to work on a research
project. However, t-test analysis showed no statistical significance between groups on all items
except for frequency of letter of recommendation. Receipt of letter of recommendation was rated
lowest of all faculty support items (between ‘1- not at all’ and ‘2 - occasionally’) and
significantly lower for Non-HW (M=1.55, SD = .664) than for HW (M=1.71, SD=.708; t(590)=-
2.269, p = .024, two-tailed). The magnitude of differences in the means (mean difference = -
.161, 95% CI: -.301 to -.022) was a very small effect (eta squared = .009). Both the non-HW and
HW group reported equivalent and highest levels of faculty support in terms of “respect” of 2.55
ranging between 2-occasionally and 3-frequently.
Factor analysis (PCA with Oblimin rotation) revealed the potential to split scale into two
subscales with one component grouped around academic support and the second component
grouped around affective and emotional support. Table 23 displays the pattern matrix showing
the loadings onto the two subscales.
Reliability analyses were conducted to evaluate the two faculty support subscales and a
composite 9-item scale. A Cronbach’s alpha of .797 was found for the 5-item subscale based on
professional support demonstrating good internal consistency. A Cronbach’s alpha of .687 was
found for the 4-item subscale based on affective and personal support. No improvements to
internal consistency for the two subscales would be made if any item were deleted. A list-wise
deletion based on all variables in the procedure excluded 72 missing cases for an N of 566. A
Cronbach’s alpha of 0.826 for the 9-item scale demonstrated good internal consistency. To
remain consistent with prior studies using the 9-item faculty support scale, the composite
measure was chosen for overall faculty support measure.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

92
Table 23
Factor Analysis Pattern Matrix for Faculty Support
Component
Variables 1 2
Opportunity to work on a research project .836 -.138
Advice about educational program .815 -.012
Encouragement for graduate school .784 -.031
Help in achieving professional goals .564 .393
Letter of recommendation .489 .190
Intellectual challenge and stimulation -.067 .789
Respect -.063 .743
Opportunity to discuss coursework outside of class .057 .716
Emotional support/development .308 .481

Sense of Belonging
On measures of sense of belonging, respondents rated items, generally, in high
agreement. Highest rated items were “enjoy going to school here” with 90.2% of the sample in
agreement or strong agreement, followed by “feel accepted in my major” and “feel comfortable
in my major” over 83% of the sample in agreement or strong agreement. Respondents felt
highest sense of belonging to items dealing with major, lower sense of belonging to items
dealing with school/institution, and the lowest sense of belonging to campus community.
Approximately 45% of the respondents did not see themselves as a part of the campus
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

93
community, did not feel they are a member of the campus community, and did not feel a sense of
belonging to the campus community.
No significant differences were found for scores in sense of belonging between non-HW
and HW groups. It is noted that the HW group exhibited higher mean scores for belonging to
school, and belonging to major than the Non-HW group and a lower means score for sense of
belonging to campus community than the Non-HW group.

Table 24
Descriptive Statistics for Sense of Belonging Items
All non-HW HW
Sense of Belonging Item
% agree or
agree strongly M (SD) M (SD)
Enjoy going to school here 90.2 3.14 (.60) 3.18 (.56)
Feel like I really belong at this school 77.8 2.92 (.72) 3.01 (.63)
Wish I had gone to a different school
a
59.4 2.58 (.90) 2.65 (.88)
Feel accepted in my major 84.6 3.10 (.70) 3.16 (.65)
Feel comfortable in my major 83.0 3.10 (.70) 3.14 (.69)
Feel a part of my major 72.8 2.92 (.75) 2.92 (.78)
See myself as a part of the campus community 54.8 2.58 (.83) 2.51 (.78)
Feel I am a member of the campus community 54.0 2.56 (.83) 2.48 (.79)
Feel a sense of belonging to the campus community 54.5 2.56 (.83) 2.50 (.76)
Note.
a
item scores were reverse-coded.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

94
The 9-item sense of belonging scale was subjected to principal component analysis
(PCA) using SPSS version 22. Prior to performing PCA, the suitability of data for factor analysis
was assessed by inspection of the correlation matrix for coefficients above .3, the Kaiser-Meyer-
Olkin value (.835), and statistical significance of Bartlett’s Test of Sphericity.
PCA found three components with eigenvalues above 1, explaining 50.4%, 18.0%, and
12.7% of the variance respectively. Inspection of the scree plot revealed a break after the third
component. Parallel Analysis showed three components with eigenvalues above the
corresponding criterion values for a randomly generated data matrix on the same size (9
variables x 560 subjects). Following the results of analysis, it was decided to extract three
components.
The three-component solution explained a total of 81.2% of the total variance. The
correlation coefficients for the three components were .331, .376, and .383 supporting the use of
the Oblimin rotation solution. The pattern matrix is displayed in Table 25.
Consistent with the theoretical model, factor analysis revealed three, 3-item subscales
consisting of a sense of belonging to school, to major, and to campus community. Reliability
analysis found strong measures of internal consistency with Cronbach’s α values to school,
major, and campus community of .731, .885, and .952 respectively.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

95
Table 25
Factor Analysis Pattern Matrix for Sense of Belonging
Component
Variables 1 2 3
Enjoy going to school here .110 .182 .689
Feel like I really belong at this school .316 .081 .638
Wish I had gone to a different school
a
-.117 -.089 .915
Feel accepted in my major .011 .899 -.017
Feel comfortable in my major -.125 .951 .042
Feel a part of my major .116 .847 -.023
See myself as a part of the campus community .957 .020 -.024
Feel I am a member of the campus community .984 -.050 -.014
Feel a sense of belonging to the campus community .899 .035 .072
Note.
a
item scores were reverse-coded.

Satisfaction
Table 26 presents descriptive statistics for six satisfaction items and one composite item.
Satisfaction, in general, was not very high or very low. Overall, the sample rated mean
satisfaction levels between neutral (2), and satisfied (3) with mean scores ranking from highest
for interaction with peers (2.83), STEM majors (2.64), overall quality of collegiate experience
(2.63), academic advising and student support (2.49), quality of instruction (2.48) and amount of
contact with faculty (2.44). Satisfaction with interaction with peers garnered the highest ratings
with 70.0% of respondents satisfied or very satisfied. For items contact with faculty, academic
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

96
advising/student support, STEM major, and quality of instruction only a little more than 50%
were satisfied or very satisfied, the rest being neutral or dissatisfied. The highest percentages of
dissatisfied respondents were with academic advising and student support (17.6%), quality of
instruction (15.2%), and amount of contact with faculty (12.9%). A noticeable number of non-
response/prefer not to answer responses came from “STEM major” prompt indicating a possible
hesitation to rate satisfaction with STEM major or a misunderstanding of what was being asked.
46 participants did not respond to the STEM major item, where as five or less participants did
not respond to the other five satisfaction items.

Table 26
Descriptive Statistics for Satisfaction Items
All non-HW HW
Satisfaction Item
% satisfied or
very satisfied M (SD) M (SD)
Academic advising and student support 53.9 2.49 (.92) 2.51 (.93)
Contact with faculty 48.7 2.46 (.81) 2.39 (.85)
Quality of instruction* 54.6 2.52 (.84) 2.33 (.90)
STEM major 53.3 2.62 (.79) 2.78 (.77)
Overall quality of experience 61.5 2.62 (.81) 2.72 (.77)
Interaction with peers* 70.0 2.80 (.78) 2.98 (.73)
Composite Satisfaction - 15.57 (3.69) 15.74 (3.70)
Note. Non-HW and HW groups significantly differ at: * p <.05, **p <0.01 level.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

97
Students of the HW group were slightly more likely (by inspection of means) than non-
HW group to report higher levels of satisfaction for items academic advising and student
support, STEM major, overall quality of collegiate experience, interaction with peers, and the
composite satisfaction variable (sum of all six items). Students of the HW group were less likely
to report higher levels of satisfaction for amount of contact with faculty and quality of
instruction. Significant difference was found by independent t-test between groups for quality of
instruction (t(630) = 2.156, p = .031, two-tailed) and interaction with peers ( t(626) = -2.389, p =
.018, two-tailed). The magnitude of differences in the means for quality of instruction (mean
difference = .192, 95% CI: .17 to .367) and interaction with peers (mean difference = -.185, 95%
CI: -3.44 to -.26) were very small (eta
2
= .007 and .009, respectively).
A factor analysis (PCA method) was applied to the satisfaction data with the intent to
confirm that the items were grouped appropriately. A Kaiser-Meyer-Olkin measure of .852 and
statistical significance reached of Bartlett’s test of sphericity showed the data was suitable for
factorability. The component matrix is shown in Table 27. A single component was found with
an eigenvalue of 3.359 explaining 56.0% of the variance. The results of the factor analysis
confirmed the utility of all 6-items for a single satisfaction scale.
A reliability analysis for the scale followed a listwise deletion based on all variables in
the procedure excluding 53 missing cases for an n of 585. A Cronbach’s alpha of 0.842 for the 6-
item scale demonstrated good internal consistency. A composite measure for Satisfaction ranging
from 6 to 24 consisted of a sum of the six items (M=15.56, SD=3.698).

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

98
Table 27
Factor Analysis Component Matrix for Satisfaction
Variables Component #1
Overall quality of collegiate experience so far .823
STEM major .772
Amount of faculty contact .766
Quality of instruction .755
Academic advising and student support .700
Interaction with peers .662

STEM Self-Efficacy
Table 28 displays descriptive data for eight STEM self-efficacy items and one intent to
persist item. Overall, responses were very high with highest percentages of agreement or strong
agreement to items ‘succeed in major curriculum’ (83.1%), ‘can complete science requirements’
(83.0%), ‘can complete math requirements’ (82.9%), and ‘can complete major at this institution’
(82.1%). It is noted that responses to item ‘can excel this semester’ should be interpreted with
caution given that excelling in the current semester may describe a value judgment beyond that
the intended scope of persisting in a STEM major. The lowest rated item was ‘can succeed while
not giving up outside interests’ reflected 55.3% of the sample in agreement or strong agreement.
A noticeable percentage of the sample (10.8% - 16.5%) did not provide an answer to the set of
items related to persistence and ability to persist with a high of 102 respondents preferring not to
answer to the item “I can persist in my STEM major during the next academic year.” The intent
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

99
to persist item was rated very high with 83.5% of respondents indicating agreement or strong
agreement.

Table 28
Descriptive Statistics for STEM Self-Efficacy Items and Intent to Persist
All non-HW HW
STEM self-efficacy Item
% agree or
agree strongly M (SD) M (SD)
Succeed in major curriculum 83.1 3.37 (.63) 3.41 (.57)
Not giving up outside interests 55.3 2.73 (.92) 2.73 (.91)
Can complete math requirements 82.9 3.38 (.68) 3.41 (.69)
Can complete science requirements 83.0 3.35 (.65) 3.40 (.65)
Can excel this semester 73.7 3.12 (.76) 3.20 (.68)
Can persist next academic year 77.9 3.28 (.64) 3.30 (.62)
Can complete major at this institution 82.1 3.39 (.64) 3.43 (.63)
Confident in ability to complete 77.6 3.28 (.75) 3.29 (.74)
Composite Self-Efficacy - 26.02 (4.67) 26.23 (4.03)
Intent to Persist** 83.5 3.48 (.69) 3.67 (.49)
Note. Non-HW and HW groups significantly differ at: * p <.05, **p <0.01 level.

Inspection of means found HW students reported higher self-efficacy scores than non-
HW students on all measures except for ‘succeeding in STEM curriculum while not having to
give up participation in outside interests’ which was found equivalent for both groups. No
statistically significant differences were found, however, for STEM self-efficacy scores.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

100
Significant difference was found by t-test for intent to persist scores for non-HW (M =
3.48, SD = .693) and HW (M = 3.67, SD = .493; t (562) = -3.215, p = .002, two-tailed). The
magnitude of the differences in the means (mean difference = -.189, 95% CI: -.306 to -.073) was
small (eta squared = .018).
A factor analysis (PCA method) was applied to the STEM self-efficacy data to confirm
like grouping and investigate any potential subscales present. Suitability for factor analysis was
met (Kaiser-Meyer-Olkin measure = .908 and statistical significance of Bartlett’s test of
sphericity). The component matrix is shown in Table 29. A single component was found with an
eigenvalue of 5.259 explaining 65.7% of the variance. The results of the factor analysis
confirmed the utility of all 8-items for a single STEM Self-efficacy scale.

Table 29
Factor Analysis Component Matrix for STEM Self-Efficacy
Variables Component #1
Can complete STEM major at this institution .892
Feel confident in ability to complete a STEM degree at UH Manoa .891
Can persist in STEM major during next academic year .887
Can succeed in STEM major curriculum .860
Can complete Science requirements for STEM major .821
Can excel in current major this semester .808
Can complete Math requirements for STEM major .731
Can succeed in major curriculum while NOT giving up participation in
outside interests
.531

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

101
A reliability analysis followed a listwise deletion based on all variables in the procedure
excluded 144 missing cases for an n of 494. A Cronbach’s alpha of 0.914 for the 8-item scale
demonstrated very high internal consistency. The Cronbach’s alpha if item “I can succeed in my
STEM major curriculum while NOT having to give up participation in my outside interests”
were deleted would increase to 0.933, however the item was kept in the scale to be consistent
with the LAESE scale used in prior studies. A composite measure for STEM self-efficacy
ranging from 8 to 32 consisted of a sum of the eight items (M=26.02, SD=3.226).
Summary of Reliability Tests for Multi-Item Variables
Ten multi-item variables were identified from the data using principal component
analysis and reliability analysis. Existing literature discussing both theoretical rational and prior
application of scales also supported composite measures. Table 30 displays a summary of the
composite variables with the Cronbach’s α measures for internal consistency. The composite
variables were used in subsequent analysis as factors for regression to predict intent to persist,
for multiple correlation to STEM self-efficacy, and for comparison between Native Hawaiian
and non-Hawaiian groups to answer the overall research questions.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

102
Table 30
Summary of Composite Variables and Reliability Analysis
Composite variable Items Cronbach’s α
SES 3 .649
Family Support 2 .838
Peer Interaction 7 .837
Faculty Interaction 3 .609
Faculty Support 9 .826
Sense of belonging to school 3 .731
Sense of belonging to major 3 .885
Sense of belonging to campus community 3 .952
Satisfaction 6 .842
STEM self-efficacy 8 .914

Summary of Part One
The first part of the chapter highlighted the descriptive data for the items either by count
and percentage or by mean and standard deviation. The sample was found to be different than the
UHM STEM undergraduate population by having a larger percentage of females (by a factor of
1.4) and larger percentage of Native Hawaiians (by a factor of 1.5). The sample reflects
underrepresentation of College of Natural Science students, White students, and Japanese
students by factors of .9, .8, and .7 respectively. The researcher found that although the sample
did not represent the exact characteristics of the population, differences were not substantial to
limit the generalizability of the input of 638 students to the 3,592 students in the population.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

103
Further, it is noted that the results are based on a slightly higher voice from females and Native
Hawaiians.
The descriptive data set revealed a number of interesting results. The sample carried a
3.55 high school GPA (higher than the general UH Manoa freshmen class profile) and a college
GPA of 3.18. The sample self-described their family income at a level between middle and
lower-middle income with the majority reporting a mother’s education below Bachelor’s level
and the majority reporting father’s education below Bachelor’s level. Highest agreements were
to items of enjoying going to school here, sense of belonging to their major, family support,
receiving respect and intellectual challenge from faculty, and satisfaction with interaction with
peers. Lowest rated measures were to items related to sense of belonging to campus community,
frequency of faculty interaction during office hours or outside of class, tutoring other students,
faculty support related to advice, encouragement for graduate school, or research, satisfaction
with amount of contact with faculty and satisfaction with academic advising and student support.
In general, the respondents indicated high agreement on measures of the two outcomes of
interest in this study, their beliefs about their abilities to succeed in STEM and their intentions to
complete their STEM degree at UH Manoa. Over 82% of respondents stated that they could
complete their math, science, and major curriculum, could complete their major at UHM, and
intended to graduate in STEM at UH Manoa.
Similarities and differences between Hawaiian and non-Hawaiian groups were discussed.
Independent t-tests found significance difference on 12 items shown in Table 31, however the
actual differences in mean scores between the two groups were very small.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

104
Table 31
Significant t-test Results by Native Hawaiian Status
Variable
non-HW
M (SD)
HW
M (SD) t p eta
2

HS GPA* 3.58 (.43) 3.44 (.58) 2.309 .023 .009
Family: Approves of my attending** 3.43 (.58) 3.63 (.51) -3.504 .001 .020
Family: Encourages me to continue ** 3.36 (.67) 3.51 (.52) -2.617 .010 .011
Composite Family** 3.40 (.58) 3.58 (.47) -3.362 .001 .018
Faculty: Recommendation* 1.55 (.66) 1.71 (.71) -2.269 .024 .009
Peers: Outside of class* 2.44 (.62) 2.42 (.63) -2.130 .034 .007
Peers: Cooperatively on assignments* 2.38 (.61) 2.35 (.61) -1.967 .047 .006
Peers: Studied with others* 2.30 (.68) 2.26 (.68) -2.175 .030 .008
Peers: Feedback from classmates* 2.21 (.65) 2.18 (.65) -2.171 .030 .007
Peers: Group project* 2.16 (.66) 2.14 (.67) -1.967 .050 .006
Composite Peer Interaction* 15.46 (3.22) 16.26 (3.16) -2.367 .018 .009
Satisfaction, instruction* 2.52 (.84) 2.33 (.90) 2.156 .031 .007
Satisfaction, interaction with peers* 2.80 (.78) 2.98 (.73) -2.389 .018 .009
Intent to persist** 3.48 (.69) 3.67 (.49) -3.215 .002 .018
Note. * Significant at the p <0.05 level, ** Significant at the p <.01 level.

It is noted, however, that a series of t-tests may run the risk of inflated Type-I error,
finding a significant difference between groups when there in fact is no difference. The t-test
results will be further tested in the second part of this chapter by way of one-way analysis of
variance and multiple analysis of variance.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

105
The descriptive analysis presented in part one is important to situate the results that
follow for three reasons. First, it provides a clearer picture of the sample and context to evaluate
if interpretations of the results can be generalized by the researcher to provide implications for
the UHM STEM population and if results can be transferred by readers to provide insight to their
particular settings. Second, an initial comparison of Native Hawaiian and non-Hawaiian groups
can be made to recognize areas of similarity and highlight key areas of difference that can be
investigated further in subsequent analysis. Finally, the descriptive analysis, principal component
analysis, and reliability tests lay the foundation for the multivariate techniques to follow, which
in turn address the research questions.
Research Question 1: Correlation to STEM Self-Efficacy
Sequential multiple correlation was used to answer research question one, ‘what are the
personal input and environmental factors associated with STEM self-efficacy beliefs on
undergraduate students?’
The terms correlation and regression are often used interchangeably although the term
regression is used to denote that the intent of the analysis is prediction, and the term correlation
is more often used when the goal is to assess the relationships between the DV and the IVs
(Tabachnick & Fidell, 2007). Multiple correlation is the appropriate technique to answer this
question given that the DV is continuous (composite scale of 8 self-efficacy items) and the
model is interested in the relationships between a set of independent variables. Tabachnick and
Fidell (2007) describe standard multiple regression as an atheoretical, shotgun approach, and
step-wise/statistical regression as an exploratory, model-building (rather than a model-testing)
procedure, both which are not fitting for this study. Sequential multiple correlation (rather than
standard or step-wise correlation) is chosen in order for the researcher to enter and interpret IVs
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

106
based on theoretical grounds from Lent’s (2013) Social Cognitive Career Theory and Astin’s
(1977) Inputs – Environment – Outcomes model.
The aim of this sequential multiple regression is to evaluate how much variance in the
desired outcome (STEM self-efficacy) can be explained by a function of inputs (entered into
blocks 1 and 2) and environment (entered into blocks 3 and 4). Table 32 displays the blocks of
IVs and order of entry used in analysis. A second objective, along with determining the overall
predictive ability of conceptual model, is to determine the relative contribution of each
independent variable. It should be noted that there are many other factors that may contribute to
the variation in the DV (STEM self-efficacy) but the scope of this investigation is limited to the
selected independent variables.
Block 1 consists of background, input variables: gender, ethnicity (recoded into five
dichotomous variables for White, Asian, part-Hawaiian, Mixed, and non-Hawaiian
underrepresented racial minorities), SES, and high-school GPA. They represent background, pre-
college characteristics that Bean (1985) refers to as defining variables and Tinto (1987) describe
as family background and individual attributes. These enter into the regression analysis prior to
the college environment IVs so that the initial influence and additional influence of factors to the
outcome can be evaluated. Conceptually, it makes sense when investigating college effects
because students typically arrive at the University with ethnicity, gender, SES, and pre-college
schooling set and college environmental factors build on their experience. Practically, it makes
sense such that additional influence (variance on the outcome) can be determined on factors that
the college may have more control over.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

107
Table 32
Independent Variables Used for Self-Efficacy Sequential Multiple Correlation
Order Variable Block Range I-E-O
1 Gender: Male 1 0 = female, 1 = male I
2 Ethnicity: White 1 0 = no, 1 = yes I
3 Ethnicity: Hawaiian 1 0 = no, 1 = yes I
4 Ethnicity: Asian 1 0 = no, 1 = yes I
5 Ethnicity: Mixed 1 0 = no, 1 = yes I
6 Ethnicity: URM (non-HW) 1 0 = no, 1 = yes I
7 SES 1 0.00 (low) to 1.00 (high) I
8 High School GPA (reflect log 10) 1 1.3 to 4.0 I
9 Incoming status: high school 2 0 = no, 1 = yes I
10 Incoming status: 2-year transfer 2 0 = no, 1 = yes I
11 Incoming status: 4-year transfer 2 0 = no, 1 = yes I
12 Incoming status: Non-traditional 2 0 = no, 1 = yes I
13 College: CTAHR 2 0 = no, 1 = yes E
14 College: Engineering 2 0 = no, 1 = yes E
15 College: Natural Sciences 2 0 = no, 1 = yes E
16 College: SOEST 2 0 = no, 1 = yes E
17 Educational level 2 1 = freshmen to 5 = 5
th
year Senior E
18 Family support 3 1 to 4 E
19 Program participation 3 0 = none, 1 = participation in one or more E
20 Peer Interaction 3 7 to 21 E
21 Faculty Interaction 3 3 to 9 E
22 Faculty Support 3 9 to 27 E
23 College GPA 3 1.3 to 4.0 E
24 Belonging to School 4 3 to 12 E
25 Belonging to Major 4 3 to 12 E
26 Belonging to Campus Community 4 3 to 12 E
27 Satisfaction 4 6 to 24 E
28 Financial ability 4 1 to 4 E
DV STEM self-efficacy - 8 to 32 O

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

108
Block 2 consists of input academic classification variables: incoming student status
(recoded into four dichotomous variables for first-time clean freshmen, community college
transfers, four-year institution transfers, and returning or non-traditional students), College
(recoded into four dichotomous variables for Engineering, Natural Science, CTAHR, and
SOEST), and educational level (freshmen to 5
th
year senior). These variables represent
characteristics that define the student after college entry but prior to other environmental factors.
Blocks 3 and 4 consist of environment variables ordered based on a conceptual
progression of student experience. For example, the model theorizes that a college student
experiences levels of family support, peer-interaction, and faculty support prior to defining levels
of sense of belonging and overall satisfaction. Financial ability (reverse scored item assessing
levels of financial concern) fits into Lent’s (2013) SCCT model as an environmental contextual
variable proximal to choice behavior and was placed in block 4. This information will be used to
assess each independent variable in terms of what it adds to the prediction of STEM self-efficacy
after the previous variables have been controlled for.
Preliminary and post-analysis were conducted to test assumptions of normality, linearity,
multicollinearity, and homoscedasticity. Univariate outliers were found in independent variables
HS GPA, Family Support, and College GPA, and within dependent variable STEM Self-
Efficacy. Outliers to HS GPA (very low GPA), College GPA (very low GPA), Family support
(very low family support), and Self-efficacy were found to be true cases and kept in the sample.
Independent variable HS GPA was transformed by reflection and log10 function to address
substantial negative skewness and positive kurtosis. Multivariate outliers were inspected
following regression using Mahalannobis distances and Cook’s Distance. Pallant (2013) and
Tabachnick and Fidell (2007) indicate that Cook’s Distance values larger than 1 pose potential
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

109
problems. In this analysis, the maximum Cook’s Distance was 0.137 suggesting no undue
influence from strange cases on the results of the model as a whole. Therefore, college GPA,
family GPA, and Self-efficacy were not transformed and kept in the model.
Size of sample was found sufficient for generalizability with N=341 following listwise
deletion for all missing values. Tabachnick and Fidell (2007) recommend a minimum sample
size of N > 50 + 8m (where m is the number of IVs) for testing multiple correlation and N > 104
+ m for testing individual predictors. In this analysis of 28 IVs, the sample size of 341 exceeded
minimum requirements of 50 + 8(28) = 224 and 104 + 28 =132.
Table 33 displays the variance explained by the sequential multiple regression with the
inclusion of independent variable blocks into each model. Input, background variables entered in
block 1 explained 4.3% of the variance in STEM self-efficacy and input college status variables
in block 2, explained an additional 6.6% of the variance. Environmental variables in block 3
explained an additional 23.3% of the variance and sense of belonging, satisfaction, and financial
ability explained an additional 8% after controlling for prior measures. R was significantly
different from zero at the end of each step. The total variance explained by the model as a whole
was 42.2%, F(25, 315) = 9.20, p < .001.

Table 33
Sequential Multiple Regression Predicting Self-Efficacy
Model R R
2
Sig. R
2
Change F Sig F Change
1 .208 .043 .037 .043 2.161 .037
2 .330 .109 .000 .066 2.851 .001
3 .585 .342 .000 .233 8.324 .000
4 .650 .422 .000 .080 9.196 .000
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

110
The statistical significance of the overall model indicate that the null hypothesis, that
STEM self-efficacy is not affected by the set of independent variables, should be rejected. The
research hypothesis, that the set of independent variables contribute to the prediction of STEM
self-efficacy, was accepted. Table 34 displays the final model unstandardized regression
coefficients (B) with standard errors (SE B), standardized coefficients (β), and significance for
each variable demonstrate the relative contributions to the predictive ability of the model of
STEM self-efficacy was assessed for each independent variable.
Model 1 tested background input variables alone and was found to be statistically
significant in predicting 4.3% of the variance in STEM self-efficacy. In the final model, none of
the background input variables reached statistical significance and Ethnicity: Native Hawaiian
variable dropped out of the model. Negative β values for Gender: Male and Ethnicity: URM-
nonHW suggest a prediction of lower STEM self-efficacy for males than females and a lower
STEM self-efficacy for URM-nonHW relative to Hawaiians. The highest weighted standardized
coefficient in Block 1 was Asian (.119) suggesting Asian ethnicity variable predicts a higher
STEM self-efficacy than Hawaiians.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

111
Table 34
Regression Coefficients Predicting STEM Self-Efficacy
Variable B SE B β p
Gender: Male -.704 .444 -.078 .114
Ethnicity: White .261 .713 .021 .714
Ethnicity: Asian 1.077 .590 .119 .069
Ethnicity: Mixed .887 .666 .075 .184
Ethnicity: URM (non-HW) -.256 1.031 -.012 .804
SES 2.109 1.184 .086 .076
High School GPA (reflect, log10) -.075 1.826 -.002 9.67
Incoming status: 2-year transfer .472 .569 .043 .408
Incoming status: 4-year transfer .359 .718 .023 .618
Incoming status: Non-traditional or returning 1.852 .966 .090 .056
College: CTAHR .627 .688 .043 .363
College: Engineering* 1.234 .498 .132 .014
College: SOEST -.927 .955 -.045 .332
Educational level** .472 .183 .129 .010
Family support*** 1.438 .382 .186 .000
Program participation* 1.158 .465 .126 .013
Peer Interaction -.024 .070 -.017 .734
Faculty Interaction -.097 .169 -.030 .567
Faculty Support .056 .070 .048 .422
College GPA*** 2.302 .444 .267 .000
Belonging to School .007 .146 .003 .960
Belonging to Major*** .625 .133 .268 .000
Belonging to Campus Community** -.292 .110 -.151 .008
Satisfaction .208 .248 .041 .403
Financial ability .208 .248 .041 .403
(Constant) 2.563 2.162 .237
Note. Variables Ethnicity: Hawaiian, Incoming Status: High School, and College: Natural
Sciences excluded from equation.
* p < .05, ** p < .01, *** p < .001.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

112
Model 2 tested input academic classification variables added to the block 1 input
variables. Model 2 was found to be statistically significant explaining a total of 10.9% of the
variance in STEM self-efficacy. In the final model, incoming status: high school and College:
Natural Sciences were excluded from the equation. Two factors, College: Engineering and
educational level were found to be statistically significant. Positive standardized coefficients
show that Engineering students relative to Natural Science students would predict higher STEM
self-efficacy and that STEM self-efficacy would increase as students progress in educational
level. Other variables did not reach statistical significance, which was unexpected, were SES
and high school GPA cited as strong predictors of STEM self-efficacy (Cady et al., 2009; Perna,
2000).
Model 3 assessed the predictive ability of block 1 and block 2 variables with the addition
of college environmental factors. Model 3 was found to be statistically significant contributing
23.3% of variance predictive ability for a total of 34.2%. In the final model, family support,
program participation, and college GPA reached statistical significance. Positive β values
suggest that increase in family approval and encouragement, participation in at least one
program, and increase in college GPA contributes to increased STEM self-efficacy beliefs.
College GPA was found to be one of the strongest factors with a standardized coefficient of .267.
Although not found to be statistically significant, unexpected results were negative relationships
for peer interaction and faculty interaction to the dependent variable. It should be noted that
college environment factors in block 3 contributed the majority of the predictive ability of the
overall self-efficacy model.
Model 4 added environmental variables sense of belonging, satisfaction, and financial
ability to the previous factors and contributed an additional 8% to the predictive ability when
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

113
controlling for other factors. It is noted these factors may have a larger influence on the
dependent variable but given their order in the sequential multiple regression technique and the
correlation with other independent variables, may be credited with a lesser, unique contribution.
In block 4, only two variables, sense of belonging to major and sense of belonging to campus
community, were found to be statistically significant. An unexpected result was that sense of
belonging to campus community displayed a negative effect on self-efficacy in the model. This
could be related to the distinction students make between sense of belonging to major and to
school versus their views of the general campus community.
From the overall model, sense of belonging to major was found to be the most influential
significant predictor (β = .268), followed by College GPA (.267), family support (.186), sense of
belonging to campus community (-.151), College: Engineering (.132), educational level (.129)
and program participation (.126). Many IVs unexpectedly did not reach statistical significance.
Research Question 2: Correlation to Intent to Persist
Logistic regression was used to answer research question two, ‘what are the personal
input and environmental factors associated with intent to persist on undergraduate students?’ The
outcome of interest in this analysis is intent to persist, which indirectly informs related outcomes
including intent to leave, actual persistence, and actual dropout. Astin’s I-E-O model framework,
used to guide this study, views college outcomes as functions of inputs and environment. The
aim of this analysis is to predict, from a set of 29 independent input and college environment
variables, the students that report an interest in completing their STEM degree at UH Manoa as
opposed to those that do not. Analysis should help investigate how well the complete set of IVs
in the model explain the outcome as well as provide assessment for individual variables in terms
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

114
of their likelihood of increasing, decreasing, or having no effect on the probability of the
outcome.
The methods for answering question two (intent to persist) were very similar to the
methods for answering question one (STEM self-efficacy). A sequential logistic regression was
chosen (as opposed to standard or statistical) to allow for entry of independent variables in
blocks in a manner consistent with the conceptual model. 29 independent variables were utilized,
including all 28 IVs shown in Table 32 previously used for multiple correlation as predictors in
blocks 1 through 4. STEM self-efficacy was tested as the 29
th
independent variable, to predict
intent to persist. The four point likert-like measure of intent to persist, was coded into a
dichotomous measure of all respondents whom strongly disagree, disagree, or prefer not to
answer into group “0” and all respondents whom indicated strongly agree or agree into group
“1.” Logistic regression is the preferred technique to predict group membership given a set of
independent variables (Pallant, 2013).
The full logistic regression model containing all predictors was statistically significant,
χ
2
(26, N = 341) = 61.24, p < .001, indicating that the model was able to distinguish between
respondents who reported and did not report an intent to persist. The model as a whole explained
between 16.4% (Cox & Snell R Square) and 44.4% (Nagelkerke R Square) of the variance in
intent to persist status and correctly classified 95.6% of cases.
Table 35 shows regression coefficients, Wald statistics, odds ratios, and 95% confidence
intervals for each of the independent variables. According to the Wald criterion, only two
variables reliably predicted intent to persist status, educational level (p = .005) and STEM self-
efficacy (p = .002). The stronger predictor of intent to persist was STEM self-efficacy recording
an odds ratio of 1.27. Educational level recorded an odds ratio of 2.46. The odds ratio represents
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

115
the change in the odds of being in one of the outcome categories when the value of the predictor
increases by one unit (Tabachnick & Fidell, 2007). For every one-year increase in educational
level (i.e. sophomore to junior) the odds of the student reporting intent to persist increases by a
factor of 2.5. The significance of the relation makes sense, that students closer to completing
their STEM degree will more likely indicate their intent to complete their STEM degree, but the
magnitude of the difference appears very high. This finding supports the literature identifying the
first three semesters of college as being critical to affect student retention.
The composite STEM self-efficacy measure was found to have the highest predictive
ability on intent to persist. If a student increased their level of agreement by one unit (from
disagree to agree or from agree to strongly agree) to any one of the eight STEM self-efficacy
measures (i.e. I can complete the math requirements, I can excel this semester) then the
probability of them belonging to the intent to persist group would increase by a factor of 1.27.
Although no significance was found in the predictive ability of the other variables in the
model, the analysis provided some data that we expected in some areas, and unsupported by the
literature in others. The parameter estimates showed negative coefficients for SES, financial
ability, satisfaction, program participation, and sense of belonging to major indicating an indirect
relationship between the independent variables and the predicted outcome group.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

116
Table 35
Sequential Logistic Regression Predicting Intent to Persist
95% C.I.
Variable B SE B Wald p
Odds
Ratio Lower Upper
Gender: Male -.236 .693 .116 .733 .790 .203 3.069
Ethnicity: White 2.83 1.075 .069 .792 1.328 .161 10.919
Ethnicity: Hawaiian 1.645 1.402 1.377 .241 5.183 .332 80.945
Ethnicity: Asian .312 1.104 .080 .777 1.367 .157 11.891
Ethnicity: Mixed .197 1.291 .023 .879 1.218 .097 15.281
SES -.566 1.796 .099 .753 .568 .017 19.193
High School GPA -.689 .834 .684 .408 .502 .098 2.571
Incoming status: high school 2.420 1.412 2.937 .087 11.249 .706 179.197
Incoming status: 2-year transfer 1.547 1.451 1.137 .286 .213 .012 3.659
Incoming status: 4-year transfer 1.952 1.702 1.316 .251 7.043 .251 197.734
College: CTAHR 1.246 1.786 .487 .485 3.478 .105 115.142
College: Engineering 1.109 1.498 .548 .459 3.032 .161 57.116
College: Natural Sciences .161 1.344 .014 .905 1.175 .084 16.359
Educational level** .901 .322 7.839 .005 2.462 1.310 4.627
Family support .810 .574 1.993 .158 2.247 .730 6.916
Program participation -.015 .781 .000 .985 .985 .213 4.554
Peer Interaction .068 .101 .457 .499 1.071 .878 1.305
Faculty Interaction .167 .273 .374 .541 1.182 .692 2.016
Faculty Support .003 .105 .001 .975 1.003 .817 1.233
College GPA .679 .688 .972 .324 1.971 .511 7.598
Belonging to School .031 .240 .017 .896 1.032 .644 1.653
Belonging to Major -.011 .210 .003 .958 .989 .656 1.492
Belonging to Campus Community .041 .174 .056 .813 1.042 .741 1.465
Satisfaction -.042 .110 .146 .703 .959 .772 1.190
Financial ability -.297 .410 .522 .470 .743 .333 1.661
Self-Efficacy** .238 .077 9.465 .002 1.269 1.090 1.476
(Constant) -.354 7.412 .002 .962 .702
Note. Variables Ethnicity: URM non-HW, Incoming Status: Non-traditional, and College:
SOEST excluded from the equation.
** p < .01.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

117
Research Question 3: Native Hawaiian vs. Non-Hawaiian Respondents
Research question three asked, “how do the background, environmental, and outcome
characteristics differ, if at all, among Native Hawaiian and non-Hawaiian students?” The related
null hypothesis predicts that there are no differences for each variable between Native Hawaiian
and non-Hawaiian respondents. To answer this question, multiple techniques were utilized. First,
non-parametric Chi-Square tests were utilized to test for significant differences between Non-
HW and HW groups across input categorical variables (gender, college, academic level, and
incoming student status). A series of independent t-tests and one-way ANOVAs utilized as an
screening prior to more sophisticated analysis to explore for significant differences across select
input and environmental continuous variables (high school GPA, College GPA, level of parent
education, self-reported income, and financial ability). Finally, multivariate analysis of variance
was conducted on all variables.
Chi-square tests for independence (with Yates Continuity Correction for gender) were
utilized to evaluate if the proportions of categorical variables varied significantly by Native
Hawaiian status as seen in Table 11. Significant association between Native Hawaiian status was
found for gender, Engineering, Natural Sciences, and participation in one or more programs. No
significant difference was found for educational level and pre-UH Manoa student status. Relative
to non-Hawaiians, Hawaiians were more highly enrolled in Engineering, fewer enrolled in
Natural Sciences, higher percentage male, and higher participation in one or more programs.
A series of one-way between-groups analysis of variance was conducted to explore
further if the significant variables found by t-test results between the non-Hawaiian and
Hawaiian groups were due to Hawaiian grouping variability or due to chance. Table 36 displays
results of the series of ANOVAs. Some of the variables tested did not pass the test for
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

118
homogeneity of variance and are presented in a separate Table 37, Robust tests of equality of
means.

Table 36
ANOVA of Select Characteristics by Native Hawaiian Status
Variable
non-HW
M (SD)
HW
M (SD) df2 F p eta
2

Faculty support: letter of recommendation 1.55 (.66) 1.71 (.71) 593 5.787* .016 .010
Peer: outside of class 2.42 (.63) 2.55 (.59) 633 3.959* .047 .006
Peer: cooperate on assignments 2.36 (.61) 2.48 (.62) 631 3.267 .071 .006
Peer: Studied with others 2.27 (.68) 2.41 (.66) 631 3.893* .049 .006
Peer: feedback from classmates 2.18 (.65) 2.32 (.67) 633 4.093* .043 .006
Peer: group project 2.14 (.67) 2.27 (.65) 630 3.269 .071 .005
Composite Peer Interaction 15.49 (3.23) 16.23 (3.16) 624 4.727* .030 .008
Satisfaction: quality of instruction 2.51 (.84) 2.34 (.90) 634 3.494 .062 .006
Note. * p < .05, ** p < .01, *** p < .001.

Table 37
Robust Tests of Equality of Means of Select Characteristics by Native Hawaiian Status
Variable
non-HW
M (SD)
HW
M (SD) df2
Brown-
Forsythe p
High school GPA 3.58 (.43) 3.44 (.58) 126.946 5.231* .024
Family: approves of my attending 3.43 (.58) 3.62 (.51) 166.652 12.001** .001
Family: encourages me to continue 3.35 (.67) 3.51 (.52) 177.968 6.929** .009
Composite Family 3.39 (.58) 3.58 (.47) 171.313 11.301** .001
Satisfaction: interaction with peers 2.80 (.78) 2.97 (.73) 164.694 4.996* .027
Intent to persist 3.49 (.69) 3.68 (.49) 189.564 10.407** .001
Note. * p < .05, ** p < .01, *** p < .001.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

119
Statistically significant differences were found on measures of peer interaction, faculty
letter of recommendation, high school GPA, family approval and encouragement to continue,
satisfaction with interaction with peers, and intent to persist. Three items found to be
significantly different by t-test, satisfaction with quality of instruction, worked on group project,
and worked cooperatively with other students on course assignments failed to reach significance
in the ANOVA. The results revealed a theme around items of peer interaction frequency (with
peers outside of class, studied with other students, get feedback from peers) and satisfaction with
peer interaction found higher for Hawaiians than non-Hawaiians. It is noted that Hawaiians
report higher levels of family approval and family encouragement to continue attending UH
Manoa, and higher levels of intent to persist than non-Hawaiians.
Multivariate Analysis of Variance
Multivariate analysis of variance (MANOVA) is a technique to compare groups on a
range of different characteristics. It is an ideal statistical technique to answer this research
question to investigate the data based on two independent groups (Hawaiian and non-Hawaiian)
against a set of variables. MANOVA is the preferred analysis technique over running a series of
separate t-tests or ANOVAs testing each variable, the latter which runs the risk of an inflated
Type 1 error (Pallant, 2013; Tabachnick & Fidell, 2007). However, Tabachnick and Fidell
(2007) also note that often MANOVA is considerably less powerful than ANOVA, particularly
in finding significant group differences for a particular DV, which risk Type 2 error. Therefore,
the investigation of the data by way of ANOVA and MANOVA was used.
A one-way between-groups multivariate analysis of variance was performed to
investigate if college environmental characteristics vary by Native Hawaiian or non-Hawaiian
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

120
status. Fourteen dependent variables were selected for MANOVA testing dependent variable was
Native Hawaiian status.

Table 38
MANOVA of Selected Characteristics by Hawaiian and Non-Hawaiian Groups (N = 345)
Dependent variable F p partial eta
2

SES .107 .744 .000
High School GPA (reflect and log10) .553 .457 .002
Family support 3.742 .054 .011
Peer Interaction .831 .363 .002
Faculty Interaction .205 .651 .001
Faculty Support .091 .763 .000
College GPA .456 .500 .001
Belonging to School .337 .562 .001
Belonging to Major .074 .786 .000
Belonging to Campus Community 1.084 .299 .003
Satisfaction .019 .892 .000
Financial ability .593 .442 .002
STEM self-efficacy .000 .993 .000
Intent to Persist 4.595 .033 .013

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

121
A statistically significant difference between non-Hawaiian and Hawaiian groups was not
found on the overall MANOVA using linear combinations of dependent variables, F(14, 345) =
1.108, p = .349; Wilks’ Lambda = .96; partial eta squared = .045. When the results for the
dependent variables were considered separately, the only difference to reach statistical
significance was intent to persist, F (1, 343) = 4.595, p = .033, partial eta squared = .013. An
inspection of the mean scores indicated that Hawaiian students reported higher levels (M = 3.69,
SD = .50) than non-Hawaiians (M = 3.50, SD = .70) indicating their agreement with the
statement intend to complete a STEM degree at UH Manoa. A second independent variable in
the model, family support, approached statistical significance, F (1, 343) = 3.742, p = .054,
partial eta squared = .011. An inspection of the mean scores indicated that Hawaiian students
reported higher levels (M = 3.53, SD = .47) than non-Hawaiians (M = 3.38, SD = .60) in
agreement to statements that their family approves of their attending UH Manoa and encourages
them to continue to attend UH Manoa.
The results of the MANOVA show that no significant differences between groups were
found based on the set of 14 dependent college environment variables. Intent to persist was
found to be significantly higher for Native Hawaiians than non-Hawaiians in both the MANOVA
and ANOVA techniques. Composite family support approached significance (p=.054) and
composite peer interaction and high school GPA failed to reach significance in the MANOVA,
though variables were found to be significant in the one-way ANOVA. Differences in significant
results could be due to increased family-wise error by ANOVA or increase in Type I-error by
MANOVA due to correlations between a large set of dependent variables.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

122
Summary of Results
The major results of the regression analyses were that the set of input and environment
variables were successful in predicting 42.2% (F(25, 315) = 9.20, p < .001) of the variance in
STEM self-efficacy and successful in predicting between 16.4% (Cox & Snell R Square) and
44.4% (Nagelkerke R Square) of the variance in intent to persist status correctly classifying
95.6% of cases (χ
2
(26, N = 341) = 61.24, p < .001). Native Hawaiian and non-Hawaiian groups
were found to vary by measures of intent to persist, family support, program participation, and
frequency and satisfaction with peer interaction. The major results are presented in Table 39.

Table 39
Significant Predictors and Native Hawaiian Differences
STEM Self-efficacy Intent to Persist Native Hawaiian vs. Non-Hawaiian
Belonging to major STEM self-efficacy Intent to persist
College GPA Educational level Family support
Family support Program participation
College: Engineering
a
Frequency of peer interaction
Belonging to campus community
b
Satisfaction with peer interaction
Educational level
Program participation
Note.
a
Relative to reference college Natural Science

b
Negative predictor

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

123
For STEM self-efficacy, sense of belonging to major was found to be the strongest
significant predictor (β = .268), followed by College GPA (.267), family support (.186), sense of
belonging to campus community (-.151), College: Engineering (.132), and program participation
(.126). For Intent to Persist, the strongest significant predictor was STEM self-efficacy (p = .002)
recording an odds ratio of 1.27. The only other significant predictor of intent to persist found was
educational level (p = .005) with an odds ratio of 2.46.
Significant difference between Native Hawaiian and non-Hawaiian groups were found on
only one item, intent to persist (p = .033) by MANOVA and on an additional 7 items related to
family support, frequency and satisfaction of peer interaction, and faculty letter of
recommendation by ANOVA. Program participation was also found to be significantly higher,
by Chi-square test, for Hawaiians.
Discussion of these results as they relate to the literature and to future research and
practice will be presented in Chapter 5. A synthesis of the results of the three research questions
will attempt to address an overarching research question: How can the University of Hawaii at
Manoa increase the persistence and degree completion of undergraduate STEM majors as a
whole and Native Hawaiian undergraduate STEM majors in particular.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

124
CHAPTER 5
DISCUSSION
The broader context for the motivation of this study is that the U.S. educational system is
failing to keep pace with global competitors to produce a citizenship literate in STEM and a
STEM workforce that is well-trained and well-educated. The National Academies (NRC, 2007,
2010) strategize that part of the solution lies in improving college outcomes for underrepresented
minority and indigenous student in STEM. The challenge then falls to local educational systems
to address areas of improvement in a leaky STEM pipeline. The specific context for the
motivation of this study is the goal of improving college outcomes for Native Hawaiian STEM
majors at the University of Hawaii at Manoa.
The purpose of this study was to investigate two outcomes of interest — self-efficacy
beliefs and intentions to persist — for Native Hawaiian and non-Hawaiian STEM majors at the
University of Hawaii at Manoa. These two outcomes were chosen because of their influence on
choice goals, motivation, and actual persistence (Bandura, 1997; Lent, 2013; Cabrera et al.,
1992; Bean, 1980). It was important to study Native Hawaiians because they have historically
been an underserved group in post-secondary education and, as the indigenous people of Hawaii,
are the subject of commitment for improved educational attainment and participation at the
University of Hawaii, the only provider of public higher education in Hawaii (UoHBR, 2012). It
was also important to study non-Hawaiian STEM majors in order to provide a reference group to
understand differences from Native Hawaiian STEM majors, and in order derive meaningful
recommendations for all students and STEM programs.
This single-institution, cross-sectional study gathered survey data from 638
undergraduate STEM majors (17% response rate of all STEM majors at UHM) on measures of
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

125
pre-college, college environment, and outcome variables. Using the framework of Lent’s (2013)
Social Cognitive Career Theory and Astin’s (1999) Inputs – Environment – Outcomes model,
descriptive, multiple regression, logistics regression, and MANOVA techniques were used to
answer this study’s three research questions:
1. What are the personal input and environmental factors associated with STEM self-
efficacy beliefs of undergraduate STEM students?
2. What are the personal input and environmental factors associated with intent to
persist in STEM of undergraduate STEM students?
3. How do these factors and outcomes differ, if at all, amongst Native Hawaiian and
non-Hawaiian students?
Discussion of Findings
The major results of the study (presented in Table 39) are organized into seven overall
findings, three addressing STEM self-efficacy, two addressing intent to persist, and two
addressing differences for Native Hawaiian students. First, STEM self-efficacy beliefs increased
with higher sense of belonging to major but decreased with higher sense of belonging to campus
community. Second, STEM self-efficacy increased with positive past performance including
higher College GPA and higher educational level. Third, environmental factors of family
support, program participation, and engineering college were found to increase STEM self-
efficacy. Fourth, higher STEM self-efficacy explained higher intent to persist. Fifth, higher
education level predicted higher intent to persist. Sixth, Native Hawaiians were found to report
higher levels relative to non-Hawaiians of commitment to completing their STEM major at UH
Manoa as shown by intent to persist and family support. Finally, Native Hawaiian STEM majors
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

126
exhibited higher levels of peer interaction and program involvement. Discussion of these seven
findings are presented by research question.
Research Question 1 Findings: Predictors of STEM Self-Efficacy
The complete set of background input, academic classification, and college environment
variables was found to be significant in predicting STEM self-efficacy explaining 42.2% of total
variance (F(25, 315) = 9.20, p < .001). The study found seven significant predictors for STEM
self-efficacy out of the set of 28 tested. Sense of belonging to major was found to be the
strongest significant predictor (β = .268), followed by College GPA (.267), family support
(.186), sense of belonging to campus community (-.151), College: Engineering (.132),
educational level (.129), and program participation (.126). Many IVs unexpectedly did not reach
statistical significance.
The results in response to question one highlight three major findings. First, STEM self-
efficacy beliefs increased with higher sense of belonging to major but decreased with higher
sense of belonging to campus community. Second, STEM self-efficacy increased with positive
past performance including higher College GPA and higher educational level. Third,
environmental factors of family support, program participation, and engineering college were
found to increase STEM self-efficacy.
The first finding is that for UHM students’ sense of belonging to major was found to be
the strongest predictor for STEM self-efficacy (β = .268, p < .001) while sense of belonging to
campus community (β = -.151, p = .008) was the only significant negative predictor found.
Existing research supports the positive relationship between sense of belonging to major to self-
efficacy but the indirect relationship between sense of belonging to campus community to self-
efficacy was unusual.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

127
Possible interpretation of these results are that students with high STEM self-efficacy
have a stronger identity in STEM (relating to their major peers, faculty, and academic study), but
do not connect, frequently interact, of feel a part of the non-STEM or overall community. They
may be expressing their perception of a STEM and non-STEM cultural division found at many
large public universities. Or highly effacious students could be expressing a disappointment in
the level of or longing to get involved in general campus community activities, school spirit, and
extracurricular activities outside of their major. When surveyed if students think they can be
successful in their STEM major without giving up their participation in outside interests almost
half did not agree. Similarly, the interpretation is that students with low STEM self-efficacy do
not strongly connect with their STEM major community and have a stronger sense of being a
part of the general campus community. Cole and Espinoza (2008) suggest active campus
involvement outside of STEM can have negative effects on the persistence of URM students
within STEM majors due to a conflict between the values within their STEM major and the
respective disciplines of their peers.
The findings reinforce the notion that different senses of belonging exist for students and
they can relate to outcomes in different ways. Descriptive data showed that this sample reported
sense of belonging to major highest, slightly lower on sense of belonging to school (which did
not reach statistical significance in the regression analysis), and lowest on belonging to campus
community. The implication for researchers is that different senses of belonging exist for
students and unique analysis can provide clearer detail on the dynamics of college effects. The
findings support Hurtado and Carter’s (1997) assessment that sense of belonging is an important
but under studied variable.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

128
Second, STEM self-efficacy increased with positive past performance including higher
College GPA and higher educational level. College GPA was found to be the second strongest
(β = .267, p < .001) predictor of STEM self-efficacy. Positive association with College GPA and
self-efficacy were consistent with the research (Bandura, 1977; Lent, 2013; Pajares, 1996) as
self-efficacy beliefs are, for most people, based on the interpreted result of one’s own past
performance. Those with higher GPAs reported higher self-abilities in completing their STEM
degree. Moreso, those with lower GPAs who may need to most help, may exhibit behaviors
associated with lower self-efficacy beliefs such as reduced effort and commitment to future tasks
such as seeking tutorial support and studying and preparation for coursework.
Educational level was found to be the sixth strongest factor in predicting student beliefs
about their ability to complete a STEM degree at their current institution. The results indicate
that student’s level of STEM self-efficacy increase as they progress from first-year to second-
year and so on. This is consistent with the theoretical research (Bandura, 1997; Lent, 2013)
suggesting an increase in self-efficacy beliefs from mastery experiences or prior, personal
success at a similar task. The assumption here is that students at higher education levels, have
had some prior personal success, and have persisted onto the next educational level. Empirical
research on academic level and Mathematics self-efficacy (Jordan, Sorby, & Amato-Henderson,
2012) also found consistent results.
Finally, environment factors of family support, program participation, and engineering
college were found to increase STEM self-efficacy. Classification in Engineering was found
significant in the regression model (whereas Tropical Agriculture, Ocean & Earth Sciences, and
reference college Natural Sciences did not reach statistically significant findings). The results
indicate that Engineering students (β = .132) show higher levels of STEM self-efficacy than
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

129
reference college Natural Sciences. The researcher did not have an initial hypothesis in testing
College associations with the outcomes variable, though the finding for Engineering was
unexpected. The National Center for Education Statistics (U.S. Government Accountability
Office, 2012) indicate a lower degree completion rates (in any STEM degree) and a higher rate
of leaving college without earning a degree for students entering engineering/engineering
technologies and computer sciences compared to students who entered physical sciences, natural
sciences, and biological/agricultural sciences. In other words, the national data show persistence
in engineering to be lower relative to other STEM fields. In this study at UH Manoa, respondents
in were found to be have higher levels of self-beliefs about completing their engineering degree
relative to other STEM fields.
Family support (β = .186, p < .001) was found to be another environmental predictor for
STEM self-efficacy. This finding supported the literature that found lack of family support to be
a barrier to success in STEM, whereas ongoing encouragement from parents positively
influenced self-efficacy (Sandler, 1999; Swail & Perna, 2002). Bandura (1977, 1997) identifies
social persuasion as a key source of self-efficacy beliefs, especially when feedback comes from
influential others. Family approval and encouragement to continue was found to be the third
strongest factor in explaining STEM self-efficacy beliefs after sense of belonging to major and
GPA.
Participation in at least one academic/student support program was found to explain
higher levels of STEM self-efficacy. Of the study sample, 217 students (34%) identified
participation in one or more programs, the most frequent being the UHM Honors Program (105),
the Native Hawaiian Science & Engineering Mentorship Program (82), and the American Indian
Science & Engineering Society (42). It is noted that this study listed primarily minority support
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

130
programs in the survey (with the option of selecting “other” and an open response category)
because an interest was in investigating college effects especially for Native Hawaiian STEM
students. The positive relationship between program participation and STEM self-efficacy was
consistent with the theoretical research, such as Astin’s theory of student involvement, and
empirical review (Clewell et al., 2006; Leggon & Pearson, 2009). This result demonstrated
consistency between the positive influence of Native Hawaiian-serving programs and minority
support programs.
Research Question 2 Findings: Predictors of Intent to Persist
Research on self-efficacy is often examined in concert with outcome expectations,
interest, or choice goals to connect the beliefs of what one can do with what one will do. In
research question two, this study investigated the same conceptual model of background input,
academic classification, and college environment variables with the addition of independent
variable STEM self-efficacy to predict intent to persist. The full logistic regression model
containing all predictors was statistically significant, χ
2
(26, N = 341) = 61.24, p < .001,
indicating that the model was able to distinguish between respondents who reported and did not
report an intent to persist. The model as a whole explained between 16.4% (Cox & Snell R
Square) and 44.4% (Nagelkerke R Square) of the variance in intent to persist status and correctly
classified 95.6% of cases.
There were two major findings in response to research question two. First, increased
STEM self-efficacy and higher education levels explained higher intent to persist. Second, higher
education levels explained higher commitment to degree completion. These findings are
discussed in order of predictive strength.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

131
Consistent with the research literature, this study found STEM self-efficacy significant (p
= .002) in explaining Intent to Persist. The interpretation is that students that report high self-
beliefs in their ability to complete their STEM degree at the institution also report intent to do so.
If they think that they can do it, then they are more likely to pursue it. Conversely, students with
low self-efficacy are less likely to attempt to engage in future task related activity and are less
likely to expend as much energy. This confirms the literature (Lent, 2013; Pajares, 1996;
Zimmerman, 2000) arguing the importance of self-efficacy as a central construct in mediating
choice goal, motivation, and behavior outcomes.
Educational level was the only other factor found significant (p = .005) in the model. An
odds ratio of 2.46 implies that students at higher levels are much more likely to have positive
intent to complete their STEM major at the institution. Sophomores are about 2.5 times more
likely than freshmen and Juniors are about 6 times more likely than freshmen to be found in the
intent to persist group. It is noted that this result is derived from cross-sectional data where the
sample of upperclassmen are past persisters as opposed to longitudinal data that tracks all
students, including stop-outs and leavers, as they increase in educational level.
Educational level was found to be significant in predicting both STEM self-efficacy and
intent to persist. Although this study makes no claims regarding causality, the “chicken-or-egg”
(which came first?) question is relevant in trying to better understand the dynamics of increased
educational level (indicating successful year to year persistence), STEM self-efficacy, and intent
to persist. Do students have a higher commitment to continue on to graduation because they
have made it to a certain educational level, or have they made it this far because they have a
higher commitment to continue on to graduation? Do students have a higher commitment to
graduation because they think they have the ability to do it or has their level of commitment
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

132
influenced their choice activities, past performance (such as year to year persistence), and their
perceptions of their abilities? Due to the reciprocal nature of human motivation and behavior,
researchers (Pajares, 1996; Bandura, 1997) believe such questions are unlikely to be resolved.
Research Question 3 Findings: Differences by Native Hawaiian Ethnicity
There were two major findings in response to research questions three. First, Native
Hawaiians were found to report higher levels relative to non-Hawaiians of commitment to
completing their STEM major at UH Manoa as shown by intent to persist and family support.
Second, Native Hawaiian STEM majors exhibited higher levels of peer interaction and program
involvement.
This study found Native Hawaiians to report high levels of commitment and family
support to complete their STEM degree at the institution. First, Native Hawaiians reported higher
agreement than non-Hawaiians on the query ‘I intend to complete a STEM degree at UH
Manoa.’ This variable was found statistically significant when analyzed independently by
ANOVA (p = .001) and when analyzed collectively by MANOVA (p = .033), however the effect
size (partial eta
2
= .013) was small.
The intent to persist result sheds new light on the beliefs of Native Hawaiian students.
Because of the scarcity of literature investigating beliefs of Native Hawaiian students on college
outcomes relative to other ethnicities, the researcher did not have a hypothesis on the potential
findings. However, studies have shown that Hawaiians students face higher barriers than
majority students on measures of pre-college achievement, high school graduation, college
enrollment, and financial ability (KSP, 2009; Hokoana, 2010; Oliveira, 2005). By Lent’s
conceptual model, a lower intent to persist finding for Native Hawaiians than non-Hawaiians
would be expected.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

133
In addition, family support, specifically a student’s feeling that their family approves of
their attending the institution and encourages them to continue attending, was found to be a
statistically significant difference by ANOVA (p = .001) and approached significance by
MANOVA (F = 3.742, p = .054). Although the relative effect was found to be small (partial eta
2

= .011), Native Hawaiians reported higher levels than non-Hawaiians to both items.
This result was, in part, supported by literature finding resilience, as enabled by parental
support, particularly important for minority and indigenous students who may possess less social
and cultural capital than others (Speck & Keahiola-Karasuda, 2011). Much of the literature on
Native Hawaiians identify the importance of building the learning process inclusive of home,
family, and broader community (Benham, 2006) and in situating learning in connection with
family members (Kawakami, 1999).
Much of the conceptual literature focused on the influence of family support on academic
outcomes, however it was unclear as to how family support varied by ethnicity and for
Hawaiians in particular. Family responsibilities, job responsibilities, and lack of financial ability
were identified by college administrators (PPRC, 2010) as prevalent obstacles for minority
students and it has been shown that these factors are high among the Native Hawaiian population
(Hagedorn & Tibbetts, 2003). Therefore, it was unexpected and encouraging to find sense of
family encouragement higher for Native Hawaiians than their non-Hawaiian counterparts.
Despite individual and family barriers that have been found to disadvantage Native Hawaiian
students, this study found higher commitment to completing a STEM degree and support from
family among Native Hawaiians relative to non-Hawaiians.
The second major finding in response to differences between Native Hawaiian and non-
Hawaiian STEM students was higher levels of program involvement and peer interaction. Native
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

134
Hawaiians were found to be twice as likely (67.9% to 26.9%) to participate in at least one
program than non-Hawaiians (χ
2
(1, n=637) = 67.76, p < 0.001). It is encouraging to recognize
that two out of three Native Hawaiian STEM students participate in some form of STEM support
program.
Measures of peer interaction (satisfaction with interaction with peers (p = .027),
frequency of peer interaction outside of class (p = .047), studying with other students (p = .049),
getting feedback from classmates (p = .043)) were found significantly higher for Native
Hawaiians. The statistical claim on this finding is not as high (found significant by ANOVA, but
not by MANOVA) as others discussed, but presented for discussion and future investigation.
These findings are supported by the literature on Native Hawaiians as social-cultural learners
that make meaning out of the relevance to community.
In addition, Native Hawaiians were found significantly different than non-Hawaiians in
that their STEM majors were more male (56% to 42%), more likely to be found in engineering
(43% to 32%), and less likely to be found in natural sciences (36% to 51%). These differences
were found for the study sample and could be further explored using total enrollment data
available to verify if the differences found in the sample are true to differences found in the
population. This initial findings leads to further questions including, do Native Hawaiians enroll
in and pursue STEM fields differently than non-Hawaiians? Are Native Hawaiian females less
likely than non-Hawaiian females to pursue STEM fields? and if so, why?
Limitations
There are a number of limitations to this single-institution, quantitative study. Creswell
(2008) defines a limitation as a design weakness that could potentially reduce the study’s scope
and validity. Limitations regarding external validity, internal validity, and bias were identified.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

135
First, the sample was found to differ from the total UHM STEM population by way of
higher representations of females and Hawaiians. It should be noted that the results of this study
reflect a larger voice from females (1.4X) and Native Hawaiians (1.5X) and threatens the
external validity of generalizing findings for the intended (UHM STEM) population. The sample
was derived from a single, public four-year institution primarily serving residents of the state of
Hawaii and was limited to Bachelor’s degree seeking students in academic disciplines of Natural
Science, Ocean & Earth Sciences, Engineering, and Tropical Agriculture. Therefore, this study is
not claimed to be generalized to out-of-state institutions that may serve different student
demographics or for non-stem or community college populations.
Second, limitations regarding the internal validity of the research should be noted. The
data was collected via a close-ended, self-report questionnaire. The researcher had to rely on the
assumption of truthful and honest responses, which may threaten the internal validity. While
honest self-report data had the advantage of gathering respondents’ perceptions of themselves
and their environment, the data may not be accurate to “reality” or as seen by others. Additional
assumptions were made that the respondents understood the questions and interpreted the close-
ended response categories in the same way. Due to the complex, reciprocal nature of human
behavior, motivation, and environmental effects, in this study no claims of causality are made.
Finally, limitations regarding bias affected this study. Bias describes systematic and
unknown error in results or inferences (Creswell, 2008). The study employed close-ended survey
responses and quantitative design and analysis in attempt to minimize bias although it is
acknowledged that completely eliminating bias is unlikely or even impossible. Efforts were
made to reduce the effects of researcher bias, given that the researcher is employed at the
institution in the capacity of Minority Engineering Program coordinator and NHSEMP director.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

136
In efforts to stem influence on potential study participants, the researcher took sabbatical leave
over the semester that data collection took place and had little to no contact via email, phone,
course instruction, or face-to-face interaction with potential participants.
Response bias, if students reported higher or lower measures consciously or
subconsciously based on their expectations of the study, are a limitation. Overrepresentation of
Native Hawaiians in the sample may have been a result of awareness of the researcher or it may
be the case that more Native Hawaiians were drawn to the study given that the title identified
Native Hawaiian students as a particular group of interest. Different levels of participation
among Native Hawaiians and non-Hawaiians could influence the results such as if the Native
Hawaiian sample included more students that normally would not complete the survey or if
Native Hawaiians exhibited a more pronounced response bias. The study attempted to minimize
bias, where possible, by research design and approach.
Ho’okahua Conceptual Framework
Based on the major findings of this study, a conceptual framework is proposed to inform
policy, practice, and research discussions in Native Hawaiian education, especially in regard to
NH STEM post-secondary education. The ho’okahua, or foundation building, framework is
modeled after Native Hawaiian construction of dry-stacking volcanic rock to form walls
retaining soil platforms as kahua or foundations. The traditional foundations provided space for
residential, religious, agricultural, and recreational purposes. Likewise, the proposed ho’okahua
framework provides a method to support future activities, practices, and action items.
Educational initiatives or action items should be based on sound theory or data.
Likewise, dry-stacking volcanic rock upon each other without mortar or joints requires different
pieces to fit and for each successive stone to rest by gravity on at least three points of contact.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

137
The analogy follows that the recommendations are built on three overarching findings of the
study. First, STEM self-efficacy leads to intention and motivation to persist in STEM. Second,
student self-beliefs about their abilities in STEM are derived from their sense of belonging in
STEM. Third, Native Hawaiian STEM students engage in high levels of involvement and
interaction with family, peers, and STEM programs. Figure 2 displays the Ho’okahua
framework.

Figure 2. Ho’okahua conceptual framework

Action items at the center of the Ho’okahua model are akin to a stone placed on the
second level of the wall. The three points of contact proposed for each successive stone selected
are the three overarching ideas derived from this study describing key beliefs and behaviors of
Native Hawaiian STEM undergraduate majors. This framework guides the successive
recommendations for practice and future research.
!
Sense of
Belonging
in STEM
Family, Peer,
and Major
Involvement
Self-Efficacy
Action
Item
Figure 5.1. Ho’okahua Conceptual Framework
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

138
Implications for Practice
The findings provide insight for four implications for practice. As the specific motivation
of this study was the goal of improving college outcomes for Native Hawaiian STEM majors, the
recommendations will focus on implications for the University of Hawaii with potential for
transferability to other settings to be determined by the reader. The four recommendations relate
to academic community, first-year learning communities, Native Hawaiian STEM programs, and
decentralized advising and student support.
Academic Community
The first implication is that the institution and STEM departments should strive to build
academic communities for students. Although GPA was a close second, the strongest
determinant of a student’s self-perceived ability to complete their degree was if they felt a part of
their major. More so, sense of belonging to major was the strongest determinant for STEM major
self-efficacy, which in turn was the strongest determinant for decision to persist at the institution
in their STEM major. This was found true for both Native Hawaiian and non-Hawaiian students.
Academic community is built around relationships and communication between students,
a supportive peer network, faculty, staff, and academic themes or goals. Good, Rattan, and
Dweck (2012) describe sense of belonging in an academic domain as viewing oneself as being
inside a discipline rather than on the fringes of it and a sense of being valued and accepted by
fellow members of the discipline. Although early research on sense of belonging focused on
campus climate (Hurtado & Carter, 1997) and social integration to the institution (Tinto, 1993),
this study argues that academic community at the major or department level is most important
for undergraduate STEM students. The overall recommendation is to promote student
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

139
involvement in the major. This recommendation follows the ho’okahua framework considering
the impact of self-efficacy, sense of belonging, and involvement.
First Year Learning Communities
Specific academic community should be encouraged at the first-year, academic major
level. Co-registration or block scheduling that enables students to take courses together can
change the way students experience the curriculum, their sense of belonging to a major, common
theme, and academic community. Tinto (2003) identifies three things that all variations of
learning communities have in common: shared knowledge; shared knowing; and shared
responsibility. This is particularly important for first-year students given this study’s findings
that the first year students are likely to have the lowest levels of STEM self-efficacy and lowest
levels of intentions to persist. More so, existing curricula are structured such that many STEM
students have limited or no courses or contact with their STEM major in their first year.
The recommendation is for the institution to offer and encourage first-year students to
participate in structured learning communities including designated learning communities for
Native Hawaiian, high-risk students, and advanced students. The UHM College Opportunities
Program (COP), a state-funded program that serves primarily Native Hawaiian, Filipino, first-
generation, and other students with low admissions credentials but high potential, co-enrolls
cohorts of students the summer prior to freshmen year and requires students to meet with staff
and program mentors to during throughout their freshmen year. The federally funded Native
Hawaiian Science & Engineering Mentorship Program (NHSEMP) requires co-enrollment (in
Math, Chemistry, Hawaiian Studies, and Introduction to Engineering) for their first-year Native
Hawaiian engineering students. Both programs have reported successful outcomes related to
student involvement and in-major persistence. Bridge programs, learning communities, and first
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

140
year student services support are recommended strategies for improving Native Hawaiian STEM
outcomes.
Native Hawaiian STEM Programs
The University can improve college outcomes by supporting and implementing Native
Hawaiian STEM programs. Program participation was found to be significant in higher STEM
self-efficacy (which in turn predicted higher persistence) and significantly higher among Native
Hawaiians. The programs that support undergraduate research, such as the Honors Program,
NHSEMP, C-MORE Scholars Program, and Undergraduate research and mentoring, promote
faculty interaction and provide students an important view of their academic discipline beyond a
provider of classroom instruction. Ethnic enclaves in STEM higher education were found to be
important for URM students to find support within unknown or chilly climates (Cole &
Espinoza, 2008; Hurtado & Carter, 1997). Ortiz and Santos (2009) found ethnic membership
identity (sense of worth derived from one’s ability to contribute to the ethnic group) significantly
correlated to college efficacy, social efficacy, academic efficacy, and self-esteem.
Native Hawaiian programs should place particular emphasis on the social-cultural aspects
of learning given the finding that Native Hawaiian STEM students display higher levels of
interaction with their peers and in promoting behavior that encourages sense of belonging to their
academic discipline. This can be achieved through mentoring programs (by faculty,
upperclassmen, or role models), collaborative learning experiences such as projects or research,
and in clarifying the values of the discipline and its connection to the Native Hawaiian and
broader community.
At UHM, Native Hawaiian underrepresentation persists in all STEM colleges but has
particular opportunity for improvement in Natural Sciences where Native Hawaiians in this study
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

141
were found to be lower enrolled. As Engineering was found to be a positive indicator of STEM
self-efficacy and higher enrolled among Native Hawaiians, there may be an emerging critical
mass. SOEST, which relative to the other STEM colleges, has a low overall undergraduate
enrollment are challenged to grow a Native Hawaiian community of peers with low numbers of
students.
Decentralized Advising, Student Support, and STEM Curriculum
The recommendation, for increasing self-beliefs of ability, involvement, and sense of
belonging, is for each STEM college or department to provide regular or mandatory advising and
student services for their majors and pre-majors. In addition, staff and faculty should receive the
training and support necessary to be effective in developing a connection with students. Hovland
et al. (1997) argues the best way to keep students stimulated, challenged, and progressing toward
a meaningful goal is through informed academic advising.
Student services support, via Native Hawaiian programs or other, at the college or
departmental level can also enhance the student’s network and sense of community in their
STEM major. Enrollment management tools can be used to better facilitate communication with
between majors and departments for purposes such as advising, scheduling tutors or other
interventions, tracking, and two-way communication. For example, enrollment management
software utilized at the UHM Athletics department allows for regular email, text messaging, and
group conversation between advisors, faculty, tutors, and scholar athletes that changes the style
and frequency of communication, involvement, and engagement at the University.
Additional conversation regarding common STEM curriculum is recommended. This
study choose to investigate intent to persist and self-efficacy on a general STEM level such that
for a student that start as a mathematics major but changes to physics major were considered a
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

142
persisting STEM student. Some researchers (Ohland, Sheppard, Lichtenstein, Eris, Chachra, &
Layton, 2008) argue that institutional retention rates in STEM majors can be improved by better
aligning student migration between like majors. For example, Ohland et al. (2008) found that
engineering majors persisted at or above the same rate as all other majors studied, but the field
suffered from the lowest rate of inward migration. Common STEM curriculum between like
majors could benefit the overall persistence of undergraduate students in the STEM pathways.
Future Research
There exists a limited number of studies focusing on Native Hawaiians in post-secondary
education and less so in STEM education. Future research is needed to provide a better
understanding of the dynamics of Native Hawaiian education and outcomes. Based on the
findings and limitations of this study, future research is proposed in varying research design,
non-Hawaii students, pre-college Native Hawaiian STEM education, interaction effects,
community college STEM pathways, Native Hawaiian identity in STEM, and post-baccalaureate
transition and training. The importance of self-efficacy, sense of belonging, and involvement has
been derived from NH undergraduate STEM students, but further research is needed to evaluate
the applicability of the ho’okahua framework for other populations.
Interaction Effects and Assessment of Findings
College behaviors, beliefs, and school outcomes operate in an interdependent, complex
manner. Future research can explore interaction effects or structure variables to investigate a
range of important and interesting questions. For example, what factors best contribute to STEM
self-efficacy for females in engineering, physics, and computer sciences? How do the dynamics
of behaviors, attitudes, and outcomes for Native Hawaiians differ for Engineering, Ocean and
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

143
Earth Science, Natural Science, and Tropical Agriculture majors? What input or environmental
factors explain sense of belonging?
Three results of this study in particular warrant further investigation: a higher amount of
family support among Native Hawaiians; program participation; and sense of belonging. Are
these findings consistent with Native Hawaiians in non-STEM majors? If Native Hawaiian
families show higher approval and encouragement of student’s post-secondary goals and
completion, then how can educational institutions utilize this information? How does sense of
belonging to campus community interact with sense of belonging to major? These questions, for
example, offer potential for intriguing research.
Qualitative, Mixed-Method, and Longitudinal Design
This study chose to approach the research topic from a quantitative method. Future
qualitative or mixed-method research is needed to provide more vivid description and
understanding of the issues facing Native Hawaiians in STEM. For example, family support and
program participation were found significant in increasing STEM self-efficacy and higher for
Native Hawaiians. Qualitative study could further investigate how Native Hawaiian students
experience these factors.
In addition, longitudinal research that tracks individuals or cohorts can provide a valuable
perspective and understanding of changes over time. How and why do internal (motivation,
satisfaction, self-efficacy) and external (peer/faculty interaction, involvement) factors change as
a student experiences college? Educational level was found significant in this cross-sectional
study and assumptions had to be made to interpret past and future events as well as comparison
between students at different educational levels. Longitudinal research has the opportunity to
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

144
provide more conclusive results by investigating the same students or cohorts over time and
utilizing data from confirmed events such as actual persistence and actual leavers.
Native Hawaiian Education in the Continental U.S.
To better inform Native Hawaiian education, further research is needed investigating
Native Hawaiian education in the continental U.S. The majority (88.7%) of Native Hawaiian
college and graduate students in the state of Hawaii attend the University of Hawaii (KSP, 2005).
Therefore, the University of Hawaii is a logical setting to research Native Hawaiian college
outcomes for a large percentage of the Native Hawaiian college-going population. However,
many Native Hawaiian residents chose to attend post-secondary institutions out of state. In
addition, the 2010 U.S. Census data show that 45% of the domestic Native Hawaiian population
resides in Alaska or the continental U.S. Studies that can capture findings for this understudied
group will address a void in the research.
There are challenges for researchers in Native Hawaiian education. Many post-secondary
institutions on the continental U.S. do not have enough Native Hawaiian STEM student
enrollments to provide adequate sample sizes to make powerful statistical claims. Studies that
focus on the experience of URM students in general may not easily transfer to indigenous or
Native Hawaiian students. It is important, however, that Native Hawaiian education and policy
be informed by the best data available.
The Pre-College STEM Pipeline
Research focusing on existing Native Hawaiian college students is similar to fishing with
a broken throw net. Even with improved technique, only a fraction of the potential outcomes can
be realized. Improvements to Native Hawaiian education at the pre-college level especially
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

145
within math and science education are foundational to improve the quantity and quality of Native
Hawaiian college-going students.
This study offers some potential direction for research at the pre-college level. Peer
interaction, family support, and intent to persist were found to significantly higher for Native
Hawaiians than their non-Hawaiian counterparts. Do these findings translate into elementary or
secondary settings? How can these findings inform potential STEM majors?
The key elements of the ho’okahua framework related to self-efficacy, sense of
belonging, and involvement can be vetted at the pre-college level. Outside of the scope of this
study are critical questions needed to address the shortage of Native Hawaiians (and others) in
the STEM pipeline. What influences the motivations of pre-college Native Hawaiian students to
pursue STEM? What influences STEM self-efficacy at the Pre-college levels?
Community College Pathways
Although this study focused on students enrolled the University of Hawaii’s four-year
campus, a large and growing percentage of the Native Hawaiians college-going population are
found at the University of Hawaii community college system. Many Native Hawaiians earn
certificates and associates degrees at the community colleges that prepare them for technology
jobs and many begin at the community colleges, earn pre-STEM college credits, transfer, and
graduate with a bachelors degree in STEM. Future research can test or extend this study’s
findings with the community college and/or transfer student population to better understand the
matriculation and persistence of the community college to workforce and community college to
BS degree pathways.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

146
Native Hawaiian Congruence
Cultural congruence of Native Hawaiian identity in STEM is certainly an exciting area of
research (Kanaʻiaupuni, Ledward, & Keohokalole, 2012). The premise is that education is a
cultural process. When educational institutions are rooted in a cultural worldview in which
mainstream Western values, knowledge, and practices are the norm, then non-Western and
indigenous students are disadvantaged and educational disparities may result. Native Hawaiian
identity and cultural congruence were not investigated in this study though it is likely to be
associated with key factors of the ho’okahua framework such as sense of belonging, self-
efficacy, and family and peer interaction. Future research on Native Hawaiian cultural
congruence at all levels (early childhood, math and science primary and secondary education,
higher education, and informal education) may provide insight into teaching, learning, and
outcomes.
Post-Baccalaureate Transition and Preparation
Completing a STEM major does not, by itself, lead to persistence into the STEM
workforce. Similarly, STEM education and STEM employment does not always lead to
innovation or economic benefit as is desired to increase the nation’s competitiveness. Future
research can address questions such as what factors lead to STEM employment? Why do STEM
graduates pursue certain career and life trajectories such as teaching, graduate education,
industry, entrepreneurship, or non-STEM careers? What skills, knowledge, and preparation do
STEM graduates need for these trajectories and how well are educational institutions aligned
with providing them? What is the role of industry, non-profit sector, and informal education in
meeting the needs of the STEM workforce? These questions relate the broader problem to
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

147
address national competitiveness and workforce outcomes that extend beyond the scope of this
study.
Conclusion
The National Research Council (2007, 2010) argues that if U.S. educational institutions
improved the recruitment, retention and success rates of minority students in STEM, then the
country would be better equipped to innovate, compete, and problem solve. Beyond
international competitiveness, many of the world’s current and future challenges such as energy
dependence, climate change, and scarcity of natural resources will require STEM solutions. To
address these global issues on a local scale, the educational system plays a leading role in
developing the talent of all students. Indeed, educators are as vital to future opportunity as the
scientist or the engineer.
This study attempted to address the broad challenge of America’s ‘quiet crisis’ by
examining college outcomes for Native Hawaiian and non-Hawaiian STEM majors at the
University of Hawaii at Manoa. Significant predictors of STEM self-efficacy and intent to
persist were found for all students and the significant differences between Native Hawaiian and
non-Hawaiian students were presented. The ho’okahua framework presents the key concepts of
the findings of STEM self-efficacy, sense of belonging, and involvement. It is hoped that this
study makes some contribution to the literature supporting educational policy, practice, and
future research especially in regard to Native Hawaiian education. The additional translations of
ho’okahua refer to the community and values that are needed to build a firm foundation as well
as a degree of commitment to settle down to a task with determination to see it through (Pukui &
Elbert, 1986). Native Hawaiians come from a system of beliefs, values, and traditions that
demonstrate excellence in science, engineering, and education and can no longer be underserved
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

148
or underperform. The charge to educators and the kuleana (responsibility) to Native Hawaiians
are to effect a positive change in the readiness, self-efficacy, and achievement of Native
Hawaiians in the STEM pathways.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

149
REFERENCES
Adelman, C. (1999). Answers in the toolbox: Academic intensity, attendance patterns, and
Bachelor’s degree attainment. Washington, D.C.: U.S. Department of Education, Office
of Educational Research and Improvement.
Ah Sam, A. L., & Robinson, N. B. (1998). Pacific Islanders in higher education: Barriers to
recruitment and retention. Pacific Educational Research Journal, 9(1), 39-49.
Alu Like, Inc. (1988). Native Hawaiian students at the University of Hawaii: Implications for
vocational and higher education. Honolulu, HI: Author.
American Educational Research Association, American Psychological Association, & National
Council on Measurement in Education. (1999). Standards for educational and
psychological testing. Washington, D.C.: American Educational Research Association.
Astin, A. (1975). Preventing students from dropping out. San Francisco, CA: Jossey-Bass.
Astin, A. (1977). Four critical years. San Francisco, CA: Jossey-Bass.
Astin, A. (1993). What matters in college? Four critical years revisited. San Francisco, CA:
Jossey-Bass.
Astin, A. (1999). Student involvement: A developmental theory for higher education. Journal of
College Student Development, 40(5), 518-529.
Babbie, E. (1990). Survey research methods. Belmont, CA: Wadsworth.
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change.
Psychological Review, 84, 191-215.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Englewood Cliffs, NJ: Prentice-Hall.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York, NY: Freeman.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

150
Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual Review of
Psychology, 52, 1-26.
Bartlett, M. S. (1954). A note on the multiplying factors for various χ2 approximations. Journal
of the Royal Statistical Society, Series B (Methodological), 16(2), 296-298.
Bean, J. P. (1980). Dropouts and turnover: The synthesis and test of a causal model of student
attrition. Research in Higher Education, 12(2), 155-187.
Bean, J. P. (1985). Interaction effects based on class level in an explanatory model of college
dropout syndrome. American Educational Research Journal, 22(1), 35-64.
Benham, M. K. P. (2006). A challenge to Native Hawaiian and Pacific Islander scholars: What
the research literature teaches us about our work. Race Ethnicity and Education, 9(1), 29-
50.
Besterfield-Sacre, M., Altman, C., & Shuman, L. (1997). Characteristics of freshmen
engineering students: Models for determining student attrition in engineering, Journal of
Engineering Education, 86(1), 139-149.
Betz, N. E., & Hackett, G. (1981). The relationship of career-related self-efficacy expectations to
perceived career options in college women and men. Journal of Counseling Psychology,
28, 399-410.
Betz, N. E., & Schifano, R. (2000). Evaluation of an intervention to increase realistic self-
efficacy and interests in college women. Journal of Vocational Behavior, 56(1), 35-52.
Boskin, M. J., & Lau, L. J. (1992). Capital, technology, and economic growth. In N. Rosenberg,
R. Landau, & D. Mowery (Eds.), Technology and the wealth of nations (pp. 17-55).
Stanford, CA: Stanford University Press.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

151
Briggs, S. R., & Cheek, J. M. (1986). The role of factor analysis in the development and
evaluation of personality scales. Journal of Personality, 54(1), 106-148.
Burke, R., & Mattis, M. (2007). Women and minorities in science, technology, engineering and
mathematics upping the numbers. Northampton, MA: Edward Elgar Publishing.
Byars, A., & Hackett, G. (1998). Applications of social cognitive theory to the career
development of women of color. Applied and Preventive Psychology, 7, 255-267.
Cabrera, A., Castaneda, M., Nora, A., & Hengstler, D. (1992). The convergence between two
theories. The Journal of Higher Education, 63(2), 143-164.
Cady, E., Fortenberry, N., Drewery, M., & Bjorklund, S. (2009). Development and validation of
surveys measuring student engagement in engineering, part 2. Proceedings of the
Research in Engineering Education Symposium 2009, Palm Cove, QLD, Australia.
Choy, S. (2001). Students whose parents did not go to college: Postsecondary access,
persistence, and attainment (NCES 2001-126). Washington, D.C.: U.S. Department of
Education, National Center for Education Statistics.
Clark, L. A., & Watson, D. (1995). Constructing validity: Basic issues in objective scale
development. Psychological Assessment, 7(3), 309-319.
Clewell, B. C., de Cohen, C., Tsui, L., & Deterding, N. (2006). Revitalizing the nation’s talent
pool in STEM. Arlington, VA: National Science Foundation, Directorate for Education
and Human Resources.
Cole, D. (2007). Do interracial interactions matter? An examination of student-faculty contact
and intellectual self-concept. Journal of Higher Education, 78(3), 248-272.
Cole, D., & Ahmadi, S. (2010). Reconsidering campus diversity: An examination of Muslim
students’ experiences. The Journal of Higher Education, 81(2), 121-139.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

152
Cole, D., & Espinoza, A. (2008). Examining the academic success of latino students in science,
technology, engineering and mathematics (STEM) majors. Journal of College Student
Development, 49(4), 285-300.
Cole, D., & Espinoza, A. (2009). When gender is considered: Racial ethnic minority students in
STEM majors. Journal of Women and Minorities in Science and Engineering, 15, 263-
277.
Committee on Science, Engineering, and Public Policy. (2011). Expanding underrepresented
minority participation: America’s science and technology talent at the crossroads.
Washington, D.C.: The National Academies Press.
Creswell, J. W. (2008). Research design: Qualitative, quantitative, and mixed methods
approaches. Thousand Oaks, CA: Sage.
Day, J., & Newburger, E. (2002). The big payoff: Educational attainment and synthetic estimates
of work-life earnings (Current Population Reports, P23-210). Washington, D.C.: U.S.
Census Bureau.
DeVellis, R. F. (2012) Scale development: Theory and applications. Washington, D.C.: Sage.
Donaldson, K., Lichtenstein, G., & Sheppard, S. (2008). Socioeconomic status and the
undergraduate engineering experience: Preliminary findings from four American
universities (Research brief). Seattle, WA: Center for the Advancement of Engineering
Education.
Duderstadt, J. (2008). Engineering for a changing world: A roadmap to the future of
engineering, practice, research, and education. Ann Arbor, MI: University of Michigan,
The Millennium Project.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

153
Entwisle, D. R., & Astone, N. M. (1994). Some practical guidelines for measuring youth’s
race/ethnicity and socioeconomic status. Child Development, 65(6), 1521-1540.
Fink, A. (2013). How to conduct surveys: A step-by-step guide (5th ed.). Thousand Oaks, CA:
Sage.
Fischer, M. (2007). Settling into campus life: Differences by race/ethnicity in college
involvement and outcomes. The Journal of Higher Education, 78(2), 125-156.
Flores, L. Y., & O’Brien, K. M. (2002). The career development of Mexican American
adolescent women: A test of social cognitive career theory. Journal of Counseling
Psychology, 49(1), 14.
Friedman, T. (2006). The world is flat: A brief history of the twenty-first century. New York,
NY: Farrar, Straus, & Giroux.
Good, C., Rattan, A., & Dweck, C. (2012). Why do women opt out? Sense of belonging and
women’s representation in mathematics. Journal of Personality and Social Psychology,
102(4), 700-717.
Hagedorn, L., & Tibbetts, K. (2003, September). Factors contributing to college retention in the
Native Hawaiian population. Paper presented at Kamehameha Schools Research
Conference, Honolulu, HI.
Hausman, L., Schofield, J., & Woods, R. (2007). Sense of belonging as a predictor of intentions
to persist among African American and white first-year college students. Research in
Higher Education, 48(7), 803-839.
Hokoana, L. (2010). Native Hawaiians and college success: Does culture matter? (Unpublished
doctoral dissertation). University of Southern California, Los Angeles, CA.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

154
Hovland, M. J., Anderson, E. C., McGuire, W. G., Crockett, D., Kaufmann, J., & Woodward, D.
(1997). Academic advising for student success and retention. Iowa City, IA: Noel/Levitz.
Hsu, P., & Nielson, J. (2010). Population update 2010: The R&E annual update series.
Honolulu, HI: Kamehameha Schools Research & Evaluation.
Huang, G., Taddese, N., Walter, E., Peng, S. S. (2000). Entry and persistence of women and
minorities in college science and engineering education (NCES 2000-601). Washington,
D.C.: U.S. Department of Education, Office of Educational Research and Improvement,
National Center for Education Statistics. Retrieved from
http://nces.ed.gov/pubs2000/2000601.pdf
Hurtado, S., & Carter, D. (1997). Effects of college transition and perceptions of the campus
racial climate on Latino college students’ sense of belonging. Sociology of Education,
70(4), 324-345.
Hurtado, S., Newman, C. B., Tran, M. C., & Chang, M. J. (2010). Improving the rate of success
for underrepresented racial minorities in STEM fields: Insights from a national project.
New Directions for Institutional Research, 2010(148), 5-15.
Jackson, S. (2004). The quiet crisis: Falling short in producing American scientific and technical
talent. San Diego, CA: Building Engineering & Science Talent. Retrieved from
http://www.bestworkforce.org/publications.htm
Jordan, K., Sorby, S., & Amato-Henderson, S. (2012). Pilot intervention to improve “sense of
belonging” of minorities in engineering. Proceedings of American Society of Engineering
Education, 2012. Washington, D.C.: American Society of Engineering Education.
Kaiser, H. F. (1970). A second generation Little Jiffy. Psychometrika, 35(4), 401-415.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

155
Kaiser, H. F., & Rice, J. (1974). Little Jiffy, Mark IV. Educational and Psychological
Measurement, 34, 111-117.
Kamehameha Schools Press. (2005). Ka Huaka’i: 2005 Native Hawaiian educational
assessment. Honolulu, HI: Author.
Kamehameha Schools Press. (2009). Native Hawaiian educational assessment update, 2009: A
supplement to Ka Huaka‘i, 2005. Honolulu, HI: Author.
Kame’eleihiwa, L. (1992). Native land and foreign desires: Pehea la e pono ai? Honolulu, HI:
Bishop Museum Press.
Kanaʻiaupuni, S., Ledward, B., & Keohokalole, K. (2012). New research on the impact of
cultural influences in education on Native Hawaiian student outcomes. AAPI Nexus:
Asian Americans & Pacific Islanders Policy, Practice, and Community, 9(1), 221-229.
Kawakami, A. J. (1999). Sense of place, community, and identity: Bridging the gap between
home and school for Hawaiian students. Education and Urban Society, 32(1), 18-40.
Kuh, G. (1993). In their own words: What students learn outside the classroom. American
Educational Research Journal, 30(2), 277-304.
Kuh, G., & Hu, S. (2001). The effects of student-faculty interaction in the 1990s. The Review of
Higher Education, 24(3), 309-332.
Kumashiro, K. K. (2006). Toward an anti-oppressive theory of Asian Americans and Pacific
Islanders in education. Race Ethnicity and Education, 9(1), 129-135.
Landis, R. (Ed.) (1985). Handbook on improving the retention and graduation of minorities in
engineering. White Plains, NY: National Action Council for Minorities in Engineering.
Landivar, L. (2013). Disparities in STEM employment by sex, race, and Hispanic origin.
American Community Survey Reports (ACS-24). Washington, D.C.: U.S. Census Bureau.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

156
Leggon, C., & Pearson, Jr., W. (2009). Assessing programs to improve minority participation in
STEM fields: What we know and what we need to know. Retrieved from
http://www.advance.rackham.umich.edu/ncid/assessing_programs.pdf
Lent, R. W. (2013). Social cognitive career theory. In S. Brown & R. Lent (Eds.), Career
development and counseling: Putting theory and research to work. Somerset, NJ: Wiley.
Lent, R. W., Brown, S. D., & Hackett, G. (1994). Toward a unifying social cognitive theory of
career and academic interest, choice, and performance. Journal of Vocational Behavior,
45, 79-122.
Lent, R. W., Brown, S. D., & Hackett, G. (2000). Contextual supports and barriers to career
choice: A social cognitive analysis. Journal of Counseling Psychology, 47, 36-49.
Makuakane-Drechsel, T., & Hagedorn, L. S. (2000). Correlates of retention among Asian Pacific
Americans in community colleges: The case for Hawaiian students. Community College
Journal of Research and Practice, 24(8), 639-655.
Marra, R., & Bogue, B. (2006). Women engineering students’ self-efficacy — A longitudinal
multi-institution study. Proceeding of the 2006 Women in Engineering Programs and
Advocates Network (WEPAN) Conference, Pittsburgh, PA.
Matsumoto, D. (2010). Factors affecting Native Hawaiian student persistence in higher
education. Retrieved from ProQuest Dissertations. (UMI No. 3403612)
Multon, K. D., Brown, S. D., & Lent, R. W. (1991). Relation of self-efficacy beliefs to academic
outcomes: A meta-analytic investigation. Journal of Counseling Psychology, 38, 30-38.
Museus, S., & Liverman, D. (2010). High-performing institutions and their implications for
studying underrepresented minority students in STEM. New Directions for Institutional
Research, 148, 17-27. doi:10.1002/ir.358
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

157
National Research Council. (2007). Rising above the gathering storm: Energizing and employing
America for a brighter future. Washington, D.C.: The National Academies Press.
National Research Council. (2010). Rising above the gathering storm, revisited: Rapidly
approaching category 5. Washington, D.C.: The National Academies Press.
National Science Board. (2010). Science and engineering indicators, 2010 (NSB 10-01).
Arlington, VA: National Science Foundation.
National Science Foundation. (2011). Women, minorities, and persons with disabilities in
science and engineering: 2011 (Special Report NSF 11-309). Arlington, VA: National
Science Foundation, Division of Science Resources Statistics. Retrieved from
http://www.nsf.gov/statistics/wmpd/
Nunnally, J. (1978). Psychometric methods. New York, NY: McGraw.
Obama, B. (2011, January 25). Remarks by the President in the State of Union Address, United
States Capitol, Washington, D.C. Whitehouse.gov. Retrieved from
http://www.whitehouse.gov/the-press-office/2011/01/25/remarks-president-state-union-
address
Ohland, M., Sheppard, S., Lichenstein, G., Eris, O., Chachra, D., & Layton, R. (2008).
Persistence, engagement, and migration in engineering programs. Retrieved from
http://digitalcommons.olin.edu/facpub_2008/3
Oliveira, J. (2005). Native Hawaiians’ success in higher education: Predictive factors and
Bachelor’s degree completion (Unpublished doctoral dissertation). University of
Southern California, Los Angeles, CA.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

158
Organization for Economic Cooperation and Development. (2009). Education at a glance, 2009:
OECD Indicators. Paris: Author. Retrieved from: http://www.oecd.org/education/skills-
beyond-school/43636332.pdf
Ortiz, A. M., & Santos, S. J. (2009). Ethnicity in college: Advancing theory and improving
diversity practices on campus. Arlington, VA: Stylus.
Pacific Policy Research Center. (2010). Influential factors in degree attainment and persistence
to career or further education for at-risk/high educational needs students. Honolulu, HI:
Kamehameha Schools, Research & Evaluation Division.
Pajares, F. (1996). Self-efficacy beliefs in academic settings. Review of Educational Research,
66(4), 543-578.
Pajares, F., & Urdan, T. (Eds.) (2006) Adolescence and education, Vol. 5: Self-efficacy beliefs of
adolescents. Greenwich, CT: Information Age.
Pallant, J. (2013). SPSS survival manual: A step by step guide to data analysis using IBM SPSS
(5th ed.). New York, NY: McGraw Hill Education.
Pascarella, E. T., & Chapman, D. (1983). A multiinstitutional path analytic validation of Tinto’s
model of college withdrawal. American Educational Research Journal, 20(1), 87-102.
Pascarella, E. T., & Terenzini, P. T. (1991). How college affects students: Findings and insights
from twenty years of research. San Francisco, CA: Jossey-Bass.
Pascarella, E. T., & Terenzini, P. T. (2005). How college affects students: A third decade of
research (Vol. 2). San Francisco, CA: Jossey-Bass.
Perna, L. (2000). Differences in the decision to attend college among African Americans,
Hispanics, and Whites. The Journal of Higher Education, 71(2), 117-141.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

159
Pintrich, P. (2003). A motivational science perspective on the role of student motivation in
learning and teaching contexts. Journal of Educational Psychology, 95(4), 667-686.
Pukui, M., & Elbert, S. (1986). Hawaiian dictionary. Honolulu: University of Hawaii Press.
Rittmayer, M. A., & Beier, M. E. (2009). Self-efficacy in stem. In B. Bogue & E. Cady (Eds.).
Applying Research to Practice (ARP) Resources. Retrieved from
http://www.engr.psu.edu/AWE/ARPresources.aspx
Robbins, S. B., Lauver, K., Le, H., Davis, D., Langley, R., & Carlstrom, A. (2004). Do
psychosocial and study skill factors predict college outcomes? A meta-analysis.
Psychological Bulletin, 130(2), 261-288.
Rottinghaus, P., Larson, L., & Borgen, F. (2003). The relation of self-efficacy and interests: A
meta-analysis of 60 samples. Journal of Vocational Behavior, 2, 221-236.
Sandler, B. R. (1999). The chilly climate: Subtle ways in which women are often treated
differently at work and in classrooms. Retrieved from
http://www.bernicesandler.com/id23.htm
Schunk, D. H., & Pajares, F. (2002). The development of academic self-efficacy. In A. Wigfield
& J. S. Eccles (Eds.), Development of achievement motivation: A volume in the
educational psychology series (pp. 15-31). San Diego, CA: Academic Press.
Sheppard, S., Gilmartin, S., Chen, H., Donaldson, K., Lichtenstein, G., Eris, O., . . . Toye, G.
(2010). Exploring the engineering student experience: Findings from the Academic
Pathways of People Learning Engineering survey (APPLES) (TR-10-01). Seattle, WA:
Center for the Advancement for Engineering Education.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

160
Sheu, H., Lent, R., Brown, S., Miller, M., Hennessy, K., & Duffy, R. (2010). Testing the choice
model of social cognitive career theory across Holland themes: A meta-analytic path
analysis. Journal of Vocational Behavior, 76, 252-264.
Sirin, S. R. (2005). Socioeconomic status and academic achievement: A meta-analytic review of
research. Review of Educational Research, 75(3), 417-453.
Slaughter, J. B. (2008). Confronting the “new” American dilemma. Washington, D.C.: National
Action Council for Minorities in Engineering.
Smith, T., Wilson, D., Jones, D., Plett, M., Bates, R., & Veilleux, N. (2012). Investigation of
belonging for engineering and science undergraduates by year in school (AC 2012-
5111). Washington, D.C.: American Society for Engineering Education.
Speck, B., & Keahiola-Karasuda, R. (2011). Post-high update 2011: Native Hawaiians in post-
secondary education. Honolulu, HI: Kamehameha Schools Research & Evaluation.
Stannard, D. E. (1988). Before the horror: The population of Hawaii on the eve of western
contact. Honolulu, HI: University of Hawaii Press.
Suskie, L. A. (1996). Questionnnaire survey research: What works (2nd ed.). Tallahassee, FL:
The Association for Institutional Research.
Swail, W. S., & Perna, L. W. (2002). Pre-college outreach programs: A national perspective. In
W. G. Tierney & L. S. Hagedorn (Eds.), Extending their reach: Strategies for increasing
access to college (pp. 15-34). Albany, NY: SUNY Press.
Tabachnick, B., & Fidell, L. (2007). Using multivariate statistics. Boston, MA: Pearson
Education.
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

161
Tierney, W. (2004). Power, identity, and the dilemma of college student departure. In J. M.
Braxton (Ed.), Reworking the student departure puzzle. Nashville, TN: Vanderbilt
University Press.
Tinto, V. (1975). Dropout from higher education: A theoretical synthesis of recent research.
Review of Educational Research, 45(1), 89-125.
Tinto, V. (1987). Leaving college: Rethinking the causes and cures of student attrition. Chicago,
IL: University of Chicago Press.
Tinto, V. (1993). Leaving college: Rethinking the causes and cures of student attrition (2nd ed.).
Chicago, IL: The University of Chicago Press.
Tinto, V. (2003). Learning better together: The impact of learning communities on student
success (Higher education monograph series). Syracuse University, NY: School of
Education, Higher Education Program.
University of Hawaii Board of Regents. (2012). Bylaws of the Board of Regents of the University
of Hawaii. Honolulu, HI: Author. Retrieved from
http://www.hawaii.edu/offices/bor/policy/index.html
University of Hawai’i Hawaiian Studies Task Force. (1986). Ka’u: University of Hawai’i
Hawaiian Studies task force report (Unpublished manuscript). Honolulu, HI: University
of Hawai’i at Manoa.
University of Hawaii Institutional Research Office. (2009). Peer and benchmark comparison
groups, University of Hawaii 2009. Honolulu, HI: Author. Retrieved from
http://www.hawaii.edu/iro/
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

162
University of Hawaii Institutional Research Office. (2010). Hawaiian student headcount
enrollment by college and major. Honolulu, HI: Author. Retrieved from
http://manoa.hawaii.edu/ovcaa/mir/?page=fac_enr
University of Hawaii Institutional Research Office. (2012). Fall 2012 enrollment report.
Honolulu, HI: Author. Retrieved from:
https://www.hawaii.edu/institutionalresearch/home.action
University of Hawaii at Manoa. (2011). Achieving our destiny: the University of Hawaii at
Manoa 2011-2013 strategic plan. Honolulu, HI: Author. Retrieved from
http://manoa.hawaii.edu/vision/pdf/achieving-our-destiny.pdf
University of Hawaii at Manoa Native Hawaiian Advancement Task Force. (2012). Ke au hou: A
new beginning. Honolulu, HI: Author. Retrieved from
http://manoa.hawaii.edu/chancellor/NHATF/
University of Hawaii Office of the Vice President for Academic Planning & Policy. (2002).
University of Hawaiʻi system strategic plan: Entering the University’s second century,
2002-2010. Honolulu, HI: Author. Retrieved from
http://www.hawaii.edu/ovppp/stratplansys.html
University of Hawaii Office of the Vice President for Academic Planning & Policy. (2008).
Serving the State of Hawaii: University of Hawaii System strategic outcomes and
performance measures, 2008–2015. Honolulu, HI: Author. Retrieved from
http://www.hawaii.edu/ovppp/uhplan/
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

163
University of Hawaii Office of the Vice President for Student Affairs. (1999). College student
experiences at the University of Hawaii at Manoa in 1990, 1993, 1996, and 1999.
Honolulu, HI: Author. Retrieved from
http://studentaffairs.manoa.hawaii.edu/about/research_reports.php#undergraduates
U.S. Government Accountability Office. (2012). Science, technology, engineering, and
mathematics education: Strategic planning needed to better manage overlapping
programs across multiple agencies (GAO-12-108). Washington, D.C.: Author.
Walpole, M. (2003). Socioeconomic status and college: How SES affects college experiences
and outcomes. The Review of Higher Education, 27(1), 45-73.
World Economic Forum. (2010). The global information technology report, 2009-2010.
Retrieved from http://www.weforum.org/reports/global-information-technology-
report/index.htm
Zhang, G., Anderson, T., Ohland, M., Carter, R., & Thorndyke, B. (2002). Identifying factors
influencing engineering student graduation and retention: A longitudinal and cross-
institutional study. Proceedings of the American Society for Engineering Education
National Conference, Montreal, Canada.
Zimmerman, B. J. (2000). Self-efficacy: An essential motive to learn. Contemporary
Educational Psychology, 25, 82-91.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

164
APPENDIX A
UH HUMAN STUDIES PROGRAM APPROVAL

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

165
APPENDIX B
INFORMED CONSENT AND SURVEY INSTRUMENT

2/28/2014 Qualtrics Survey Software
https://s.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview&T=48FRi8 1/12
I n t r o d u c t i o n a n d I n f o r m e d C o n s e n t
U n i v e r s i t y o f S o u t h e r n C a l i f o r n i a
C o n s e n t t o P a r t i c i p a t e i n R e s e a r c h
S e l f - E f f i ca cy a n d I n t e n t i o n s t o P e r si st o f N a t i ve H a w a i i a n a n d N o n - H a w a i i a n
S ci e n ce , T e ch n o l o g y, E n g i n e e r i n g , a n d M a t h e m a t i cs M a j o r s I n t r o d u c t i o n
T h e p u r p o se o f t h i s st u d y i s t o a sse ss h o w st u d e n t s f e e l a b o u t co m p l e t i n g a S T E M d e g r e e a t U H
M a n o a . Y o u a r e i n vi t e d t o p a r t i ci p a t e i n a r e se a r ch st u d y co n d u ct e d b y g r a d u a t e st u d e n t Jo sh u a
K a a ku a a n d f a cu l t y a d vi so r D a r n e l l C o l e o f t h e U n i ve r si t y o f S o u t h e r n C a l i f o r n i a . Y o u a r e e l i g i b l e t o
p a r t i ci p a t e i n t h i s p r o j e ct b e ca u se yo u a r e a t l e a st 1 8 ye a r s o l d a n d e n r o l l e d a s a st u d e n t a t U H
M a n o a i n a S T E M co l l e g e / sch o o l . Y o u r p a r t i ci p a t i o n i s vo l u n t a r y. Y o u sh o u l d r e a d t h e i n f o r m a t i o n
b e l o w , a n d a sk q u e st i o n s a b o u t a n yt h i n g yo u d o n o t u n d e r st a n d , b e f o r e d e ci d i n g w h e t h e r t o
p a r t i ci p a t e .
Y o u r r e l a t i o n sh i p w i t h U S C o r U H w i l l n o t b e a f f e ct e d , w h e t h e r o r n o t yo u p a r t i ci p a t e i n t h i s st u d y.
P r o c e d u r e s
I f yo u d e c i d e t o t a ke p a r t i n t h i s st u d y, yo u w i l l b e a ske d t o f i l l o u t a n o n l i n e su r ve y. T h e su r ve y
q u e st i o n s a r e m a i n l y m u l t i p l e ch o i ce . H o w e ve r , t h e r e w i l l b e a f e w q u e st i o n s w h e r e yo u m a y a d d a n
o p e n - e n d e d r e sp o n se . Y o u d o n o t h a ve t o a n sw e r a n y q u e st i o n s t h a t yo u d o n ’ t w a n t t o , cl i ck “ p r e f e r
n o t t o a n sw e r ” t o m o ve t o t h e n e xt q u e st i o n . C o m p l e t i n g t h e su r ve y w i l l t a ke a p p r o xi m a t e l y 1 0 - 1 5
m i n u t e s.
B e n e f i t s T h e r e a r e n o d i r e ct b e n e f i t s f o r p a r t i ci p a n t s. H o w e ve r , t h e f i n d i n g s f r o m t h i s st u d y m a y h e l p cr e a t e a
b e t t e r u n d e r st a n d i n g o f t h e w i sh e s a n d n e e d s o f cu r r e n t a n d f u t u r e U H M a n o a S T E M st u d e n t s. R i s k s
T h e r e a r e n o a n t i ci p a t e d r i sks f o r p a r t i ci p a n t s.
C o n f i d e n t i a l i t y a n d P r i v a c y
A l l d a t a o b t a i n e d f r o m p a r t i ci p a n t s w i l l b e ke p t co n f i d e n t i a l a n d w i l l o n l y b e r e p o r t e d i n a n a g g r e g a t e
f o r m a t ( b y r e p o r t i n g o n l y co m b i n e d r e su l t s a n d n e v e r r e p o r t i n g i n d i vi d u a l o n e s) . T h e d a t a co l l e ct e d
w i l l b e st o r e d i n t h e H I P P A - co m p l i a n t , Q u a l t r i cs- se cu r e d a t a b a se u n t i l i t h a s b e e n d e l e t e d b y t h e
p r i m a r y i n ve st i g a t o r . T h e r e sp o n se s w i l l b e ke p t i n a l o cke d o f f i ce f o r t h e d u r a t i o n o f t h e st u d y. A l l
i n f o r m a t i o n w i l l b e d e st r o ye d u p o n co m p l e t i o n o f t h e r e se a r ch st u d y a n t i ci p a t e d i n Ju l y 2 0 1 4 .
S e ve r a l p u b l i c a g e n ci e s w i t h r e sp o n s i b i l i t y f o r r e se a r ch o ve r si g h t , i n cl u d i n g t h e U n i ve r si t y o f H a w a i i
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

166

2/28/2014 Qualtrics Survey Software
https://s.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview&T=48FRi8 2/12
Y e s, I w o u l d l i k e t o co n t i n u e o n t o t h e s u r v e y .
N o , I d o n o t w a n t t o p a r t i ci p a t e .
U H M C o l l e g e o f E n g i n e e r i n g
U H M C o l l e g e o f N a t u r a l S c i e n c e s
U H M C o l l e g e o f T r o p i c a l A g r i c u l t u r e a n d H u m a n R e s o u r c e s ( C T A H R )
U H M S ch o o l o f O c e a n a n d E a r t h S c i e n c e & T e c h n o l o g y ( S O E S T )
O t h e r :
I p r e f e r n o t t o a n s w e r
C i v i l & E n v i r o n m e n t a l E n g i n e e r i n g
C o m p u t e r E n g i n e e r i n g
E l e c t r i ca l E n g i n e e r i n g
M e ch a n i c a l E n g i n e e r i n g
H u m a n S t u d i e s P r o g r a m ( U H H S P ) a n d t h e U n i v e r si t y o f S o u t h e r n C a l i f o r n i a ’ s H u m a n S u b j e c t s
P r o t e ct i o n P r o g r a m ( U S C H S P P ) , m a y a c ce ss t h e d a t a . T h e H S P P r e vi e w s a n d m o n i t o r s r e s e a r ch
st u d i e s t o p r o t e ct t h e r i g h t s a n d w e l f a r e o f r e se a r ch s u b j e c t s .
Q u e s t i o n s
I f yo u h a ve a n y q u e s t i o n s o r c o n c e r n s a b o u t t h i s s t u d y, p l e a s e f e e l f r e e t o co n t a ct t h e r e s e a r c h t e a m :
Jo sh u a K a a ku a a t ( 8 0 8 ) 9 5 6 - 2 2 8 9 , j k a a k u a @ h a w a i i . e d u o r D r . D a r n e l l C o l e a t ( 2 1 3 ) 8 2 1 - 4 3 6 3 ,
d a r n e l l c@ u sc . e d u . I f yo u h a v e q u e s t i o n s a b o u t yo u r r i g h t s a s a r e s e a r ch p a r t i ci p a n t i n g e n e r a l o r i f
yo u w a n t t o t a l k t o s o m e o n e i n d e p e n d e n t o f t h e r e s e a r ch t e a m , p l e a se co n t a ct t h e U H H u m a n
S t u d i e s P r o g r a m a t ( 8 0 8 ) 9 5 6 - 5 0 0 7 o r u h i r b @ h a w a i i . e d u o r t h e U S C U n i ve r si t y P a r k I n s t i t u t i o n a l
R e vi e w B o a r d ( U P I R B ) , 3 7 2 0 S o u t h F l o w e r S t r e e t # 3 0 1 , L o s A n g e l e s , C A 9 0 0 8 9 - 0 7 0 2 , ( 2 1 3 ) 8 2 1 - 5 2 7 2 o r u p i r b @ u s c. e d u .
V o l u n t a r y P a r t i c i p a t i o n
Y o u c a n f r e e l y c h o o s e t o t a ke p a r t o r t o n o t t a ke p a r t i n t h i s s u r ve y. T h e r e w i l l b e n o p e n a l t y o r l o ss o f
b e n e f i t s f o r e i t h e r d e ci si o n . I f y o u d o a g r e e t o p a r t i c i p a t e , y o u c a n st o p a t a n y t i m e .
B l o c k 1
W h i c h a ca d e m i c C o l l e g e / S ch o o l a r e yo u e n r o l l e d i n ?
W h a t i s y o u r cu r r e n t m a j o r ?
C o l l e g e o f E n g i n e e r i n g
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

167

2/28/2014 Qualtrics Survey Software
https://s.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview&T=48FRi8 3/12
P r e - E n g i n e e r i n g
B i o l o g y
B i o c h e m i st r y
B o t a n y
C h e m i st r y
C o m p u t e r S ci e n c e
E t h n o b o t a n y
I n f o r m a t i o n & C o m p u t e r S c i e n ce s
M a r i n e B i o l o g y
M a t h e m a t i c s
M i cr o b i o l o g y
M o l e cu l a r C e l l B i o l o g y
P h ys i cs
Z o o l o g y
A n i m a l S ci e n ce s
B i o l o g i ca l E n g i n e e r i n g
F o o d S ci e n ce & H u m a n N u t r i t i o n
M o l e cu l a r B i o sci e n c e s a n d B i o t e ch n o l o g y
N R E M
P l a n t & E n v i r o n m e n t a l B i o t e ch n o l o g y
P l a n t a n d E n v i r o n m e n t a l P r o t e ct i o n S ci e n c e s
T r o p i ca l P l a n t a n d S o i l S ci e n ce s
G e o l o g y a n d G e o p h ysi cs
G e o l o g y ( B A )
G l o b a l E n v i r o n m e n t a l S ci e n ce
M e t e o r o l o g y
E n vi r o n m e n t a l S t u d i e s
I n t e r d i s ci p l i n a r y S t u d i e s
P r e - M e d i c i n e
P r e - P h ysi c a l T h e r a p y
O t h e r
I p r e f e r n o t t o a n sw e r
C o l l e g e o f N a t u r a l S ci e n ce s
C T A H R
S O E S T
O t h e r
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

168

2/28/2014 Qualtrics Survey Software
https://s.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview&T=48FRi8 4/12
F r e sh m e n
S o p h o m o r e
Ju n i o r
S e n i o r
5 t h ye a r S e n i o r
A l u m n i
I p r e f e r n o t t o a n s w e r
H i g h S ch o o l
T w o - ye a r co l l e g e
F o u r - y e a r co l l e g e
V o ca t i o n a l / T e c h n i ca l s c h o o l
M i l i t a r y
W o r k i n g a f u l l - t i m e j o b
O t h e r
I p r e f e r n o t t o a n s w e r
F i r st - t i m e c o l l e g e st u d e n t
R e t u r n i n g o r n o n - t r a d i t i o n a l c o l l e g e s t u d e n t
T r a n sf e r st u d e n t f r o m a t w o - ye a r c o l l e g e
T r a n sf e r st u d e n t f r o m a f o u r - y e a r c o l l e g e
I p r e f e r n o t t o a n s w e r
W h a t i s y o u r cu r r e n t a c a d e m i c s t a n d i n g ?
W h e r e w e r e y o u i m m e d i a t e l y b e f o r e st a r t i n g a t t h i s i n st i t u t i o n ?
W h e n yo u f i r st e n t e r e d t h i s i n s t i t u t i o n , w e r e yo u : ( M a r k o n e )
S e n s e o f B e l o n g i n g
P l e a s e i n d i ca t e y o u r a g r e e m e n t o r d i s a g r e e m e n t w i t h t h e f o l l o w i n g st a t e m e n t s:
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

169

2/28/2014 Qualtrics Survey Software
https://s.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview&T=48FRi8 5/12
D i s a g r e e S t r o n g l y D i s a g r e e A g r e e A g r e e S t r o n g l y
I p r e f e r n o t t o
a n s w e r
" I e n j o y g o i n g t o s c h o o l h e r e " " I f e e l l i ke I r e a l l y b e l o n g a t t h i s
sch o o l "
" I w i sh I h a d g o n e t o a d i f f e r e n t
sch o o l "
" I f e e l a cce p t e d i n m y m a j o r " " I f e e l co m f o r t a b l e i n m y m a j o r " " I f e e l t h a t I a m a p a r t o f m y
m a j o r "
" I se e m yse l f a s a p a r t o f t h e
ca m p u s co m m u n i t y"
" I f e e l t h a t I a m a m e m b e r o f
t h e ca m p u s c o m m u n i t y "
" I f e e l a se n se o f b e l o n g i n g t o
t h e ca m p u s c o m m u n i t y "
f a c - s t u d e n t i n t e r a c t i o n
H o w o f t e n d o y o u i n t e r a ct w i t h y o u r i n s t r u c t o r s ( f a cu l t y, t e a ch i n g a ssi st a n t s) e . g . b y p h o n e , e m a i l , i n
p e r so n , o r o t h e r ) ?
N o t a t a l l O cca s i o n a l l y F r e q u e n t l y I p r e f e r n o t t o a n s w e r
I n st r u ct o r s d u r i n g cl a s s I n st r u ct o r s d u r i n g o f f i c e h o u r s I n st r u ct o r s o u t si d e o f cl a s s o r
o f f i ce h o u r s
H o w o f t e n d o y o u r e c e i v e t h e f o l l o w i n g f r o m y o u r i n s t r u c t o r s?
N o t a t a l l O c ca si o n a l l y F r e q u e n t l y I p r e f e r n o t t o a n s w e r
E n co u r a g e m e n t f o r g r a d u a t e
sch o o l
O p p o r t u n i t y t o w o r k o n a
r e se a r ch p r o j e ct
A d vi ce a b o u t e d u c a t i o n a l
p r o g r a m
R e sp e ct E m o t i o n a l
su p p o r t / d e ve l o p m e n t
L e t t e r o f r e co m m e n d a t i o n I n t e l l e ct u a l ch a l l e n g e a n d
st i m u l a t i o n

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

170

2/28/2014 Qualtrics Survey Software
https://s.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview&T=48FRi8 6/12
O p p o r t u n i t y t o d i s c u s s
co u r se w o r k o u t s i d e o f cl a s s
H e l p i n a ch i e v i n g p r o f e s s i o n a l
g o a l s
H o w o f t e n d o y o u d o t h e f o l l o w i n g a c t i vi t i e s ?
N o t a t a l l O cca s i o n a l l y F r e q u e n t l y I p r e f e r n o t t o a n s w e r
S t u d i e d w i t h o t h e r s t u d e n t s T u t o r e d a n o t h e r c o l l e g e
st u d e n t
W o r ke d o n a g r o u p p r o j e ct W o r ke d co o p e r a t i v e l y w i t h
o t h e r st u d e n t s o n co u r se
a ssi g n m e n t s
D i scu sse d i d e a s w i t h
cl a ssm a t e s ( i n d i vi d u a l s o r
g r o u p s)
G o t f e e d b a c k o n m y w o r k a n d
i d e a s f r o m c l a ssm a t e s
I n t e r a ct e d w i t h cl a ssm a t e s
o u t si d e o f cl a s s
f a m i l y a p p r o v a l , s a t i s f a c t i o n , p r o g r a m s , G P A
R a t e yo u r a g r e e m e n t t o t h e f o l l o w i n g st a t e m e n t s:
D i s a g r e e s t r o n g l y D i s a g r e e A g r e e A g r e e S t r o n g l y
I p r e f e r n o t t h e
a n s w e r
" M y f a m i l y a p p r o v e s o f m y
a t t e n d i n g t h i s u n i ve r si t y"
" M y f a m i l y e n c o u r a g e s m e t o
co n t i n u e a t t e n d i n g t h i s
u n i ve r si t y"
R a t e yo u r sa t i s f a c t i o n w i t h t h i s i n s t i t u t i o n o n e a ch a s p e c t o f ca m p u s l i f e l i st e d b e l o w :
D i s s a t i s f i e d N e u t r a l S a t i sf i e d V e r y S a t i sf i e d
I p r e f e r n o t t o
a n s w e r
Q u a l i t y o f i n st r u ct i o n A m o u n t o f co n t a ct w i t h f a cu l t y
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

171

2/28/2014 Qualtrics Survey Software
https://s.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview&T=48FRi8 7/12
N o n e
C - M O R E S ch o l a r s P r o g r a m
H u i M a n a w a K u p o n o N a t i v e H a w a i i a n S c h o l a r sh i p P r o g r a m
K u a a n a N a t i ve H a w a i i a n S t u d e n t S e r v i c e s
M A R C
N a P u a N o ' e a u
N H S E M P
P I P E S
U n d e r g r a d u a t e R e se a r ch a n d M e n t o r i n g ( U R M ) i n t h e B i o l o g i c a l S c i e n ce s
U H M a n o a H o n o r s P r o g r a m
S A C N A S
A I S E S
O t h e r :
I p r e f e r n o t t o a n s w e r
A o r A + ( i . e . 3 . 9 o r a b o ve o n a 4 . 0 s c a l e )
A - ( 3 . 5 - 3 . 8 )
B + ( 3 . 2 - 3 . 4 )
B ( 2 . 9 - 3 . 1 )
B - ( 2 . 5 - 2 . 8 )
C + ( 2 . 2 - 2 . 4 )
C ( 1 . 9 - 2 . 1 )
C - ( 1 . 5 - 1 . 8 )
D + o r l o w e r ( l e ss t h a n 1 . 4 )
I n t e r a ct i o n w i t h p e e r s A ca d e m i c a d v i si n g a n d
st u d e n t su p p o r t
S T E M M a j o r O ve r a l l q u a l i t y o f y o u r
co l l e g i a t e e xp e r i e n ce s o f a r
T h e f o l l o w i n g i s a l i st o f a c a d e m i c a n d / o r a c a d e m i c p r e p a r a t i o n p r o g r a m s . C h e c k a l l t h e a ct i vi t i e s
t h a t yo u h a ve p a r t i c i p a t e d i n :
W h a t i s y o u r a p p r o xi m a t e c u m u l a t i v e g r a d e p o i n t a ve r a g e ?
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

172

2/28/2014 Qualtrics Survey Software
https://s.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview&T=48FRi8 8/12
I p r e f e r n o t t o a n s w e r
A m e r i ca n I n d i a n o r A l a s k a n N a t i v e
C h i n e se
F i l i p i n o
A si a n I n d i a n
J a p a n e s e
S T E M s e l f - e f f i c a c y / I n t e n t t o p e r s i s t
R a t e yo u r a g r e e m e n t t o t h e f o l l o w i n g st a t e m e n t s:
D i s a g r e e S t r o n g l y D i sa g r e e A g r e e A g r e e S t r o n g l y
I p r e f e r n o t t o
a n s w e r / N o t
a p p l i c a b l e
I i n t e n d t o co m p l e t e a S T E M
d e g r e e a t U H M a n o a
I ca n su c c e e d i n m y S T E M
m a j o r cu r r i cu l u m
I ca n su c c e e d i n m y S T E M
m a j o r cu r r i cu l u m w h i l e N O T
h a vi n g t o g i ve u p p a r t i c i p a t i o n
i n m y o u t si d e i n t e r e s t s ( e . g .
e xt r a cu r r i cu l a r a ct i v i t i e s, f a m i l y ,
sp o r t s, e t c. )
I ca n co m p l e t e t h e m a t h
r e q u i r e m e n t s f o r m y S T E M
m a j o r
I ca n co m p l e t e t h e s c i e n c e
r e q u i r e m e n t s f o r m y S T E M
m a j o r
I ca n e xc e l i n m y cu r r e n t S T E M
m a j o r t h i s se m e st e r
I ca n p e r s i st i n m y S T E M m a j o r
d u r i n g t h e n e x t a ca d e m i c ye a r
I ca n co m p l e t e m y S T E M m a j o r
a t t h i s i n st i t u t i o n
I f e e l co n f i d e n t i n m y a b i l i t y t o
co m p l e t e a S T E M d e g r e e a t
U H M a n o a
D e m o g r a p h i c s
E t h n i ci t y ( s e l e c t o n e o r m o r e ) :
A si a n
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

173

2/28/2014 Qualtrics Survey Software
https://s.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview&T=48FRi8 9/12
K o r e a n
L a o t i a n
O t h e r A s i a n
T h a i
V i e t n a m e se
B l a ck o r A f r i c a n A m e r i ca
C a u c a si a n o r W h i t e
G u a m a n i a n o r C h a m o r r o
N a t i ve H a w a i i a n o r p a r t - H a w a i i a n
M i c r o n e si a n
S a m o a n
T o n g a n
O t h e r P a c i f i c I sl a n d e r
O t h e r
I p r e f e r n o t t o a n s w e r
Y e s
N o
I p r e f e r n o t t o a n s w e r
M a l e
F e m a l e
I p r e f e r n o t t o a n s w e r
H i g h i n co m e
U p p e r - m i d d l e i n c o m e
N a t i ve H a w a i i a n o r P a ci f i c I s l a n d e r
W e r e a n y o f yo u r a n c e st o r s H a w a i i a n ?
G e n d e r :
W o u l d yo u d e s c r i b e yo u r f a m i l y a s : ( m a r k o n e )
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

174

2/28/2014 Qualtrics Survey Software
https://s.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview&T=48FRi8 10/12
M i d d l e i n co m e
L o w e r - m i d d l e i n c o m e
L o w i n co m e
I p r e f e r n o t t o a n s w e r
L e ss t h a n h i g h s c h o o l
G r a d u a t e d f r o m h i g h sch o o l
A t t e n d e d c o l l e g e b u t d i d n o t co m p l e t e d e g r e e
C o m p l e t e d a n A s s o c i a t e d e g r e e ( A A , A S , e t c. )
C o m p l e t e d a B a c h e l o r d e g r e e ( B A , B S , e t c . )
C o m p l e t e d a M a st e r d e g r e e ( M A , M S , e t c . )
C o m p l e t e d a P r o f e ssi o n a l o r D o c t o r a l d e g r e e ( JD , M D , P h D , e t c. )
U n s u r e o r n o t a p p l i ca b l e
I p r e f e r n o t t o a n s w e r
L e ss t h a n h i g h s c h o o l
G r a d u a t e d f r o m h i g h sch o o l
A t t e n d e d c o l l e g e b u t d i d n o t co m p l e t e d e g r e e
C o m p l e t e d a n A s s o c i a t e d e g r e e ( A A , A S , e t c. )
C o m p l e t e d a B a c h e l o r d e g r e e ( B A , B S , e t c . )
C o m p l e t e d a M a st e r d e g r e e ( M A , M S , e t c . )
C o m p l e t e d a P r o f e ssi o n a l o r D o c t o r a l d e g r e e ( JD , M D , P h D , e t c. )
U n s u r e o r n o t a p p l i ca b l e
I p r e f e r n o t t o a n s w e r
N o n e ( I a m c o n f i d e n t t h a t I w i l l h a v e s u f f i c i e n t f u n d s)
W h a t w a s t h e h i g h e st l e ve l o f e d u c a t i o n co m p l e t e d b y yo u r M o t h e r ?
W h a t w a s t h e h i g h e st l e ve l o f e d u c a t i o n co m p l e t e d b y yo u r F a t h e r ?
D o yo u h a ve a n y c o n c e r n s a b o u t y o u r a b i l i t y t o f i n a n ce yo u r c o l l e g e e d u c a t i o n ?
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

175

2/28/2014 Qualtrics Survey Software
https://s.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview&T=48FRi8 11/12
S o m e ( I p r o b a b l y w i l l h a ve s u f f i c i e n t f u n d s)
M a j o r ( I h a v e f u n d s b u t w i l l g r a d u a t e w i t h s i g n i f i ca n t d e b t )
E xt r e m e ( N o t su r e i f I w i l l h a ve s u f f i c i e n t f u n d s t o co m p l e t e c o l l e g e )
I p r e f e r n o t t o a n s w e r
A o r A + ( i . e . 3 . 9 o r a b o ve o n a 4 . 0 s c a l e )
A - ( 3 . 5 - 3 . 8 )
B + ( 3 . 2 - 3 . 4 )
B ( 2 . 9 - 3 . 1 )
B - ( 2 . 5 - 2 . 8 )
C + ( 2 . 2 - 2 . 4 )
C ( 1 . 9 - 2 . 1 )
C - ( 1 . 5 - 1 . 8 )
D + o r l o w e r ( l e ss t h a n 1 . 4 )
I p r e f e r n o t t o a n s w e r
N o , I d o n o t w a n t a su m m a r y o f t h e s t u d y ' s r e s u l t s
Y e s, p l e a s e se n d m e a su m m a r y o f t h e s t u d y' s r e su l t s t o t h i s e m a i l a d d r e s s ( o p t i o n a l ) :
I p r e f e r n o t t o a n s w e r
W h a t w a s yo u r a p p r o x i m a t e h i g h s c h o o l g r a d e p o i n t a v e r a g e ?
D e b r i e f
I s t h e r e a n yt h i n g yo u w a n t t o t e l l u s a b o u t y o u r e xp e r i e n ce s i n co l l e g e t h a t w e h a ve n ' t a l r e a d y a ske d
a b o u t ?
W o u l d yo u l i k e a s u m m a r y o f t h i s s t u d y ' s r e su l t s w h e n t h e y b e co m e a v a i l a b l e ?
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

176

2/28/2014 Qualtrics Survey Software
https://s.qualtrics.com/ControlPanel/Ajax.php?action=GetSurveyPrintPreview&T=48FRi8 12/12
M a h a l o . T h a n k y o u f o r c o m p l e t i n g t h i s su r ve y! Y o u r t i m e a n d i n p u t a r e g r e a t l y a p p r e ci a t e d .
SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

177
APPENDIX C
UNIVERSITY OF HAWAII AT MANOA STEM MAJORS
College of Engineering
Civil & Environmental Engineering
Computer Engineering
Electrical Engineering
Mechanical Engineering
Pre-Engineering

College of Natural Sciences
Biology
Biochemistry
Botany
Chemistry
Computer Science
Ethnobotany
Information & Computer Sciences
Marine Biology
Mathematics
Microbiology
Molecular Cell Biology
Physics
Zoology

College of Tropical Agriculture & Human Resources (CTAHR)
Animal Sciences
Biological Engineering
Food Science & Human Nutrition
Molecular Biosciences and Biotechnology
Natural Resources and Environmental Management
Plant & Environmental Biotechnology
Plant and Environmental Protection Sciences
Tropical Plant and Soil Sciences

School of Ocean and Earth Science & Technology (SOEST)
Geology and Geophysics
Geology (BA)
Global Environmental Science
Meteorology

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

178
APPENDIX D
DESCRIPTIVE STATISTICS BY MAJOR, LEVEL, AND PRE-INSTITUTION STATUS
Table 40
Sample and Population by STEM Academic Major
Sample (N=638) Population (N=3592)
Major Frequency % Frequency %
Biology 111 17.4 741 20.6
Civil & Environmental Engineering 71 11.1 298 8.3
Mechanical Engineering 69 10.8 309 8.6
Electrical Engineering 43 4.7 219 6.1
Marine Biology 35 5.5 286 8.0
Computer Science 24 3.8 216 6.0
Animal Science 24 3.8 122 3.4
Food Sciences & Nutrition 24 3.8 110 3.1
Microbiology 23 3.6 85 2.4
Chemistry 19 3.0 97 2.7
Mathematics 19 3.0 77 2.1
GES 18 2.8 49 1.4
Pre-Engineering 17 2.7 252 7.0
ICS 16 2.5 126 3.5
Biochemistry 13 2.0 52 1.4
Molecular Cell Biology 13 2.0 59 1.6
Physics 13 2.0 54 1.5
NREM 13 2.0 73 2.1
Zoology 11 1.7 58 1.6
Computer Engineering 9 1.4 76 2.1
Other 9 1.4 0 0
Biological Engineering 6 .9 41 1.1
Meteorology 6 .9 23 .6
PEPS 5 .8 21 .6

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

179
Table 40, continued
Sample (N=638) Population (N=3592)
Major Frequency % Frequency %
Geology & Geophysics 5 .8 35 1.0
Plant & Environmental Biotechnology 4 .6 18 .5
TPSS 4 .6 42 1.2
Pre-Medicine 4 .6 0 0
Botany 3 .5 25 .7
Ethnobotany 3 .5 20 .6
Molecular Biosciences & Biotechnology 2 .3 0 0
Geology 0 0 8 .2
Prefer not to answer 2 .3 0 0.0
Total 638 100.0 3592 100.0
Note. Students enrolled in one or more academic majors (double majors) were classified into
their primary major for this study. GES = Global Environmental Sciences; ICS = Information &
Computer Sciences; NREM = Natural Resources & Environmental Management; PEPS = Plant
and Environmental Protection Sciences; TPSS = Tropical Plant and Soil Sciences.

SELF-EFFICACY BELIEFS AND INTENTIONS TO PERSIST

180
Table 41
Frequency Counts by Educational Level and Pre-Institution Status (N = 638)
Variable Category n %
Educational Level
Freshmen 93 14.6
Sophomore 113 17.7
Junior 194 30.4
Senior 134 21.0
5
th
year Senior 87 13.6
Prefer not to answer 17 2.6
Incoming Student Status
First-time college student 395 61.9
Returning or non-traditional 36 5.6
Transfer: 4-year college 62 9.7
Transfer: 2-year college 141 22.1
Prefer not to answer 4 0.6
Prior Institution
High School 390 61.1
2-year college 141 22.1
4-year college 57 8.9
Full-time employment 32 5.0
Military 4 .6
Vocational or technical school 2 .3
Other 6 .9
Prefer not to answer 6 .9

Abstract (if available)

Abstract This study applies the framework of Social Cognitive Career Theory and Astin’s (1999) Inputs—Environment—Outcomes model to investigate the personal input and environmental factors associated with self-efficacy beliefs and intentions to persist in Science, Technology, Engineering, and Mathematics (STEM) and to examine the differences of these factors and outcomes between Native Hawaiian and non-Hawaiian students. Conducted at a large, public tier-one research institution in Hawaii, this cross-sectional study gathered survey data from 638 undergraduate STEM majors and analyzed data through factor analysis, regression, and MANOVA techniques. The findings indicate that sense of belonging to major, past performance, and family support explained STEM self-efficacy. Self-efficacy, in turn, predicted intent to complete a STEM degree at the institution. This study also found higher levels of peer interaction, program involvement, family support, and intentions to persist for Native Hawaiians relative to non-Hawaiians. A Ho’okahua or foundation building framework is presented based on self-efficacy, sense of belonging, and involvement to guide educational practice and theory. The implication for practice is that academic communities at the department or discipline level, especially for underclassmen and Native Hawaiians, are important to improve degree completion in STEM. The findings provide direction for Native Hawaiian education research to further investigate socio-cultural aspects of learning and Native Hawaiian congruence in STEM.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Raising cultural self-efficacy among faculty and staff of a private native Hawaiian school system

PDF

Factors affecting native Hawaiian student persistence in higher education

PDF

Institutional diversity's impact on Latinx students' self-efficacy and sense of belonging

PDF

Women's self-efficacy perceptions in mathematics and science: investigating USC-MESA students

PDF

Impact of culturally responsive education on college choice

PDF

Ma ka hana ka ike perpetuating excellence in Native Hawaiian education: Native Hawaiian Education Council members' approaches to supporting the needs of Native Hawaiians

PDF

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

PDF

The Relationship Between Institutional Marketing and Communications and Black Student Intent to Persist in Private Universities

PDF

Persistence interventions for Native Hawaiian students

PDF

The relationship among gender, race/ethnicity, sense of validation, science identity, science self-efficacy, persistence, and academic performance of biomedical undergraduates

PDF

A comparison of student motivation by program delivery method: self-efficacy, goal orientation, and belongingness in a synchronous online and traditional face-to-face environment

PDF

An exploration of principal self-efficacy beliefs about transformational leadership behaviors

PDF

The relationship between ethnicity, self-efficacy, and beliefs about diversity to instructional and transformational leadership practices of urban school principals

PDF

Motivational, parental, and cultural influences on achievement and persistence in basic skills mathematics at the community college

PDF

Mathematics Engineering Science Achievement and persistence in science, technology, engineering, and mathematics majors: the influence of MESA on the retention of first generation females in STEM...

PDF

From cultural roots to worldview: Cultural connectedness of teachers for native Hawaiian students

PDF

The relationship of students' self-regulation and self-efficacy in an online learning environment

PDF

Nursing students' perceptions of formal faculty mentoring

PDF

The impact of student-faculty interaction on undergraduate international students' academic outcome

PDF

The underrepresentation of women in science, technology, engineering, and mathematics (STEM) leadership positions

Asset Metadata

Creator Kaakua, Joshua Kananimauloa (author)

Core Title Self-efficacy beliefs and intentions to persist of Native Hawaiian and non-Hawaiian science, technology, engineering, and mathematics majors

School Rossier School of Education

Degree Doctor of Education

Degree Program Education (Leadership)

Publication Date 10/23/2014

Defense Date 05/09/2014

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

Tag Native Hawaiian,OAI-PMH Harvest,self-efficacy,STEM

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Cole, Darnell G. (committee chair), Benham, Maenette (committee member), Sundt, Melora A. (committee member)

Creator Email kaakua@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-508956

Unique identifier UC11298666

Identifier etd-KaakuaJosh-3026.pdf (filename),usctheses-c3-508956 (legacy record id)

Legacy Identifier etd-KaakuaJosh-3026.pdf

Dmrecord 508956

Document Type Dissertation

Format application/pdf (imt)

Rights Kaakua, Joshua Kananimauloa

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