University of Southern California Dissertations and Theses

The effects of a math summer bridge program on college self-efficacy and other student success measures in community college students

(USC Thesis Other)

The effects of a math summer bridge program on college self-efficacy and other student success measures in community college students

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Running&head:&MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF <EFFICACY&&&&&&&&&&&&&&&&& & & & THE EFFECTS OF A MATH SUMMER BRIDGE PROGRAM ON COLLEGE SELF- EFFICACY AND OTHER STUDENT SUCCESS MEASURES IN COMMUNITY COLLEGE STUDENTS by Sarah Emiko Akina 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 May 2016 Copyright 2016 Sarah Emiko Akina MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & ii& DEDICATION This dissertation is dedicated to my husband, Regal K. Akina and my parents, Theresa and Roy Inouye. MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & iii& ACKNOWLEDGEMENTS I wish to thank my committee chair, Dr. Robert Keim, who has helped me with not only coursework and writing my dissertation, but also stress management as I concurrently got engaged, married, and had my first child. Thank you to Dr. Dennis Hocevar, for your guidance from 2,479 miles away. Dr. Ardis Eschenberg, thank you for being an amazing support and mentor for me over the past 5 years, always providing opportunities for growth. And to Dr. Judy Oliveira, thank you for being a role model of what a local girl can become! USC Hawaii Cohort 2013, I cannot believe our journey is coming to an end. I know that my experience wouldn’t have been as positive and productive without each and every one of you! I would like to acknowledge the two people who have pushed and supported me throughout my entire life, my mom and dad. Mom, thank you for always being my cheerleader, believing in me when I didn’t think things were possible. Dad, you are the epitome of positive peer (Dad) pressure, always pushing me to set my expectations higher than I’ve ever imagined, and then putting in just as many hours as I did as my master editor. I would especially like to acknowledge my best friend, my love, and my rock: Kawai. None of this would have been possible without your love and support. Raizo and I are so blessed to have you in our lives. You are an incredible husband and an even more amazing father. Finally, to my son, Raizo, mommy can’t wait to spend all of her newly found free time playing with you! MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & iv& TABLE OF CONTENTS Dedication ii Acknowledgements iii List of Tables vi List of Figures xii Abstract xiii Chapter 1 1 Introduction 1 Background of the Problem 2 Statement of the Problem 10 Purpose of the Study 11 Research Questions 12 Methodology 14 Importance of the Study 15 Organization of the Dissertation 16 Chapter 2 17 Literature Review 17 Astin’s Theory of Student Involvement 20 Inputs-Environment-Outcomes Model 22 Bandura’s Social Cognitive Theory 23 Conceptual Model 31 Background Factors (Input) 32 Environmental Factors (Environment) 37 Outcome Variables (Output) 43 Summary 48 Chapter 3 49 Methodology 49 Research Design 52 Setting, Context, Environment 55 Population and Sample 59 Data Collection 60 Quantitative Instrumentation 62 Quantitative Analysis 70 Summary 70 Chapter 4 72 Results and Analysis 72 Part One: Descriptive Statistics 75 Participant Characteristics 75 Input Characteristics by Native Hawaiian Status 80 Input Characteristics by Age Categories 82 MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & v& College Self-Efficacy Scale Factor Analysis and Reliability Analysis 85 Sense of Belonging Inventory Factor Analysis and Reliability Analysis 98 Instructor Interaction Inventory Factor Analysis and Reliability Analysis 109 Peer Interaction Inventory Factor Analysis and Reliability Analysis 111 Instructor Satisfaction Inventory Factor Analysis and Reliability Analysis 117 Support Services Satisfaction Inventory Factor Analysis and Reliability Analysis 119 Overall Windward Community College Satisfaction Inventory Factor Analysis and Reliability Analysis 121 Summary of Reliability Tests for Multi-Item Variables 123 Summary of Part One 123 Research Question 1 Results: College Self-Efficacy 131 Research Question 2 Results: Sense of Belonging 139 Research Question 3 Results: Student Engagement and Satisfaction 147 Research Question 4 Results: Persistence 178 Research Question 6 Results: Cumulative GPA 199 Summary of Results 209 Chapter 5 213 Discussion 213 Discussion of Findings 217 Research Question 1 Findings: Impact on College Self-Efficacy 219 Research Question 2 Findings: Impact on Sense of Belonging 219 Research Question 3 Findings: Impact on Student Engagement and Satisfaction 221 Research Question 4 Findings: Impact on Persistence 222 Research Question 5 Findings: Impact on Graduation 224 Research Question 6 Findings: Impact on Cumulative GPA 225 Limitations 226 Alternative Framework 228 Implications for Practice 231 Future Research 234 Conclusion 237 References 239 Appendices 256 Appendix A: Student Survey 256 Appendix B: College Self-Efficacy Inventory 278 Appendix C: Sense Of Belonging Inventory Questions 279 MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & vi& LIST OF TABLES Table 1. Key Components of the WCC Math Summer Bridge Program 41 Table 2. Windward Community College 2014 Developmental Math Pass Rates 56 Table 3. Input-Independent Variable Questions 63 Table 4. Environment-Independent Variable Questions 65 Table 5. Outputs-Dependent Variable Questions 66 Table 6. Participant Sex 76 Table 7. Participant Ethnicity 78 Table 8. Participant Age 78 Table 9. Participant Enrollment Status 79 Table 10. Participant Compass Placement Score 80 Table 11. Input Characteristics by Native Hawaiian Status 81 Table 12. Input Characteristics by Native Hawaiian Status for Experiment Group 82 Table 13. Input Characteristics by Age Categories 84 Table 14. Input Characteristics by Age Categories for Experiment Group 85 Table 15. Descriptive Statistics for College Self-Efficacy Items Pre/Post 2015 MSB 88 with Ethnicity Breakdown Table 16. Descriptive Statistics for College Self-Efficacy Items Post MSB 2015, 90 2014, 2013, 2012 with Ethnicity Breakdown Table 17. Descriptive Statistics for College Self-Efficacy Items Pre/Post 2015 MSB 94 with Age Group Breakdown Table 18. Descriptive Statistics for College Self-Efficacy Items Post MSB 2015, 96 2014, 2013, 2012 with Age Group Breakdown Table 19. Descriptive Statistics for Sense of Belonging-Antecedents (SOBI-A) 99 Items Pre/Post 2015 MSB with Ethnicity Breakdown MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & vii& Table 20. Descriptive Statistics for Sense of Belonging-Antecedents (SOBI-A) 100 Items Post MSB 2015, 2014, 2013, 2012 with Ethnicity Breakdown Table 21. Descriptive Statistics for Sense of Belonging-Antecedents (SOBI-A) 102 Items Pre/Post 2015 MSB with Age Group Breakdown Table 22. Descriptive Statistics for Sense of Belonging-Antecedents (SOBI-A) 103 Items Post MSB 2015, 2014, 2013, 2012 with Age Group Breakdown Table 23. Descriptive Statistics for Sense of Belonging-Psychological (SOBI-P) 105 Items Pre/Post 2015 MSB with Ethnicity Breakdown Table 24. Descriptive Statistics for Sense of Belonging-Psychological (SOBI-P) 106 Items Post MSB 2015, 2014, 2013, 2012 with Ethnicity Breakdown Table 25. Descriptive Statistics for Sense of Belonging-Psychological (SOBI-P) 107 Items Pre/Post 2015 MSB with Age Group Breakdown Table 26. Descriptive Statistics for Sense of Belonging-Psychological (SOBI-P) 108 Items Post MSB 2015, 2014, 2013, 2012 with Age Group Breakdown Table 27. Descriptive Statistics for Instructor Interaction-In Class Pre/Post 2015, 113 2014, 2013, 2012 MSB with Ethnicity Breakdown Table 28. Descriptive Statistics for Instructor Interaction-In Class Pre/Post 2015, 113 2014, 2013, 2012 MSB with Age Group Breakdown Table 29. Descriptive Statistics for Instructor Interaction-Outside of Class Pre/Post 114 2015, 2014, 2013, 2012 MSB with Ethnicity Breakdown Table 30. Descriptive Statistics for Instructor Interaction-Outside of Class Pre/Post 114 2015, 2014, 2013, 2012 MSB with Age Group Breakdown Table 31. Descriptive Statistics for Peer Interaction-School Pre/Post 2015, 2014, 115 2013, 2012 MSB with Ethnicity Breakdown Table 32. Descriptive Statistics for Peer Interaction-School Pre/Post 2015, 2014, 115 2013, 2012 MSB with Age Group Breakdown Table 33. Descriptive Statistics for Peer Interaction-Social Pre/Post 2015, 2014, 116 2013, 2012 MSB with Ethnicity Breakdown Table 34. Descriptive Statistics for Peer Interaction-Social Pre/Post 2015, 2014, 116 2013, 2012 MSB with Age Group Breakdown MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & viii& Table 35. Descriptive Statistics for Instructor Satisfaction Pre/Post 2015, 2014, 118 2013, 2012 MSB with Ethnicity Breakdown Table 36. Descriptive Statistics for Instructor Satisfaction Pre/Post 2015, 2014, 118 2013, 2012 MSB with Age Group Breakdown Table 37. Descriptive Statistics for Support Services Satisfaction Pre/Post 2015, 120 2014, 2013, 2012 MSB with Ethnicity Breakdown Table 38. Descriptive Statistics for Support Services Satisfaction Pre/Post 2015, 120 2014, 2013, 2012 MSB with Age Group Breakdown Table 39. Descriptive Statistics for Overall Windward Community College 122 Satisfaction Pre/Post 2015, 2014, 2013, 2012 MSB with Ethnicity Breakdown Table 40. Descriptive Statistics for Overall Windward Community College 122 Satisfaction Pre/Post 2015, 2014, 2013, 2012 MSB with Age Group Breakdown Table 41. Summary of Composite Variables and Reliability Analysis 123 Table 42. Significant t-test Results by Ethnicity (Non-Native Hawaiian v. 128 Native Hawaiian) Table 43. Significant t-test Results by Age Group (Traditional v. Nontraditional) 129 Table 44. Descriptive Statistics for College Self-Efficacy Scores for Time 1 and 133 Time 2-Total Table 45. Descriptive Statistics for College Self-Efficacy Scores for Post MSB 133 2015, 2014, 2013, 2012 Table 46. Descriptive Statistics for College Self-Efficacy Scores for Time 1 and 135 Time 2-Non-Native Hawaiian v. Native Hawaiian Table 47. Descriptive Statistics for College Self-Efficacy Scores for Post MSB 136 2015, 2014, 2013, 2012-Non-Native Hawaiian v. Native Hawaiian Table 48. Descriptive Statistics for College Self-Efficacy Scores for Time 1 and 137 Time 2-Traditional v. Nontraditional Table 49. Descriptive Statistics for College Self-Efficacy Scores for Post MSB 138 2015, 2014, 2013, 2012-Traditional v. Nontraditional MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & ix& Table 50. Descriptive Statistics for Sense of Belonging (SOBI-A and SOBI-P) 141 Scores for Time 1 and Time 2-Total Table 51. Descriptive Statistics for Sense of Belonging (SOBI-A and SOBI-P) 143 Scores for Time 1 and Time 2-Non-Native Hawaiian v. Native Hawaiian Table 52. Descriptive Statistics for Sense of Belonging (SOBI-A and SOBI-P) 145 for Time 1 Time 2-Traditional v. Nontraditional Table 53. Engagement and Satisfaction Questions with Coding 149 Table 54. Descriptive Statistics Student Engagement Scores for Time 1 and 152 Time 2-Total Table 55. One-Way Between Groups ANOVA Student Engagement Scores for 154 2015, 2014, 2013, and 2012 Cohorts Table 56. One-Way Between Groups ANOVA Significant Differences-Student 155 Engagement Scores for 2015, 2014, 2013, and 2012 Cohorts Table 57. Descriptive Statistics Student Satisfaction Scores for Time 1 and 156 Time 2-Total Table 58. One-Way Between-Groups ANOVA Student Satisfaction Scores for 157 2015, 2014, 2013, and 2012 Cohorts Table 59. Descriptive Statistics Student Engagement Scores for Time 1 and 160 Time 2-Total Table 60. Ethnicity Grouping-One-Way Between Groups ANOVA Student 162 Engagement Scores for 2015, 2014, 2013, and 2012 Cohorts Table 61. Ethnicity Grouping-One-Way Between Groups ANOVA Significant 163 Differences-Student Engagement Scores for 2015, 2014, 2013, and 2012 Cohorts Table 62. Descriptive Statistics Student Satisfaction Scores for Time 1 and 165 Time 2-Total Table 63. Ethnicity Grouping-One-Way Between Groups ANOVA Student 167 Satisfaction Scores for 2015, 2014, 2013, and 2012 Cohorts Table 64. Ethnicity Grouping-One-Way Between Groups ANOVA Significant 167 Differences-Student Satisfaction Scores for 2015, 2014, 2013, and 2012 Cohorts MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & x& Table 65. Age Grouping-Descriptive Statistics Student Engagement Scores for 170 Time 1 and Time 2-Total Table 66. Age Grouping-One-Way Between Groups ANOVA Student Engagement 172 Scores for 2015, 2014, 2013, and 2012 Cohorts Table 67. Age Grouping-One-Way Between Groups ANOVA Significant 173 Differences-Student Engagement Scores for 2015, 2014, 2013, and 2012 Cohorts Table 68. Age Grouping-One-Way Between Groups ANOVA Significant 175 Differences-Student Engagement Scores for 2015, 2014, 2013, and 2012 Cohorts Table 69. One-Way Between Groups ANOVA Student Satisfaction Scores for 177 2015, 2014, 2013, and 2012 Cohorts Table 70. Research Question 3: Student Engagement and Satisfaction, Statistically 178 Significant Results Table 71. One-Way Between-Groups ANOVA Persistence Rates 181 (Experiment v. Control A) Table 72. One-Way Between-Groups ANOVA Persistence Rates 182 (Experiment v. Control B) Table 73. Ethnicity Grouping-One-Way Between-Groups ANOVA Persistence 184 Rates (Experiment v. Control A) Table 74. Ethnicity Grouping-One-Way Between-Groups ANOVA Persistence 184 Rates (Experiment v. Control B) Table 75. Age Grouping-One-Way Between-Groups ANOVA Persistence Rates 186 (Experiment v. Control A) Table 76. Age Grouping-One-Way Between-Groups ANOVA Persistence Rates 186 (Experiment v. Control B) Table 77. Complete-One-Way Between-Groups ANOVA Persistence Rates 188 (Experiment v. Control A) Table 78. Complete-One-Way Between-Groups ANOVA Persistence Rates 189 (Experiment v. Control B) Table 79. One-Way Between-Groups ANOVA Graduation Rates 192 (Experiment v. Control A) MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & xi& Table 80. One-Way Between-Groups ANOVA Graduation Rates 192 (Experiment v. Control B) Table 81. Ethnicity Grouping-One-Way Between-Groups ANOVA Graduation 194 Rates (Experiment v. Control A) Table 82. Ethnicity Grouping-One-Way Between-Groups ANOVA Graduation 194 Rates (Experiment v. Control B) Table 83. Age Grouping-One-Way Between-Groups ANOVA Graduation Rates 196 (Experiment v. Control A) Table 84. Age Grouping-One-Way Between-Groups ANOVA Graduation Rates 197 (Experiment v. Control B) Table 85. Complete- One-Way Between-Groups ANOVA Graduation Rates 198 (Experiment v. Control A) Table 86. Complete-One-Way Between-Groups ANOVA Graduation Rates 199 (Experiment v. Control B) Table 87. One-Way Between-Groups ANOVA Cumulative GPA 201 (Experiment v. Control A) Table 88. One-Way Between-Groups ANOVA Cumulative GPA 202 (Experiment v. Control B) Table 89. Ethnicity Grouping-One-Way Between-Groups ANOVA 204 Cumulative GPA (Experiment v. Control A) Table 90. Ethnicity Grouping-One-Way Between-Groups ANOVA 204 Cumulative GPA (Experiment v. Control B) Table 91. Age Grouping-One-Way Between-Groups ANOVA Cumulative GPA 206 (Experiment v. Control A) Table 92. Age Grouping-One-Way Between-Groups ANOVA Cumulative GPA 206 (Experiment v. Control B) Table 93. Complete-One-Way Between-Groups ANOVA Cumulative GPA 208 (Experiment v. Control A) Table 94. Complete-One-Way Between-Groups ANOVA Cumulative GPA 209 (Experiment v. Control B) Table 95. Total Study Results 218 MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & xii& LIST OF FIGURES Figure 1 Astin’s Input-Environment-Outcomes Model 33 Figure 2 Research Design 54 Figure 3 Windward Community College Math Pathways 58 Figure 4 Alternative Framework 230 MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & xiii& ABSTRACT This study uses Astin’s (1999) Inputs-Environment-Outcomes (I-E-O) framework to investigate if a developmental math summer bridge program had a significant effect on six (6) student success measures; three pertaining to mindset/behaviors leading to student success: (1) college self-efficacy (CSE); (2) sense of belonging; and (3) engagement and satisfaction; and three measuring student success: (4) persistence; (5) graduation; and (6) cumulative GPA for students enrolled in a community college in Hawaii. This study had three parts to its design: (1) pre/post experiment; (2) cross-sectional; and (3) comparison group. Nine hundred and fifty-seven (957) students were included in this study and their survey responses and/or student records were measured through repeated measures and between-groups ANOVA techniques. The findings indicate that the WCC Math Summer Bridge (MSB) Program has a significant and long-lasting effect on the mindset/behaviors leading to student success (CSE and engagement and satisfaction), which translates to these students having higher persistence and graduation rates, as well as higher cumulative GPAs than non-MSB students. A new framework was created to better understand the interconnected relationship that each of these outcomes have with one another and how a multi-faceted intervention can lead to exponential growth, setting students on a higher trajectory for success. The implication for practice is that academic communities and multi- faceted interventions are integral for helping students who place into developmental math courses not only pass these courses, but also do well in their other coursework, persist, and graduate. It is suggested that further research be done in identifying a more valid inventory to measure sense of belonging for community college students, as well as conduct qualitative and mixed-methods research to better understand the students’ perspective/experience while going through the MSB program and how they perceived its impact on their success. MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 1& CHAPTER 1 INTRODUCTION Living in the 21 st century, earning a college degree is essential for making a living wage in the United States. Especially after the Great Recession in 2007, the job market for individuals without a college degree not only stagnated, but also drastically depleted, while college graduates’ job opportunities held strong (Baum, Ma, & Payea, 2013; Carnevale, Smith, & Strohl, 2013). Community colleges (CCs) provide an opportunity for individuals who have fewer resources (academic, social, and economic) to enroll at a higher education institution. The community colleges’ open-door policy and low-cost tuition have been providing individuals access to higher education since its creation at the turn of the 20 th century (Jurgens, 2010). However, the focus now has evolved from how to get more students into college (access), to how to keep these students successfully on-track through graduation. Self-efficacy, the belief in one’s “capability to exercise some measure of control over (one’s) own functioning and over environmental events” (Bandura, 2001, p. 10) has been highly researched and identified as one of the most important factors in student persistence and graduation (Brown, Tramayne, Hoxha, Telander, Fan, & Lent, 2008; Gore, 2006; Robbins, Lauver, Le, Davis, Langley, & Carlstrom, 2004; Wright, Jenkins-Guarnieri, & Murdock, 2012). When reviewing one program in particular, this study addresses the question: as an institution, are there impactful academic and non-academic programs that could help boost college self-efficacy in students and therefore improve their chances of persistence and graduation? First, this chapter will highlight the importance of earning a college degree and how community colleges provide an opportunity to many underprivileged and underrepresented populations in an attempt to bring forth equity throughout society (especially at the different MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 2& socioeconomic levels). Then, in Chapter 2, the review of the literature will discuss self-efficacy and its conceptual framework will be introduced as an important factor in student persistence and success in higher education. In using Astin’s (1993) Student Involvement Theory and I-E-O Framework in conjunction with the Social Cognitive Theory (Bandura, 1977, 1986), this study will identify key factors in student success: (1) college self-efficacy; (2) sense of belonging; (3) engagement and satisfaction; (4) persistence; (5) graduation; and (6) cumulative grade point average (GPA) to provide a lens of investigation for students at Windward Community College and measuring the impact of the Math Summer Bridge Program on said factors. Then, in Chapter 3, the purpose, research questions, and methodology of the study will be described. Chapter 4 will present the findings and analysis, followed by Chapter 5 summarizing the findings and providing suggestions for future research, as well as the conclusion. References and appendices conclude this dissertation. Background of the Problem In today’s economy, the importance of earning a degree in higher education is at an all- time high. This concept has grown exponentially since the mid 1900’s and was exemplified during the Great Recession. Since 2007, individuals who have earned a college degree have not only earned more money, but have also been provided the greatest opportunity (Baum, et al., 2013). Job growth has only been seen for those who have earned a bachelor’s degree or higher (8% gain), while individuals who have earned a high school diploma or less experienced a net loss of 10% (Carnevale, et al., 2013). However, the value of earning a college degree does not just benefit individuals economically. College graduates are less likely to go to jail, more likely to live healthier lifestyles, follow habits that are beneficial to society in general (e.g. more likely to vote and pay taxes), and provide better opportunities for their children, perpetuating MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 3& generations of more prosperous lifestyles (Baum, et al., 2013). With these impactful statistics, the United States government has also taken interest in promoting a college education. In 2012 during his State of the Union Address, President Barack Obama created a nationwide initiative for the US to have the largest proportion of college graduates in the world by 2020 (State of the Union Address, 2012). Cost of earning a degree With the increased value of a college degree has come a record high cost of earning it. When controlling for inflation, the cost of tuition, room, and board at public institutions rose 42% between the 2000-2001 and 2010-2011 school years (National Center for Education Statistics, 2012). Community colleges provide a less expensive opportunity for individuals to earn a 2-year degree, which if it’s a liberal arts degree, generally accounts for the first two years of a bachelor’s degree. During the 2014-2015 school year, the average in-state tuition for community colleges was $3,347 compared to $9,139 at its 4-year public institution counterparts (The College Board, 2014). By community colleges offering tuition at a lower cost, higher education becomes more accessible for people from lower income backgrounds, and is reflected in its student population. One in five community college students live in poverty (Mullin, 2012) and 47% received federal, state, or institutional financial aid and jumps to 58% when including students who receive federal student loans (American Association of Community Colleges, 2015). Equal Access Community colleges are considered the democracy’s college, providing higher educational opportunities for all. The United States’ Community College platform is to provide affordable access to a quality education (American Association of Community Colleges, 2013). MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 4& This has been the foundation for community colleges since its origins when President Lincoln signed the Morrill Act of 1890 (creating a place to provide higher education opportunities for all, including women and minorities through open-door admissions), and has continued to be reflected through many other policies throughout history (Jurgens, 2010; Hendrick, Hightower, & Gregory, 2006; Lucas, 2006). At its inception, these institutions were created for two reasons: 1) to take away the “burden” of providing general education courses from the universities, and 2) provide vocational and technical training and education (Jurgens, 2010), both of which are still seen today. However, there is an additional task put upon the present day community college institutions. With the decline in quality of the K-12 public education system (Darling- Hammond, 2007; Eger & McDonald, 2012), community colleges have adapted to meet the needs of potential students by providing developmental/remedial education courses for those who lack the academic skills to perform at the college level. Variance in Academic Skills The current open-door policy for community colleges generally means that all students are accepted to the institution as long as they meet one of the three requirements: (1) 18 years old or older; (2) earned a high school diploma; or (3) earned a high school diploma equivalency (e.g. general education diploma-GED). However, this means that there are no screening or denying admissions based on academic standards or requirements; with “greater access has come broader variability in students’ readiness for college-level work” (Handel & Williams, 2011, p. 29). Most community college students are not academically prepared for higher education, especially in mathematics and reading and/or writing. According to Adams, Adams, Franklin, Gulick, Gulick, Shearn, & Mireles (2012), 63% of all community college students place into courses lower than college-level compared to 19.9% of all students at 4-year institutions (Adams, et al., MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 5& 2012; Gilroy, 2010), and some CCs developmental placement rates exceed 90% (Kerrigan & Slater, 2010). The discrepancy in competence is recorded in high school as well. Only 28% of all students who took the ACT standardized test met all college benchmark scores in English composition, algebra, college social science, and basic biology (ACT, 2012). In a study conducted by Achieving the Dream including over 250,000 students at 57 CCs, over half (59%) of all first-time, full-time freshmen placed into developmental math (Bailey, Jeong, & Cho, 2009). Clearly, there is a disconnect between K-12 preparation and college readiness. When completion of high levels of math (e.g. at least Algebra II) in high school is seen as a significant predictor of earning a bachelor’s degree, it is important for all of higher education, but especially community colleges, to understand what academic knowledge sets students are entering college with (Adelman, 2006). Other barriers CC students face Besides having a wide variance in academic skill sets, community college students differ from 4-year students in many other ways. Community college students on average have lower (if any) college admissions test scores (e.g. SAT or ACT) and lower high school academic achievement (Silver-Pacuilla, Perin, & Miller, 2013). There is also a larger nontraditional student population at the CCs (63%, AACC, 2015) where students have been out of school or formal education for a longer period of time, and have more difficulty recalling academic content that they learned in high school (Silver-Pacuilla, et al., 2013). Besides academic barriers, community college students also have other obligations and responsibilities to juggle. According to the American Association of Community Colleges (2015), 69% of the community college student body works at least a part-time job. Community college students are also more likely to come from lower socioeconomic backgrounds and are more likely to be first-generation college MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 6& students (Berger, 2000). Because of many of these factors, community college students are less likely to have social or cultural capital to help them thrive in a college environment. CCs have a student population that is more at risk for experiencing adversities (e.g. poverty, homelessness, childcare issues, and lack of academic preparedness), which in turn affects student success (American Association of Community Colleges, 2013; Bahr, 2010; Calcagno, Bailey, Jenkins, Kienzl, & Leinbach, 2008). Nontraditional Students in Higher Education The definition of nontraditional student varies from study to study, however it generally includes a delay in college enrollment following high school, part-time/full-time enrollment status, working more than 35 hours/week, financially independent, and/or having dependents (children) (Philibert, Allen, & Elleven, 2008). Nontraditional students are more likely to face additional barriers that their traditional counterparts do not. Mercer (1993) categorized these barriers into three groups: situational, dispositional, and institutional. Situational barriers refer to specific circumstances that nontraditional students commonly face (e.g. family, employment) and time/responsibility conflicts between these barriers and school. Dispositional barriers include attitudinal or psychological obstacles that nontraditional students might have (e.g. adjustment or self-esteem). Finally, institutional barriers include structures that the school has created that might prove difficult for the nontraditional student when trying to earn a degree (e.g. class time/availability). All three of these barriers make it more difficult for nontraditional students to attend school full-time and/or persist (Keith, 2007; Mercer, 1993; Taniguchi & Kaufman, 2005). However, nontraditional students who attend a community college are the majority on campus and have the potential to experience a unique peer support system and sense of belonging (Clark, 2012). For the purposes of this study, nontraditional students will be identified if they meet the MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 7& requirement: did not attend college right after graduating high school. This will be measured by age. Students who are 26+ years old will be considered nontraditional. Total number of students affected The total number of students who attend community college is quite significant. Forty- six percent (46%) of all undergraduate students are a part of a community college system. These students are enrolled in the 1123 total community colleges across the United States of America (American Association of Community Colleges, 2015). Community colleges have larger increases in growth than their 4-year counterparts (Jurgens, 2010) and provide education to the majority of minority students (61% of total Native Americans, 57% of total Hispanics, 52% of total Blacks) (American Association of Community Colleges, 2015). Student Success Factors Many higher education institutions use prior academic performance to indicate how well a potential student will do in college. Especially at 4-year institutions, they use high school grade point average (GPA) and college readiness tests (e.g. SAT or ACT) to make admissions decisions. However, as mentioned before, community colleges do not use the same admissions criteria. Although prior academic performance indicators have been found to have some predictive value for student success (Ishitani, 2003; Johnson, 2008), it has not been found to be the most valuable way to measure a student’s potential for success (Brown, et al., 2008). Self- efficacy, and more specifically, college or academic self-efficacy has been shown to be a better indicator of student success (Brown, et al., 2008; Gore, 2006; Robbins, et al., 2004; Wright et al., 2012) and because of this, will be the focus of this study. Encompassing all of these findings, Barbatis (2010) found four themes contributing to student persistence: (1) precollege characteristics; (2) external college support/community MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 8& influences; (3) social involvement; and (4) academic integration. Precollege characteristics (sense of responsibility, resourcefulness, persistence-qualities similar to self-efficacy) and external college support and community influence (parents, familial relationships, friends) were found to be the most influential for ethnically diverse developmental students. College interventions Community colleges are already implementing practices to help enhance opportunities for student success. According to the American Association of Community Colleges (2013), on average, community colleges have smaller class sizes (25-35 students per class) compared to 4- year institutions that can have class sizes of 200-400 students. Smaller class sizes increase the chances of more interaction with instructors and peers, therefore increasing the opportunity for a quality learning environment and enriching learning experiences (De Paola, Ponzo, & Scoppa, 2013). Almost all community colleges offer developmental courses. However, after the Great Recession, with current federal and state budget restraints and decline in funding for higher education institutions (Boyd, 2009), legislatures and boards of higher education are continuing to increase their standards of accountability for higher education institutions (Perna, Klein, & McLendon, 2014), making it even more critical to identify programs that successfully help students complete remediation (completion of their developmental/remedial courses as well as their corresponding college courses) (Bahr, 2010). One of the barriers of remediation is the amount of courses required to be eligible for college coursework. Developmental math usually consists of three levels: arithmetic and pre-algebra, algebra I and geometry, and algebra II (Bahr, 2010). In many community colleges, students who place into developmental math courses must complete these courses before becoming eligible for college-level math and college degree MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 9& requirements. However, successful remediation, or even developmental math completion is very low. Only 20% of students who place into developmental math successfully complete remediation in three years (Bailey, et al., 2009). Therefore, although community colleges have attempted to address this issue, there is still much work to be done. Math summer bridge programs have recently been introduced to address this issue and help improve student success (Sablan, 2014). Summer bridge programs are usually implemented to help students catch up academically and gain other supplemental skills (e.g. study skills) that aid individuals in becoming successful students (Achieving the Dream, 2014; Calcagno, et al., 2008; Sablan, 2014). Especially in community colleges where the majority of students do not dorm, summer bridge programs are viewed as one way to improve cohort experiences and a sense of place due to the lack of “college life” experience (e.g. dorming), and nurture student engagement (Calcagno, et al., 2008). This supports research showing that students who feel more connected to the campus are more likely to spend time on campus, attend classes, and generally engage in college (Bahr, 2010; Calcagno, et al., 2008; Crisp & Nora, 2010; Goldrick-Rab, 2010; Barbatis, 2010, Tinto, 1993). Tinto’s Model of Student Integration (1993) is built on the importance of social integration (e.g. participation in campus activities, interaction with peers) in strengthening students’ commitment to earning a college degree. How to measure student success Student grade point average (GPA), persistence, and graduation are used frequently to measure student success. These measures show the quality of effort a student is putting towards their education as well as the level of mastery of their coursework (GPA), if they are continuing MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 10& to work towards their goal of earning a degree (persistence), and if they accomplish their goal (graduation). College Self-Efficacy and Math Summer Bridge This study aims to address the gap in the literature examining the changes in college self- efficacy after experiencing a specific intervention (e.g. math summer bridge program) (Wright, et al., 2012). Although there is a significant amount of literature supporting academic/college self-efficacy as a significant factor in prediction and supporting student success (Brown, et al., 2008; Gore, 2006; Robbins, et al., 2004; Wright, et al., 2012; Zajacova, Lynch, & Espenshade, 2005), this study will identify one specific intervention with a specific focus of developmental mathematics, and measure how it affects college self-efficacy for all students, as well as specific focus to a) Native Hawaiian students, and b) nontraditional community college students. The Windward Community College Math Summer Bridge (MSB) Program is an intervention (that students participate in at any point throughout their college career) to help students complete a developmental math course and boost college self-efficacy, therefore improving student success. Statement of the Problem Retention in community college systems is a continual problem. Almost half (43.6%) of all CC students stop out without earning any degree or transferring to another institution (Shapiro & Dundar, 2012), which is strongly related to the statistic of up to 90% of community college students place into developmental education (Kerrigan & Slater, 2010). Beyond having to take extra courses, extending time to earning a degree, developmental education students are also more likely to drop out of school due to low self-esteem and/or weak motivation (Bahr, 2012). Community colleges need to find innovative ways to address multiple barriers that its students face. It is hypothesized, math summer bridge programs that address not only the developmental MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 11& math component, but also college self-efficacy, sense of belonging, and engagement and satisfaction could be a more holistic way to address student success. Purpose of the Study The purpose of the study is to examine the effects of the Windward Community College Math Summer Bridge Program on students’ college self-efficacy, sense of belonging, engagement and satisfaction, and long-term student successes (GPA, persistence, and graduation). The null hypothesis is that the WCC Math Summer Bridge Program does not positively impact students’ attitudes, beliefs, and behaviors (college self-efficacy, sense of belonging, engagement and satisfaction) or successes (GPA, persistence, and graduation). The Student Involvement Theory and I-E-O Framework (Astin, 1993) will be used to understand community college student inputs, environment, and outcomes, as well as gain insight into how these variables interact with one another. Social Cognitive Theory (Bandura, 1977, 1986), and Social Cognitive Career Theory (Lent, Brown, & Hackett, 1994) frameworks will also be applied to understand each characteristic within the Inputs (Ethnicity, Sex, Socioeconomic Status, Compass Test Score, Age, and Cohort Year), Environment (Employment and Full-time/Part-time College Status), and Outcomes (College Self-Efficacy, Sense of Belonging, Engagement and Satisfaction, Persistence, Graduation, and Cumulative GPA). College self-efficacy will be measured to determine how community college students self identify their beliefs on their capabilities to complete the math summer bridge course and to persist through graduation and/or transfer. MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 12& Research Questions In order to contribute to the literature on community college student college self-efficacy within the framework of a developmental math summer bridge program, this study is guided by six research questions. The first three measure mindset and/or behaviors that lead to student success (college self-efficacy, sense of belonging, and engagement and satisfaction). Data was collected from the experiment group (n = 165) to address these three questions: •! RQ1. Is there a significant change in college self-efficacy for students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 1A. Is there a significant change in college self-efficacy for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 1B. Is there a significant change in college self-efficacy for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? •! RQ 2. Is there a significant change in sense of belonging for students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 2A. Is there a significant change in sense of belonging for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 2B. Is there a significant change in sense of belonging for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 13& •! RQ 3. Is there a significant change in engagement and satisfaction for students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 3A. Is there a significant change in student engagement and satisfaction for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 3B. Is there a significant change in student engagement and satisfaction for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? The following three questions measure student success through persistence, graduation, and cumulative grade point average (GPA). Data was collected from the experimental group (n = 294), Control Group A (n = 203), and Control Group B (n = 460) to address these questions: •! RQ 4. Do the Math Summer Bridge Program students persist at a greater rate than non-Math Summer Bridge Program students? o! RQ 4A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students persist at a greater rate than their non-Math Summer Bridge Program counterparts? o! RQ 4B. Do the nontraditional or traditional Math Summer Bridge Program students persist at a greater rate than their non-Math Summer Bridge Program counterparts? •! RQ 5. Do the Math Summer Bridge Program students graduate at a greater rate than non- Math Summer Bridge Program students? o! RQ 5A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students graduate at a greater rate than their non-Math Summer Bridge Program counterparts? MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 14& o! RQ 5B. Do the nontraditional or traditional Math Summer Bridge Program students graduate at a higher rate than their non-Math Summer Bridge Program counterparts? •! RQ 6. Do the Math Summer Bridge Program students have a higher cumulative grade point average (GPA) than non-Math Summer Bridge Program students? o! RQ 6A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students have higher cumulative GPAs than their non-Math Summer Bridge Program counterparts? o! RQ 6B. Do the nontraditional or traditional Math Summer Bridge Program students have higher cumulative GPAs than their non-Math Summer Bridge Program counterparts? Methodology This dissertation will consist of two parts, 1) pre-post experiment group paired with a cross-sectional longitudinal design, and 2) comparison group design. This study consists of 957 total students (experiment and two control groups). The experiment group has 294 students (N=294), 82 of which were in the 2015 cohort. Control Group A consists of 203 students who took the same developmental math courses during the same summers as each of the experiment group students, but did not participate in the Math Summer Bridge Program. Control Group B did not take any summer math, but instead were selected because they match 126 individual MSB students on seven (7) characteristics: (1) First semester enrolled; (2) Current enrollment status (full- or part-time); (3) Age; (4) Ethnicity (Native Hawaiian or non-Native Hawaiian); (5) Sex; (6) Pell Grant recipient (or not); and (7) Compass placement test score. As of Fall 2015, there have been four cohorts (2012, 2013, 2014, 2015) of students and each cohort will serve as different points in time before or after the intervention (WCC Math Summer Bridge). This study MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 15& is a quantitative study, measuring quantitative scores on the College Self-Efficacy Inventory (Solberg, O’Brien, Villareal, Kennel, & Davis, 1993), Sense of Belonging Inventory (Hagerty & Patusky, 1995), engagement, satisfaction, persistence, graduation, and cumulative GPA. Quantitative measures including multi-linear regressions, One-Way Analyses of Variance (ANOVA) tests, and Post-Hoc t-tests will be used. Importance of the Study In today’s economy, there is a great need for a college education; both for individuals’ and the larger job market’s benefit (Baum, et al., 2013; Carnevale, et al., 2013). However, there are also a staggering number of individuals who are not prepared to enter and succeed in college (Adams, et al., 2012; ACT, 2012). Public high school students aren’t gaining the academic and study skills necessary to enter at, and be successful in college-level courses (Almy & Theokas, 2010; Darling-Hammond, 2007; Long, Latarola, & Conger, 2009), and individuals who are thinking of entering college at a nontraditional age are also often lacking academic and technological skills needed to succeed (Silver-Pacuilla, et al., 2013). Community colleges serve as a critical bridge in helping a larger population gain access, and succeed in obtaining a higher education. These institutions provide quality educational opportunities for a lower cost than other public and private institutions, as well as pre-college level math and English courses to help individuals gain the skills necessary to be prepared for college level academics if they have not yet acquired these skills. Because academic proficiency is not one of the acceptance criteria, there is a wide spectrum of skill levels for community college students when they enter. The majority of these students enter underprepared for core subjects (math and English), and also lack the confidence in their abilities to learn the required materials. MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 16& The Windward Community College Math Summer Bridge program aims to address both of these issues (lack of math preparedness and lack of self-efficacy) by 1) providing high quality, rigorous curriculum and support throughout a developmental math course, and 2) helping increase overall college self-efficacy to promote persistence and graduation with a college degree. By exposing students to a positive and engaging experience while taking a developmental math course, it is hypothesized that they will have an increase, not only in math self-efficacy, but also college self-efficacy in general, and therefore positively impacts their chances for student success (persistence/graduation and cumulative GPA). Organization of the Dissertation Chapter 1 of this study presents the background and statement of the problem, research questions, methodology, and importance of the study. Chapter 2 presents a review of the literature. Chapter 3 presents the methodology of the study, research design, and population sample. Chapter 4 presents the results of the study. Chapter 5 provides an analysis of the findings as well as recommendations for future studies and conclusions. Finally, this study concludes with references and appendices. MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 17& CHAPTER 2 LITERATURE REVIEW Introduction This study will use Astin’s Student Involvement Theory (1975, 1977, 1999) and Inputs- Environment-Outcomes Framework (1993) to categorize inputs, environment, and outcomes that impact and help identify student success. There are a total of six (6) outcomes being measured; three are research-based characteristics and behaviors that are linked with student success: (1) college self-efficacy; (2) sense of belonging; and (3) engagement and satisfaction; and three are measures used by institutions to determine student success: (4) persistence; (5) graduation; and (6) cumulative GPA. Since the literature has shown the power of college self-efficacy on student success, this study will also use Bandura’s Social Cognitive Theory with an emphasis on self- efficacy. A conceptual model presented based on the literature has been created to show the interactions between each framework/theory and characteristics that have been identified for the study. There have been six (6) input characteristics identified: ethnicity, sex, socioeconomic status, Compass placement test score, age, and Math Summer Bridge (MSB) Program cohort year. Three (3) environmental factors were identified; the most important being the Windward Community College Math Summer Bridge program, as well as employment, and full-time/part- time college status. Through two control (non-Math Summer Bridge/no summer math) and experimental (Math Summer Bridge) groups, six (6) outcomes were measured: (1) college self- efficacy; (2) sense of belonging; (3) engagement and satisfaction; (4) persistence; (5) graduation; and (6) cumulative GPA. MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 18& This study will provide the background information for the reader to understand the purpose of this study and six research questions guiding the study. The first three measure mindset and/or behaviors that lead to student success (college self-efficacy, sense of belonging, and engagement and satisfaction). Data was collected from the experiment group (n = 165) to address these three questions: •! RQ1. Is there a significant change in college self-efficacy for students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 1A. Is there a significant change in college self-efficacy for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 1B. Is there a significant change in college self-efficacy for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? •! RQ 2. Is there a significant change in sense of belonging for students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 2A. Is there a significant change in sense of belonging for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 2B. Is there a significant change in sense of belonging for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? •! RQ 3. Is there a significant change in engagement and satisfaction for students who participated in the Math Summer Bridge Program? Is this change maintained over time? MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 19& o! RQ 3A. Is there a significant change in student engagement and satisfaction for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 3B. Is there a significant change in student engagement and satisfaction for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? The following three questions measure student success through persistence, graduation, and cumulative grade point average (GPA). Data was collected from the experimental group (n = 294), Control Group A (n = 203), and Control Group B (n = 460) to address these questions: •! RQ 4. Do the Math Summer Bridge Program students persist at a greater rate than non-Math Summer Bridge Program students? o! RQ 4A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students persist at a greater rate than their non-Math Summer Bridge Program counterparts? o! RQ 4B. Do the nontraditional or traditional Math Summer Bridge Program students persist at a greater rate than their non-Math Summer Bridge Program counterparts? •! RQ 5. Do the Math Summer Bridge Program students graduate at a greater rate than non- Math Summer Bridge Program students? o! RQ 5A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students graduate at a greater rate than their non-Math Summer Bridge Program counterparts? o! RQ 5B. Do the nontraditional or traditional Math Summer Bridge Program students graduate at a higher rate than their non-Math Summer Bridge Program counterparts? MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 20& •! RQ 6. Do the Math Summer Bridge Program students have a higher cumulative grade point average (GPA) than non-Math Summer Bridge Program students? o! RQ 6A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students have higher cumulative GPAs than their non-Math Summer Bridge Program counterparts? o! RQ 6B. Do the nontraditional or traditional Math Summer Bridge Program students have higher cumulative GPAs than their non-Math Summer Bridge Program counterparts? Astin’s Theory of Student Involvement Astin’s Theory of Student Involvement (1975, 1977, 1999) was created through research of a longitudinal study on college student dropout rates and reasons, aimed at identifying every significant effect that could be rationalized in terms of the involvement concept. Astin (1999) regards student involvement as the “amount of physical and psychological energy that the student devotes to the academic experience” (p. 518). Because the emphasis is on how the student spends their time and energy, Astin (1999), postulates the importance of behavioral aspects of the individual, in comparison to thoughts or feelings about academics. There are five basic postulates included in this theory (Astin, 1999, p. 519): 1.! Involvement is the investment of physical and psychological energy in various objects. 2.! Involvement occurs along a continuum; different students invest in a specific subject in various degrees and the same student invests different amounts of energy into different subjects. 3.! Involvement has both quantitative and qualitative features (not just how long you study, but also the quality of study time). MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 21& 4.! Student learning and personal development is directly proportional to the quality and quantity of student involvement in the program. 5.! Effectiveness of any educational policy or practice is directly related to the capacity of that policy or practice to increase student involvement. Especially in regards to number four, the Windward Community College Math Summer Bridge program aims to encourage student engagement (both time on task and effort) by scaffolding a large amount of guided study time as well as setting high standards (mandatory test retake for scores less than 95%) with the support systems in place to help the students achieve these goals (e.g. ample tutor support, instructor engagement outside of class time, close communication between tutor, instructor, staff, and counselors). This theory of student involvement provides a link between the variables emphasized in three theories (Subject Matter Theory, Resource Theory, and Individualized (Eclectic) Theory) and the learning outcomes desired by the student and professor (Astin, 1999). The more time a student spends on activities, the more they will learn (Astin, 1999). Astin’s Theory of Student Involvement (1975, 1977, 1999) has very similar content to learning theories’ time-on-task or vigilance concepts, which provides support for the paradigm shift for educators: more energy must be focused on what the student does, instead of what the instructor is teaching/doing. Student involvement (and how to get students fully engaged) should be their focus instead of blaming the student for not learning through the way the instructor teaches (Astin, 1999). According to Astin (1999), “To achieve the effects intended, (one) must elicit sufficient student effort and investment of energy to bring about the desired learning and development” (p. 522). The WCC Math Summer Bridge is set up in a way that forces the students to a) put in the time MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 22& (quantity) to study and learn, and b) elicits the quality by affording numerous trained tutors that are actively looking to help individuals. Through his longitudinal research, Astin also found that students who work full-time off- campus were more likely to drop out of college, whereas students who worked part-time on campus were less likely to drop out. This supports his theory of the importance of student involvement/engagement and is very important when understanding the community college environment. As noted earlier, the student population at community colleges are older (average age is 28 years old) and are more likely to work at least part-time (70.3% of all community college students currently work at least part-time), and the majority of these jobs are off campus (American Association of Community Colleges, 2015). Like in many other factors, such as academic preparedness, in terms of student engagement, most community colleges are already starting at a disadvantage. Inputs-Environment-Outcomes Model The framework of this study is based on Astin’s Inputs-Environments-Outcomes (I-E-O) Model (1993), which is a “conceptual guide for assessment activities in higher education” (Astin & Antonio, 2012, p. 17). For this theory, outcomes/outputs (college self-efficacy, sense of belonging, engagement and satisfaction, persistence, graduation, and cumulative GPA) must be assessed in terms of inputs and environmental factors. Outcomes are what the researcher is trying to develop or achieve, inputs are personal qualities that the students bring with them, and environmental factors include experiences that the student has while in the program/institution (e.g. WCC Math Summer Bridge). For each category (Input, Environment, or Outcome), how a variable is identified is completely relative to the nature of the problem. Given different circumstances, each component (e.g. self-efficacy) could be included in each of the three MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 23& categories (Astin & Antonio, 2012). The environment is the most critical for the instructor and/or institution because this is the one element that they can control (Astin, 1999). This theory can be applied to a vast array of disciplines and fields when trying to understand development of humans and what impacts their development (Astin & Antonio, 2012). It is a “tool for trying to understand why things are the way they are and for learning what might be done to make things different if we feel the need to change them” (Astin & Antonio, 2012, p. 22). Because of its adaptability in application, the I-E-O Model can be used for quantitative or qualitative studies and is ideal for this particular research design. Bandura’s Social Cognitive Theory The foundation of Bandura’s Social Cognitive Theory (1977, 1986) is human agency, which is “characterized by a number of core features that operate through phenomenal and functional consciousness” (Bandura, 2001, p. 1). This theory rejects dualisms about individualism and collectiveness and instead believes in the “bidirectional view of evolutionary processes” (Bandura, 2001, p. 20) where biology shapes behavior and social and technological advances impact biology (e.g. medicine or birth control). In this theory, there are three modes of agency: direct personal agency (the individual can do it themselves), proxy agency (the individual can have someone else complete goals/objectives/actions for them-e.g. politicians enact laws for the public), and collective agency (the collective group can make a societal impact when they believe in the abilities of the group as a whole). There are also four core features of human agency: (a) Intentionality; (b) Forethought; (c) Self-reactiveness; and (d) Self-reflectiveness (Bandura, 2001). Intentionality is defined as the “proactive commitment to bringing (intentions) about. Intentions and actions are different aspects of a functional relation separated in time” (Bandura, MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 24& 2001, p. 6). However, it is important to caution that although actions might have intentions of achieving a desired outcome, this might not actually happen. In order for someone to plan for and/or adapt to obstacles, individuals with agency need to also possess the second core feature, forethought. As the second core feature, forethought “motivates and guides actions in anticipation of future events (Bandura, 2001, p. 7). However, in regards to agency, future events aren’t actionable motivation (motivation that directly causes action). Instead, forethought influences cognitive motivation; the potential of future events are seen or understood as current motivation. The third core feature, self-reactiveness, includes self-evaluation and readjustments to reaching an individual’s goals or plans. This adjustment is made through comparing personal performance with goals and standards (both proximal goals, impacting what someone does in the present, and distal goals, a guide impacting a general path and/or shapes what a person does as a whole). Therefore, for the first three core features of human agency, in individual must first make action plans (intentionality), then think about what will happen (forethought), and adjust accordingly to continue motivation to follow through with the plan (self-reactiveness). Finally, the fourth core feature of human agency is self-reflectiveness. This is when the individual “evaluate(s) motivation, values, and the meaning of their life pursuits” (Bandura, 2001, p. 10). It is within this feature that self-efficacy comes into fruition. Self-efficacy is a very popular psychosocial variable used to examine human behavior in a wide variety of settings, especially in education. Eleven percent of all articles published in the Journal of Counseling Psychology, Journal of Vocational Behavior, and Journal of Career Assessment between 2001 and 2005 were related to self-efficacy (Gore, 2006). MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 25& Self-Efficacy Bandura (2001) calls self-efficacy the “foundation of human agency” (Bandura, 2001, p. 10) because people need to believe that their actions can achieve goals and/or produce desired results, which in turn motivates them to start, persist, and achieve those outcomes; self-efficacy has a “central role in the self-regulation of motivation” (Bandura, 2001, p. 10). Self- reflectiveness and self-efficacy helps individuals select which obstacles they attempt to overcome, how much effort to exude, and how long to persist in trying to overcome those obstacles in attempts of achieving a goal. These two concepts also frame how an individual views failure: as “motivation or demoralizing” (Bandura, 2001, p. 10). This type of efficacy, coping efficacy, not only serves as a navigator/predictor of future endeavors, but individuals who have high coping efficacy have a decreased chance of stress and depression, and also increases in resiliency when faced with difficult situations (Bandura, 2001). When working with college students, and more specifically developmental math community college students, this is why self-efficacy (coping and general) is so important. It can be a pivotal quality that helps students with less previous experience, opportunity, and academic achievement overcome these obstacles to achieve their goals of earning a college degree. As these students are faced with adversities and/or nuanced stress presented when enrolling in college and taking classes, they are more likely to have heightened anxiety towards attending and persisting in school (Bahr, 2012). However, if they have stronger college, coping, and personal self-efficacy, they will be more likely to persist and succeed. Self-efficacy not only helps individuals thrive in high-stress situations, but it also influences what people engage in and what environments they surround themselves with (Bandura, 2001). If a student believes that they can succeed in college, they are more likely to MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 26& stay on or around campus and socialize with their college peers and faculty (DeWitz, Woosley, & Walsh, 2009). This also has a reciprocal relationship, in which an increase in students’ feelings/sense of belonging to a community or connectedness with others (e.g. the community college) increases their self-efficacy as well (Martin & Dowson, 2009). In this generation, students are fortunate to have increased and easier access to larger bodies of knowledge. Technology and society have especially been pivotal in improving this accessibility to knowledge (e.g. Google and the internet). By having the ability to search for knowledge at home, at any hour of the day, seven days a week, students are given more control over self-learning. This in turn, impacts self-efficacy (usually positively) for development and self-renewal (Bandura, 2001). Although, for students who do not possess the technological skills needed to access these innovations (e.g. older/nontraditional students), they could have lower college self-efficacy. Self-efficacy is domain specific (Bandura, 1977, 1986, 1997). An individual has different levels of self-efficacy depending on the task or situation. This can be seen not only in academic versus non-academic situations, but can also differ between subjects within an academic setting (e.g. high English self-efficacy but low math self-efficacy) (Jameson & Fusco, 2014; Pajares, 1996; Wright, et al., 2012). Self-efficacy comes from four sources of information: (1) personal accomplishments; (2) verbal persuasion (i.e. encouragement from faculty or peers); (3) vicarious learning experiences; and (4) physiological and affective reactions (Bandura, 1986). Self-efficacy and student success. Many studies have shown a direct link between self- efficacy and academic performance, academic achievement, and academic goals (Gore, 2006; Hall & Ponton, 2005; Jameson & Fusco, 2014; Lent, Brown, & Hackett, 1994; Majer, 2009; Multon, Brown, & Lent, 1991; Pajares, 1996; Wright, et al., 2012; Zimmerman, Bandura, & MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 27& Martinez-Pons, 1992). Gore (2006) found a strong relationship between efficacy beliefs of self- regulatory behaviors (e.g. finishing homework) and efficacy beliefs relating to academic achievement. Out of the four sources of information that an individual derives self-efficacy from (personal performance accomplishments, verbal persuasion, vicarious learning experiences, and physiological and affective reactions), personal performance accomplishments is the “most influential” (Gore, 2006, p. 110). Therefore, as practitioners and administrators, community college faculty and staff need to view student success through the lens of creating situations where students have positive personal experiences that build their college self-efficacy. College self-efficacy. First called “Academic Self-Efficacy”, this concept referred to a student’s confidence in their abilities to be successful with academic tasks (Chemers, Hu, & Garcia, 2001; Gore, 2006). It later evolved into “College Self-Efficacy” pertaining to the level of confidence in one’s abilities to effectively complete tasks related to college success as a whole (Solberg, et al., 1993). College self-efficacy encompasses the belief in success in a student’s overall college experience, including social and academic integration constructs from Astin (1999) and Tinto (1993) providing a all-encompassing concept in understanding college student development (Chemers, et al., 2001; Gore, 2006). Due to many pertinent studies looking at academic or college self-efficacy, for the purposes of this study, both will be considered in the literature review. Even after controlling for relevant variables (i.e. gender, ethnicity, first generation status), increased college self-efficacy at the end of a student’s first semester is “associated with significantly higher odds of persisting” (Wright, et al., 2012, p. 302; Brown, et al., 2008; Gore, 2006; Robbins, Lauver, Le, Davis, Langley, & Carlstrom, 2004). Academic self-efficacy MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 28& accounts for more variance in college outcomes (performance and persistence) than more traditional predictors (e.g. ACT/SAT scores or high school GPA) (Brown, et al., 2008; Gore, 2006; Kahn & Nauta, 2001; Lopez, Lent, Brown, & Gore, 1997; Multon, et al., 1991; Pajares & Miller, 1995) and could account for up to 14% of the variance in college students’ GPA (Brown, et al., 2008). Gore (2006) showed a significant correlation between academic self-efficacy and college persistence (mean correlation r = .26). However, Gore (2006) also points out that the predictive value of academic self-efficacy is dependent on three factors: (1) when self-efficacy beliefs are being measured; (2) what aspect of self-efficacy is being measured; and (3) what college outcome one wishes to predict (p. 112). Academic self-efficacy is not a good predictor of student success before entering college, but is significant after the 1 st and 2 nd semesters of college (Gore, 2006). Studies have shown that college students are more likely to give an accurate account of college self-efficacy after they have had first-hand experience in attending college. Before college, students are “college naïve” (Gore, 2006, p. 110), however after the first semester, they are more likely to have an accurate idea of what college is really like and can give a more educated assessment of what they believe they’re capable of (Gore, 2006; Gore, Leuwerke, & Turley, 2005; Wright, et al., 2012; Zajacova, et al., 2005). The significance of how self-efficacy can impact college student success, paired with the idea that a student doesn’t necessarily need a high college self-efficacy before entering school (because it is more important that it increases while they’re enrolled), shows that there can be interventions within the student’s first year of college that can significantly increase their chances of persisting and succeeding in college. It also concludes that no matter what stages a student is in (freshman, sophomore, etc.), an intervention that abruptly increases college self- MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 29& efficacy (e.g. math summer bridge) could put a student on a higher trajectory for success. With strong student success prediction and correlation, measuring students’ college self-efficacy pre and post interventions (such as First Year Experience programs or the Windward Community College’s Math Summer Bridge program), can be a good way of measuring program effectiveness and chances of college success (Gore, 2006). The first semester of college is a critical time for enhancing students’ self-efficacy beliefs (Gore, 2006). Self-efficacy and developmental students. The importance of college self-efficacy on student success can be seen when making a comparison between developmental and college- ready students. Contrary to what might seem like common sense, academic aptitude is not a strong predictor of college success. According to Brown, et al. (2008), students who place into college-level courses are no more likely to earn a college degree than those who place into developmental courses unless they develop “strong confidence in their college academic capabilities and robust goals for college completion” (p. 306). These studies provide strong support for the community college open-door policy and the importance of being able to provide students opportunities to thrive and succeed, no matter what their SAT/ACT or GPA were in high school. These opportunities can help students rewrite how they perceive themselves as students and create a new baseline of past performance accomplishments (Brown, et al., 2008), especially through special programs such as the Windward Community College Math Summer Bridge. This is the purpose of this study: to provide students with low college and/or math self-efficacy an opportunity to create new and successful experiences, which will in turn help to boost college self-efficacy and increase the chances of student success. MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 30& Self-efficacy and math anxiety. Community colleges (and higher education institutions in general) are at a disadvantage when it comes to math anxiety because students enter college with preconceived notions and experiences about math and math classes (Hekimoglu & Kittrell, 2010). Changing these beliefs is the ideal starting point in not only helping students alter the math aversion path they’re currently on, but also when effectively teaching math (Hekimoglu & Kittrell, 2010). Math self-efficacy, like other types of self-efficacy, not only affect how the student views their capabilities and potential for success in math courses/curricula, but also guide how much effort, motivation, and persistence one exudes when trying to overcome mathematic obstacles (Hekimoglu & Kittrell, 2010). Cognitive dissonance, including providing opportunities where students’ prior assumptions about their mathematic capabilities are proven wrong, allows the educator to help students overcome math anxiety and improve math self- efficacy levels. Self-efficacy and traditional v. nontraditional students. As mentioned in Chapter 1, the definition of nontraditional student varies and can include: (a) a delay in enrollment following high school; (b) part-time/full-time enrollment status; (c) working more than 35 hours/week; (d) financially independent (e.g. orphan or foster child); or has (e) dependents (children) (Philibert, et al., 2008). These students face unique barriers. Mercer (1993) categorized these barriers into three groups: situational, dispositional, and institutional. Ritt (2008) had similar categories: personal (personal/family commitments, work and family schedules, financial limitations, general fear of returning to school), professional (lack of support by employer, time release, etc.), and institutional (tuition cost, class availability/day/time). Beyond barriers, many nontraditional students also lack confidence in their abilities to succeed in college, and paired with self-consciousness about their “nontraditional” status can MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 31& lead to lower performance rates and academic self-efficacy in comparison to their traditional peers (Schlossberg, Lynch, & Chickering, 1989; Lundberg, McIntire, & Creasman, 2008). This is also true when looking specifically at math self-efficacy in nontraditional students. According to Jameson and Fusco (2014), adult learners have higher levels of math anxiety and lower levels of math self-efficacy than students of traditional ages. These attitudes could be heightened by other students’ stereotypes of nontraditional students’ lack of mathematic abilities (Hollis- Sawyer, 2011). Conversely, adult learners with higher levels of self-efficacy are more satisfied with college experience and more likely to persist when faced with obstacles (Kemp, 2002; Lim, 2001; Tyler-Smith, 2006). Therefore, college self-efficacy is an important factor to understand within the scope of the adult learner. For the purposes of this study, college self-efficacy will encompass the student’s perception of how capable they are to achieve various tasks associated with academic success (e.g. completing a developmental math course, interacting with faculty/staff, taking good notes). Self-efficacy is a domain-specific, socio-cognitive concept that has consistently shown positive correlations with academic success. This study will look at how college self-efficacy specifically correlates with academic persistence, graduation, and cumulative GPA in community college students. Conceptual Model The impact of self-efficacy on community college student success is a topic that has gained momentum for research over the past 20 years. So has the impact of developmental math placement on community college student success (Bailey, et al., 2010; Bahr, 2010; Bahr, 2012; Long, Latarola, & Conger, 2009). However, there has been little research on successful interventions to help address self-efficacy and developmental math issues in community college MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 32& students. This study aims to examine the long- and short-term significance of a specific intervention (Windward Community College Math Summer Bridge program) on college self- efficacy, sense of belonging, engagement and satisfaction, and college success (persistence, graduation, and cumulative GPA) of community college students. This section of the study will identify the conceptual model used, as well as each of the factors identified within Astin’s (1993) Inputs-Environment-Outcomes Model. It is hypothesized that each input and environmental factor has an effect on the six identified outcomes (college self-efficacy, sense of belonging, engagement and satisfaction, persistence, graduation, and cumulative GPA), however the Windward Community College Math Summer Bridge will have the largest positive effect. This study will also critically examine the difference in effects on a) non-Native Hawaiian vs. Native Hawaiian, and b) traditional vs. nontraditional Windward Community College students. Background Factors (Input) The first component of the I-E-O framework is the Input or background factors. All six factors [(1) ethnicity (Almy & Theokas, 2010; Bahr, 2010; Brayboy, Castagno, & Maughan, 2007; Condron & Roscigno, 2003; Flores & Park, 2013; Long, Latarola, & Conger, 2009; Melguizo, 2010; The Civil Rights Project, 2012; White & Lowenthal, 2011); (2) sex (American Association of University Women, 2013; Buchman & DiPrete, 2006; Goldin, Katz, & Kuziemko, 2006); (3) socioeconomic status (Robbins, et al., 2004; Rubin, 2012); (4) Compass placement test score; (5) age; and (6) cohort year] have been proven to have significant effects and/or correlations with community college student success . The following section provides a brief explanation of each factor and how this information will be collected. MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 33& Figure 1. Astin’s Inputs-Environment-Outcomes Model Inputs:( Ethnicity Sex Compass& Placement& Test& Score Age Cohort& Year Environment( (Control( Group) & Employment & Full<time/Part< time& Enrollment Environment( (Experimental( Group)( WCC(Math( Summer(Bridge( Employment& Full<time/Part< time& Enrollment Outcomes:( College& Self<Efficacy& Sense& of& Belonging& Engagement& && Satisfaction& Persistence& Graduation& Cumulative& GPA & & Figure 1. Conceptual Model of the Impacts of the WCC Math Summer Bridge Program on college self-efficacy, sense of belonging, engagement and satisfaction, persistence, graduation, and cumulative GPA MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 34& Ethnicity Ethnicity does not cause a student to be successful in college or not. However, numerous studies have shown a correlation between minorities and lack of college student success (persistence, graduation, GPA) (Almy & Theokas, 2010; Bahr, 2010; Brayboy, Castagno, & Maughan, 2007; Condron & Roscigno, 2003; Flores & Park, 2013; Long, Latarola, & Conger, 2009; Melguizo, 2010; The Civil Rights Project, 2012; White & Lowenthal, 2011). This correlation has been shown to have direct links to other input factors included in this study. Minority students are more likely to come from low-socioeconomic backgrounds (Almy & Theokas, 2010; Bahr, 2010; Brayboy et al., 2007; Condron & Roscigno, 2003; Long, et al., 2009), low-quality K-12 education systems (Almy & Theokas, 2010; Long, et al., 2009), and are less likely to complete a college education (The Civil Rights Project, 2012; White & Lowenthal, 2011). Given that overall, half of the community college population is comprised of minority students (American Association of Community Colleges, 2015) and 80% of Windward Community College’s student population is made up of minority students (Windward Community College, 2014), it is important for this study to collect this background information. This study will give specific attention to Native Hawaiian students. Throughout the University of Hawaii System, Native Hawaiian students are overrepresented in remedial/developmental math courses (37.3% in math versus 23% of total population) (Office of Hawaiian Affairs, 2014), which the researcher believes emphasizes the importance of identifying developmental/remedial math programs that especially benefit Native Hawaiian students. Sex Similar to ethnicity, sex/gender is not a causational factor of student success, but there have been studies linked to the correlation of college student success based on sex/gender. As a MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 35& whole, more women are attending, persisting, and graduating from higher education institutions than men. At 4 year institutions, women outnumber men 56% to 44% overall in undergraduate enrollment (National Center for Education Statistics [NCES], 2014) and at the community colleges, 57% to 43% (American Association of Community Colleges, 2015). Women are also slightly more likely to earn their associate degree than men (American Association of University Women, 2013; Buchman & DiPrete, 2006; Goldin, Katz, & Kuziemko, 2006). For the purposes of this study, sex (not gender) will be surveyed. Socioeconomic Status Socioeconomic status has been strongly correlated to college student success. Students who come from lower socioeconomic backgrounds are (a) less likely to have access to high- quality K-12 education systems (Almy & Theokas, 2010; Darling-Hammond, 2007; Long, et al., 2009) and (b) more likely to place into developmental education when in college (ACT, 2012; Bahr, 2010). The effects of these two barriers can be seen throughout the research. Students who come from lower socioeconomic backgrounds are less likely to attend, persist, and graduate from college (Robbins, et al., 2004; Rubin, 2012). Since the majority of students are receiving financial aid and come from lower socioeconomic backgrounds, community colleges need to be sensitive to the specific challenges that their students face (Pascarella, 2006; Rubin, Denson, Kilpatrick, Matthews, Stehlik, Zyngier, 2014). For the purposes of this study, socioeconomic status will be self-reported by the student and recorded on the demographic survey. Rubin, et al. (2014) states that self-reporting of socioeconomic status is not only a valid measure, but also provides insight as to how the student views himself or herself. MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 36& Compass Placement Test Score At Windward Community College the placement test used to identify which math and English courses students will take is the Compass test created by ACT. This score is used to identify which level of math each student must start at. Students are not allowed to take any math (or English) courses higher than what they placed into. Because the course numbers have changed over the past 5 years, the researcher instead identifies student placement by the level (below college math) they place into: 1 level below (Elementary Algebra II), 2 levels below (Elementary Algebra I), and 3 levels below (Pre-Algebra). This information will be collected through records obtained through the WCC Institutional Research Office. Age Student’s age will be collected through the Windward Community College institutional research office. This is to identify students as either traditional or nontraditional students. For the purpose of this study, traditional students are considered to be 25 or younger, and nontraditional students 26+ years old. Literature shows that age has a direct link with college self-efficacy (Jameson & Fusco, 2014; Lundberg, et al., 2008). Cohort Year Finally, the year that each student participated in the WCC Math Summer Bridge program will be collected. This is to measure how much time has lapsed since they have participated in the intervention and to show the long-term effects (if any) of the intervention. Students from the 2012 cohort are 3 years post-intervention, 2013 are 2 years post-intervention, 2014 are 1 year post-intervention, and 2015 cohort took both a pre- and post-test to measure the immediate effects of the WCC Math Summer Bridge Program (intervention). This information MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 37& will be collected through student self-report as well as through information obtained through the WCC Institutional Research Office. Environmental Factors (Environment) Three environmental factors have been identified for this study: (a) the Windward Community College Math Summer Bridge program; (b) employment; and (c) full-time/part-time student status. Windward Community College Math Summer Bridge Program The Windward Community College (WCC) Math Summer Bridge (MSB) program began in 2005 and evolved to its present form in 2012. It is grant funded and provides developmental math courses at all three levels that are offered at WCC (pre-algebra, elementary algebra I, and elementary algebra II). Over the past four years (2012-2015), this program has offered 19 developmental math courses to 294 students. Their success rates are much higher than the 16- week semester developmental math courses: 45.2% earned an “A”, 80.7% earned an “A” or “B”, and 88.3% passed with an “A” “B” or “C”. This bridge program is seven weeks long, Monday through Friday. Students who participate must attend all classes (2 hours per day on Monday, Wednesday, and Friday) as well as a minimum of two hours of study hall every day (Monday through Friday). There are at least 5-10 tutors on staff at all times (7:00am-6:00pm) to help students when they need assistance, and many times, also open on Saturday and Sunday for extra study hall hours, tutoring, and retesting. The mandatory study hall hours are not only to ensure that they complete their homework everyday before leaving campus, it also helps students practice good study habits, and strongly encourages student involvement with the college, instructors, faculty and staff, and peers. MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 38& This program provides opportunities for students to access all four of Bandura’s core features of human agency: (1) intentionality; (2) forethought; (3) self-reactiveness; and (4) self- reflectiveness. In order for students to join this program, they have to fill out an application and meet with the academic counselor running the program. During this meeting, students are made aware of the expectations and obligations of the program (e.g. mandatory study hall, test retake policy) and after this orientation, they sign a contract (intentionality), which is a “proactive commitment to bringing (their intentions) about” (Bandura, 2001, p. 6). As previously stated, when students are forewarned about the time commitment of the program and level of dedication expected of them, it provokes Bandura’s socio-cognitive concept of human agency, forethought, to motivate and guide their actions for the future MSB and math course success. This includes spending about 20 hours/week on campus and retaking chapter tests if they receive anything lower than a 95%. Students will not pass the class if they have a test score lower than 70% and are only able to take 2 retakes. However, these high expectations are paired with both academic and non-academic support systems to help each student achieve these goals. Weekly tests and daily homework assignments (with consistent and immediate feedback) provide numerous opportunities for self-reactiveness, where students can evaluate and adjust their actions in order to reach their goals (e.g. passing course or future college major). And finally, in conjunction with these opportunities for self-reactiveness, students can look inward, as well as talk with the academic counselors, tutors, and staff for self-reflectiveness. Throughout this program, sometimes on a daily basis, students are re-evaluating, adjusting, and reflecting on what their goals are, the strategies they are using, and better ways of doing so. Math instructors for this program are selected based on their approachable, engaging, and hands-on teaching styles, as well as their willingness to commit to providing extra tutoring MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 39& support and working well outside their normally scheduled office hours. Tutors for this program are also selected for similar qualities and many students who participated in the program come back as tutors for following summers. These tutors are trained to actively float between different students in the room, making sure to elicit interaction from all students (not just those who speak up and ask for help). The large number of tutors not only helps ensure there is enough coverage, but also increases the chances of diverse tutoring styles to address diverse student learning styles. Communication is another key piece of this program. There is constant communication between students and their academic counselors (specific to the Math Summer Bridge Program), students and their instructors, and academic counselors and instructors. Academic counselors employ intrusive advising, which “provides a mechanism to nurture students, assist them with academic plans, build relationships, and create connections with the institution” (Jones & Hansen, p. 89-90; Ryan, 2013; Smith, 2007; White & Schulenberg, 2012), and has been shown to increase student completion and retention (Orozco, Alvarez, & Gutkin, 2010). Intrusive advising not only provides a direct and up to date link between student, instructor, and counselor around student progress, it also shows the student that each person involved has a genuine concern for their progress. Since almost all of the students come from low-income backgrounds, the program is free and books are provided. There is also free lunch every day and free breakfast on test days. This helps students focus more on learning the mathematic content and less on other barriers (e.g. lack of funds for supplies or food, hunger), and also helps build a sense of community and belonging. Especially in Native Hawaiian and Hawaii’s local cultures, food is seen as an important aspect of gatherings and meetings and reinforces the specific cultural practices that these students are familiar with. Providing food also creates an environment where students socialize not only with MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 40& their peers, but also with their academic counselors, instructors, staff, and administration. All aforementioned parties come to the TRiO Student Support Services office on a daily basis to sit and eat with the students. By providing students the opportunity to interact with various key members of the campus, students will be more likely to be comfortable in asking for help, and engage with key faculty, staff, and administration, as well as become better informed about where to seek help or find various resources, thus increasing student involvement with the campus and student success. Success of the MSB program will be measured by the students’ final grades for the math course. This program takes a holistic approach to servicing all aspects of the developmental student: personal, environmental, social, and emotional. Research shows in order to be successful, developmental programs must address all of these facets of the developmental community college student, which in turn “influence students’ experiences and perceptions about learning” (Di Tommaso, 2012, p. 942; Grimes & David, 1999; McCabe, 2003). Table 1 shows the key components of the Math Summer Bridge Program and how each relates to the key concepts of student success. MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 41& Table 1 Key Components of the WCC Math Summer Bridge Program Type Component Structure 7 weeks, 5 days/week, ≥ 2 hours/day Structure Application/orientation/contract Structure Mandatory study hall ≥2 hours/day. Extra study hall hours Saturday and Sunday Structure Mandatory test retake for <95% Instructor Approachable, engaging, and hands-on teaching styles. Provides extra tutoring support, works well outside normal office hours Support Communication between: -students and academic counselors -students and instructors -academic counselors and instructors Support Academic counselors employ intrusive advising Support Free tuition and books Support Free lunch every day Support Free breakfast on test days Support ≥ 5-10 tutors on staff at all times (7:00am-6:00pm) Support Tutors trained to actively float between students, eliciting interaction from all students (not just those who ask for help). Many tutors are previous students of the MSB Program Employment As previously mentioned, in comparison to their 4 year counterparts, community college students are more likely to hold a job while attending school. According to the American Association of Community Colleges (2015), 69% of community college students work at least a part-time job. This is correlated with, the majority of the community college population (63%) being over the age of 21 and come from lower socioeconomic backgrounds (American Association of Community Colleges, 2015). Number of hours a student works per week has MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 42& been directly correlated with student persistence. The more hours a student works, the less likely they are to persist (Lanni, 1997; Nakajima, et al., 2012). This is due to the fact that students who are spending more time working have less time to focus on studying and completing school related tasks and are more likely to be sleep deprived (Nakajima, et al., 2012). Nakajima et al. (2012) also found that a higher number of hours worked directly correlates to family and financial responsibilities, “indicating that (these students) had increased barriers to overcome compared to those who did not work” (p. 604). In the scope of paying bills versus attending school, students with family and financial responsibilities naturally chose work in order to uphold these responsibilities. Student employment status will be collected through self-report. Students will be asked three questions: (1) if they currently work; (2) if they work on or off campus (or both); and (3) on average, how many hours they work per week. Full-time/Part-time Student Status With a similar trend to employment, community college students are more likely to attend college part-time. In fact, the majority of community college students (61%) attend school only part-time (American Association of Community Colleges, 2015). The first issue that part-time students face is that it simply takes more time to graduate. And although one might think that when given the extra time to complete degree requirements, part-time students should be just as successful, unfortunately, that is not the case (Complete College America, 2011). There are other facets of a part-time student’s life that dramatically affect their abilities to succeed. Part-time students are more likely to have other responsibilities (e.g. family or financial), which take up more time, leaving less for academics and research has shown that their lack of available time as the reason for their lack of success (MacCann, Fogarty, & Roberts, 2012). These barriers are reflected in the data. When giving part-time students double the time MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 43& (4 years) to complete a 2-year associate degree, only 7.8% of part-time students were successful (compared to 18.8% of their full-time counterparts) (Complete College America, 2011). Although it is not the focus of this study, 52% of the Windward Community College student body are part-time students (Windward Community College, 2014) and part-time/full-time status is an important piece in understanding external (environmental) factors that affect community college student success. Part-time/full-time status will be collected both at the time of attending the Math Summer Bridge and what their current status is (if they completed the MSB in previous years) through student self-report and internal checks with data collected from the WCC Institutional Research Office. Outcome Variables (Output) Six outcome variables have been identified for this study: (1) college self-efficacy; (2) sense of belonging; (3) engagement and satisfaction; (4) persistence; (5) graduation; and (6) cumulative GPA. College self-efficacy will be measured by using the College Self-Efficacy Inventory, created by Solberg, O’Brien, Villareal, Kennel, & Davis (1993). Sense of belonging will be measured with a modified version of the Sense of Belonging Inventory, created by Hagerty & Patusky (1995). Engagement (instructor and peer) and satisfaction will be measured by nine multiple-choice questions on the survey. Persistence, graduation, and cumulative GPA are the student success measures analyzed in this study. College Self-Efficacy Evolving from Bandura’s Social Cognitive Theory (1977, 1986), college self-efficacy pertains to how confident the student is in their abilities to complete tasks related to being successful in college (Solberg, et al., 1993). This includes completing homework assignments, interacting with faculty and peers, and comprehending content. Self-efficacy pertaining to MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 44& college success and goal setting are two qualities that have been strongly linked to higher GPAs, and therefore higher persistence rates (Nakajima, et al., 2012). As mentioned earlier, having higher college self-efficacy has been shown to increased chances of persistence and graduation (Brown, et al., 2008; Gore, 2006; Majer, 2009; Pajares, 1996; Robbins, et al., 2004; Wright, et al., 2012) and is a valid measure for this study. Sense of Belonging In the college context, Strayhorn (2012) defines sense of belonging as “students’ perceived social support on campus, a feeling or sensation of connectedness, the experience of mattering or feeling cared about, accepted, respected, valued by, and important to the group (e.g., campus community) or others on campus (e.g., faculty, peers)” (p. 3). This concept has been linked to motivation, achievement, and influence on behavior (Bozak, 2013; Hagerty, Lynch- Sauer, Patusky, Bouwsema, & Collier, 1992; Hagerty & Patusky, 1995; Strayhorn, 2012). The two main attributes of sense of belonging are 1) valued involvement and 2) fit of individual to the environment (Hagerty, et al., 1992). The importance of sense of belonging has been shown as important for not only traditional aged college students, but for returning adult college students as well (Harris, 2006-2007). For the WCC Math Summer Bridge, valued involvement is addressed through the constant feedback and support from instructors, tutors, counselors, and staff, and by creating an environment where students are strongly encouraged to connect with one another in order to succeed. Students also are surrounded with peers who have placed into developmental math and many of whom have higher math aversion/anxiety, which addresses their individual fit to the environment. By providing non-academic services (e.g. free breakfast and lunch), the MSB program aims to provide a heightened sense of belonging not only to the program, but also to the college as a whole. MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 45& Sense of belonging will be measured with Hagerty and Patusky’s (1995) Sense of Belonging Inventory (SOBI). With a total of 18 Likert scale questions, this inventory has two parts (SOBI-A and SOBI-P), addressing the proposed antecedents and psychological tenets of sense of belonging. When this inventory was created, it was first tested on community college students with an internal consistency reliability of .93 and .72 and test-retest reliability over an 8- week period of .84 and .66 for SOBI-P and SOBI-A respectively. Faculty Interaction Faculty and student interaction (FSI) can occur both inside and outside of the classroom. FSI has been shown to be an important environmental factor for student success. Astin (1985) states that the “best way” to engage students is to increase interactions between faculty and students (p. 162). The WCC Math Summer Bridge program does this by having increased office hours, provides free breakfast and lunch, and mandatory interaction with tutors and instructors. As students’ interactions with faculty members increase both in and outside of classrooms, their development and satisfaction also increase (Astin, 1985). For community college students, the opportunities for FSI are less likely outside of the classroom due to lack of extracurricular activities and events. However, when they do interact, there is also a strong positive correlation between faculty-student interaction (FSI) and community college GPA and other aspects of student learning (Lundberg, 2014; Thompson, 2001; Wirt & Jaeger, 2014). Thompson (2001) found that students who report having more interaction with faculty members also report having more concern about academic achievement and exert more effort in reaching academic goals. Faculty-Student Interaction will be measured by two quantifying questions: “How many times do you talk to/interact with your instructor during class?” “How many times do you talk to/interact with your instructor outside of class?” both questions are multiple choice: (a) 0 MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 46& times/week; (b) 1-3 times/week; (c) 4+ times/week. This is to measure the students’ perception of how often they interact with faculty members. Peer Interaction For the purposes of this study, peer interaction pertains to interaction with fellow students at Windward Community College (studying, socializing, etc.). Peer interactions, when purely social, can sometimes have negative effects on academic performance (Di Tommaso, 2012). Some developmental community college students even view peer interaction as a hindrance to their education because of negative experiences in high school (Di Tommaso, 2012). However, research also shows that when students interact with like-minded peers, this can provide positive support networks that can help students achieve their academic goals (Di Tommaso, 2012; Lundberg, 2014; Prado, 2012). During the Math Summer Bridge program, peer interaction is built in through peer tutoring, small group tutoring, and four different rooms (varying on noise level) where students can choose what area is best for them to study in. Peer interaction will be measured by two (2) multiple choice questions: “How many times do you interact with your peers at school for academic purposes (e.g. study groups, work on projects)?” and “How many times do you interact with your peers at school for social purposes (e.g. eat lunch, drink coffee, socialize)?” Both questions are multiple choice: (a) 0 times/week; (b) 1-3 times/week; (c) 4-6 times/week; or (d) 7+ times/week. Satisfaction with College Environment A student’s satisfaction with his/her college environment is strongly related to student success. This includes performance (Low, 2000), persistence (Starr, Betz, & Menne, 1972; Walker, 2008), and overall achievement (Astin, 1993; Oja, 2011; Tinto, 1993). Students who are satisfied with their institution are also more likely to engage in and be satisfied with pro- MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 47& academic activities and services (e.g. academic counseling, instructor effectiveness, campus climate, registration effectiveness) (Oja, 2011). Also, students who are more satisfied with their institution are more likely to become and stay engaged on campus, and therefore, more likely to succeed (Astin, 1993). This will be measured by three Likert scale questions: “I am very satisfied with the academic instruction I’ve received at Windward Community College so far” “I am very satisfied with the student support services (e.g. academic advising, financial aid office) I’ve received at Windward Community College so far” and “Overall, I am very satisfied with Windward Community College as a whole.” Persistence If the object of attending college is to earn a degree, continued enrollment in college and/or graduation from college is another way that this study is determining student success. Persistence (also referred to as retention) and graduation are used across studies and institutional measurements as a way of determining college success. For the purposes of this study, persistence will be measured by whether or not a student is still currently (Fall 2015) enrolled at a college or university. If a student transferred to another institution (whether they earned a degree at WCC or not), they will be considered as persisting. Data will be collected via the University of Hawaii (UH) System database. Graduation As previously mentioned, graduation, is one of the final measures of student success. For this study, graduation will be measured by whether or not a student has earned any 2- or 4-year degree before the Fall 2015 semester. Common practice of students enrolled at a primarily liberal arts community college is to transfer to a 4-year university after completing their degree. If a student earned their Associates of Arts (AA) or Associates of Science (AS), transferred to a MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 48& 4-year university, but has not completed their 4-year degree yet, they will still be counted as “graduated” because they earned one degree. Data will be collected via the University of Hawaii System database. Cumulative GPA Cumulative college GPA is a consistent way of measuring student success (Robbins, et al., & Carlstrom, 2004; Robbins, Allen, Casillas, Peterson, & Le, 2006). One study found that GPA was the strongest predicting variable for student persistence (in comparison to age, work hours, and financial assistance) (Nakajima, Dembo, & Mossler, 2012). Since success in each course is measured by a course grade earned, cumulative GPA measures how successful students have been throughout their entire college career by quantifying their grades earned for all coursework. For the purposes of this study, we will be measuring cumulative GPA on a 4.0 scale and data will be collected from the UH System database. Summary The literature shows a need to further investigate the importance of college self-efficacy (CSE) and address the gap in specific interventions that target CSE. By using Astin’s (1993) Inputs-Environment-Outcomes framework and two guiding theories (Astin’s Theory of Student Involvement, and Social Cognitive Theory), this study aims to measure the significance that the Windward Community College Math Summer Bridge program has on six student outcomes (college self-efficacy, sense of belonging, engagement and satisfaction, persistence, graduation, and cumulative GPA). MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 49& CHAPTER 3 METHODOLOGY The purpose of this two part study [(1) pre-post experiment group paired with a cross- sectional longitudinal design and (2) comparison group design] is to investigate the effects of the Windward Community College Math Summer Bridge program on students’ cognitive, affective, and behavioral traits linked with college success (college self-efficacy, sense of belonging, and engagement and satisfaction) as well as success measures (persistence, graduation, and cumulative GPA). Based on the Social Cognitive Theory (Bandura, 1977, 1986), Student Involvement Theory (Astin, 1975, 1977, 1999) and I-E-O framework (Astin, 1993), this study will explore the significance of various background and environmental factors on community college student college self-efficacy, sense of belonging, engagement and satisfaction, persistence, graduation, and cumulative GPA. There are six (6) questions identified for this study. The first three measure mindset and/or behaviors that lead to student success (college self-efficacy, sense of belonging, and engagement and satisfaction). Data was collected from the experiment group (n = 165) to address these three questions: •! RQ1. Is there a significant change in college self-efficacy for students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 1A. Is there a significant change in college self-efficacy for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 1B. Is there a significant change in college self-efficacy for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 50& •! RQ 2. Is there a significant change in sense of belonging for students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 2A. Is there a significant change in sense of belonging for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 2B. Is there a significant change in sense of belonging for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? •! RQ 3. Is there a significant change in engagement and satisfaction for students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 3A. Is there a significant change in student engagement and satisfaction for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 3B. Is there a significant change in student engagement and satisfaction for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? The following three questions measure student success through persistence, graduation, and cumulative grade point average (GPA). Data was collected from the experimental group (n = 294), Control Group A (n = 203), and Control Group B (n = 460) to address these questions: •! RQ 4. Do the Math Summer Bridge Program students persist at a greater rate than non-Math Summer Bridge Program students? MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 51& o! RQ 4A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students persist at a greater rate than their non-Math Summer Bridge Program counterparts? o! RQ 4B. Do the nontraditional or traditional Math Summer Bridge Program students persist at a greater rate than their non-Math Summer Bridge Program counterparts? •! RQ 5. Do the Math Summer Bridge Program students graduate at a greater rate than non- Math Summer Bridge Program students? o! RQ 5A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students graduate at a greater rate than their non-Math Summer Bridge Program counterparts? o! RQ 5B. Do the nontraditional or traditional Math Summer Bridge Program students graduate at a higher rate than their non-Math Summer Bridge Program counterparts? •! RQ 6. Do the Math Summer Bridge Program students have a higher cumulative grade point average (GPA) than non-Math Summer Bridge Program students? o! RQ 6A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students have higher cumulative GPAs than their non-Math Summer Bridge Program counterparts? o! RQ 6B. Do the nontraditional or traditional Math Summer Bridge Program students have higher cumulative GPAs than their non-Math Summer Bridge Program counterparts? This chapter contains an overview of the research design, setting, context and environment, and population and sample. The quantitative and qualitative data collection and MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 52& analysis procedures will also be presented. Each of the scales, along with their adaption for this study will be addressed along with their reliability and validity measures. Research Design This study included both a pre/post and cross-sectional quantitative design at a single community college institution. Since there have been four cohorts who have gone through the Math Summer Bridge program, and qualifications for the program maintain similar proportions of student characteristics (e.g. sex, ethnicity, age), each cohort will serve as different points in time for the typical MSB student (pre-MSB and post-MSB, one year later, two years later, and three years later). Inputs, environmental factors, and outcome measures will be collected through a self-report survey and including the modified College Self-Efficacy Inventory (Solberg, O’Brien, Villareal, Kennel, & Davis, 1993) and modified Sense of Belonging Inventory (Hagerty and Patusky, 1995) paired with institutional data (persistence, graduation, and cumulative GPA). Surveys will be administered to all students who have participated in the WCC Math Summer Bridge Program (experiment group). First, this study will analyze both pre/post and cross-sectional data to identify any differences in college self-efficacy, sense of belonging, and engagement and satisfaction. Then, experiment group student success measures (persistence, graduation, and cumulative GPA) will be compared to two different control groups to identify if they perform at higher levels than students who have not participated in the WCC Math Summer Bridge program. Control Group A consists of students who took the same developmental math course over the same summer, but did not participate in the WCC MSB Program. Control Group B is composed of Windward Community College students who did not take any summer math course, but instead were matched to individual MSB students on seven (7) characteristics: (1) First semester enrolled; (2) MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY & & & 53& Current enrollment status (full- or part-time); (3) Age; (4) Ethnicity (Native Hawaiian or non- Native Hawaiian); (5) Sex; (6) Pell Grant recipient (or not); and (7) Compass placement test score. Figure 2 provides a visual of this design. If there are significant and long-lasting positive effects on the three psychological outcome measures (CSE, SBI, and engagement and satisfaction), and the experiment group does better than the two control groups on the three student success measures (persistence, graduation, cumulative GPA), the researcher will be able to infer that the Windward Community College Math Summer Bridge Program has positive effects on developmental student success. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 54% Control Group A 2015 Cohort 2014 Cohort 2013 Cohort 2012 Cohort Other Summer Math Other Summer Math Other Summer Math Other Summer Math Experiment Group (MSB) 2015 Cohort 2014 Cohort 2013 Cohort 2012 Cohort Pre-Math Summer Bridge Program Post-Math Summer Bridge Program Post-MSB 1 year Post-MSB 2 years Post-MSB 3 years Control Group B 2015 Cohort 2014 Cohort 2013 Cohort 2012 Cohort No Math in Summer No Math in Summer No Math in Summer No Math in Summer Figure 2. This design consists of three parts: (1) pre/post experimental; (2) cross-sectional; and (3) comparison. The 2015 Experiment Group Cohort will first be measured using a pre/post design. Then, a cross-sectional design will be used to determine if there is a change over time for RQ1-RQ3 using the four experiment group cohorts’ post-survey responses. Finally, to answer RQ 4-RQ6, the Experiment group will be compared to each of the two control groups: (a) Experiment group with Control Group A, students will be matched by their course taken during that specific summer; (b) In comparing the Experiment group with Control Group B, students will be matched based on seven (7) characteristics: (1) First semester enrolled; (2) Current enrollment status (full- or part-time); (3) Age; (4) Ethnicity (Native Hawaiian or non-Native Hawaiian); (5) Sex; (6) Pell Grant recipient (or not); and (7) Compass placement test score. Figure 2. Research Design% MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 55% Setting, Context, Environment This single-institution study focused on Windward Community College (WCC) Math Summer Bridge (MSB) students. Located in Kaneohe on Oahu, Windward Community College is a small, 2-year institution that is a part of the 10-campus University of Hawaii System. It has a total of 2799 students with similar demographics as the average community college; 61% female/39% male; 66% part-time/34% full-time; and 80% minority (higher than the national average) (Windward Community College, 2014). Windward Community College offers 4 Associate Degrees (AA-Liberal Arts, AA-Hawaiian Studies, AS-Natural Science, and AS- Veterinary Technology) along with two certificates of achievement (CA), eight certificates of competence (CO), and five academic subject certificates (ASC). Eighty-one percent (81%) of all WCC students place into developmental/remedial mathematics (Windward Community College, 2015). Community College Ethnic Diversity As mentioned previously, Windward Community College primarily consists of minority students (80%). Unlike the majority of mainland community colleges, Windward Community College serves a low percentage of African Americans (0.86%) and Hispanics (2.36%). The largest ethnic group at the WCC is Native Hawaiian or part Native Hawaiian, constituting 41.78% of the total population (Windward Community College, 2014). In the University of Hawaii System, consisting of four 4-year institutions and 7 community colleges, Native Hawaiian students are primarily enrolled at the community college level (68.3%) (Office of Hawaiian Affairs, 2014), further solidifying need for this study’s focus on Native Hawaiians at a University of Hawaii community college. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 56% Developmental Mathematics As do most community colleges nationwide, Windward Community College offers developmental classes in English and math. Specifically for math, there are three levels below college-level math (Pre-Algebra, Elementary Algebra I, and Elementary Algebra II). Students take the Compass placement test in order to determine which course they need to take. This placement test is mandatory and students are not able to advance to a higher mathematics course until they complete the course that they place into (or retake the test and place into a higher math class). Windward Community College also has a Math Graduation Requirement for their AA degree in Liberal Arts. This requirement entails that students either place out of or complete their developmental math sequence. Currently there are three courses which fulfill the developmental math graduation requirement: Math 28, Math 25, and Math 29. In addition, students must also fulfill the college level symbolic reasoning requirement to earn an AA degree. As of 2012, WCC has two different methods for implementation of these developmental math courses: 1) traditional in-class lecture, and 2) a computer-based, self-paced track. Although the computer-based method is newer (introduced in Spring 2012), the success rates are significantly lower than its traditional lecture-based counterpart. Pass rates for the traditional lecture track are 72%, 56%, and 63% compared to the computer track’s 45%, 22%, and 53% pass rates for Pre-Algebra, Elementary Algebra I, and Elementary Algebra II respectively (see Table 2). Table 2 Windward Community College 2014 Developmental Math Pass Rates Pre-Algebra Elementary Algebra I Elementary Algebra II Traditional Lecture-Based Courses 72% 56% 63% Self-Paced, Computer-Based Courses 45% 22% 53% MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 57% One caveat is that the second level of the computer-based track (Math 28) includes a little more content than its traditional lecture-based counterpart (Math 24), and therefore fulfills WCC’s math graduation requirement. Further explanation of the different developmental math courses and relationship between each is shown in Figure 3. For the purposes of the WCC Math Summer Bridge program, since the inception of the computer-based track in 2012, the instructors offer the content of the computer-track for the two lower developmental math courses (Math 19 and 28) but implement it in the traditional, lecture-based style of teaching. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 58% Math%21%% Math%24% Math%19% Math%29*% Math%28*% Math%25*% ComputerFBased%Track% Traditional%LectureFBased%Track% Math%100**% Math%103**% Math%112**% Math%101**%% *%Fulfills%the%Developmental%Math%Graduation%Requirement%for%AA%degree% **%Fulfills%the%Symbolic%Reasoning%Requirement%for%AA%degree% % Math%101:%Vet%Tech%Math% Math%112:%Elementary%Education%Math% Math%103:%College%Algebra%(STEM/Business%Track)% Math%100:%Fulfills%Symbolic%requirement%(deadFend,%nonFSTEM/Business%Track)% Figure 3. Windward Community College Math Pathways Figure%3.%Pathways%of%all%developmental%math%courses%offered%at%Windward%Community% College,%what%type%of%instruction%model%each%follows,%and%how%the%developmental%math% tracks%feed%into%collegeFlevel%math%courses.% MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 59% Population and Sample The population of interest for this study is all Windward Community College students who place into developmental mathematics, and ideally, all community college students (nationwide) who place into developmental mathematics. This is because the hopes are to be able to identify a specific intervention (WCC Math Summer Bridge), as well as key components of these interventions and apply them to the larger community college student population in order to improve developmental mathematics remediation, college self-efficacy, sense of belonging, engagement and satisfaction, persistence, graduation, and cumulative GPA; and therefore student success. The sample used for this study includes all students who have completed the WCC Math Summer Bridge program as of August 2015 (N=294). This study consists of an experiment group (WCC Math Summer Bridge students) and two control groups, totaling 957 students in all. There were 294 WCC Math Summer Bridge participants (N=294) of which 83 completed the survey from the 2012 through 2014 cohorts, and 67 completed the pre- and post-survey for the 2015 cohort. Control Group A (students who took the same developmental math course at Windward Community College over the same summer) had 203 students and Control Group B (students who did not take any summer math, but instead were matched to individual experiment students based on 7 characteristics) consisted of 460 students. Students included in the experiment group were given a survey. This survey included demographic questions and environmental factor questions, as well as a modified College Self- Efficacy Inventory (Solberg, O’Brien, Villareal, Kennel, & Davis, 1993) and Sense of Belonging Inventory (Hagerty and Patusky, 1995). College performance and aptitude information was collected through the institution’s research office and then matched with their survey answers. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 60% Data collection and initial analysis was done through a double blind pairing completed by the researcher’s faculty supervisor. Data Collection As a quantitative study, data collected was analyzed with specific focus to “quantitative or numeric description of trends, attitudes, or opinions of a population by studying a sample of that population” (Creswell, 2014, p. 155). This study will use survey research to get both pre/post intervention, and cross-sectional views of students’ Inputs, Environment, and Outputs factors based on Astin’s I-E-O model. Then, the study will collect student success measures (persistence, graduation, cumulative GPA) on all experiment and control group students in order to see if there is a difference in performance for Math Summer Bridge participants. The survey used in this study includes input (demographic) information (e.g. sex, age, dependents) as well as environmental factors (e.g. employment status, full-time/part-time student status). Engagement and satisfaction (one outcome measure) is also included in the first part of the survey. The last part of the survey addresses the other two outcome measures related to metacognitive and affective states. Each question in this section of the survey was either in the form of multiple choice, numerical answers (e.g. age: 37), or Likert scale. The second part of the survey consisted of two modified inventories, the College Self- Efficacy Inventory, and Sense of Belonging Inventory. The original College Self-Efficacy Inventory has been modified into 19 questions addressing self-efficacy issues specific to community college students and developmental math students. Internal consistency was validated in the data analysis (pre-survey, Cronbach’s α = .959; post-survey, Cronbach’s α = .926). Similar procedures were used in creating the Sense of Belonging Inventory for this study. Questions were adapted from the larger Sense of Belonging Inventory (Hagerty and Patusky, MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 61% 1995) to customize a shorter, 18-question inventory with specific attention to college students. Internal consistency was validated for each of these scales as well (SOBI-A pre-survey, Cronbach’s α = .829; SOBI-A post-survey, Cronbach’s α = .591; SOBI-P pre-survey, Cronbach’s α = .683; SOBI-P post-survey, Cronbach’s α = .715). Surveys were administered using Survey Monkey, an online survey collector. For the 2015 cohort, students took the pre- and post-surveys in a computer lab on campus before they started the course in order to receive a textbook, and right before they took their final exam. It is important to note that the timing of when the post-survey was administered was not ideal, but practical in order to ensure students have internet access, and the majority of students complete the survey. The second part of the data collection process included collecting student academic achievement information (e.g. Compass placement test scores, persistence rates, graduation rates, cumulative GPA, Math Summer Bridge course grades) and were accessed through the institution’s research office. In order to protect the integrity of the study, an outside researcher collected the student surveys and academic achievement information, matched them up, and then assigned them a random number (instead of identifying name). After this was completed, the data was given to the researcher for further investigation. Procedures After IRB approval, data collection began in June 2015 before the start of the 2015 Math Summer Bridge program. Notice of the proposed study was given to the Chancellor, Vice Chancellor of Academic Affairs, Vice Chancellor of Student Affairs, and MSB instructors and coordinators at Windward Community College. A list of email addresses and academic achievement information were obtained through the institution’s research office for each student who participated in the WCC Math Summer Bridge program (Experiment Group) as well as MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 62% students who took the same level of developmental math (Control Group A) and those with similar input and environmental factors who did not participate in any summer developmental math course (Control Group B). Request for student participation was sent out via email. Survey collection was complete as of August 1, 2015. Quantitative Instrumentation The survey used in this study consisted of researcher-developed demographics and survey questions, as well as modified (and verified valid) instruments from published studies. The first modified instrument used is the College Self-Efficacy Inventory (Solberg, et al., 1993). Due to the lack of relevance for certain parts of the inventory (e.g. housing/living on campus, dorm roommates), the researcher removed these questions and replaced them with questions specific to the community college student (e.g. developmental math, time management in relation to balancing school and family/financial responsibilities). The second instrument modified for this survey is the Sense of Belonging Inventory (Hagerty and Patusky, 1995). Since sense of belonging is only one environmental factor of the study, this inventory has been shortened for student convenience. Independent Variables: Input Factors Data will be collected on eight (8) independent variables. The first six (6) consist of the Input or background information. This information will be collected through a combination of demographic survey questions created by the researcher and information collected directly from Windward Community College’s Institutional Research Office. The input variables include: (1) Ethnicity; (2) Sex; (3) Socioeconomic Status; (4) Compass Placement Test Scores; (5) Age; and (6) MSB Cohort Year, as seen in Table 3. The other two (employment and college enrollment status), are environmental factors and will be discussed in the next section. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 63% Table 3 Input-Independent Variable Questions No. Variable Items Source 1 Ethnicity 2 Survey Demographics, Student Records 2 Sex 1 Survey Demographics, Student Records 3 Socioeconomic Status 4 Survey 4 Compass Placement Test Scores 1 Student Records 5 Age 1 Student Records 6 Math Summer Bridge Cohort Year 1 Student Records Ethnicity. Ethnicity will be collected through two survey questions. The first will be a question asking the student to choose the ethnicity that they most identify with (only allowed one choice). To keep in consistency with the University of Hawaii System application, this study will use the same ethnicity options. The second question on the survey will ask students “Were any of your ancestors Native Hawaiian?” This question is also identical to the UH System application and will be used to identify any students who are Native Hawaiian or part Native Hawaiian who might not primarily identify with this ethnicity (and therefore do not select it in question 1). One of the missions of Windward Community College and the University of Hawaii System as a whole is to especially serve the Native Hawaiian community, and therefore important information to identify. Refer to question 2 and 3 on the Student Survey (Appendix A). Sex. Student’s biological sex will be identified through one dichotomous survey question (Male or Female). Refer to question 4 on the Student Survey (Appendix A). Socioeconomic status. Students who come from lower socioeconomic backgrounds are less likely to succeed in college. Part of this is due to financial constraints, but another key factor is social capital. This is especially true for first-generation college students. Students who have parents that did not earn a college degree often enter higher education institutions with MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 64% fewer resources and lacking a preparedness and realistic expectations of what college consists of and what it takes to become successful (Atherton, 2014; Cushman, 2007; McCarron & Inkelas, 2006). Socioeconomic status will be determined through four self-reported questions on the survey. Refer to questions 5-8 in the Student Survey (Appendix A). Compass placement test scores. Before students are able to take Math or English courses at Windward Community College, they have to complete the Compass placement test. The student’s math placement score will be recorded through information received from the WCC Institutional Research Office. If the student took multiple tests within the 2-year period where the test is valid, the researcher will record the highest score. This is compliant with the placement rules used by the academic advisors at the institution. Age. The student’s age will be recorded as a measure to determine traditional or nontraditional student designation at the time of WCC Math Summer Bridge participation. This information will be confirmed through student information received by the WCC Institutional Research Office. Math Summer Bridge cohort year. This independent variable will be collected through data obtained from the WCC Institutional Research Office. Independent Variables: Environment This study will investigate two (2) independent environmental variables that contribute to/affect the study’s student outcomes/dependent variables (college self-efficacy, sense of belonging, engagement and satisfaction, persistence, graduation, and cumulative GPA). Independent variables will be collected through a demographic survey created by the researcher. The two (2) variables include: 1) Employment, and 2) Full-time/Part-time College Status as seen in Table 4. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 65% Table 4 Environment-Independent Variable Questions No. Variable Items Source 7 Employment 3 Survey 8 Full-time/Part-time College Status 1 Survey Employment. This variable will be answered through the student’s self-report on the demographic survey created by the researcher. It will be answered through three questions. Refer to questions 39-41on the Student Survey (Appendix A). Full-time/part-time college status. The student’s college enrollment status will be measured through self-report on the demographic survey created by the researcher (see question 28 in the Student Survey of Appendix A). This information will be confirmed through data received from the Windward Community College Institutional Research Office. Intervention: Windward Community College Math Summer Bridge Program For this study, the intervention used is the Windward Community College Math Summer Bridge program. Students who participate in the program (including those who withdraw from the program or receive an N grade-no credit) are considered the experimental group (Group A). These students’ six outcome measures will be recorded (college self-efficacy, sense of belonging, engagement and satisfaction, persistence rates, graduation rates, and cumulative GPA). Then, the experiment groups’ success measures (persistence rates, graduation rates, and cumulative GPA) will be compared to each of the two control groups: students who took a non- MSB developmental math course over the same summer (Control Group A) and also to students who did not take any summer developmental math, but match individual experiment group students on 7 characteristics (Control Group B) (see Figure 2 on p. 54). MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 66% Dependent Variables: Outcome Measures This study will measure the effects of input and environmental factors (including the WCC Math Summer Bridge program intervention) on six different outcome measures. These measures include: (1) college self-efficacy; (2) sense of belonging; (3) engagement and satisfaction; (4) persistence; (5) graduation; and (6) cumulative GPA. These six dependent variables will be measured through: (a) modified published inventory; (b) researcher-created multiple-choice questions; or (c) academic achievement information obtained through the WCC Institutional Research Office. Table 5 reflects each outcome measure and the source of each type of information. Table 5 Outputs-Dependent Variable Questions No. Variable Items Source 1 College Self-Efficacy 1 College Self-Efficacy Inventory 2 Sense of Belonging 1 Sense of Belonging Inventory 3 Engagement and Satisfaction 7 Researcher-created questions 4 Persistence 1 Student Records 5 Graduation 1 Student Records 6 Cumulative GPA 1 Student Records College self-efficacy. College self-efficacy was measured through a summative scale questionnaire modified from Solberg, et al.’s College Self-Efficacy Inventory (1993). The original inventory consisted of 20 questions formatted as 10-point Likert scale questions. The published version of this inventory had a Cronbach’s α of .88 (Gore, Leuwerke, & Turley, 2005; Solberg, et al., 1993). The modified version used for this study kept the same 10-point Likert scale format with a total of 19 questions. An internal analysis has been conducted due to the alteration of original questions to best fit the current study and has been found to have strong MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 67% internal consistency (pre-survey, Cronbach’s α = .959; post-survey, Cronbach’s α = .926). See Appendix B for the College Self-Efficacy Inventory. Sense of Belonging. The second dependent variable, sense of belonging, was measured through a modified version of the Sense of Belonging Inventory (Hagerty and Patusky, 1995). This inventory has been condensed to 18 Likert scale questions. Similar to the original version, this modified inventory has two parts (SOBI-A and SOBI-P), addressing proposed antecedents and psychological beliefs of sense of belonging (See Appendix C). SOBI-Antecedents focus aims to measure the precursors of sense of belonging. In other words, the social conditions that an individual needs to be aware of before they can establish a sense of belonging. SOBI- Psychological focus aims to measure the psychological state of sense of belonging by having the student rank (strongly disagree to strongly agree) different aspects of sense of belonging. When this inventory was created, it was first tested on community college students with an internal consistency reliability of Cronbach’s α: .93 and .72 and test-retest reliability over an 8-week period Cronbach’s α: .84 and .66 for SOBI-P and SOBI-A respectively. The modified scales used in this study had similar internal consistency marks: Cronbach’s α: .83 and .68 for the pre- test and Cronbach’s α: .59 and .72 for the post-test. Appendix C lists all of the questions along with coding for SOBI-A and SOBI-P designations. Engagement and Satisfaction. The third dependent variable measured in this study is composed of two major components. Student engagement was measured by 1) how many times the student interacts with their instructors a) during class, and b) outside of class (e.g. office hours), and 2) how many times the student interacts with their peers, both in a) an academic and b) social capacity. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 68% Faculty interaction. Faculty/Student Interaction was measured by two multiple-choice questions. This was to measure the students’ perception of how often and why they interact with faculty members. The two questions were, “How many times do you talk to/interact with your instructor during class?” and “How many times do you talk to/interact with your instructor outside of class?” For both questions, students had three different answer options: (a) 0 times/week; (b) 1-3 times/week; or (c) 4+ times/week. Refer to questions 20 and 20 in the Student Survey (Appendix A) for these questions. Peer interaction. Peer interaction was measured by two multiple-choice questions on the questionnaire. This was to measure the students’ perception of how often and why they interact with peers. The two questions were, “How many times do you interact with your peers at school for social purposes (e.g. eat lunch, drink coffee, socialize)?” and “How many times do you interact with your peers at school for academic purposes (e.g. study groups, work on projects)?” For both questions, students had four different answer options: (a) 0 times/week; (b) 1-3 times/week; (c) 4-6 times/week; (d) 7+ times/week. Refer to questions 22 and 23 in the Student Survey (Appendix A) for these questions. Satisfaction with college environment. Student’s satisfaction with their college environment was measured through three Likert scale questions developed by the researcher. These questions measured satisfaction with instructors, student support services, and Windward Community College overall. The three questions are: “I am very satisfied with the academic instruction I’ve received at Windward Community College so far” “I am very satisfied with the student support services (e.g. academic advising, financial aid office) I’ve received at Windward Community College so far” and “Overall, I am very satisfied with Windward Community College as a whole”. Each of these questions are 5-point Likert scale questions ranging from MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 69% “strongly disagree” to “strongly agree”. Refer to questions 20-22 in the Student Survey (Appendix A) for these questions. Persistence. Persistence will be identified by whether or not the student is still enrolled in college for the Fall 2015 semester after completing the WCC Math Summer Bridge program. Persistence will include both WCC enrollment and students’ transfer/enrollment at another institution and will be recorded through information obtained through the WCC Institutional Research Office. Graduation. Graduation will include earning a 2- or 4-year degree from Windward Community College or any other institution. Certificates will not count. If a student has earned a 2-year degree, but still enrolled in school towards another degree, they will count as having graduated. Therefore, if a student has earned a degree (e.g. Associates of Arts in Liberal Arts) and then transferred to another institution (and is still enrolled), they will count for both persisting and graduating. This information will be obtained through the WCC Institutional Research. Cumulative GPA. Cumulative GPA will be measured for all courses taken at Windward Community College, including remedial/developmental courses. This information will be obtained through the WCC institutional Research Office. Validity and Reliability Measures In order to ensure the validity and reliability of the measures taken during this study, two specific steps have been taken. The first is that inventories used were taken and modified from other published studies. An internal analysis has been done on each of the modified inventories to confirm its reliability. The second step is that each part of the research and methods were reviewed and assessed by experts. These experts inspected all parts of the study before it is MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 70% implemented and provided critiques and feedback, including measurements used and wording of questions. Expert review is a popular method for validating survey instruments and widely accepted in the academic community (American Educational Research Association, 1999). Quantitative Analysis Data collected was analyzed through inferential and descriptive statistics. Inferential statistics were used to make educated deductions about how the Windward Community College Math Summer Bridge program could positively impact the rest of the developmental math students at the institution and provide impetus for future studies at other 2 year colleges with similar demographics and resources. Descriptive statistics were used to identify the significance of the Windward Community College Math Summer Bridge Program (intervention) on each dependent variable (outputs). First, a one-way repeated measures analysis of variance was used to determine if there was an increase in any of the first three output measures (college self- efficacy, sense of belonging, and engagement and satisfaction). Then a one-way between- measures analysis of variance (ANOVA) was used to determine if the change (or lack there of) acquired after the intervention was sustained over the next three years (Cohorts 2012-2014). Finally, one-way between-measures ANOVA tests were used to determine if there is a significant difference between the MSB students (experiment group) and Control Group A, and the MSB students (experiment group) and Control Group B on each of the three student success measures (persistence, graduation, and cumulative GPA). Summary This chapter detailed the methodology of this quantitative study involving participants and non-participants of the Windward Community College Math Summer Bridge program. Quantitative survey procedures and analysis procedures were described and validity and MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 71% reliability measures addressed. The statistical information and further analysis will be provided in Chapter 4. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 72% CHAPTER 4 RESULTS AND ANALYSIS The purpose of this study was to determine whether or not the Windward Community College Math Summer Bridge program had an effect on students’ college self efficacy, sense of belonging, and/or engagement and satisfaction, as well as compare the experiment group (WCC Math Summer Bridge participants) to two different control groups (Control Group A: non-MSB summer math students; Control Group B: non-summer math students matched on 7 characteristics) on three (3) different student success measures: (1) Cumulative GPA; (2) Persistence; and (3) Graduation. This study sought out to do so by answering six (6) research questions. The first three measure mindset and/or behaviors that lead to student success (college self-efficacy, sense of belonging, and engagement and satisfaction). Data was collected from the experiment group (n = 165) to address these three questions: •! RQ1. Is there a significant change in college self-efficacy for students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 1A. Is there a significant change in college self-efficacy for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 1B. Is there a significant change in college self-efficacy for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? •! RQ 2. Is there a significant change in sense of belonging for students who participated in the Math Summer Bridge Program? Is this change maintained over time? MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 73% o! RQ 2A. Is there a significant change in sense of belonging for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 2B. Is there a significant change in sense of belonging for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? •! RQ 3. Is there a significant change in engagement and satisfaction for students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 3A. Is there a significant change in student engagement and satisfaction for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 3B. Is there a significant change in student engagement and satisfaction for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? The following three questions measure student success through persistence, graduation, and cumulative grade point average (GPA). Data was collected from the experimental group (n = 294), Control Group A (n = 203), and Control Group B (n = 460) to address these questions: •! RQ 4. Do the Math Summer Bridge Program students persist at a greater rate than non-Math Summer Bridge Program students? o! RQ 4A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students persist at a greater rate than their non-Math Summer Bridge Program counterparts? MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 74% o! RQ 4B. Do the nontraditional or traditional Math Summer Bridge Program students persist at a greater rate than their non-Math Summer Bridge Program counterparts? •! RQ 5. Do the Math Summer Bridge Program students graduate at a greater rate than non- Math Summer Bridge Program students? o! RQ 5A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students graduate at a greater rate than their non-Math Summer Bridge Program counterparts? o! RQ 5B. Do the nontraditional or traditional Math Summer Bridge Program students graduate at a higher rate than their non-Math Summer Bridge Program counterparts? •! RQ 6. Do the Math Summer Bridge Program students have a higher cumulative grade point average (GPA) than non-Math Summer Bridge Program students? o! RQ 6A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students have higher cumulative GPAs than their non-Math Summer Bridge Program counterparts? o! RQ 6B. Do the nontraditional or traditional Math Summer Bridge Program students have higher cumulative GPAs than their non-Math Summer Bridge Program counterparts? Chapter 4 will present the data collected and analyzed to address each of the six research questions. This chapter consists of two sections. Part 1 describes the results of the study; the sample used, variables, and data collected. Descriptive statistics were used to identify sample and larger population proportions as well as ethnicity and age group itemizations. Factor analysis & reliability tests on multi-item variables are also presented in this section. The second section of MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 75% this chapter is the analysis in relation to the six research questions. First, to identify if there was an initial increase in college self efficacy and/or sense of belonging and/or engagement and satisfaction in the pre/post tests for the 2015 MSB cohort (Research Question 1, 2, and 3), a one- way repeated measures ANOVA was used, followed by a one-way between-groups ANOVA to measure if the change was long-lasting (post 2015 versus each of the previous cohorts). Then, to compare student success measures (cumulative GPA, persistence, and graduation) between the experiment group and each of the two control groups, one-way between-groups ANOVA tests were used. For each of these tests, results were also broken down by ethnicity (Native Hawaiian v. non-Native Hawaiian) and age (traditional v. nontraditional). Finally, this chapter concludes with a summary of the key results. Part One: Descriptive Statistics Participant Characteristics A total of 957 students were included in this study. The experiment group consisted of 294 students. Eighty-two (82) of which were in the 2015 cohort and 212 were in the 2012 through 2014 cohorts. Of the 82 students in the 2015 cohort, sixty-seven (67) completed the pre- and post-survey. Eighty-three (83) of the 2012-2014 cohorts completed the post-survey. Of the 663 students in the control groups, 203 took a non-MSB summer math course (Control Group A), and 460 were in Control Group B, selected because they matched MSB students on seven (7) characteristics: (1) First semester enrolled; (2) Current enrollment status (full or part-time); (3) Age; (4) Ethnicity (Native Hawaiian or non-Native Hawaiian); (5) Sex; (6) Pell Grant recipient (or not); and (7) Compass placement test score. It is noted that since the sample size is larger than 100, the power of this analysis is not an issue (Stevens, 1996, p. 6). MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 76% Sex. Table 6 displays the experiment (n=294) and control groups’ (n=203; n=460) populations by sex as well as the Windward Community College student population (N=2661). A chi square goodness of fit test was performed showing that there were no significant differences in the proportion of males and females identified in the Experiment (χ 2 (1, n=294) = .85, p = .358) and Control Group A (χ 2 (1, n=203) = 1.519, p = .218) groups. However, for Control Group B, there was a significant difference in proportion of females and males, χ 2 (1, n=460) = 71.32, p = .000. Since Control Group B is only used to match experiment group participants and the Experiment group is proportional to the total Windward Community College student population, the researcher concluded it was valid to use this group in the study. Table 6 Participant Sex (n=957; N=2661) Sex Experiment Control A Control B WCC Freq. % Freq. % Freq. % Freq. % Female 185 62.9 131 64.5 366 79.6 1576 59.2 Male 109 37.1 72 35.5 94 20.4 1036 39.0 Total 294 100.0 203 100.0 460 100.0 2661 100.0 Note. 49 students did not report their sex in WCC’s report. Similar to the national average of community college students, there are more women than men enrolled. According to the American Association of Community Colleges (2016), 57% of the total community college population are women, versus 43% men. Windward Community College’s population is a little more skewed than the national average (59.23% female, 38.93% male) (Windward Community College, 2014) and the total sample population used for this study generally follows this trend. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 77% Ethnicity. Table 7 displays the experiment and control groups, as well as the total Windward Community College population categorized by if they are Native Hawaiian or not. Like many other minorities, Native Hawaiians are underrepresented in higher education (Office of Hawaiian Affairs, 2014). Nationally, Native Hawaiians are grouped in a larger category with Asians and other Pacific Islanders. However with over 30 different ethnic groups included (Census, 2016), this type of grouping has been shown to skew the data (Benham, 2006). Even included in this larger population, Native Hawaiians, Asians, and other Pacific Islanders only make up 6% of the total community college population nationwide (AACC, 2016). The ethnicity demographics for all three groups in the sample population greatly differ from the nationwide statistic, and when conducting a chi square goodness of fit test, both the Experiment (χ 2 (1, n=294) = 52.792, p = .000) and Control B (χ 2 (1, n=460) = 45.944, p = .000) groups statistically differed from the Windward Community College total student population. Control Group A did not significantly differ from the total Windward Community College student population, χ 2 (1, n=203) = 3.049, p = .081. The students involved in the Math Summer Bridge Program were also involved with the WCC TRiO Student Support Services program, Paipai o Koʻolau Project, and/or Hūlili Transfer Program, which has higher proportions of Native Hawaiian students. Therefore, the researcher expected there to be a larger proportion of Native Hawaiian students in the experiment group (and larger proportion of Native Hawaiian students in Control Group B since they were collected based on the experiment group’s characteristics). MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 78% Table 7 Participant Ethnicity (n=957; N=2661) Ethnicity Experiment Control A Control B WCC Freq. % Freq. % Freq. % Freq. % Native Hawaiian 185 62.9 73 36.0 265 57.6 1118 42 Non-NH 109 37.1 130 64.0 195 42.4 1543 58 Total 294 100.0 203 100.0 460 100.0 2661 100 Age. The researcher was unable to attain the specific breakdown of Windward Community College student ages to check the sample’s similarity of distribution. However, the average age at WCC is 26 years old, meaning that there are more nontraditional students enrolled. This is similar to/than the national average (63% are 22 years or older, 37% are 21 or younger) (NCES, 2015). Table 8 shows the distribution of students based on age for experimental and control groups as well as the total Windward Community College population. Table 8 Participant Age (N=957) Age Experiment Control A Control B Freq. % Freq. % Freq. % Traditional 166 56.5 112 55.2 363 79.0 23 and below 148 50.3 69 34.0 356 77.4 24-25 18 6.2 43 21.2 7 1.6 Nontraditional 128 43.5 91 44.8 97 21.0 26-35 53 18.0 55 27.1 61 13.3 36-45 29 10.0 20 9.9 23 5.0 46+ 46 15.5 16 7.8 13 2.7 Total 294 100.0 203 100.0 460 100.0 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 79% Full-time/Part-time. Nationally, community college students are much more likely to be part-time (less than 12 credits/semester) than students enrolled at a 4-year institution. According to the National Center for Education Statistics (2014), 88% of public 4-year institution students are full-time, compared to only 38% at community colleges. At Windward Community College, the characteristics are similar to the national community college average with 35.2% being full-time and 64.8% of the total population being part-time (Windward Community College, 2014). All three sample populations for this study are different than both the Windward Community College and national averages (Experiment, χ 2 (1, n=294) = 45.952, p = .000; Control 1, χ 2 (1, n=203) = 14.092, p = .000; Control 2, χ 2 (1, n=460) = 15.188, p = .000). Table 9 breaks down the experiment and control groups as well as the total WCC population by full-time/part-time status. Table 9 Participant Enrollment Status (n=957; N=2661) Credits Experiment Control A Control B WCC Freq. % Freq. % Freq. % Freq. % Full-Time (12+ credits) 159 54.1 97 47.8 122 26.5 936 35.2 Part-Time (1- 11 credits) 135 45.9 106 52.2 338 73.5 1725 64.8 Total 294 100.0 203 100.0 460 100.0 2661 100.0 Since the WCC Math Summer Bridge Program was held for four (4) hours per day, five (5) days per week, for seven (7) weeks, it was expected that not many part-time students would be able to attend, given they probably have other commitments (e.g. work), which keep them from attending school full time. Also, since the non-MSB summer math courses were held MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 80% during the day, the researcher also hypothesized similar reasons why part-time students who work during those hours would not be able to take these courses either. Compass Placement Test Score. The researcher was unable to attain compass placement test score information for the total Windward Community College student population (and therefore, was unable to run statistics on distribution equivalence for sample to total populations. However, it is known that 80% of all Windward Community College students place into developmental or remedial math courses. Having the majority of community college place into developmental/remedial coursework is a statistic seen nationwide, with over 66% of community college students estimated to place into remedial math (Bailey, Jeong, & Cho, 2010). Table 10 shows the experiment and control groups’ math placement in comparison to the total Windward Community College student population. Table 10 Participant Compass Placement Score (N=957) Compass Experiment Control A Control B Freq. % Freq. % Freq. % 1 level below 41 14.1 44 21.7 42 9.1 2 levels below 67 23.0 73 36.0 79 17.2 3 levels below 183 62.9 86 42.3 339 73.7 Total 291 100.0 203 100.0 460 100.0 Note. Three (3) WCC MSB students did not take the placement test. Input Characteristics by Native Hawaiian Status As previously discussed, Native Hawaiians are an ethnic minority underrepresented in higher education. Not only are Native Hawaiians not enrolling in higher educational institutions at a proportional rate to the total population, those individuals who do enroll, do poorer in MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 81% student success measures than their non-Native Hawaiian counterparts (Office of Hawaiian Affairs, 2014). This is why it was important for this study to identify the experiment, control, and total population through this lens. Table 11 shows the breakdown of each group by input variables and ethnicity for the entire sample used (N=957) and Table 12 shows the breakdown for each group by input variables and ethnicity for the experiment group (N=294). Table 11 Input Characteristics by Native Hawaiian Status (N=957) Variable Non-Native Hawaiian Native Hawaiian n % n % Sex Female 285 65.7 397 75.9 Male 149 34.3 126 24.1 Age Traditional 295 68.0 346 66.2 23 and below 255 58.8 318 60.8 24-25 40 9.2 28 5.4 Nontraditional 139 32.0 177 33.8 26-35 61 14.1 108 20.7 36-45 34 7.8 38 7.3 46+ 44 10.1 31 5.9 Enrollment Status Full-Time (12+ credits) 145 33.4 233 44.6 Part-Time (1-11 credits) 289 66.6 290 55.4 Compass Score 1 level below 82 18.9 45 8.8 2 levels below 125 28.8 94 18.0 3 levels below 226 52.1 382 73.0 N/A 1 0.2 2 0.2 Total 434 100.0 523 100.0 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 82% Table 12 Input Characteristics by Native Hawaiian Status for Experiment Group (N=294) Variable Non-Native Hawaiian Native Hawaiian n % n % Sex Female 64 58.7 121 65.4 Male 45 41.3 64 34.6 Age Traditional 64 58.7 102 55.1 23 and below 56 51.4 92 49.7 24-25 8 7.3 10 5.4 Nontraditional 45 41.3 83 44.9 26-35 14 12.8 39 21.1 36-45 7 6.4 22 11.9 46+ 24 22.0 22 11.9 Enrollment Status Full-Time (12+ credits) 48 44 111 60.0 Part-Time (1-11 credits) 61 56 74 40.0 Compass Score 1 level below 18 16.5 23 12.4 2 levels below 30 27.5 37 20.0 3 levels below 60 55.0 123 66.5 N/A 1 1.0 2 1.1 Total 109 37.1 185 62.9 Input Characteristics by Age Categories Age (traditional versus nontraditional) is another important lens to analyze data pertaining to community college students. As discussed in Chapter 2, there are different challenges that nontraditional students face when attending college and they can exponentially grow depending on other variables (e.g. Compass placement test, full- or part-time status) MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 83% (Jameson & Fusco, 2014; Lundberg, et al., 2008; Schlossberg, Lynch, & Chickering, 1989). Table 13 highlights input variables based on traditional (≤ 25 years old) or nontraditional ages (26+ years old) for the entire study sample and Table 14 highlights the same qualities of the experiment group. There were a few nuances found when reviewing the distribution of students included in the study. For the total sample population (both experiment and control groups), it was found that the nontraditional population’s credit enrollment (37.3% full-time/62.7% part-time) has a more similar distribution to the national average (38%, NCES, 2014) than the total study’s distribution. It is thought that this is because most of the students involved in the study did not possess the same general characteristics as the average community college student (full-time job and part-time credit load) because they were able to attend the program 5 days/week, 4 hours/day, for 7 week, which would be hard to do if they were only able to attend school part- time, but of those who did work full-time (and wouldn’t be able to attend school full-time), it would be the nontraditional students. Also for the total sample population, both traditional (72.2% female v. 27.8% male) and nontraditional (69.3% female v. 30.7% male) groups were skewed more than the school average with 69.3-72.2% of the population being female. Specifically for the experiment group, both traditional and nontraditional students have higher proportions of Native Hawaiian students (61.4% and 64.8% respectively). Also, nontraditional students in the experiment group have the largest proportion of students placing into the lowest level of remedial math (69.5%). This is important to note because the sample population, especially the experiment group, there are higher proportions of “at risk” groups (e.g. minority, nontraditional, remedial math). Therefore, having positive effects in college self- MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 84% efficacy, sense of belonging, engagement and satisfaction, and student success measures, would give strong indications for applicability in larger community college populations. Table 13 Input Characteristics by Age Category (N = 957) Variable Traditional Nontraditional n % n % Sex Female 463 72.2 219 69.3 Male 178 27.8 97 30.7 Ethnicity Non-Native Hawaiian 295 46.0 139 44.0 Native Hawaiian 346 54.0 177 56.0 Enrollment Status Full-Time (12+ credits) 260 40.6 118 37.3 Part-Time (1-11 credits) 381 59.4 198 62.7 Compass Score 1 level below 96 15.0 31 9.8 2 levels below 139 21.6 80 25.3 3 levels below 403 62.9 205 64.9 Missing (Unknown) 3 0.5 0 0.0 Total 641 56.5 316 43.5 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 85% Table 14 Input Characteristics by Age Categories for Experiment Group (N = 294) Variable Traditional Nontraditional n % n % Sex Female 107 64.5 78 60.9 Male 59 35.5 50 39.1 Ethnicity Non-Native Hawaiian 64 38.6 45 35.2 Native Hawaiian 102 61.4 83 64.8 Enrollment Status Full-Time (12+ credits) 89 53.6 70 54.7 Part-Time (1-11 credits) 77 46.4 58 45.3 Compass Score 1 level below 29 17.5 12 9.4 2 levels below 40 24.1 27 21.1 3 levels below 94 56.6 89 69.5 Missing (Unknown) 3 1.8 0 0.0 Total 166 56.5 128 43.5 College Self-Efficacy Scale Factor Analysis and Reliability Analysis Every survey returned (N=150) was 100% complete. Therefore, the three multi-question scales (including the College Self-Efficacy Scale), had no missing data. The College Self- Efficacy Scale consisted of 19 questions on a 10-point scale, 1 being “not confident at all” and 10 being “extremely confident”. The researcher considered scores 8-10 to be valid scores when calculating the percent of students ranking themselves as “extremely confident”. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 86% College Self-Efficacy Scale factor analysis and reliability analysis-Ethnicity breakdown. Table 15 and Table 16 show the percentage of students who reported themselves to be “extremely confident” on each of the College Self-Efficacy Scale items with ethnicity breakdowns for the 2015 cohort (pre/post) and post MSB intervention (2015, 2014, 2013, and 2012 cohorts) respectively. For many of the items, percentages of students who were “extremely confident” (score of 8-10) went down, however not by a statistically significant margin. This is thought to be because the post-test was administered to the students right before they took their final exam. The test was given at that period of time to ensure all students complete the survey before they left for the summer break. Because students had higher levels of anxiety before their final exam and their final grade was unknown at that point, it is hypothesized that the students scored themselves lower than they would have after finishing their final exam and completing the Math Summer Bridge Program. There was only one item that had a statistically significant increase for non-Native Hawaiians: CSE 1 “Make new friends in college”, Time 1 (pre-survey) (M = 6.96, SD = 2.57), Time 2 (post-survey) (M = 7.96, SD = 2.27), t (25) = -3.469, p = .002 (two-tailed). There were two items that had significant changes for Native Hawaiian students: 1) CSE 3 “Talk to college staff/faculty”, Time 1 (M = 8.21, SD = 2.34), Time 2 (M = 8.83, SD = 1.87), t (41) = -2.172, p = .036 (two-tailed); 2) CSE 7 “Participate in class discussions”, Time 1 (M = 8.14, SD = 2.34), Time 2 (M = 8.95, SD = 1.56), t (41) = -2.397, p = .021 (two-tailed). Native Hawaiian students also scored higher on 10 out of 19 scale items compared to their non-Native Hawaiian counterparts. As a whole, item 10, “Join a student organization” is hypothesized to be low MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY % % % 87% because Windward Community College is typical of a commuter CC campus and there aren’t many opportunities to join student organizations on WCC’s campus. For the second set of data (Table 16), post surveys were given to students who were a part of the 2015 cohort (immediately post intervention), 2014 cohort (1 year post intervention), 2013 cohort (2 years post intervention), and 2012 cohort (3 years post intervention). When observing the cross-sectional patterns for cohorts 2012-2015, there were two distinct trends. For non-Native Hawaiian students, the scores for each of the College Self-Efficacy items to drop in score for the 2013 and 2014 cohorts, where they would be in the middle of their schooling and probably incurring more difficulty courses, leading to higher stress levels and lower self- efficacy, and then increase again for 2012 cohort, where students are probably at the end of (or finished with) schooling and therefore feeling better and more confident about their capabilities (efficacy) in college. There were seven items for non-Native Hawaiian students that had significant changes over time (all of them increasing over time): 1) CSE 3 “Talk to college staff/faculty” (F (3, 46) = .904, no variance); 2) CSE 4 “Talk to college administration” (F (3, 46) = .995, p = .011); 3) CSE 7 “Participate in class discussions” (F (3, 46) = 2.831, p = .049); 4) CSE 12 “Ask a professor a question about math coursework” (F (3, 46) = 1.231, no variance); 5) CSE 14 “Complete the Math requirement (pass Math 25 or Math 28)” (F (3, 46) = .772, no variance); 6) CSE 17 “Keep up to date with my schoolwork” (F (3, 46) = 1.810, no variance); 7) CSE 18 “ Write course papers” (F (3, 46) = 1.575, p = .004). Native Hawaiian student trends were different. Looking at the different cohorts, over time, it looks like their scores for each College Self-Efficacy item go down (although not at statistically significant levels). MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 88% Table 15 Descriptive Statistics for College Self-Efficacy Items Pre/Post 2015 MSB with Ethnicity Breakdown (n = 67) All % extremely confident (8-10) Non-NH M (SD) NH M (SD) College Self-Efficacy Item Pre Post Pre Post Pre Post Make new friends at college 60.3 64.2 6.96 (2.569)** 7.96 (2.271) ** 8.00 (2.368) 8.55 (2.319) Successfully complete my current math course 79.4 72.3 8.77 (1.366) 8.73 (1.710) 8.43 (2.199) 9.02 (1.774) Talk to college staff/faculty 73.5 73 8.27 (2.070) 8.54 (1.923) 8.21 (2.343)* 8.83 (1.873)* Talk to college administration 72.1 66.2 7.96 (2.441) 8.12 (2.179) 8.07 (2.473) 8.17 (2.478) Manage time effectively 69.1 58.8 7.92 (2.432) 8.31 (1.850) 7.64 (2.397) 8.02 (2.225) Ask a question in class 70.6 70.3 8.27 (2.183) 8.69 (1.761) 8.19 (2.587) 8.64 (1.872) Participate in class discussions 66.2 71.6 8.23 (2.197) 8.54 (1.726) 8.14 (2.343)* 8.95 (1.561)* Research a term paper 57.4 59.5 7.65 (2.058) 7.88 (2.065) 7.45 (2.804) 7.71 (2.234) Do well on my exams 58.8 63.5 7.92 (1.937) 8.31 (1.871) 7.50 (2.350) 8.10 (2.128) Join a student organization 32.4 40.5 6.23 (2.875) 6.58 (2.996) 6.38 (2.631) 7.02 (3.127) Talk to my professors outside of class 61.8 65.5 8.15(2.203) 8.23 (1.966) 7.43 (2.881) 8.07 (2.299) *p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 89% Table 15 Descriptive Statistics for College Self-Efficacy Items Pre/Post 2015 MSB with Ethnicity Breakdown Cont. (n = 67) All % extremely confident (8-10) Non-NH M (SD) NH M (SD) College Self-Efficacy Item Pre Post Pre Post Pre Post Ask a professor a question about math coursework 76.5 74.3 8.35 (2.019) 8.65 (1.696) 8.55 (2.098) 8.64 (2.082) Take good class notes 72.1 70.9 8.38 (1.813) 8.65 (1.648) 8.17 (2.326) 8.57 (1.990) Complete the Math Requirement (pass Math 25 or Math 28) 83.8 79.1 8.50 (1.703) 8.54 (1.902) 8.69 (1.919( 8.64 (2.082) Seek help from my tutors and instructors to better understand math concepts 82.4 80.4 8.58 (1.677) 8.85 (1.642) 8.60 (1.795) 9.12 (1.656) Understand my textbook 70.6 65.5 8.08 (2.134) 7.96 (2.254) 8.02 (2.101) 8.40 (2.037) Keep up to date with my schoolwork 86.8 75.7 8.65 (1.765) 8.96 (1.536) 8.71 (1.812) 8.76 (1.819) Write course papers 67.6 64.2 8.12 (2.251) 8.27 (2.127) 7.52 (2.839) 8.14 (2.019) Complete my math homework 91.2 79.7 9.04 (1.907) 9.00 (1.549) 9.05 (1.752) 9.07 (1.659) * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 90% Table 16 Descriptive Statistics for College Self-Efficacy Items Post MSB 2015, 2014, 2013, 2012 with Ethnicity Breakdown (n = 150) All Non-NH NH % extremely confident (8-10) M (SD) M (SD) College Self-Efficacy Item 15 14 13 12 15 14 13 12 15 14 13 12 Make new friends at college 70.6 53.6 68.4 50 8.15 (2.034) 8.00 (2.330) 7.17 (2.406) 8.75 (2.500) 8.55 (2.319) 7.55 (2.038) 8.42 (2.301) 6.67 (2.807) Successfully complete my current math course 80.9 57.1 65.8 68.8 8.54 (1.838) 7.50 (3.117) 7.83 (2.368) 9.25 (1.500) 8.90 (1.872) 7.90 (2.174) 8.15 (2.111) 8.00 (2.486) Talk to college staff/faculty 77.9 64.3 68.4 68.8 8.38 (2.118) 8.38 (1.768) 8.00 (2.558) 10 (0.000) 8.71 (1.954) 8.20 (2.397) 8.00 (2.683) 7.73 (2.864) Talk to college administration 72.1 57.1 60.5 68.8 8.23** (2.084) 8.38** (1.768) 7.58** (2.875) 9.75** (.500) 8.14 (2.504) 7.65 (2.661) 7.69 (2.768) 7.33 (2.934) Manage time effectively 64.7 57.1 50 50 8.23 (1.986) 8.00 (1.069) 7.25 (1.485) 7.75 (2.630) 7.90 (2.250) 7.25 (2.403) 7.69 (2.131) 6.75 (2.598) Ask a question in class 77.9 67.9 63.2 62.5 8.69 (1.761) 8.38 (1.408) 7.67 (2.640) 9.25 (1.500) 8.64 (1.872) 7.80 (2.285) 8.23 (2.338) 7.67 (2.309) Participate in class discussions 82.4 64.3 60.5 62.5 8.65* (1.648) 8.38* (1.408) 7.42* (1.621) 9.75* (.500) 8.88 (1.580) 7.80 (2.215) 8.23 (2.215) 7.67 (2.270) Research a term paper 60.3 57.1 52.6 68.8 7.85 (2.130) 6.75 (2.550) 7.08 (2.429) 9.75 (.500) 7.64 (2.272) 7.55 (1.932) 7.69 (2.379) 6.92 (2.503) Do well on my exams 72.1 50 57.9 56.3 8.23 (2.006) 7.63 (1.768) 7.00 (2.449) 8.50 (1.732) 7.98 (2.322) 6.40 (2.501) 7.62 (2.246) 6.67 (2.535) * p < .05. ** p < .01 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 91% Table 16 Descriptive Statistics for College Self-Efficacy Items Post MSB 2015, 2014, 2013, 2012 with Ethnicity Breakdown Cont. (n = 150) All Non-NH NH % extremely confident (8-10) M (SD) M (SD) College Self-Efficacy Item 15 14 13 12 15 14 13 12 15 14 13 12 Join a student organization 47.1 25 36.8 50 6.85 (2.781) 7.00 (2.000) 4.83 (3.099) 7.00 (2.582) 6.98 (3.143) 5.45 (2.819) 6.62 (2.729) 7.00 (3.191) Talk to my professors outside of class 67.6 71.4 63.2 50 8.23 (1.966) 8.25 (1.488) 7.50 (2.276) 8.50 (2.380) 7.95 (2.326) 8.00 (2.248) 7.85 (2.477) 7.17 (2.791) Ask a professor a question about math coursework 76.5 78.6 71.1 62.5 8.58 (1.677) 8.88 (1.356) 8.25 (1.865) 10.00 (0.000) 8.52 (2.144) 8.10 (2.693) 8.46 (2.102) 7.67 (2.605) Take good class notes 76.5 67.9 68.4 50 8.54 (1.860) 7.88 (1.553) 7.75 (2.137) 7.75 (2.872) 8.50 (1.991) 8.05 (1.959) 8.27 (2.201) 7.25 (2.896) Complete the Math Requirement (pass Math 25 or Math 28) 77.9 71.4 84.2 75 8.50 (1.985) 8.38 (2.560) 8.25 (2.137) 10.00 (0.000) 8.52 (2.144) 8.05 (2.856) 9.35 (1.294) 8.17 (2.725) Seek help from my tutors and instructors to better understand math concepts 85.3 71.4 81.6 68.8 8.77 (1.632) 8.88 (1.553) 8.58 (1.730) 9.75 (.500) 9.00 (1.767) 8.30 (2.203) 8.81 (2.117) 7.75 (2.667) Understand my textbook 70.6 64.3 57.9 56.3 7.85 (2.378) 8.00 (2.330) 7.17 (2.552) 8.25 (2.363) 8.26 (2.131) 7.50 (2.544) 8.04 (2.163) 7.08 (2.275) Keep up to date with my schoolwork 83.8 75 65.8 62.5 8.92 (1.547) 8.63 (1.188) 8.00 (2.089) 10.00 (.000) 8.64 (1.898) 8.00 (2.176) 8.27 (2.051) 7.33 (2.871) Write course papers 69.1 67.9 52.6 56.3 8.19** (2.245) 6.88** (3.271) 7.33** (2.570) 9.75** (.500) 7.98 (2.147) 7.80 (2.462) 7.58 (2.686) 6.92 (2.843) Complete my math homework 85.3 75 71.1 75 8.85 (1.826) 8.88 (1.356) 8.00 (1.954) 9.75 (.500) 8.95 (1.766) 8.05 (2.523) 8.69 (1.871) 8.00 (2.923) *p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 92% College Self-Efficacy Scale factor analysis and reliability analysis-Age group breakdown. As previously mentioned, this was a 10-point scale, 1 being “not confident at all” and 10 being “extremely confident”. The researcher considered scores 8-10 to be “extremely confident”. Table 17 and Table 18 show the percentage of students who reported themselves to be “extremely confident” on each of the College Self-Efficacy Scale items with age group breakdowns for the 2015 cohort (pre/post) and post MSB intervention (2015, 2014, 2013, and 2012 cohorts). When analyzing the data while grouping answers by age (traditional v. nontraditional), different trends emerged. Overall, there were many more statistically significant differences for College Self-Efficacy (CSE) pre- and post-test items compared to when the data was segregated by ethnicity. As a whole, nontraditional students had higher means than traditional students for the CSE items. However, there were more statistically significant increases in post MSB scores for traditional students (six compared to two). Statistically significant increases for traditional students include: CSE 1 “Make new friends at college” Time 1 (pre-survey) (M = 6.85, SD = 2.73), Time 2 (post-survey) (M = 7.56, SD = 2.46), t (33) = -2.167, p = .038 (two-tailed); CSE 3 “Talk to college staff/faculty” Time 1 (M = 7.06, SD = 2.51), Time 2 (M = 8.06, SD = 2.46), t (33) = .008 (two-tailed); CSE 6 “Ask a question in class” Time 1 (M = 7.41, SD = 2.65), Time 2 (M = 8.35, SD = 2.03), t (33) = -2.788, p = .009 (two-tailed); CSE 7 “Participate in class discussions” Time 1 (M = 7.35, SD = 2.41), Time 2 (M = 8.29, SD = 1.98), t (33) = -2.448, p = .020 (two-tailed); CSE 9 “Do well on my exams” Time 1 (M = 7.38, SD = 2.32), Time 2 (M = 8.06, SD = 2.22), t (33) = -2.228, p = .033 (two-tailed); CSE 11 “Talk to my professors outside of class” Time 1 (M = 6.79, SD = 2.83), Time 2 (M = 7.79, SD = 2.24), t (33) = -2.753, p = .010 (two-tailed). Nontraditional scores are as follows 1) CSE 1 “Make new friends in college” Time MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 93% 1 (M = 8.35, SD = 1.97), Time 2 (M = 9.09, SD = 1.87), t (33) = -2.62, p = .013 (two-tailed); 2) CSE 15 “Seek help from my tutors and instructors to better understand math concepts” Time 1 (M = 8.85, SD = 1.62), Time 2 (M = 9.50, SD = .90), t (33) = -2.956, p = .006 (two-tailed). Similar to the non-Native Hawaiian student trends, for traditional students, there was a pattern of scores dropping for the 2013 and 2014 cohorts for both traditional and nontraditional students and then increasing again for the 2012 cohort. There were two items in the scale that had statistically significant trends, both for traditional students 1) CSE 9 “Do well on my exams” (F (3, 63) = 2.769, p = .049), 2) CSE 16 “Understand my textbook” (F (3, 63) = 2.931, p = .040). MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 94% Table 17 Descriptive Statistics for College Self-Efficacy Items Pre/Post 2015 MSB with Age Group Breakdown (n = 67) All Traditional Nontraditional % extremely confident (8-10) M (SD) M (SD) College Self-Efficacy Item Pre Post Pre Post Pre Post Make new friends at college 60.3 64.2 6.85(2.732)* 7.56(2.464)* 8.35(1.968)* 9.09(1.865)* Successfully complete my current math course 79.4 72.3 8.62(1.970) 8.71(1.962) 8.50(1.895) 9.12(1.493) Talk to college staff/faculty 73.5 73 7.06(2.510)** 8.06(2.242)** 9.41(.957) 9.38(1.129) Talk to college administration 72.1 66.2 7.06(2.785) 7.62(2.640) 9.00(1.557) 8.68(1.918) Manage time effectively 69.1 58.8 7.47(2.525) 7.91(2.151) 8.03(2.263) 8.35(2.013) Ask a question in class 70.6 70.3 7.41(2.653)** 8.35(2.028)** 9.03(1.883) 8.97(1.547) Participate in class discussions 66.2 71.6 7.35(2.411)* 8.29(1.987)* 9.00(1.809) 9.29(.970) Research a term paper 57.4 59.5 6.91(2.723) 7.41(2.376) 8.15(2.190) 8.15(1.877) Do well on my exams 58.8 63.5 7.38(2.323)* 8.06(2.215)* 7.94(2.059) 8.29(1.835) Join a student organization 32.4 40.5 5.62(2.697)** 7.00(3.094)** 7.03(2.564) 6.71(3.070) Talk to my professors outside of class 61.8 65.5 6.79(2.826)** 7.79(2.240)** 8.62(2.132) 8.47(2.063) * p < .05. ** p < .01 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 95% Table 17 Descriptive Statistics for College Self-Efficacy Items Pre/Post MSB with Age Group Breakdown Cont. (n = 67) All Traditional Nontraditional % extremely confident (8-10) M (SD) M (SD) College Self-Efficacy Item Pre Post Pre Post Pre Post Ask a professor a question about math coursework 76.5 74.3 7.94(2.282) 8.38(2.174) 9.00(1.670) 8.91(1.640) Take good class notes 72.1 70.9 8.09(2.165) 8.47(1.895) 8.41(2.119) 8.74(1.831) Complete the Math Requirement (pass Math 25 or Math 28) 83.8 79.1 8.32(2.142) 8.56(1.957) 8.91(1,422) 8.65(2.073) Seek help from my tutors and instructors to better understand math concepts 82.4 80.4 8.32(1.838) 8.53(2.048) 8.85(1.617)** 9.50(.896)** Understand my textbook 70.6 65.5 7.79(2.185) 8.29(2.053) 8.29(2.008) 8.18(2.208) Keep up to date with my schoolwork 86.8 75.7 8.29(1.978) 8.50(1.973) 9.09(1.485) 9.18(1.336) Write course papers 67.6 64.2 7.26(2.821) 7.74(2.287) 8.24(2.362) 8.65(1.686) Complete my math homework 91.2 79.7 8.94(1.825) 8.71(2.008) 9.15(1.794) 9.38(.985) * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 96% Table 18 Descriptive Statistics for College Self-Efficacy Items Post MSB 2015, 2014, 2013, 2012 with Age Group Breakdown (n = 150) All Traditional Nontraditional % extremely confident (8-10) M (SD) M (SD) College Self-Efficacy Item 15 14 13 12 15 14 13 12 15 14 13 12 Make new friends at college 70.6 53.6 68.4 50 7.70 (2.365) 7.00 (2.179) 7.33 (2.469) 7.30 (2.627) 9.06 (1.846) 8.00 (2.028) 8.48 (2.254) 7.00 (3.347) Successfully complete my current math course 80.9 57.1 65.8 68.8 8.67 (1.979) 7.00 (2.872) 8.00 (2.171) 8.00 (2.000) 8.86 (1.751) 8.16 (2.167) 8.09 (2.214) 8.83 (2.858) Talk to college staff/faculty 77.9 64.3 68.4 68.8 8.06 (2.277) 7.11 (2.759) 6.80 (3.212) 8.10 (2.378) 9.09 (1.597) 8.79 (1.718) 8.78 (1.808) 8.67 (3.266) Talk to college administration 72.1 57.1 60.5 68.8 7.70 (2.640) 6.22 (3.073) 6.00 (3.162) 8.20 (1.989) 8.63 (1.942) 8.63 (1.640) 8.74 (1.839) 7.50 (3.886) Manage time effectively 64.7 57.1 50 50 7.97 (2.158) 6.67 (2.449) 7.00 (1.813) 7.00 (2.449) 8.09 (2.161) 7.84 (1.893) 7.91 (1.975) 7.00 (2.966) Ask a question in class 77.9 67.9 63.2 62.5 8.33 (2.056) 7.67 (2.000) 7.93 (2.154) 8.10 (1.969) 8.97 (1.524) 8.11 (2.132) 8.13 (2.616) 8.00 (2.757) Participate in class discussions 82.4 64.3 60.5 62.5 8.36 (1.966) 7.44 (2.068) 7.47 (2.167) 8.10 (1.912) 9.20 (1.023) 8.21 (1.988) 8.30 (1.964) 8.33 (2.733) Research a term paper 60.3 57.1 52.6 68.8 7.48 (2.373) 6.89 (2.147) 6.67 (2.554) 7.80 (2.348) 7.94 (2.043) 7.53 (2.118) 8.04 (2.142) 7.33 (2.944) Do well on my exams 72.1 50 57.9 56.3 8.12* (2.219) 5.78* (2.489) 6.87* (2.503) 7.30* (2.312) 8.03 (2.203) 7.21 (2.200) 7.78 (2.131) 6.83 (2.858) * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 97% Table 18 Descriptive Statistics for College Self-Efficacy Items Post MSB 2015, 2014, 2013, 2012 with Age Group Breakdown Cont. (n = 150) All Traditional Nontraditional % extremely confident (8-10) M (SD) M (SD) College Self-Efficacy Item 15 14 13 12 15 14 13 12 15 14 13 12 Join a student organization 47.1 25 36.8 50 7.18 (2.952) 5.00 (2.121) 5.27 (3.305) 6.50 (2.383) 6.69 (3.046) 6.32 (2.849) 6.57 (2.608) 7.83 (3.251) Talk to my professors outside of class 67.6 71.4 63.2 50 7.79 (2.274) 7.67 (2.398) 6.87 (2.800) 7.10 (2.378) 8.31 (2.097) 8.26 (1.881) 8.30 (1.941) 8.17 (3.251) Ask a professor a question about math coursework 76.5 78.6 71.1 62.5 8.33 (2.189) 7.67 (2.739) 7.67 (1.988) 8.10 (2.079) 8.74 (1.738) 8.63 (2.216) 8.87 (1.914) 8.50 (3.209) Take good class notes 76.5 67.9 68.4 50 8.52 (1.906) 7.67 (2.121) 7.27 (2.463) 8.00 (2.160) 8.51 (1.976) 8.16 (1.708) 8.65 (1.799) 6.33 (3.615) Complete the Math Requirement (pass Math 25 or Math 28) 77.9 71.4 84.2 75 8.67 (1.882) 7.67 (2.693) 8.67 (1.877) 8.70 (2.111) 8.37 (2.250) 8.37 (2.793) 9.22 (1.506) 8.50 (3.209) Seek help from my tutors and instructors to better understand math concepts 85.3 71.4 81.6 68.8 8.48 (2.063) 8.00 (2.236) 7.93 (2.520) 8.10 (2.079) 9.31 (1.183) 8.68 (1.945) 9.26 (1.356) 8.50 (3.209) Understand my textbook 70.6 64.3 57.9 56.3 8.33* (2.072) 6.11* (2.571) 6.80* (2.704) 7.50* (2.273) 7.89 (2.361) 8.37 (2.087) 8.39 (1.777) 7.17 (2.483) Keep up to date with my schoolwork 83.8 75 65.8 62.5 8.48 (2.002) 7.89 (2.315) 7.60 (2.165) 7.90 (2.558) 9.00 (1.495) 8.32 (1.797) 8.57 (1.903) 8.17 (3.251) Write course papers 69.1 67.9 52.6 56.3 7.79 (2.302) 6.67 (2.915) 6.60 (3.225) 7.80 (2.348) 8.31 (2.040) 7.95 (2.549) 8.09 (1.998) 7.33 (3.559) Complete my math homework 85.3 75 71.1 75 8.73 (2.035) 7.89 (2.028) 7.93 (1.944) 8.30 (2.359) 9.09 (1.502) 8.47 (2.389) 8.83 (1.825) 8.67 (3.266) * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 98% Sense of Belonging Inventory Factor Analysis and Reliability Analysis This study used an altered form of Hagerty and Patusky’s (1995) Sense of Belonging Inventory. This scale has been condensed to 18 questions with a specific focus on sense of belonging in college. This scale measures two different types of sense of belonging: Antecedent and Psychological properties. For the purposes of this study, each part was separated out to investigate nuances in the data: SOBI-A (Antecedents) and SOBI-P (Psychological). The following section will look at each part of the test with special consideration for ethnicity (Native Hawaiian or non-Native Hawaiian) and age group (traditional or nontraditional). There were sixty-seven (67) completed pre/post tests for the 2015 cohort and eighty-three (83) completed surveys from the 2012-2014 cohorts. Sense of Belonging-Antecedents (SOBI-A) inventory factor analysis and reliability analysis-Ethnicity breakdown. The Sense of Belonging Inventory-Antecedents focus aims to measure the precursors of sense of belonging. In other words, the social conditions that an individual needs to be aware of before they can establish a sense of belonging. This SOBI-A scale consisted of eight questions, numbered 2, 4, 6, 8, 11, 12, 14, and 17 on the larger SOBI 18- question scale. Table 19 (Pre/Post 2015 MSB) and Table 20 (Post 2015, 2014, 2013, 2012) display the SOBI-A total percent that agree/strongly agree as well as the mean (M) and standard deviation (SD) for Non-Native Hawaiian and Native Hawaiian students. Table 21 (Pre/Post 2015 MSB) and Table 22 (Post 2015, 2014, 2013, 2012) display SOBI-A total percentage of students that agree/strongly agree with each item as well as the mean (M) and standard deviation (SD) for traditional and nontraditional students. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 99% Table 19 Descriptive Statistics for Sense of Belonging-Antecedents (SOBI-A) Items Pre/Post 2015 MSB with Ethnicity Breakdown (n = 67) Sense of Belonging- Antecedents Items All Non-NH NH % agree or strongly agree M (SD) M (SD) Pre Post Pre Post Pre Post It is important to be valued by others, especially at school 75 70.9 2.92(.796) 2.77(.908) 2.95(.795) 2.86(.926) I have felt valued in the past 82.4 83.1 3.19(.694) 2.96(.999) 2.95(.661) 3.10(.850) It is important that I fit in school 57.4 48 2.77(.951) 2.65(1.018) 2.64(.759) 2.52(.943) I have qualities 94.1 94.6 3.35(.485) 3.46(.508) 3.14(.566) 3.31(.715) I want to be a part of things 88.2 87.8 3.27(.604) 3.08(.845) 3.14(.647) 3.21(.606) It is important that my opinions are valued 72.1 81.1 3.00(.938) 3.27(.778) 2.88(.670) 2.93(.712) Others recognize my strengths 86.8 81.1 3.19(.567) 2.92(.796) 2.95(.623) 2.95(.795) I make myself fit in at school 73.5 62.8 3.00(.693) 2.77(.951) 2.88(.670) 2.83(.881) Note. a item scores were reverse-coded for total SOBI-A score. * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 100% Table 20 Descriptive Statistics for Sense of Belonging-Antecedents (SOBI-A) Items Post MSB 2015, 2014, 2013, 2012 with Ethnicity Breakdown (n = 150) Sense of Belonging-Antecedents Items All Non-NH NH % agree or strongly agree M (SD) M (SD) 15 14 13 12 15 14 13 12 15 14 13 12 It is important to be valued by others, especially at school 64.7 85.7 76.3 56.3 2.69 (.970) 2.75 (.886) 3.00 (.853) 2.25 (.957) 2.83 (.908) 3.25 (.639) 2.73 (.919) 2.67 (1.073) I have felt valued in the past 76.5 92.9 86.8 87.5 2.96 (.999) 3.50 (.535) 3.17 (.835) 3.75 (.500) 3.07 (.838) 3.10 (.553) 3.15 (.784) 3.17 (.937) It is important that I fit in school 50 42.9 47.4 43.8 2.62 (1.023) 2.50 (.756) 2.58 (.900) 2.50 (1.291) 2.50 (.944) 2.25 (.910) 2.23 (.992) 2.25 (.965) I have qualities 94.1 100 89.5 100 3.46 (.508) 3.50 (.535) 3.42 (.793) 3.75 (.500) 3.29 (.708) 3.45 (.510) 3.35 (.745) 3.58 (.515) I want to be a part of things 88.2 85.7 86.8 93.8 3.04 (.824) 3.38 (.518) 3.33 (.651) 3.75 (.500) 3.19 (.594) 3.25 (.786) 3.12 (.684) 3.42 (.669) It is important that my opinions are valued 79.4 85.7 81.6 75 3.19 (.801) 3.00 (.535) 3.17 (.835) 2.50 (1.000) 2.90 (.692) 3.10 (.641) 3.04 (.824) 3.08 (.793) Others recognize my strengths 79.4 82.1 81.6 81.3 3.00 (.693) 3.00 (.535) 3.00 (1.044) 3.00 (.816) 2.98 (.749) 3.00 (.918) 3.08 (.796) 3.17 (.207) I make myself fit in at school 66.2 64.3 52.6 68.8 2.85 (.881) 2.88 (.641) 2.25 (.965) 2.75 (.957) 2.79 (.871) 2.55 (.887) 2.73 (1.002) 3.00 (.953) Note. a item scores were reverse-coded for total SOBI-A score. * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 101% No statistically significant changes for any pre/post SOBI-A items were found, although it was interesting to note that many of the scores went down from pre to post-test. As previously noted, the researcher hypothesizes this is because of the timing of when the post-test was administered (right before their final exam) and students had higher levels of anxiety and stress. This might not be important because students could have found their niche and do not need to necessarily “fit in” as long as they are comfortable at school. The same could be true for “It is important to be valued by others, especially at school” where the students are more focused on their academics than their social life (Di Tommaso, 2012). There were also no significantly different scores when comparing cohorts (Table 20). Sense of Belonging-Antecedents (SOBI-A) inventory factor analysis and reliability analysis-Age group breakdown. When grouping the data by age group, there were no statistically significant changes for any item answered by nontraditional students and only one statistically significant change for traditional students was discovered, SOBI-A 4 “I have qualities” Time 1 (pre-survey) (M = 3.06, SD = .489), Time 2 (post-survey) (M = 3.35, SD = .544), t (33) = -3.273, p = .002 (two-tailed). It is also interesting that as a whole, traditional students scored themselves higher than their nontraditional counterparts on six out of eight scale items. There were no statistically significant changes in scores when comparing the four cohorts (2012, 2013, 2014, and 2015) and there weren’t any visible trends or patterns in the data. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 102% Table 21 Descriptive Statistics for Sense of Belonging-Antecedents (SOBI-A) Items Pre/Post 2015 MSB with Age Group Breakdown (n = 67) Sense of Belonging- Antecedents Items All Traditional Nontraditional % agree or strongly agree M (SD) M (SD) Pre Post Pre Post Pre Post It is important to be valued by others, especially at school 75 70.9 3.00(.603) 2.88(.808) 2.88(.946) 2.76(1.017) I have felt valued in the past 82.4 83.1 2.88(.640) 3.00(.853) 3.21(.687) 3.09(.965) It is important that I fit in school 57.4 48 2.71(.760) 2.56(.960) 2.68(.912) 2.59(.988) I have qualities 94.1 94.6 3.06(.489)** 3.35(.544)** 3.38(.551) 3.38(.739) I want to be a part of things 88.2 87.8 3.09(.570) 3.06(.736) 3.29(.676) 3.26(.666) It is important that my opinions are valued 72.1 81.1 2.91(.753) 3.00(.816) 2.94(.814) 3.12(.686) Others recognize my strengths 86.8 81.1 2.94(.600) 2.94(.776) 3.15(.610) 2.94(.814) I make myself fit in at school 73.5 62.8 2.88(.640) 2.68(.843) 2.97(.717) 2.94(.952) Note. a item scores were reverse-coded for total SOBI-A score. * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 103% Table 22 Descriptive Statistics for Sense of Belonging-Antecedents (SOBI-A) Items Post MSB 2015, 2014, 2013, 2012 with Age Group Breakdown (n = 150) Sense of Belonging- Antecedents Items All Traditional Nontraditional % agree or strongly agree M (SD) M (SD) 15 14 13 12 15 14 13 12 15 14 13 12 It is important to be valued by others, especially at school 64.7 85.7 76.3 56.3 2.88 (.820) 3.22 (.667) 2.93 (.799) 2.70 (1.160) 2.69 (1.02 2) 3.05 (.780) 2.74 (.964) 2.33 (.816) I have felt valued in the past 76.5 92.9 86.8 87.5 3.00 (.866) 3.44 (.527) 2.87 (.743) 3.30 (.675) 3.06 (.938) 3.11 (.567) 3.35 (.775) 3.33 (1.211) It is important that I fit in school 50 42.9 47.4 43.8 2.55 (.971) 2.11 (.928) 2.27 (.799) 2.30 (.949) 2.54 (.980) 2.42 (.838) 2.39 (1.076) 2.33 (1.211) I have qualities 94.1 100 89.5 100 3.36 (.549) 3.44 (.527) 3.33 (.617) 3.40 (.516) 3.34 (.725) 3.47 (.513) 3.39 (.839) 4.00 (.000) I want to be a part of things 88.2 85.7 86.8 93.8 3.03 (.728) 3.56 (.726) 3.27 (.799) 3.40 (.699) 3.23 (.646) 3.16 (.688) 3.13 (.815) 3.67 (.516) It is important that my opinions are valued 79.4 85.7 81.6 75 2.97 (.810) 3.22 (.441) 3.20 (.676) 2.90 (.994) 3.06 (.684) 3.00 (.667) 3.00 (.905) 3.00 (.632) Others recognize my strengths 79.4 82.1 81.6 81.3 3.00 (.707) 2.78 (.972) 3.13 (.743) 2.80 (.632) 2.97 (.747) 3.11 (.737) 3.00 (.953) 3.67 (.516) I make myself fit in at school 66.2 64.3 52.6 68.8 2.73 (.801) 2.33 (1.00) 2.40 (.910) 3.00 (.816) 2.89 (.932) 2.79 (.713) 2.70 (1.063) 2.83 (1.169) Note. a item scores were reverse-coded for total SOBI-A score. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 104% Sense of Belonging-Psychological (SOBI-P) inventory factor analysis and reliability analysis-Ethnicity breakdown. The Sense of Belonging Inventory-Psychological focus aims to measure the psychological state of sense of belonging by having the individual rank (strongly disagree to strongly agree) different aspects of sense of belonging. This SOBI-P scale consisted of ten questions, numbered 1, 3, 5, 7, 9, 10, 13, 15, 16, and 18 on the larger SOBI 18-question scale. Table 23 (Pre/Post 2015 MSB) and Table 24 (Post 2015, 2014, 2013, 2012) display the SOBI-P total percent that agree/strongly agree with each scale item as well as the mean (M) and standard deviation (SD) for Non-NH and NH students. For the 2015 MSB pre/post SOBI-P scores grouped by ethnicity, there were no statistically significant changes except for Item 5 (“What I offer is valued by others at school (peers, faculty, instructors”). For Native Hawaiian students, their scores significantly increased at the p = .01 level. Also, for the scores that were not statistically significant, students’ scores went in the direction that would show improvement (e.g. scores went up for Item 4, “People accept me”, and down for Item 6, “I have no place in this world”). When comparing the different cohorts’ scores, there were no significant changes over time. There were also no identifiable trends or patterns. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 105% Table 23 Descriptive Statistics for Sense of Belonging-Psychological (SOBI-P) Items Pre/Post 2015 MSB with Ethnicity Breakdown (n = 67) Note. a item scores were reverse-coded for total SOBI-P score. * p < .05. ** p < .01. Sense of Belonging- Psychological Items All Non-NH NH % agree or strongly agree M (SD) M (SD) Pre Post Pre Post Pre Post I wonder if I really fit in college a 55.9 51.4 2.38(.898) 2.54(1.140) 2.67(.979) 2.64(1.032) I am not sure if I fit with my student peers a 32.4 29.7 2.46(.811) 2.23(.815) 2.10(.850) 2.07(.894) I describe myself as a misfit a 30.9 20.3 2.19(.981) 2.08(.977) 1.93(.808) 1.88(.916) People accept me 89.7 85.1 3.15(.543) 3.19(.749) 3.14(.608) 3.21(.645) What I offer is valued by others at school (peers, faculty, instructors) 82.4 87.2 2.96(.662) 3.04(.774) 2.93(.558) ** 3.24(.617) ** I have no place in this world a 14.7 8.8 1.88(1.033) 1.58(.987) 1.57(.737) 1.50(.773) I’d rather observe college life than participate in it a 29.4 20.3 2.35(.936) 2.19(.981) 2.00(.733) 1.98(.869) I don’t really fit in at school a 13.2 12.2 1.88(.816) 1.77(.765) 1.79(.750) 1.71(.742) I feel left out at school a 17.6 13.5 1.88(.864) 1.85(.784) 1.86(.783) 1.76(.759) I do not feel valued or important at school a 8.8 10.1 1.73(.778) 1.73(.778) 1.71(6.73) 1.71(.805) MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 106% Table 24 Descriptive Statistics for Sense of Belonging-Psychological (SOBI-P) Items Post MSB 2015, 2014, 2013, 2012 with Ethnicity Breakdown (n = 150) Sense of Belonging-Psychological Items All Non-NH NH % agree or strongly agree M (SD) M (SD) 15 14 13 12 15 14 13 12 15 14 13 12 I wonder if I really fit in college a 51.5 57.4 50 37.5 2.42 (1.137) 2.25 (.886) 2.50 (1.243) 2.00 (.816) 2.60 (1.014) 2.65 (.988) 2.42 (.902) 2.42 (.996) I am not sure if I fit with my student peers a 25 28.6 36.8 31.3 2.15 (.834) 2.13 (.835) 2.08 (.996) 2.25 (.957) 2.07 (.894) 2.00 (.725) 2.04 (.871) 2.00 (.953) I describe myself as a misfit a 23.5 14.3 21.1 18.6 2.04 (.916) 1.50 (.756) 1.83 (1.030) 2.00 (1.155) 1.90 (.906) 1.65 (.745) 1.81 (.939) 1.83 (.835) People accept me 85.3 92.9 78.9 81.3 3.19 (.749) 3.13 (.641) 3.17 (.718) 3.00 (.816) 3.21 (.645) 3.15 (.489) 3.04 (.999) 2.92 (.793) What I offer is valued by others at school (peers, faculty, instructors) 83.8 100 78.9 87.5 3.00 (.800) 3.13 (3.54) 3.25 (.866) 3.00 (.816) 3.19 (.634) 3.40 (.503) 3.00 (.849) 3.17 (.835) I have no place in this world a 8.8 7.1 13.2 6.32 1.54 (905) 1.25 (.463) 1.75 (1.215) 1.25 (.500) 1.50 (.773) 1.20 (.616) 1.38 (.637) 1.42 (.900) I’d rather observe college life than participate in it a 26.5 14.3 18.4 12.5 2.19 (.981) 1.88 (.835) 1.83 (.937) 1.25 (.500) 2.00 (.855) 1.60 (.681) 1.62 (.804) 1.92 (.669) I don’t really fit in at school a 7.4 7.1 21 18.7 1.73 (.724) 1.75 (.707) 1.92 (.900) 2.50 (1.291) 1.74 (.734) 1.45 (.605) 1.62 (.852) 1.75 (.866) I feel left out at school a 13.2 7.1 18.4 12.5 1.81 (.749) 1.63 (.744) 1.75 (.965) 2.00 (.816) 1.76 (.759) 1.40 (.598) 1.58 (.809) 1.67 (.651) I do not feel valued or important at school a 11.8 7.1 10.5 6.3 1.73 (.778) 1.38 (.518) 1.42 (.669) 1.75 (.500) 1.71 (.805) 1.40 (.681) 1.58 (.902) 1.58 (.669) Note. a item scores were reverse-coded for total SOBI-P score MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 107% Sense of Belonging-Psychological (SOBI-P) inventory factor analysis and reliability analysis-Age group breakdown. Table 25 (Pre/Post 2015 MSB) and Table 26 (Post 2015, 2014, 2013, 2012) display the SOBI-P total percent that agree/strongly agree with each scale item as well as the mean (M) and standard deviation (SD) for traditional and nontraditional students. Table 25 Descriptive Statistics for Sense of Belonging-Psychological (SOBI-P) Items Pre/Post 2015 MSB with Age Group Breakdown (n = 67) Sense of Belonging- Psychological Items All Traditional Nontraditional % agree or strongly agree M (SD) M (SD) Pre Post Pre Post Pre Post I wonder if I really fit in college a 55.9 51.4 2.38(.739) 2.41(.857) 2.74(1.109) 2.79(1.225) I am not sure if I fit with my student peers a 32.4 29.7 2.18(.797) 2.24(.781) 2.29(.906) 2.03(.937) I describe myself as a misfit a 30.9 20.3 2.06(.814) 2.03(.870) 2.00(.953) 1.88(1.008) People accept me 89.7 85.1 3.06(.547) 3.15(.744) 3.24(.606) 3.26(.618) What I offer is valued by others at school (peers, faculty, instructors) 82.4 87.2 2.74(.567)** 3.15(.610)** 3.15(.558) 3.18(.758) I have no place in this world a 14.7 8.8 1.68(.768) 1.59(.821) 1.71(.970) 1.47(.896) I’d rather observe college life than participate in it a 29.4 20.3 2.06(.776) 2.26(.790) 2.21(.880) 1.85(.989) I don’t really fit in at school a 13.2 12.2 1.79(.641) 1.79(.592) 1.85(.892) 1.68(.878) I feel left out at school a 17.6 13.5 1.79(.641) 1.88(.640) 1.94(.952) 1.71(.871) I do not feel valued or important at school a 8.8 10.1 1.71(.629) 1.68(.638) 1.74(.790) 1.76(.923) Note. a item scores were reverse-coded for total SOBI-P score. * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 108% Table 26 Descriptive Statistics for Sense of Belonging-Psychological (SOBI-P) Items Post MSB 2015, 2014, 2013, 2012 with Age Group Breakdown (n = 150) Sense of Belonging- Psychological Items All Traditional Nontraditional % agree or strongly agree M (SD) M (SD) 15 14 13 12 15 14 13 12 15 14 13 12 I wonder if I really fit in college a 51.5 57.4 50 37.5 2.36 (.822) 2.33 (1.118) 2.67 (.976) 2.40 (.843) 2.69 (1.231) 2.63 (.895) 2.30 (1.020) 2.17 (1.169) I am not sure if I fit with my student peers a 25 28.6 36.8 31.3 2.21 (.781) 1.89 (.782) 2.33 (.900) 2.00 (8.16) 2.00 (.939) 2.11 (.737) 1.87 (.869) 2.17 (1.169) I describe myself as a misfit a 23.5 14.3 21.1 18.6 1.97 (.810) 1.56 (.726) 2.27 (1.100) 2.30 (.823) 1.94 (.998) 1.63 (.761) 1.52 (.730) 1.17 (.408) People accept me 85.3 92.9 78.9 81.3 3.18 (.727) 3.22 (.667) 3.13 (.834) 2.70 (.675) 3.23 (.646) 3.11 (.459) 3.04 (.976) 3.33 (.816) What I offer is valued by others at school (peers, faculty, instructors) 83.8 100 78.9 87.5 3.15 (.619) 3.44 (.527) 2.93 (.884) 3.00 (.471) 3.09 (.781) 3.26 (.452) 3.17 (.834) 3.33 (1.211) I have no place in this world a 8.8 7.1 13.2 6.32 1.52 (.712) 1.00 (.000) 1.80 (1.014) 1.30 (.483) 1.51 (.919) 1.32 (.671) 1.30 (.703) 1.50 (1.225) I’d rather observe college life than participate in it a 26.5 14.3 18.4 12.5 2.24 (.792) 1.44 (.726) 2.33 (.900) 1.80 (.632) * 1.91 (.981) 1.79 (.713) 1.26 (.449) 1.67 (.816) * I don’t really fit in at school a 7.4 7.1 21 18.7 1.76 (.561) 1.44 (.527) 2.07 (.961) 2.10 (.876) 1.71 (.860) 1.58 (.692) 1.48 (.730) 1.67 (1.211) I feel left out at school a 13.2 7.1 18.4 12.5 1.85 (.619) 1.44 (.726) 1.93 (.961) 1.80 (.632) 1.71 (.860) 1.47 (.612) 1.43 (.728) 1.67 (.816) I do not feel valued or important at school a 11.8 7.1 10.5 6.3 1.67 (.645) 1.22 (.667) 1.73 (.961) 1.70 (.483) 1.77 (.910) 1.47 (.612) 1.39 (.722) 1.50 (.837) Note. a item scores were reverse-coded for total SOBI-P score. * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 109% Similar to the SOBI-P analysis with ethnicity breakdown, there were no statistically significant changes in SOBI-P items except for Item 5 (“What I offer is valued by others at school (peers, faculty, instructors”) for traditional students; their scores significantly increased at the p = .01 level. Although there were no other statistically significant changes, students’ scores went in the direction that would show improvement (e.g. scores went up for Item 4, “People accept me”, and down for Item 6, “I have no place in this world”). Table 26 shows one question having significant change over the four cohorts for both traditional and nontraditional students, “I’d rather observe college life than participate in it.” However, there wasn’t an increase or decrease in trend. Instead, for both traditional and nontraditional students, the 2013 and 2014 cohorts had the lowest score (indicating they disagreed with the statement) and the 2015 and 2012 cohorts had higher scores. It is unclear why this occurred. One hypothesis that the researcher has concluded is that for the 2012 cohort, who have been in school the longest, they might be more focused on completing their academics and not put value in extracurricular college activities. Instructor Interaction Inventory Factor Analysis and Reliability Analysis Research has shown that students who interact with their instructors have higher success rates (course grades, persistence, graduation) because they are more likely to have a greater sense of belonging and ask for help when needed (Lundberg, 2014; Thompson, 2001; Wirt & Jaeger, 2014). This study measured both instructor interaction in class and outside of class (e.g. office hours). It was measured with one question for each. The following section gives a more detailed analysis of both in class and outside of class instructor interaction with focus on ethnicity (Native Hawaiian v. non-Native Hawaiian) and age group (traditional v. nontraditional). MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 110% Instructor interaction inventory factor analysis and reliability analysis-In class. Re- spondents were asked the following question, “How many times do you talk to/interact with your instructor in class?” with three options to choose from: (1) 0 times/week; (2) 1-3 times/week; or (3) 4+ times/week. Table 27 and Table 28 display the mean (M) and standard deviation (SD) for Pre/Post 2015, 2014, 2013, 2012 cohorts’ answers with ethnicity (non-Native Hawaiian v. Native Hawaiian) and age group (traditional v. nontraditional) emphasis. Overall, for the total sample population (all respondents) and also in both groupings (ethnicity or age), students’ instructor interaction in class increased over time. For the 2012 cohort (3 years post intervention), both non-Native Hawaiians and nontraditional students had 100% response rate saying that they interacted with their instructors 4+ times per week. High instructor interaction has been associated with other student success measures, such as higher cumulative GPA, persistence, and graduation (Lundberg, 2014; Thompson, 2001; Wirt & Jaeger, 2014). A one-way ANOVA will be conducted in part two of the analysis to determine if these increases are statistically significant. Instructor interaction inventory factor analysis and reliability analysis-Outside of class. Respondents were asked the following question, “How many times do you talk to/interact with your instructor outside of class?” with three options to choose from: (1) 0 times/week; (2) 1- 3 times/week; or (3) 4+ times/week. Table 29 and Table 30 display the mean (M) and standard deviation (SD) results for Pre/Post 2015, 2014, 2013, 2012 cohorts with ethnicity (non-Native Hawaiian v. Native Hawaiian) and age group (traditional v. nontraditional) emphasis. Similar to Instructor Interaction In Class, for this item, students’ interaction with instructors outside of class increased over time. However, in every way that was analyzed (total, ethnicity, age group), the 2013 cohort had a slight decrease in scores. A one-way ANOVA will MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 111% be completed in the second half of this chapter to see if these changes were statistically significant. Peer Interaction Inventory Factor Analysis and Reliability Analysis Peer interaction has been generalized into two different categories: school and social. School interaction involves meeting with peers for school related activities (e.g. studying for a test, working on a project). Social interaction involves meeting with peers for non-school related activities (e.g. meeting for coffee or lunch). It was important to distinguish between the two because research shows that increased peer interaction is only beneficial for students if it is academically related (Di Tommaso, 2012). The following section provides respondents’ answers for each peer interaction question. Peer interaction-school. Students were given one question in the survey pertaining to peer interaction with a school focus. This question was, “How many times do you interact with your peers at school for academic purposes (e.g. study groups, work on projects)?” Students had four different options to choose from: (1) 0 times/week; (2) 1-3 times/week; (3) 4-6 times/week; or (4) 7+ times/week. Table 31 and Table 32 provide the mean (M) and standard deviation (SD) of the respondents’ answers from the pre/post 2015 cohort and the post 2015, 2014, 2013, 2012 cohorts with ethnicity and age group breakdown respectively. There was an overall pattern for peer interaction in school. The data showed a sharp increase in peer (academic) interaction from the pre to post test (right after intervention), then scores dropped for the 2013 and 2014 cohorts and then increased again for the 2012 cohort. None of the post scores were as low as the pre 2015 scores. A one-way ANOVA will be conducted in the second half of this chapter to determine whether or not these changes were significant. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 112% Peer Interaction-Social. Students were given one question in the survey pertaining to social peer interaction. This question was, “How many times do you interact with your peers at school for social purposes (e.g. eat lunch, drink coffee, socialize)?” Students had four different options to choose from: (1) 0 times/week; (2) 1-3 times/week; (3) 4-6 times/week; or (4) 7+ times/week. Table 33 and Table 34 provide the mean (M) and standard deviation (SD) of the respondents’ answers from the pre/post 2015 cohort and the post 2015, 2014, 2013, 2012 cohorts with ethnicity (non-Native Hawaiian v. Native Hawaiian) and age group (traditional v. nontraditional) emphasis. For every group except for nontraditional students, there was an increasing trend of social peer interaction over time. Nontraditional students had an increase in social peer interaction from 2015 pre- to 2015 post-test and then increased for the 2013 and 2014 cohort responses, however came back down for 2012 cohort responses. As mentioned previously, social peer interaction is not necessarily linked to student success measures and in fact can hinder student success by taking focus away from academics (Di Tommaso, 2012). Therefore, the researcher was not expecting this measure to significantly increase over time. A one-way ANOVA will be conducted in the second half of this chapter to determine whether or not there were significant changes. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 113% Table 27 Descriptive Statistics for Instructor Interaction-In Class Pre/Post 2015, 2014, 2013, 2012 MSB with Ethnicity Breakdown (n = 150) Table 28 Descriptive Statistics for Instructor Interaction-In Class Pre/Post 2015, 2014, 2013, 2012 MSB with Age Group Breakdown (n = 150) All Non-NH NH M (SD) M (SD) M (SD) Pre Post 15 14 13 12 Pre Post 15 14 13 12 Pre Post 15 14 13 12 Faculty Interaction- In Class 2.25 (.677) 2.35 (.511) 2.75 (.41) 2.61 (.638) 2.88 (.342) 2.27 (.724) 2.48 (.510) 2.88 (.354) 2.67 (.492) 3.00 (.000) 2.24 (.656) ** 2.27 (.501) ** 2.70 (.470) ** 2.58 (.703) ** 2.83 (.398) * p < .05. ** p < .01. All Traditional Nontraditional M (SD) M (SD) M (SD) Pre Post 15 14 13 12 Pre Post 15 14 13 12 Pre Post 15 14 13 12 Faculty Interaction- In Class 2.25 (.677) 2.35 (.511) 2.75 (.41) 2.61 (.638) 2.88 (.342) 2.18 (.716) 2.36 (.549) 2.78 (.441) 2.33 (.724) 2.80 (.422) 2.32 (.638) 2.33 (.479) 2.74 (.452) 2.78 (.518) 3.00 (.000) * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 114% Table 29 Descriptive Statistics for Instructor Interaction-Outside of Class Pre/Post 2015, 2014, 2013, 2012 MSB with Ethnicity Breakdown (n = 150) Table 30 Descriptive Statistics for Instructor Interaction-Outside of Class Pre/Post 2015, 2014, 2013, 2012 MSB with Age Group Breakdown (n = 150) All Non-NH NH M (SD) M (SD) M (SD) Pre Post 15 14 13 12 Pre Post 15 14 13 12 Pre Post 15 14 13 12 Faculty Interaction- Outside 1.71 (.575) 1.71 (.607) 2.14 (.756) 1.98 (.689) 2.38 (.719) 1.77 (.514) 1.92 (.640) 2.25 (.707) 2.00 (.739) 2.75 (.500) 1.67 (.612) 1.56 (.550) 2.10 (.788) 1.85 (.675) 2.25 (.754) * p < .05. ** p < .01. All Traditional Nontraditional M (SD) M (SD) M (SD) Pre Post 15 14 13 12 Pre Post 15 14 13 12 Pre Post 15 14 13 12 Faculty Interaction- Outside 1.71 (.575) 1.71 (.607) 2.14 (.756) 1.98 (.689) 2.38 (.719) 1.65 (.544) 1.70 (.637) 1.89 (.928) 1.67 (.617) 2.10 (.738) 1.76 (.606) 1.70 (.585) 2.26 (.653) 2.04 (.706) 2.83 (.408) * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 115% Table 31 Descriptive Statistics for Peer Interaction-School Pre/Post 2015, 2014, 2013, 2012 MSB with Ethnicity Breakdown (n = 150) Table 32 Descriptive Statistics for Peer Interaction-School Pre/Post 2015, 2014, 2013, 2012 MSB with Age Group Breakdown (n = 150) All Non-NH NH M (SD) M (SD) M (SD) Pre Post 15 14 13 12 Pre Post 15 14 13 12 Pre Post 15 14 13 12 Peer Interaction- School 2.13 (.845) 2.45 (.768) 2.29 (.535) 2.32 (.873) 2.44 (.892) 1.96 (.662) 2.40 (.764) 2.00 (.000) 2.17 (.835) 2.00 (.000) 2.24 (.932) 2.49 (.779) 2.40 (.598) 2.38 (.898) 2.58 (.996) * p < .05. ** p < .01. All Traditional Nontraditional M (SD) M (SD) M (SD) Pre Post 15 14 13 12 Pre Post 15 14 13 12 Pre Post 15 14 13 12 Peer Interaction- School 2.13 (.845) 2.45 (.768) 2.29 (.535) 2.32 (.873) 2.44 (.892) 2.09 (.866) 2.24 (.663) 2.22 (.441) 2.13 (.990) 2.50 (.850) 2.18 (.834) 2.67 (.816) 2.32 (.582) 2.43 (.788) 2.33 (1.033) * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 116% Table 33 Descriptive Statistics for Peer Interaction-Social Pre/Post 2015, 2014, 2013, 2012 MSB with Ethnicity Breakdown (n = 150) Table 34 Descriptive Statistics for Peer Interaction-Social Pre/Post 2015, 2014, 2013, 2012 MSB with Age Group Breakdown (n = 150) All Non-NH NH M (SD) M (SD) M (SD) Pre Post 15 14 13 12 Pre Post 15 14 13 12 Pre Post 15 14 13 12 Peer Interaction- Social 2.16 (.940) 2.30 (.894) 2.36 (.731) 2.29 (.898) 2.44 (.964) 1.96 (.871) 2.24 (.831) 2.00 (.535) 2.17 (.835) 2.50 (1.000) 2.29 (.970) 2.34 (.938) 2.50 (.761) 2.35 (.936) 2.42 (.996) * p < .05. ** p < .01. All Traditional Nontraditional M (SD) M (SD) M (SD) Pre Post 15 14 13 12 Pre Post 15 14 13 12 Pre Post 15 14 13 12 Peer Interaction- Social 2.16 (.940) 2.30 (.894) 2.36 (.731) 2.29 (.898) 2.44 (.964) 2.15 (.958) 2.24 (.902) 2.56 (.726) 2.07 (1.033) 2.60 (.966) 2.18 (.936) 2.36 (.895) 2.26 (.733) 2.43 (.788) 2.17 (.983) * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 117% Instructor Satisfaction Inventory Factor Analysis and Reliability Analysis Instructor satisfaction was measured using one question in the survey; “I am very satisfied with the academic instruction I’ve received at Windward Community College so far.” This was a five point Likert scale, 1 being “strongly disagree” and 5 being “strongly agree”. Table 35 and Table 36 provide the total percentage of students who agreed or strongly agreed as well as the mean (M) and standard deviation (SD) for ethnicity (non-Native Hawaiian v. Native Hawaiian) and age (traditional v. nontraditional) groups. Generally there was a trend for student’s instructor satisfaction to go up over time and especially right after the intervention. Non-Native Hawaiian students had generally higher satisfaction scores than Native Hawaiian students, and nontraditional students had higher satisfaction scores than traditional students. A one-way ANOVA will be conducted in the second half of this chapter to determine whether these changes are statistically significant. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 118% Table 35 Descriptive Statistics for Instructor Satisfaction Pre/Post 2015, 2014, 2013, 2012 MSB with Ethnicity Breakdown (n = 150) Table 36 Descriptive Statistics for Instructor Satisfaction Pre/Post 2015, 2014, 2013, 2012 MSB with Age Group Breakdown (n = 150) All Non-NH NH % agree or strongly agree M (SD) M (SD) Pre Post 15 14 13 12 Pre Post 15 14 13 12 Pre Post 15 14 13 12 Instructor Satisfaction 87.9 98.5 92.8 94.8 87.5 4.31 (.736) 4.52 (.510) 4.88 (.354) 4.67 (.492) 4.75 (.500) 4.22 (.988) 4.66 (.530) 4.60 (.995) 4.65 (.629) 4.58 (.793) * p < .05. ** p < .01. All Traditional Nontraditional % agree or strongly agree M (SD) M (SD) Pre Post 15 14 13 12 Pre Post 15 14 13 12 Pre Post 15 14 13 12 Instructor Satisfaction 87.9 98.5 92.8 94.8 87.5 4.09 (.914) 4.52 (.566) 4.44 (.726) 4.47 (.640) 4.60 (.699) 4.42 (.867) 4.70 (.467) 4.79 (.918) 4.78 (.518) 4.67 (.816) * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 119% Support Services Satisfaction Inventory Factor Analysis and Reliability Analysis Students who participated in the WCC Math Summer Bridge Program had increased interaction with not only math instructors, but also academic advising staff as well, through intrusive counseling. Support Services satisfaction was measured using one question in the survey; “I am very satisfied with the student support services (e.g. academic advising, financial aid office) I’ve received at Windward Community College so far.” This was a five point Likert scale, 1 being “strongly disagree” and 5 being “strongly agree”. Table 37 and Table 38 provide the total percentage of students who agreed or strongly agreed as well as the mean (M) and standard deviation (SD) for ethnicity (non-Native Hawaiian v. Native Hawaiian) and age (traditional v. nontraditional) groups. There was an increase in students’ supportive services satisfaction right after the intervention (2015 pre- v. 2015 post-test) as well as an upward trend over time for all groupings (non-Native Hawaiian v. Native Hawaiian and traditional v. nontraditional). Also, for the 2012 cohorts, both non-Native Hawaiian and nontraditional students had 100% “strongly agree” responses. Non-Native Hawaiian student responses were generally higher than Native Hawaiian students, and nontraditional students’ higher than traditional. A one-way ANOVA will be conducted in the second half of this chapter to determine whether or not these changes were statistically significant. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 120% Table 37 Descriptive Statistics for Support Services Satisfaction Pre/Post 2015, 2014, 2013, 2012 MSB with Ethnicity Breakdown (n = 150) Table 38 Descriptive Statistics for Support Services Satisfaction Pre/Post 2015, 2014, 2013, 2012 MSB with Age Group Breakdown (n = 150) All Non-NH NH % agree or strongly agree M (SD) M (SD) Pre Post 15 14 13 12 Pre Post 15 14 13 12 Pre Post 15 14 13 12 Supportive Services Satisfaction 84.8 93.9 92.8 94.8 93.8 4.28 (.678) 4.44 (.821) 4.88 (.354) 4.83 (.389) 5.00 (.000) 4.29 (1.031) 4.78 (.475) 4.60 (.995) 4.58 (.758) 4.75 (.622) * p < .05. ** p < .01. All Traditional Nontraditional % agree or strongly agree M (SD) M (SD) Pre Post 15 14 13 12 Pre Post 15 14 13 12 Pre Post 15 14 13 12 Support Services Satisfaction 84.8 93.9 92.8 94.8 93.8 4.18 (.950) 4.52 (.712) 4.44 (.726) 4.60 (.507) 4.70 (.675) 4.39 (.864) 4.79 (.545) 4.79 (.918) 4.70 (.765) 5.00 (.000) * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 121% Overall Windward Community College Satisfaction Inventory Factor Analysis and Reliability Analysis It was important to the researcher to understand how the student felt about their overall experience at Windward Community College. This was measured using one question in the survey; “I am very satisfied with Windward Community College as a whole.” Using a five point Likert scale, 1 being “strongly disagree” and 5 being “strongly agree”. Table 39 and Table 40 provide the total percentage of students who agreed or strongly agreed as well as the mean (M) and standard deviation (SD) for ethnicity (non-Native Hawaiian v. Native Hawaiian) and age (traditional v. nontraditional) groups. As with the other two satisfaction questions, overall Windward Community College satisfaction showed similar trends of an immediate increase after the intervention as well as over time for all groups except for non-Native Hawaiians. For non-Native Hawaiian students, their mean score actually dropped right after the intervention (2015 pre-test, M= 4.52 SD = .653); 2015 post-test, M = 4.40 SD = .577) but then steadily increased over time. Nontraditional students’ overall satisfaction was higher than nontraditional students. Finally, 100% of the 2012 cohort reported that they “agree” or “strongly agree” that they are very satisfied with Windward Community College as a whole. A one-way ANOVA will be conducted in the second half of this chapter to determine whether these changes are statistically significant. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 122% Table 39 Descriptive Statistics for Overall Windward Community College Satisfaction Pre/Post 2015, 2014, 2013, 2012 MSB with Ethnicity Breakdown (n = 150) Table 40 Descriptive Statistics for Overall Windward Community College Satisfaction Pre/Post 2015, 2014, 2013, 2012 MSB with Age Group Breakdown (n = 150) All Non-NH NH % agree or strongly agree M (SD) M (SD) Pre Post 15 14 13 12 Pre Post 15 14 13 12 Pre Post 15 14 13 12 Overall WCC Satisfaction 70.7 97.0 96.5 92.1 100.0 4.52 (.653) 4.40 (.577) 5.00 (.000) 4.67 (.492) 4.75 (.500) 4.39 (1.022) 4.76 (.489) 4.75 (.910) 4.58 (.703) 4.75 (.452) * p < .05. ** p < .01. All Traditional Nontraditional % agree or strongly agree M (SD) M (SD) Pre Post 15 14 13 12 Pre Post 15 14 13 12 Pre Post 15 14 13 12 Overall WCC Satisfaction 70.7 97.0 96.5 92.1 100.0 4.33 (.924) 4.58 (.561) 4.89 (.333) 4.40 (.632) 4.70 (.483) 4.55 (.869) 4.67 (.540) 4.79 (.918) 4.74 (.619) 4.83 (.408) * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 123% Summary of Reliability Tests for Multi-Item Variables This study consisted of six multi-item variables (College Self-Efficacy pre-test, College Self-Efficacy post-test, Sense of Belonging-Antecedent (SOBI-A) pre-test, Sense of Belonging- Antecedent (SOBI-A) post-test, Sense of Belonging-Psychology (SOBI-P) pre-test, and Sense of Belonging-Psychological (SOBI-P) post-test scales). Each of these scales were tested for internal consistency. All but two scales, SOBI-A Post-Test and SOBI-P Pre-Test, had a respectable Cronbach’s α scores. Table 41 reflects the Cronbach’s α for each scale. Table 41 Summary of Composite Variables and Reliability Analysis Composite Variable Items Cronbach’s α College Self-Efficacy Pre-Test 19 .959 College Self-Efficacy Post-Test 19 .926 SOBI-A Pre-Test 18 .829 SOBI-A Post-Test 18 .591 SOBI-P Pre-Test 18 .683 SOBI-P Post-Test 18 .715 Summary of Part One Part one of this chapter provided descriptive statistics on the experiment and control groups as well as the total Windward Community College student population using total numbers and percentages or means and standard deviations. When performing chi square goodness of fit tests for each of the input characteristics, the sample population used in this study had significantly different proportions of ethnicity (non-Native Hawaiian v. Native Hawaiian) Experiment (χ 2 (1, n=294) = 52.792, p = .000) and Control 2 (χ 2 (1, n=460) = 45.944, p = .000), as well as credit enrollment (part-time v. full-time) averages (Experiment, χ 2 (1, n=294) = 45.952, p = .000; Control 1, χ 2 (1, n=203) = 14.092, p = .000; Control 2, χ 2 (1, n=460) = 15.188, MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 124% p = .000). Although skewed, the fact that this study has proportionally more Native Hawaiian students is looked upon as an advantage because this group of students performs at lower rates than the general population (Office of Hawaiian Affairs, 2014) and therefore any significance would be especially beneficial for the University of Hawaii Community Colleges and beyond, where there are larger populations of minority and “at risk” students. For many of the College Self-Efficacy Inventory items, percentages of students who were “extremely confident” (score of 8-10) went down from pre- to post-survey, however not statistically significant. This is thought to be because the post-survey was administered to the students right before they took their final exam. The test was given at this time to ensure all of them complete the survey before they left for the summer. Because students had higher levels of anxiety before their final exam and their final grade was unknown at that point, it is hypothesized that the students scored themselves lower than they would have after finishing their final exam and completing the Math Summer Bridge Program. There was only one item that had a statistically significant increase for non-Native Hawaiians (“Make new friends in college” Time 1 (pre-survey) (M = 6.96, SD = 2.57), Time 2 (post-survey) (M = 7.96, SD = 2.27), t (25) = -3.469, p = .002 (two-tailed)) and two items for Native Hawaiian students (“Talk to college staff/faculty” Time 1 (M = 8.21, SD = 2.34), Time 2 (M = 8.83, SD = 1.87), t (41) = -2.172, p = .036 (two-tailed); and “Participate in class discussions” Time 1 (M = 8.14, SD = 2.34), Time 2 (M = 8.95, SD = 1.56), t (41) = -2.397, p = .021 (two-tailed)). Native Hawaiian students also scored higher on 10 out of 19 scale items compared to their non-Native Hawaiian counterparts When observing the cross-sectional patterns for cohorts 2012-2015, there were two distinct trends. For non-Native Hawaiian students, traditional, and nontraditional students, the MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 125% scores for each of the College Self-Efficacy items to drop in score for the 2013 and 2014 cohorts, where they would be in the middle of their schooling and probably incurring more difficulty courses, leading to higher stress levels and lower self-efficacy, and then increase again for 2012 cohort, where students are probably at the end of (or finished with) schooling and therefore feeling better and more confident about their capabilities (efficacy) in college. Native Hawaiian student trends were different. Looking at the different cohorts, over time, it looks like their scores for each College Self-Efficacy item go down (although not at statistically significant levels). Similar to College Self-Efficacy scores for non-Native Hawaiian, traditional, and nontraditional students, many of the Sense of Belonging Inventory-Antecedent (SOBI-A) scores went down from pre to post-survey. The researcher hypothesizes, again, that this is because of the timing of when the post-test was administered (right before their final exam) and students had higher levels of anxiety and stress. It is also interesting that as a whole, traditional students scored themselves higher than their traditional counterparts on six out of eight scale items. For the 2015 MSB pre/post SOBI-P items, the students in all four groups identified (Non- Native Hawaiian v. Native Hawaiian and Traditional v. Nontraditional) had scores that went in the direction that would show improvement (e.g. scores went up for Item 4, “People accept me”, and down for Item 6, “I have no place in this world”). Besides this, there were also no identifiable trends or patterns. The 2013 and 2014 cohorts had the lowest scores, with higher ratings bookending in the 2015 and 2012 cohorts. It is unclear why this occurred, but it is hypothesized that it is the same reason why College Self-Efficacy Inventory items had the same patterns. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 126% For the next three questions, “Instructor Interaction In Class,” “Instructor Interaction Outside of Class,” and “School-Related Peer Interaction,” the total sample population (all respondents) and in both groupings (ethnicity or age), had reported increased interaction over time (immediately after intervention to 3 years post intervention). For “Instructor Interaction In Class,” non-Native Hawaiians and nontraditional students in the 2012 cohort (3 years post intervention) had 100% response rate saying that they interacted with their instructors 4+ times per week. High instructor interaction has been associated with other student success measures, such as higher cumulative GPA, persistence, and graduation (Astin, 1985; Lundberg, 2014; Thompson, 2001; Wirt & Jaeger, 2014). Although there was an overall pattern for peer interaction in school, the data showed a sharp increase in peer (academic) interaction from the pre- to post-survey (right after intervention), then scores dropped for the 2013 and 2014 cohorts and then increased again for the 2012 cohort. For the next related item, “Social Peer Interaction,” for every group except for nontraditional students, there was an increasing trend over time. Nontraditional students’ interaction broke the trend and decreased for the 2012 cohort. As mentioned previously, social peer interaction is not necessarily linked to student success measures and in fact can hinder student success by taking focus away from academics (Di Tommaso, 2012). Therefore, the researcher was not expecting this measure to significantly increase over time. The final three items measured were: (1) I am very satisfied with the academic instruction I’ve received at Windward Community College so far; (2) I am very satisfied with the student support services (e.g. academic advising, financial aid) I’ve received at Windward Community College so far; and (3) I am very satisfied with Windward Community College as a whole. For all three questions, the general trend was to have a sharp increase in scores right after the MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 127% intervention (2015 post-test) and then continuing to gradually increase. For Instructor Satisfaction, Native Hawaiian students and nontraditional students had higher satisfaction scores than their non-Native Hawaiian and traditional counterparts. For Support Services Satisfaction, 100% of the non-Native Hawaiian and nontraditional student responses reported they “strongly agree(d)” with that statement. Also, non-Native Hawaiian and nontraditional student scores were higher than their Native Hawaiian and traditional counterparts. Finally, for overall WCC satisfaction, nontraditional students’ satisfaction was higher than traditional students and 100% of the 2012 cohort reported that they “agree(d)” or “strongly agree(d)” with this statement. It is important to note that although there were dips in the College Self-Efficacy Inventory items and Sense of Belonging Inventory items, the majority of the satisfaction scores did not waiver. This is important to further investigate because although student’s self- confidence might have gone down the last day of the MSB program (when the post-test was administered), they still had confidence in the instructional and support faculty and staff. For each descriptive statistic, the data were analyzed by Native Hawaiian/Non-Native Hawaiian breakdowns. Independent t-tests were run for each and two data sets were found to be significant, CSE1 (Pre-survey) and SBI1 (Post-survey). Interestingly, the scores for the following questions were inversely significant. Native Hawaiian students (M = 7.79, SD = 2.645) scored significantly higher than Non-Native Hawaiian students (M = 6.49, SD = 3.061; t (97.293) = -2.143, p = .035) on CSE1 “Make new friends at college” and significantly lower on SBI1 “I wonder if I fit in college”; Native Hawaiian (M = 2.55, SD = 1.183); Non-Native Hawaiian (M = 3.18, SD = 1.956; t (79.380) = 2.203, p = .030). Table 42 shows each significant score. Descriptive statistics were also analyzed by age groups (traditional versus nontraditional). When using this lens for analysis, there were 21 significant independent t-tests found. Table 43 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 128% reflects this. On every significant item, nontraditional students had higher mean scores. While for most of these items, having a higher score is better, nontraditional students also had higher scores for two post-test items, SBI13 “I am not sure if I fit in with my student peers” and SBI15 “I describe myself as a misfit.” Even though nontraditional students made up almost half of the total MSB population (43.5%), these two scores align with the notion of nontraditional students feeling out of place in the college setting (Astin, 1985). Table 42 Significant t-test Results by Ethnicity (Non-Native Hawaiian v. Native Hawaiian) Variable Non-NH NH M (SD) M (SD) t p eta 2 CSE1-Make new friends at college (Pre- test)* 6.49 (3.061) 7.79 (2.645) -2.143 .035 .05 SBI1- I wonder if I really fit in college (Post-test)* 3.18 (1.956) 2.55 (1.183) 2.203 .030 .03 * p < .05. ** p < 0.01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 129% Table 43 Significant t-test Results by Age Group (Traditional v. Nontraditional) Variable Traditional Nontraditional M (SD) M (SD) t p eta 2 WCC Support Services Satisfaction (Pre-test)* 4.05 (1.200) 4.53 (.786) -2.274 .025 .06 CSE3- Talk to college staff/faculty (Pre-test)** 7.52 (2.416) 9.22 (1.295) -4.123 .000 .16 CSE4-Talk to college administration (Pre-test)** 7.41 (2.731) 8.91 (1.649) -3.133 .003 .10 CSE7-Participate in class discussions (Pre-test)** 7.39 (2.590) 8.76 (2.058) -2.758 .007 .08 CSE11-Talk to my professors outside of class (Pre-test)* 7.25 (2.738) 8.33 (2.185) -2.061 .043 .05 SBI8-I have qualities (Pre-test)* 3.18 (.540) 3.42 (.543) -2.093 .039 .05 SBI9-What I offer is valued by others at school (Pre-test)* 2.86 (.668) 3.16 (.562) -2.233 .028 .05 Instructor Interaction Outside of Class (Post-test)* 1.78 (.692) 2.01 (.690) -2.079 .039 .03 WCC Instructor Satisfaction (Post- test) 4.51 (.612) 4.73 (.626) -2.239 .027 .03 WCC Support Services Satisfaction (Post-test)* 4.55 (.658) 4.77 (.687) -1.986 .049 .03 * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 130% Table 43 Significant t-test Results by Age Group (Traditional v. Nontraditional) Cont. Variable Traditional Nontraditional M (SD) M (SD) t p eta 2 CSE1-Make new friends at college (Post-test)** 6.97 (2.912) 8.15 (2.679) -2.652 .009 .04 CSE3-Talk to college staff/faculty (Post-test)** 7.49 (2.726) 8.66 (2.183) -2.897 .004 .05 CSE4-Talk to college administration (Post-test)** 7.19 (2.814) 8.58 (2.019) -3.385 .001 .07 CSE5-Manage time effectively (Post-test)* 7.25 (2.263) 7.95 (2.090) -2.036 .043 .03 CSE7-Participate in class discussions (Post-test)* 7.99 (2.010) 8.69 (1.721) -2.343 .021 .03 CSE11-Talk to my professors outside of class (Post-test)* 7.46 (2.397) 8.33 (2.033) -2.453 .015 .04 CSE12-Ask a professor a question about math coursework (Post- test)* 7.99 (2.242) 8.68 (2.121) -1.995 .048 .02 CSE15-Seek help from my tutors and instructors to better understand math concepts (Post-test)* 8.19 (2.160) 8.98 (1.823) -2.438 .016 .04 SBI13-I am not sure if I fit in with my student peers (Post-test)** 2.90 (.808) 3.31 (.811) -3.159 .002 .06 SBI15-I describe myself as a misfit (Post-test)* 3.13 (.768) 3.39 (.798) -2.127 .035 .03 SBI16-It is important that I fit in at school (Post-test)* 3.15 (.725) 3.41 (.756) -2.208 .029 .03 * p < .05. ** p < .01. These items will be further analyzed using one-way analysis of variance (ANOVA) tests to decrease the risk of Type-1 error in the second half of this chapter. Since the sample size for this study is fairly large (e.g. 100 or more), the power of these tests are not in question and are “not an issue” (Stevens, 1996, p.6). Part 1 of this chapter, containing the descriptive analysis of the data collected is important because it lays the foundation for similarity between sample and total Windward Community College populations, which then elude to the effectiveness of the Math Summer Bridge intervention if brought to scale. It also describes the sample population so MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 131% others at different community colleges (or universities) could predict the effectiveness and alter the program to fit their demographics. The second reason for the first half of this chapter is to highlight the differences between non-Native Hawaiian and Hawaiian, and traditional and nontraditional populations; two key characteristics that are used when comparing student success. Finally, this section provided the foundation of the multivariate analyses that will be used to address the research questions of the study. Research Question 1 Results: College Self-Efficacy Research Question 1 in this study is a multi-part, multi-level question, asking: (1) If there was a significant change in college self-efficacy for Windward Community College Math Summer Bridge students; (2) If there was a difference for (a) Native Hawaiian v. non-Native Hawaiian students, or (b) nontraditional v. traditional students; and (3) for each of these groups, was the change long-lasting. It has already been established that there are significant changes for specific questions in this measure, however for the purposes of answering Research Question 1, a total score will be used in the analysis. The highest possible score is 190 (19 questions, 1 being “strongly disagree” and 10 being “strongly agree”). In order to address this research question and determine whether or not the null hypothesis can be rejected (that the WCC Math Summer Bridge Program does not have a positive effect on college self-efficacy), the researcher completed two different types of one-way analysis of variance (ANOVA) tests. The aim of these one-way ANOVAs were to “compare the variance (variability in scores) between the different groups [pre v. post 2015 survey and post 2015 v. post 2014 v. post 2013 v. post 2012 surveys] (believed to be due to the independent variable [WCC Math Summer Bridge Program]) with the variability within each of the groups (believed to be due to chance)” (Pallant, 2013, p. 258). The first type of one-way ANOVA conducted was a one-way repeated measures MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 132% ANOVA. By conducting repeated measures analysis of variance (ANOVA) test, the researcher was able to measure the change and effect size between the 2015 pre- and 2015 post-surveys (of the same participants before and after the intervention) to determine if there was a significant change in college self-efficacy immediately after the Windward Community College Math Summer Bridge Program. The second type of one-way ANOVA used was a between-groups ANOVA with post-hoc tests. This was used to determine if there were any significant differences between the mean scores of each of the cohorts (2015, 2014, 2013, and 2012). For both types of tests, two more were conducted to compare the same scores with specific attention to ethnicity (non-Native Hawaiian v. Native Hawaiian) and age groups (traditional v. nontraditional). For the purposes of this section, data will be presented in sections by grouping of student population; (1) total student population; (2) ethnicity (non-Native Hawaiian v. Native Hawaiian); and (3) age (traditional v. nontraditional). Total Group Analysis: College Self-Efficacy First, to answer the question, “was there a significant change in college self-efficacy immediately after the Math Summer Bridge Program?” a one-way repeated measures ANOVA was conducted to compare scores on the College Self-Efficacy Scale (total score) prior to the intervention (Time 1/pre-survey) and immediately after the intervention (Time 2/post-survey). The means and standard deviations are presented in Table 44. There was a significant effect for time, Wilks’ Lambda = .927, F (1, 67) = 5.314, p = .024, multivariate partial eta squared = .073. Therefore, the null hypothesis (the Math Summer Bridge does not positively impact students’ college self-efficacy) is rejected; the Math Summer Bridge Program significantly increased participants’ college self-efficacy with a moderate effect size. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 133% Table 44 Descriptive Statistics for College Self-Efficacy Scores for Time 1 and Time 2-Total Time Period N Mean Standard Deviation Time 1 (Pre-Survey) 68 153.25 33.69 Time 2 (Post-Survey) 68 161.13 32.96 Then, to establish whether or not this change was long-lasting, a one-way between- groups ANOVA with post-hoc tests was conducted to explore the impact of time on students’ college self-efficacy after participating in the Windward Community College Math Summer Bridge Program as measured by the College Self-Efficacy (CSE) score. The means and standard deviations are presented in Table 45. A Levene’s test for homogeneity of variances was not significant, and therefore, the assumption of homogeneity of variance was not violated; the variance in scores is the same for each of the four groups. Responses were separated into four cohorts according to the year that they participated in the Math Summer Bridge Program (2012, 2013, 2014, and 2015). There were no statistically significant difference at the p < .05 level in CSE scores for the four groups: F (3, 145) = 1.15, p = .331. Therefore, the null hypothesis (that the changes incurred after the Math Summer Bridge intervention would not be maintained over time) was rejected because the increase in CSE scores between pre- and post-survey was sustained over time. Table 45 Descriptive Statistics for College Self-Efficacy Scores for Post MSB 2015, 2014, 2013, 2012 Time Period N Mean Standard Deviation 2015 Post-Survey 68 161.13 32.96 2014 Post-Survey 28 147.57 33.38 2013 Post-Survey 37 152.38 32.14 2012 Post-Survey 16 148.19 42.58 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 134% Ethnicity Analysis (Non-Native Hawaiian v. Native Hawaiian): College Self-Efficacy The second part of this question asks the same question, “was there a significant change in college self-efficacy immediately after the Math Summer Bridge Program?” and adds, “is there a difference in significance based on ethnicity (non-Native Hawaiian v. Native Hawaiian)?” A one-way repeated measures ANOVA was conducted to compare scores on the College Self-Efficacy Scale at Time 1 (pre-survey) and Time 2 (post-survey). The means and standard deviations are presented in Table 46. There was no significant effect when separating scores by ethnicity (non-Native Hawaiian v. Native Hawaiian), non-Native Hawaiian, Wilks’ Lambda = .904, F (1, 25) = 2.650, p = .116, multivariate partial eta squared = .096; Native Hawaiian, Wilks’ Lambda = .937, F (1, 41) = 2.768, p = .104, multivariate partial eta squared = .063. Therefore, the null hypothesis (the Math Summer Bridge Program does not have a significantly positive effect on college self-efficacy for non-Native Hawaiian students; the Math Summer Bridge Program does not have a significantly positive effect on college self-efficacy for Native Hawaiian students) was accepted; the WCC Math Summer Bridge Program did not have a statistically significant effect on college self-efficacy when separating students by ethnicity (non- Native Hawaiian v. Native Hawaiian). However, although not significant, the mean scores did increase after the WCC Math Summer Bridge Program for both non-Native Hawaiians and Native Hawaiians. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 135% Table 46 Descriptive Statistics for College Self-Efficacy Scores for Time 1 and Time 2-Non-Native Hawaiian v. Native Hawaiian Time Period N Mean Standard Deviation Non-Native Hawaiian Time 1 (Pre-Survey) 26 154.04 32.64 Time 2 (Post-Survey) 26 162.23 37.15 Native Hawaiian Time 1 (Pre-Survey) 42 152.76 34.71 Time 2 (Post-Survey) 42 160.45 30.53 A one-way between-groups ANOVA with post-hoc tests was completed to determine whether or not there were any significant changes in college self-efficacy over time. Responses were categorized into four groups according to the year of participation in the WCC Math Summer Bridge Program (2012, 2013, 2014, and 2015). The means and standard deviations are presented in Table 47. For both non-Native Hawaiian and Native Hawaiian groups, a Levene’s test of homogeneity of variances was not significant, indicating that the assumption that the variance in scores for each of the four groups has not been violated. Also for both ethnicity groups, there was no statistically significant change at the p < .05 level in CSE scores for the four groups: Non-Native Hawaiian, F (3, 46) = 1.22, p = .315; Native Hawaiian, F (3, 95) = 1.35, p = .263. Therefore, the null hypothesis (the change in college self-efficacy incurred after the Math Summer Bridge Program is not maintained over time for non-Native Hawaiian students; the change in college self-efficacy incurred after the Math Summer Bridge Program is not maintained over time for Native Hawaiian students) is rejected because the (lack of) change incurred between pre- and post-test was indeed maintained over time. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 136% Table 47 Descriptive Statistics for College Self-Efficacy Scores for Post MSB 2015, 2014, 2013, 2012- Non-Native Hawaiian v. Native Hawaiian Time period N Mean Standard Deviation Non-Native Hawaiian 2015 Post-Survey 26 161.69 37.75 2014 Post-Survey 8 153.00 26.22 2013 Post-Survey 12 143.50 28.73 2012 Post-Survey 4 173.50 14.62 Native Hawaiian 2015 Post-Survey 42 158.67 31.66 2014 Post-Survey 20 145.40 36.23 2013 Post-Survey 25 156.64 33.36 2012 Post-Survey 12 139.75 45.86 Age Group Analysis (Traditional v. Nontraditional): College Self-Efficacy In the third approach to answer the question, “was there a significant change in college self-efficacy immediately after the Math Summer Bridge Program?” a one-way repeated measures ANOVA was conducted to compare scores on the College Self-Efficacy Scale (total score) prior to the intervention (Time 1/pre-survey) and immediately after the intervention (Time 2/post-survey) with specific attention to traditional and nontraditional age groups. The means and standard deviations are presented in Table 48. For traditional students, there was a significant effect for time, Wilks’ Lambda = .868, F (1, 33) = 5.005, p = .032, multivariate partial eta squared = .132. For nontraditional students, there was no significant effect for time, Wilks’ Lambda = .982, F (1, 33) = .617, p = .438. Therefore, for traditional students, the null hypothesis (the Math Summer Bridge Program does not have a significantly positive effect on college self-efficacy for traditional students) is rejected; the Math Summer Bridge Program significantly increased participants’ college self-efficacy with a moderately large effect size. However, the null hypothesis (the Math Summer Bridge Program does not have a significantly MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 137% positive effect on college self-efficacy for nontraditional students) was accepted for nontraditional students. It is important to note that for both traditional and nontraditional students, college self-efficacy scores increased post intervention. Table 48 Descriptive Statistics for College Self-Efficacy Scores for Time 1 and Time 2-Traditional v. Nontraditional Time Period N Mean Standard Deviation Traditional Time 1 (Pre-Survey) 34 143.50 36.90 Time 2 (Post-Survey) 34 156.59 42.10 Nontraditional Time 1 (Pre-Survey) 34 163.00 27.34 Time 2 (Post-Survey) 34 165.68 19.77 Finally, to establish whether or not this change was long-lasting (or if there were any significant change later in time), a one-way between-groups ANOVA with post-hoc tests was conducted to explore the impact of time on students’ college self-efficacy after participating in the Windward Community College Math Summer Bridge Program as measured by the College Self-Efficacy (CSE) score, with age grouping. The means and standard deviations are presented in Table 49. A Levene’s test for homogeneity of variances was not significant, and therefore, the assumption of homogeneity of variance was not violated; the variance in scores is the same for each of the four groups. Responses were separated into four cohorts according to the year that they participated in the Math Summer Bridge program (2012, 2013, 2014, and 2015). There were no statistically significant difference at the p < .05 level in CSE scores for each of the four cohorts: Traditional, F (3, 62) = 1.13, p = .343; Nontraditional, F (3, 79) = .522, p = .668. Therefore, the null hypothesis (the change in college self-efficacy incurred after the Math Summer Bridge Program is not maintained over time for traditional students; the change in MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 138% college self-efficacy incurred after the Math Summer Bridge Program is not maintained over time for nontraditional students) was rejected for both groups; significant increases in college self-efficacy were maintained for traditional students over time and lack of significant increases in college self-efficacy was sustained for nontraditional students. Table 49 Descriptive Statistics for College Self-Efficacy Scores for Post MSB 2015, 2014, 2013, 2012 – Traditional v. Nontraditional Time Period N Mean Standard Deviation Traditional 2015 Post-Survey 33 157.42 42.46 2014 Post-Survey 9 134.00 32.05 2013 Post-Survey 14 141.50 34.26 2012 Post-Survey 10 147.90 38.03 Nontraditional 2015 Post-Survey 35 162.09 23.51 2014 Post-Survey 19 154.00 32.84 2013 Post-Survey 23 159.00 29.59 2012 Post-Survey 6 148.67 53.24 Research Question 1 Summary Research Question 1 aimed to understand if the Windward Community College Math Summer Bridge Program had any significant impact on participants’ college self-efficacy, and if there was a change, how sustainable this change is. This was measured by a 19-question Likert scale inventory using pre- and post-surveys as well as a cross sectional data collection using students from four different years (2012, 2013, 2014, and 2015). Overall, there was a significant increase in college self-efficacy after the intervention, and this increase was maintained throughout the three years post-intervention that were measured. When distinguishing groups based on ethnicity however, there were no statistically significant increases for either non-Native Hawaiian or Native Hawaiian students. Finally, when grouping students by age, there was a MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 139% significant increase in college self-efficacy for traditional aged students, but not for nontraditional students. The increase gained for traditional students was maintained over the three years measured for this study. Research Question 2 Results: Sense of Belonging Similar to Research Question 1, Research Question 2 is a multi-part, multi-level question, asking: (1) “if there was a significant change in sense of belonging for Windward Community College Math Summer Bridge students”; (2) if there was a difference for (a) Native Hawaiian (v. non-Native Hawaiian) students, or (b) nontraditional (v. traditional) students; and (3) for each of these groups, was the change long-lasting. As previously described, the Sense of Belonging Inventory administered to the students was an 18-point scale, which was broken into two separate scales for analysis. The first, Sense of Belonging-Antecedents (SOBI-A), measured the precursors of sense of belonging. The second, Sense of Belonging-Psychology (SOBI-P), measured the psychological state of sense of belonging. As noted in the first half of this chapter, eight (8) out of eighteen (18) questions were reverse coded for total score analysis. During the first section of this chapter, very minimal significance was found when analyzing each sense of belonging item for both SOBI-A and SOBI-P scales for all approaches (total, ethnicity, and age group, for pre-, post-test, and 2012-2015 cohorts). Therefore, it was not hypothesized that significance would be found when analyzing each total scale. In order to address this research question and determine whether or not we can accept or reject the null hypothesis (the Math Summer Bridge Program does not have a significantly positive effect on sense of belonging), the researcher completed two different types of one-way analysis of variance (ANOVA) tests. The aim of these one-way ANOVAs were to “compare the variance (variability in scores) between the different groups [pre v. post 2015 surveys and post 2015 v. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 140% post 2014 v. post 2013 v. post 2012 surveys] (believed to be due to the intervention [WCC Math Summer Bridge Program]) with the variability within each of the groups (believed to be due to chance)” (Pallant, 2013, p. 258). The first type of one-way ANOVA conducted was a one-way repeated measures ANOVA. By conducting repeated measures analysis of variance (ANOVA) test, the researcher was able to measure the change and effect size between the 2015 pre- and 2015 post-test (of the same participants before and after the intervention) to determine if there was a significant change in sense of belonging immediately after the Windward Community College Math Summer Bridge Program. The second type of one-way ANOVA used was a between-groups ANOVA with post-hoc tests. This was used to determine if there were any significant differences between the mean scores of each of the previous cohorts (2015, 2014, 2013, and 2012). For both types of tests, two more were conducted to compare the same scores with specific attention to ethnicity (non-Native Hawaiian v. Native Hawaiian) and age groups (traditional v. nontraditional). Total Group Analysis: Sense of Belonging For the purposes of this section, data will be presented in sections by grouping of student population: (1) total student population; (2) ethnicity (non-Native Hawaiian v. Native Hawaiian); and (3) age (traditional v. nontraditional). First, to answer the question, “was there a significant change in sense of belonging immediately after the Math Summer Bridge Program?” a one-way repeated measures ANOVA was conducted to compare scores on both the SOBI-A and SOBI-P prior to the intervention (Time 1/pre-survey) and immediately after the intervention (Time 2/post-survey). The means and standard deviations are presented in Table 50. There were no significant effects for time for both SOBI-A and SOBI-P scores; SOBI-A, Wilks’ Lambda = .997, F (1, 66) = .213, p = .646, multivariate partial eta squared = .003; SOBI-P, Wilks’ Lambda MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 141% = .980, F (1, 66) = 1.370, p = .246, multivariate partial eta squared = .020. Therefore, the null hypothesis (the Math Summer Bridge Program does not have a significantly positive effect on sense of belonging) was accepted; the WCC Math Summer Bridge Program did not have any significant effect on student’s sense of belonging (SOBI-A or SOBI-P). Table 50 Descriptive Statistics for Sense of Belonging (SOBI-A and SOBI-P) Scores for Time 1 and Time 2-Total Time period N Mean Standard deviation SOBI-A Time 1 (Pre-test) 67 23.94 3.88 Time 2 (Post-test) 67 23.75 4.04 SOBI-P Time 1 (Pre-test) 67 30.04 5.13 Time 2 (Post-test) 67 30.82 5.33 Then, to answer the second part of this question, whether or not this change was long- lasting, a one-way between-groups ANOVA with post-hoc tests was conducted to explore the impact of time on students’ sense of belonging after participating in the Windward Community College Math Summer Bridge Program as measured by the two sense of belonging scores (SOBI-A and SOBI-P). A Levene’s test for homogeneity of variances was not significant for either scale, and therefore, the assumption of homogeneity of variance was not violated and the variance in scores is the same for each of the four groups. Responses were separated into four groups according to the year that they participated in the Math Summer Bridge program (2012, 2013, 2014, and 2015). There were no significant difference at the p < .05 level in either SOBI- A or SOBI-P scores for the four groups: SOBI-A, F (3, 146) = .222, p = .881; SOBI-P, F (3, 146) = 1.097, p = .352. Therefore, the null hypothesis (the (lack of) change in sense of belonging MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 142% incurred after the Math Summer Bridge Program is not maintained over time) for was rejected since the lack of change for SOBI-A and SOBI-P scales was maintained over time. Ethnicity Analysis (Non-Native Hawaiian v. Native Hawaiian): Sense of Belonging The second part of this question asks the same question, “was there a significant change in sense of belonging immediately after the Math Summer Bridge Program?” and adds, “is there a significant difference based on ethnicity (non-Native Hawaiian v. Native Hawaiian)?” A one- way repeated measures ANOVA was conducted to compare scores on the Sense of Belonging Scales (SOBI-A and SOBI-P) at Time 1 (pre-survey) and Time 2 (post-survey). The means and standard deviations are presented in Table 51. There was no significant increase for SOBI-A scale scores when separating scores by ethnicity (non-Native Hawaiian v. Native Hawaiian); non-Native Hawaiian, Wilks’ Lambda = .944, F (1, 25) = 1.472, p = .236; Native Hawaiian, Wilks’ Lambda = .997, F (1, 40) = .131, p = .719. Therefore, the null hypothesis (the Math Summer Bridge Program does not have a significantly positive effect on sense of belonging- antecedents for non-Native Hawaiian students; the Math Summer Bridge Program does not have a significantly positive effect on sense of belonging-antecedents for Native Hawaiian students) was accepted; the WCC Math Summer Bridge Program did not have a statistically significant effect on sense of belonging when separating students by ethnicity (non-Native Hawaiian v. Native Hawaiian). Similar results were found for SOBI-P pre/post tests. There was also no significant effect of time on SOBI-P scores when grouping by ethnicity; non-Native Hawaiian, Wilks’ Lambda = .976, F (1, 25) = .610, p = .442; Native Hawaiian, Wilks’ Lambda = .982, F (1, 40) = .739, p = .395. Therefore, the null hypothesis (the Math Summer Bridge Program does not have a significantly positive effect on sense of belonging-psychology for non-Native Hawaiian students; MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 143% the Math Summer Bridge Program does not have a significantly positive effect on sense of belonging-psychology for Native Hawaiian students) was accepted; the WCC Math Summer Bridge Program did not have a statistically significant effect on sense of belonging when separating students by ethnicity (non-Native Hawaiian v. Native Hawaiian). Table 51 Descriptive Statistics for Sense of Belonging (SOBI-A and SOBI-P) Scores for Time 1 and Time 2- Non-Native Hawaiian v. Native Hawaiian Time period N Mean Standard deviation Non-Native Hawaiian SOBI-A Time 1 (Pre-survey) 26 24.69 3.74 Time 2 (Post-survey) 26 23.88 4.04 SOBI-P Time 1 (Pre-survey) 26 29.35 5.34 Time 2 (Post-survey) 26 30.27 5.51 Native Hawaiian SOBI-A Time 1 (Pre-survey) 41 23.46 3.94 Time 2 (Post-survey) 41 23.66 4.10 SOBI-P Time 1 (Pre-survey) 41 30.49 5.01 Time 2 (Post-survey) 41 31.17 5.24 Although there was no significant increase in sense of belonging scores immediately after the intervention, a one-way between-groups ANOVA with post-hoc tests was completed to determine whether or not there were any significant changes in sense of belonging over time. Responses were categorized into four cohorts according to the year of participation in the WCC Math Summer Bridge Program (2012, 2013, 2014, and 2015). For both non-Native Hawaiian and Native Hawaiian groups, a Levene’s test of homogeneity of variances was not significant, indicating that the assumption of variance in scores for each of the four cohorts has not been MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 144% violated. For both ethnicity groups, there was no statistically significant at the p < .05 level in SOBI-A scores for the four cohorts: Non-Native Hawaiian, F (3, 46) = .077, p = .972; Native Hawaiian, F (3, 96) = .182, p = .909. There were also no statistically significant changes in means for SOBI-P scores for the four groups: Non-Native Hawaiian, F (3, 46) = .238, p = .870; Native Hawaiian, F (3, 96) = .802, p = .496. Therefore, the null hypothesis [(a) the (lack of) change in sense of belonging-antecedents incurred after the Math Summer Bridge Program is not maintained over time for non-Native Hawaiian students; (b) the (lack of) change in sense of belonging-antecedents incurred after the Math Summer Bridge Program is not maintained over time for Native Hawaiian students; (c) the (lack of) change in sense of belonging-psychological incurred after the Math Summer Bridge Program is not maintained over time for non-Native Hawaiian students; (d) the (lack of) change in sense of belonging-psychological incurred after the Math Summer Bridge Program is not maintained over time for Native Hawaiian students] is rejected because the (lack of) change incurred between pre- and post-test was sustained. Age Group Analysis (Traditional v. Nontraditional): Sense of Belonging In the third approach to answer the question, “was there a significant change in sense of belonging immediately after the Math Summer Bridge Program?” a one-way repeated measures ANOVA was conducted to compare scores on the Sense of Belonging Scales (SOBI-A and SOBI-P) prior to the intervention (Time 1/pre-survey) and immediately after the intervention (Time 2/post-survey) with specific attention to traditional and nontraditional age groups. The means and standard deviations are presented in Table 52. No significant changes were found for either traditional or nontraditional students’ SOBI-A pre- to post-test scores; Traditional, Wilks’ Lambda = 1.000, F (1, 33) = .000, p = 1.00; Nontraditional, Wilks’ Lambda = .989, F (1, 32) = .343, p = .011. No significant changes were found for both SOBI-P scores as well; Traditional, MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 145% Wilks’ Lambda = .993, F (1, 33) = .219, p = .643; Nontraditional, Wilks’ Lambda = .965, F (1, 32) = 1.144, p = .293. Therefore, for both traditional and nontraditional students, the null hypothesis [(a) the Math Summer Bridge Program does not have a significantly positive effect on sense of belonging-antecedents for traditional students; (b) the Math Summer Bridge Program does not have a significantly positive effect on sense of belonging-antecedents for nontraditional students; (c) the Math Summer Bridge Program does not have a significantly positive effect on sense of belonging-psychology for traditional students; (d) the Math Summer Bridge Program does not have a significantly positive effect on sense of belonging-psychology for nontraditional students) is accepted; the Math Summer Bridge Program did not significantly increase participants’ sense of belonging. It is important to note that for both traditional and nontraditional students, college self-efficacy scores increased post intervention. Table 52 Descriptive Statistics for Sense of Belonging (SOBI-A and SOBI-P) for Time 1 and Time 2- Traditional v. Nontraditional Time period N Mean Standard deviation Traditional SOBI-A Time 1 (Pre-survey) 34 23.47 3.18 Time 2 (Post-survey) 34 23.47 3.65 SOBI-P Time 1 (Pre-survey) 34 30.15 4.91 Time 2 (Post-survey) 34 30.41 4.86 Nontraditional SOBI-A Time 1 (Pre-survey) 33 24.42 4.49 Time 2 (Post-survey) 33 24.03 4.46 SOBI-P Time 1 (Pre-survey) 33 29.94 5.42 Time 2 (Post-survey) 33 31.24 5.82 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 146% Finally, to establish whether or not there were any significant changes later in time, a one-way between-groups ANOVA with post-hoc tests was conducted to explore the impact of time on students’ sense of belonging after participating in the Windward Community College Math Summer Bridge Program as measured by the Sense of Belonging Scores (SOBI-A and SOBI-P), with age grouping. A Levene’s test for homogeneity of variances was not significant, and therefore, the assumption of homogeneity of variance was not violated for all groups except Traditional SOBI-P scores; meaning that the variance in scores is the same for each of the four groups except the one. Responses were separated into four groups according to the year that they participated in the Math Summer Bridge program (2012, 2013, 2014, and 2015). There were no statistically significant difference at the p < .05 level in SOBI-A and SOBI-P scores for the four groups: Traditional SOBI-A, F (3, 62) = .089, p = .966; Traditional SOBI-P, F (3, 62) = 2.146, p = .103; Nontraditional SOBI-A, F (3, 79) = .226, p = .878; Nontraditional SOBI-P, F (3, 79) = 1.282, p = .286. Therefore, the null hypothesis [(a) the (lack of) change in sense of belonging-antecedents incurred after the Math Summer Bridge Program is not maintained over time for traditional students; (b) the (lack of) change in sense of belonging-antecedents incurred after the Math Summer Bridge Program is not maintained over time for nontraditional students; (c) the (lack of) change in sense of belonging-psychology incurred after the Math Summer Bridge Program is not maintained over time for traditional students; (d) the (lack of) change in sense of belonging-psychology incurred after the Math Summer Bridge Program is not maintained over time for nontraditional students] was rejected for both groups; the lack of significant increase in sense of belonging was maintained for traditional and nontraditional students over time. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 147% Research Question 2 Summary The goal of Research Question 2 was to understand if the WCC Math Summer Bridge had an impact on students’ sense of belonging. An 18-question survey was used to measure sense of belonging. For analysis purposes, the questions were split into two different scales: SOBI-Antecedents (SOBI-A), and SOBI-Psychology (SOBI-P). When analyzing student responses, it does not seem like the Math Summer Bridge program had any effect on sense of belonging, antecedents or psychological tendencies. There are a few theories as to why the null hypothesis was accepted for this research question [(a) the Math Summer Bridge Program does not have a significantly positive effect on sense of belonging-antecedents or sense of belonging- psychology for the total group; (b) for non-Native Hawaiian students; (c) for Native Hawaiian students; (d) for traditional students; and (e) for nontraditional students] and it will be discussed in Chapter 5. Research Question 3 Results: Student Engagement and Satisfaction Going beyond the metacognitive (college self-efficacy and sense of belonging), Research Question 3 aims to measure students actions and satisfaction of others asking, “After participating in the Windward Community College Math Summer Bridge Program, is there a significant change in student engagement and satisfaction?” Similar to Research Question 1 and 2, this question is a multi-part, multi-level question, asking not only if: (1) “there was a significant change in engagement and satisfaction for Windward Community College Math Summer Bridge students”; but also (2) if there was a difference for (a) Native Hawaiian (v. non- Native Hawaiian) students; or (b) nontraditional (v. traditional) students; and (3) for each of these groups, was the change long-lasting. Student engagement and satisfaction were measured by seven multiple choice survey questions. Four (4) questions measured student engagement MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 148% (with both instructors and peers) and three (3) questions measured student satisfaction (instructor, support services, and overall Windward Community College). For this part of the chapter, the researcher will analyze each individual question, as well as total engagement (instructor and peer) and total satisfaction (instructor, support, and overall WCC). For total engagement, the highest possible score is 14. Table 53 lists each question, what it’s measuring, coding/answers, and highest possible score. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 149% Table 53 Engagement and Satisfaction Questions with Coding Coding/Possible Answers Highest Possible Score Question Category 1 2 3 4 5 How many times do you talk to/interact with your instructor during class? Engagement 0 Times/ week 1-3 Times/ week 4+ Times/ week N/A N/A 3 How many times do you talk to/interact with your instructor outside of class (e.g. office hours)? Engagement 0 Times/ week 1-3 Times/ week 4+ Times/ week N/A N/A 3 On average, how many times do you interact with your peers at school for social purposes? Engagement 0 Times/ week 1-3 Times/ week 4-6 Times/ week 7+ Times/ week N/A 4 On average, how many times do you interact with your peers at school for academic purposes (e.g. study groups, work on projects)? Engagement 0 Times/ week 1-3 Times/ week 4-6 Times/ week 7+ Times/ week N/A 4 I am very satisfied with the academic instruction I’ve received at Windward Community College. Satisfaction Strongly disagree Disagree Neutral Agree Strongly agree 5 I am very satisfied with the student support services (e.g. academic advising, financial aid office) I’ve received at Windward Community College. Satisfaction Strongly disagree Disagree Neutral Agree Strongly agree 5 I am very satisfied with Windward Community College as a whole. Satisfaction Strongly disagree Disagree Neutral Agree Strongly agree 5 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 150% During the first section of this chapter, most of these questions provided significant t-test scores. Therefore, it was hypothesized that significance would be found when performing further analysis in individual and total scores. In order to address this research question and determine whether or not we can reject the null hypothesis (that the WCC Math Summer Bridge Program does not have any effect on student engagement and satisfaction), the researcher completed two different types of one-way analysis of variance (ANOVA) tests. The aim of these one-way ANOVAs were to “compare the variance (variability in scores) between the different groups [pre v. post 2015 surveys and post 2015 v. post 2014 v. post 2013 v. post 2012 surveys] (believed to be due to the independent variable [WCC Math Summer Bridge Program]) with the variability within each of the groups (believed to be due to chance)” (Pallant, 2013, p. 258). The first type of one-way ANOVA conducted was a one-way repeated measures ANOVA. By conducting repeated measures analysis of variance (ANOVA) test, the researcher was able to measure the change and effect size between the 2015 pre- and 2015 post-survey (of the same participants before and after the intervention) to determine if there was a significant change in engagement or satisfaction immediately after the Windward Community College Math Summer Bridge Program. The second type of one-way ANOVA used was a between-groups ANOVA with post-hoc tests. This was used to determine if there were any significant differences between the mean scores of each of the previous cohorts (2015, 2014, 2013, and 2012). For both types of tests, two more were conducted to compare the same scores with specific attention to ethnicity (non-Native Hawaiian v. Native Hawaiian) and age groups (traditional v. nontraditional). MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 151% Total Group Analysis For the purposes of this section, data will be presented in sections by grouping of student population: (1) total student population; (2) ethnicity (non-Native Hawaiian v. Native Hawaiian); and (3) age (traditional v. nontraditional), followed by type of dependent variable (engagement or satisfaction). Total group analysis: Student engagement pre/post MSB. There were five different one-way repeated measures tests performed to compare the means of pre-test (Time 1) and post- test (Time 2) answers for each individual student engagement question, as well as one with the total score. Table 54 displays the means and standard deviations for each. When analyzing each individual question, although all post-test scores were equal to or higher than their pre-test counterparts, three of them were found insignificant (Instructor Interaction In Class, Wilks’ Lambda = .983, F (1, 67) = 1.142, p = .289; Instructor Interaction Outside of Class, 1.00, F (1, 67) = .000, p = 1.000; Peer Interaction-Social, Wilks’ Lambda = .976, F (1, 67) = 1.628, p = .206). However, for academic peer interaction significantly increased from pre- to post-test (Wilks’ Lambda = .898, F (1, 67) = 7.584, p = .008, multivariate partial eta squared = .102. The WCC Math Summer Bridge Program significantly increased students’ interaction with their peers for academic reasons (e.g. studying for a test, group projects); the null hypothesis was rejected. Also, when looking at total engagement, students’ interaction with both instructors and peers significantly increased (Wilks’ Lambda = .935, F (1, 67) = 4.674, p = .034, multivariate partial eta squared = .065). Therefore, when measuring (total group) student engagement, the null hypothesis (the Math Summer Bridge Program does not have a significantly positive effect on student engagement) was rejected. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 152% Table 54 Descriptive Statistics Student Engagement Scores for Time 1 and Time 2-Total Time period N Mean Highest Possible Score Standard deviation Instructor Interaction In Class Time 1 (Pre-test) 68 2.25 3 .677 Time 2 (Post-test) 68 2.35 3 .512 Outside of Class Time 1 (Pre-test) 68 1.71 3 .575 Time 2 (Post-test) 68 1.71 3 .600 Peer Interaction Academic** Time 1 (Pre-test) 68 2.13 4 .845 Time 2 (Post-test) 68 2.46 4 .762 Social Time 1 (Pre-test) 68 2.16 4 .940 Time 2 (Post-test) 68 2.31 4 .885 Total Interaction* Time 1 (Pre-test) 68 8.25 14 2.14 Time 2 (Post-test) 68 8.82 14 1.86 * p < .05. ** p < .01. Total group analysis: Student engagement long-term effects for MSB. To answer the second part of this question, whether or not this change was long-lasting, a one-way between- groups ANOVA with post-hoc tests was conducted to explore the impact of time on students’ engagement after participating in the Windward Community College Math Summer Bridge Program. A Levene’s test for homogeneity of variances was found to be significant for two of the five measures (Instructor Interaction In Class, and Academic Peer Interaction); therefore significance for these two measures was determined using Welch and Brown-Forsythe tests (Robust Tests of Equality of Means). Responses were separated into four groups according to the year that they participated in the Math Summer Bridge program (2012, 2013, 2014, and 2015). Two measures were found to have significant changes over time, two were insignificant, MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 153% and a third was approaching significance with p = .051. Table 55 lists the F scores, significance, and effect size of each question and Table 56 presents the significant differences between cohorts for each question. After small increases between the pre- and post-survey, “Instructor Interaction In Class” substantially increased over time: F (3, 146) = 6.84, p = .000. The significant increase in interaction occurred first between Cohort 2015 (M = 2.35, SD = .512) and Cohort 2014 (M = 2.75, SD = .441), and then overall, between the most recent score from Cohort 2015 (M = 2.35, SD = .512) and three years post intervention for Cohort 2012 (M = 2.88, SD = .342). The difference between Cohort 2015 and Cohort 2012 had a moderately large effect size of .12. Therefore, the students’ frequency of instructor interaction in class not only maintained, but also surpassed the level it was at right after the intervention. The other significant overall difference occurred for “Instructor Interaction Outside of Class”: F (3, 146) = 5.87, p = .001. Similar to “Instructor Interaction In Class”, the significant increase in interaction occurred with a spike first between Summer (Cohort) 2015 (M = 1.71, SD = .600) and Summer (Cohort) 2014 (M = 2.14, SD = .756), and then overall, between the most recent score from Summer (Cohort) 2015 (M = 1.71, SD = .600) and Summer (Cohort) 2012 (M = 2.38, SD = .719). The difference between Cohort 2015 and Cohort 2012 had a moderate effect size of .10. Once again, students’ interaction did not just maintain the level achieved right after the intervention, but it actually significantly increased three years post-intervention. “Peer Interaction for Academic Purposes” was one of two questions that did not have statistically significant differences between means at the p < .05 level over time: F (3, 146) = .401, p = .687 (Cohort 2015, M = 2.44, SD = .761; Cohort 2014, M = 2.29, SD = .535; Cohort 2013, M = 2.32, SD = .873; Cohort 2012, M = 2.44, SD = .892). “Social Peer Interaction” was MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 154% the other question that did not have statistically significant differences between means at the p < .05 level over time; F (3, 146) = .148, p = .931 (2015 Cohort, M = 2.29, SD = .899; 2014 Cohort, M = 2.36, SD = .731; 2013 Cohort, M =2.29, SD = .898; 2012 Cohort, M =2.44, SD = .964). Therefore, the interaction level acquired immediately after the intervention (Post-test 2015) has been maintained over the following three years. Finally, for “Total Engagement”, the differences between means approached significance: F (3, 146) = 2.65, p = .051 (Cohort 2015, M = 8.79, SD = 1.87; Cohort 2014, M = 9.54, SD = 1.48; Cohort 2013, M = 9.11, SD = 2.13; Cohort 2012, M = 10.13, SD = 1.89). There was a small effect size for this question (eta squared = .052). Therefore, the null hypothesis (that the change in student engagement incurred after the Math Summer Bridge Program is not maintained over time) is rejected. Table 55 One-Way Between Groups ANOVA Student Engagement Scores for 2015, 2014, 2013, and 2012 Cohorts Time period F Sig. Eta Squared Instructor Interaction- In Class 6.838 .000 .123 Instructor Interaction- Outside of Class 5.865 .001 .108 Peer Interaction- School .401 .687 - Peer Interaction- Social .148 .931 - Total Engagement 2.650 .051 .052 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 155% Table 56 One-Way Between Groups ANOVA Significant Differences-Student Engagement Scores for 2015, 2014, 2013, and 2012 Cohorts Time period Mean Difference Sig. Instructor Interaction- In Class 2015 to 2014 .397 .005 2015 to 2012 .522 .002 Instructor Interaction- Outside of Class 2015 to 2014 .437 .021 2015 to 2012 .669 .002 Peer Interaction- School - - Peer Interaction- Social - - Total Engagement 2015 to 2012 1.33 .057 Total group analysis: Student satisfaction pre/post MSB. The were four different one- way repeated measures tests performed to compare the means of Time 1 (pre-survey) and Time 2 (post-survey) answers for each question for student satisfaction score, as well as a total satisfaction score. Table 57 displays the means and standard deviations for each. Overall WCC satisfaction did have an increase in means after the intervention, although, it was not statistically significant, Wilks’ Lambda = .970, F (1, 67) = 2.090, p = .153. The other three satisfaction measures had statistically significant improvements after the WCC Math Summer Bridge Program. The instructor satisfaction question had a significant increase and had a moderately large effect size from Time 1 (pre-survey) to Time 2 (post-survey) responses, Wilks’ Lambda = .884, F (1, 67) = 8.801, p = .004, multivariate partial eta squared = .116. Support services satisfaction also had a significant increase from pre-2015 (Time 1) to post-2015 (Time 2) responses with a moderately large effect size, Wilks’ Lambda = .874, F (1, 67) = 9.651, p = .003, multivariate partial eta squared = .126. Finally, Total Satisfaction (combination of all three satisfaction scores) had a significant improvement with a moderate MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 156% effect size, Wilks’ Lambda = .904, F (1, 67) = 7.107, p = .010, multivariate partial eta squared = .096. Therefore, since there was a significantly positive increase in student satisfaction, the null hypothesis (the Math Summer Bridge Program does not have a significantly positive effect on student satisfaction) is rejected. Table 57 Descriptive Statistics Student Satisfaction Scores for Time 1 and Time 2-Total Time period N Mean Highest Possible Score Standard deviation Instructor Satisfaction Time 1 (Pre-test) 68 4.26 5 .891 Time 2 (Post-test) 68 4.60 5 .522 Support Services Satisfaction Time 1 (Pre-test) 68 4.29 5 .899 Time 2 (Post-test) 68 4.66 5 .637 Overall WCC Satisfaction Time 1 (Pre-test) 68 4.44 5 .887 Time 2 (Post-test) 68 4.62 5 .547 Total Satisfaction Time 1 (Pre-test) 68 13.00 15 2.56 Time 2 (Post-test) 68 13.88 15 1.49 Total group analysis: Student satisfaction long-term effects for MSB. In order to address the second part of this question (whether or not the change from pre- to post-test was long-lasting), a one-way between-groups ANOVA with post-hoc tests was conducted to explore the impact of time on students’ satisfaction (instruction, support services, overall WCC, and total) after participating in the Windward Community College Math Summer Bridge Program as measured by the paired (pre/post) satisfaction questions. A Levene’s test for homogeneity of variances was found to be not significant for all of the four measures. Responses were separated into four groups according to the year that they participated in the Math Summer Bridge program MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 157% (2012, 2013, 2014, and 2015). All four measures’ changes were found to be not significant. Table 58 lists the F scores, significance, and effect size of each question. On a 1 to 5 scale, all three of the individual satisfaction questions had less than 0.2 difference between the means of the four cohorts (Instructor Satisfaction: F (3, 146) = .120, p = .949; Support Services Satisfaction: F (3, 146) = .260, p = .854; Overall WCC Satisfaction: F (3, 146) = 1.067, p = .365). For the Total Satisfaction question (scale from 1 to 16), the means for each of the four cohorts only varied by 0.3: F (3, 146) = .318, p = .812. It is a positive result when there is no significant changes post-MSB program because that means that the increase initially attained after the intervention was maintained over time. Therefore, for these three questions, the Windward Community College Math Summer Bridge program did have a significant positive effect on student satisfaction, and the null hypothesis (the change in student satisfaction incurred after the Math Summer Bridge Program is not maintained over time) was rejected. Table 58 One-Way Between-Groups ANOVA Student Satisfaction Scores for 2015, 2014, 2013, and 2012 Cohorts Time period F Sig. Eta Squared Instructor Satisfaction .120 .949 - Support Services Satisfaction .260 .854 - Overall WCC Satisfaction 1.067 .365 - Total Satisfaction .318 .812 - Ethnicity Analysis: Non-Native Hawaiian v. Native Hawaiian After examining the entire group, the second lens for analysis is by using ethnicity as a grouping factor. One-way ANOVAs were performed to check for a significant difference between means for each of the seven (7) questions (4 engagement questions and 3 satisfaction MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 158% questions) and scores for total engagement and total satisfaction. The first type of ANOVA was the one-way repeated measures ANOVA, to compute the differences between pre- and post-test means and if they are significant or not. One-way between-groups ANOVAs were the second type, used to measure the differences between means over time (Cohort 2015 to Cohort 2012). The following sections further describe these analyses and how students’ scores related to one another when grouped by ethnicity (Non-Native Hawaiian v. Native Hawaiian). Ethnicity group analysis: Student engagement pre/post MSB. There were five different one-way repeated measures tests performed to compare the means of Time 1 (pre- survey) and Time 2 (post-survey) answers for each individual student engagement question, as well as one with the total engagement score. Table 59 provides the means and standard deviations for each. Of the ten (10) pre/post measures analyzed [5 questions, 2 groups (non- Native Hawaiian, Native Hawaiian)], only two were found to have statistically significant increases between pre- and post-surveys. Academic peer interaction for non-Native Hawaiians was the first group that was found to have significant increases after the Math Summer Bridge intervention: Wilks’ Lambda = .788, F (1, 25) = 6.740, p = .016, multivariate partial eta squared = .212. Increasing .42 points on a 4-point scale, this improvement had a very large effect size. The other statistically significant increase came from the Total Engagement score for non-Native Hawaiian: Wilks’ Lambda = .755, F (1, 25) = 8.110, p = .009, multivariate partial eta squared = .245. Therefore, the null hypothesis (the Math Summer Bridge Program does not have a significantly positive effect on student engagement for non-Native Hawaiian students) was rejected. All of the Native Hawaiian groupings and the remaining three non-Native Hawaiian groupings did not have statistically significant increases in means from pre- to post-test: (a) MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 159% Instructor Interaction In Class (non-Native Hawaiian), Wilks’ Lambda = .923, F (1, 25) = 2.083, p = .161; (b) Instructor Interaction In Class (Native Hawaiian), Wilks’ Lambda = .999, F (1, 41) = .039, p = .844; (c) Instructor Interaction Outside of Class (non-Native Hawaiian), Wilks’ Lambda = .897, F (1, 25) = 2.857, p = .103; (d) Instructor Interaction Outside of Class (Native Hawaiian), Wilks’ Lambda = .981, F (1, 41) = .796, p = .377; (e) Academic Peer Interaction (Native Hawaiian), Wilks’ Lambda = .944, F (1, 41) = 2.455, p = .125; (f) Social Peer Interaction (non-Native Hawaiian), Wilks’ Lambda = .918, F (1, 25) = 2.231, p = .148; (g) Social Peer Interaction (Native Hawaiian), Wilks’ Lambda = .995, F (1, 41) = .227, p = .637; and (h) Total Engagement (Native Hawaiian), Wilks’ Lambda = .986, F (1, 41) = .596, p = .444). Although these measures did not have significant increases, it is important to note that all post scores went up except for one (Instructor Outside of Class, Native Hawaiian). MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 160% Table 59 Descriptive Statistics Student Engagement Scores for Time 1 and Time 2-Total Time period N Mean Total Possible Score Standard deviation Instructor Interaction In Class Non-Native Hawaiian Time 1 (Pre-survey) 26 2.27 3 .724 Time 2 (Post-survey) 26 2.50 3 .510 Native Hawaiian Time 1 (Pre-survey) 42 2.24 3 .656 Time 2 (Post-survey) 42 2.26 3 .497 Outside of Class Non-Native Hawaiian Time 1 (Pre-survey) 26 1.77 3 .514 Time 2 (Post-survey) 26 1.92 3 .628 Native Hawaiian Time 1 (Pre-survey) 42 1.67 3 .612 Time 2 (Post-survey) 42 1.57 3 .547 Peer Interaction Academic Non-Native Hawaiian* Time 1 (Pre-survey) 26 1.96 4 .662 Time 2 (Post-survey) 26 2.38 4 .752 Native Hawaiian Time 1 (Pre-survey) 42 2.24 4 .932 Time 2 (Post-survey) 42 2.50 4 .773 Social Non-Native Hawaiian Time 1 (Pre-survey) 26 1.96 4 .871 Time 2 (Post-survey) 26 2.23 4 .815 Native Hawaiian Time 1 (Pre-survey) 42 2.29 4 .970 Time 2 (Post-survey) 42 2.36 4 .932 Total Engagement Non-Native Hawaiian** Time 1 (Pre-survey) 26 7.96 14 1.93 Time 2 (Post-survey) 26 9.04 14 2.09 Native Hawaiian Time 1 (Pre-survey) 42 8.43 14 2.26 Time 2 (Post-survey) 42 8.69 14 1.72 * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 161% Ethnicity group analysis: Student engagement long-term effects for MSB. One-way between-groups ANOVA were performed to answer the second part of this question; whether or not these changes (or lack there of) were long lasting for non-Native Hawaiians or Native Hawaiians. A Levene’s test for homogeneity of variances was found to be significant for three of the ten measures. For two questions for non-Native Hawaiian groups, a) Instructor Interaction in Class, and b) Social Peer Interaction, Welch and Brown-Forsythe tests were not generated because there was no variance for Cohort 2012 means. Responses were separated into four groups according to the year that they participated in the Math Summer Bridge Program (2012, 2013, 2014, and 2015). Two measures were found to have significant differences in means and eight (8) were not significant. Table 60 lists the F scores, significance, and effect size of each question; Table 61 presents the significant differences between groups for each question. There was a continuing increase for Instructor Interaction In Class after the 2015 post-test (F (3, 46) = 2.257, p = .094), but it was also not statistically significant. However, for the 2012 Cohort (3 years post intervention), 100% of the students reported having the highest amount of interaction with their instructors (3 = 4+ times/week), and therefore, there was no possible way for their scores to significantly increase (M = 3.00, SD = .000). The other four interaction/engagement questions also did not have statistically significant differences (Instructor Interaction Outside, F (3, 46) = 2.078, p = .116; Academic Peer Interaction, F (3, 46) = .913, p = .442; Social Peer Interaction, F (3, 46) = .411, p = .746; Total Engagement, F (3, 46) = .546, p = .394). Therefore, for all five questions answered by non-Native Hawaiian students, levels attained at the end of the Math Summer Bridge Program were maintained, and the null hypothesis (the change in student engagement incurred after the Math Summer Bridge Program is not maintained over time for non-Native Hawaiian students) was rejected. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 162% For all five questions, scores for Native Hawaiian students increased over time. However, only two were statistically significant. The first was Instructor Interaction in Class, F (3, 96) = 5.273, p = .001, eta squared = .141, which had a large effect size. It’s significant difference was between Cohort 2015 (M = 2.26, SD = .497) and Cohort 2012 (M = 2.83, SD = .389). The second question, which also had a large effect size, was Instructor Interaction Outside Class, F (3, 96) = 4.901, p = .003, eta squared = .133. The other three questions did not have statistically significant changes in scores over time (Academic Peer Interaction, F (3, 96) = .207, p = .891; Social Peer Interaction, F (3, 96) = .211, p = .889; Total Engagement, F (3, 96) = 2.503, p = .064). Therefore, for all five questions, the level of interaction/engagement obtained after the Math Summer Bridge Program was maintained by Native Hawaiian students and the null hypothesis (the change in student engagement incurred after the Math Summer Bridge Program is not maintained over time for Native Hawaiian students) was rejected. Table 60 Ethnicity Grouping- One-Way Between Groups ANOVA Student Engagement Scores for 2015, 2014, 2013, and 2012 Cohorts Time period F Sig. Eta Squared Instructor Interaction- In Class Non-Native Hawaiian 2.257 .094 - Native Hawaiian 5.273 .001 .141 Instructor Interaction- Outside of Class Non-Native Hawaiian 2.078 .116 - Native Hawaiian 4.901 .003 .133 Peer Interaction- School Non-Native Hawaiian .913 .442 - Native Hawaiian .207 .891 - Peer Interaction- Social Non-Native Hawaiian .411 .746 - Native Hawaiian .211 .889 - Total Engagement Non-Native Hawaiian .546 .394 - Native Hawaiian 2.503 .064 - MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 163% Table 61 Ethnicity Grouping- One-Way Between Groups ANOVA Significant Differences-Student Engagement Scores for 2015, 2014, 2013, and 2012 Cohorts Time period Mean Difference Sig. Instructor Interaction- In Class NH 2015 to 2012 .571 .010 Instructor Interaction- Outside of Class NH 2015 to 2012 .679 .012 Peer Interaction- School - - Peer Interaction- Social - - Total Engagement - - Ethnicity group analysis: Student satisfaction pre/post MSB. Student participants were separated by ethnicity (non-Native Hawaiian v. Native Hawaiian) and then four different one-way repeated measures tests were performed to compare the means of Time 1 (pre-survey) and Time 2 (post-survey) answers for each individual student satisfaction score, as well as a total satisfaction score. Table 62 displays the means and standard deviations for each. Satisfaction for all satisfaction measures for non-Native Hawaiian students were not significant and all four of the measures for Native Hawaiian students had statistically significant increases after the intervention. For non-Native Hawaiian students, no satisfaction scores increased significantly: Instructor Satisfaction, Wilks’ Lambda = .936, F (1, 25) = 1.712, p = .203; Support Services Satisfaction, Wilks’ Lambda = .913, F (1, 25) = 2.395, p = .134; Overall WCC Satisfaction, Wilks’ Lambda = .969, F (1, 25) = .812, p = .376; Total Satisfaction, Wilks’ Lambda = .976, F (1, 25) = .627, p = .436. Therefore, for non-Native Hawaiian student satisfaction scores, the null hypothesis (the Math Summer Bridge Program does not have a significantly positive effect on student satisfaction) was accepted. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 164% Native Hawaiian students’ scores were the exact opposite. For all four measures, there was a statistically significant increase in student satisfaction and all of them had moderate to large effect sizes: Instructor Satisfaction, Wilks’ Lambda = .852, F (1, 41) = 7.142, p = .011, multivariate partial eta squared = .148; Support Services Satisfaction, Wilks’ Lambda = .846, F (1, 41) = 7.441, p = .009, multivariate partial eta squared = .154; Overall WCC Satisfaction, Wilks’ Lambda = .909, F (1, 41) = 4.095, p = .050, multivariate partial eta squared = .091; Total Satisfaction, Wilks’ Lambda = .859, F (1, 41) = 6.719, p = .013, multivariate partial eta squared = .141. Therefore, for Native Hawaiian students, all four measures of student satisfaction significantly increased and the null hypothesis (the Math Summer Bridge Program does not have a significantly positive effect on student satisfaction for Native Hawaiian students) was rejected. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 165% Table 62 Descriptive Statistics Student Satisfaction Scores for Time 1 and Time 2-Total Time period N Mean Highest Possible Score Standard deviation Instructor Satisfaction Non-Native Hawaiian Time 1 (Pre-survey) 26 4.31 5 .736 Time 2 (Post-survey) 26 4.50 5 .510 Native Hawaiian* Time 1 (Pre-survey) 42 4.24 5 .983 Time 2 (Post-survey) 42 4.67 5 .526 Support Services Satisfaction Non-Native Hawaiian Time 1 (Pre-survey) 26 4.27 5 .667 Time 2 (Post-survey) 26 4.46 5 .811 Native Hawaiian** Time 1 (Pre-survey) 42 4.31 5 1.024 Time 2 (Post-survey) 42 4.79 5 .470 Overall WCC Satisfaction Non-Native Hawaiian Time 1 (Pre-survey) 26 4.50 5 .648 Time 2 (Post-survey) 26 4.38 5 .571 Native Hawaiian* Time 1 (Pre-survey) 42 4.40 5 1.014 Time 2 (Post-survey) 42 4.76 5 .484 Total Satisfaction Non-Native Hawaiian Time 1 (Pre-survey) 26 13.08 15 1.85 Time 2 (Post-survey) 26 13.35 15 1.70 Native Hawaiian* Time 1 (Pre-survey) 42 12.95 15 2.93 Time 2 (Post-survey) 42 14.21 15 1.26 * p < .05. ** p < .01. Ethnicity group analysis: Student satisfaction long-term effects for MSB. In order to address the second part of this question (whether or not the change from pre- to post-test was long-lasting), a one-way between-groups ANOVA with post-hoc tests was conducted to explore the impact of time on students’ satisfaction (instruction, support services, overall WCC, and total) after participating in the Windward Community College Math Summer Bridge Program as MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 166% measured by the paired (pre/post) satisfaction questions, when grouping students by ethnicity. A Levene’s test for homogeneity of variances was found to be significant for all four measures for non-Native Hawaiian students and therefore, significance was determined by using Welch and Brown-Forsythe tests. Responses were separated into four groups according to the year that they participated in the Math Summer Bridge program (2012, 2013, 2014, and 2015). Seven of the eight measures (all but non-Native Hawaiian-Overall WCC Satisfaction) were found to be not significant. Table 63 lists the F scores, significance, and effect size of each question and Table 64 presents the significant differences between groups for each question. For Non-Native Hawaiian students, one measure significantly increased over time and had a large effect size: Overall WCC Satisfaction, F (3, 46) = 3.458, p = .024, eta squared = .184. The other three measures did not have significant increases, but did slightly increase over time: Instructor Satisfaction, F (3, 46) = 1.404, p = .254; Support Services Satisfaction, F (3, 46) = 2.137, p = .108; Total Satisfaction, F (3, 46) = 2.670, p = .059. Therefore for all four measures, since the level of satisfaction attained after the intervention was maintained over time, the null hypothesis (the (lack of) change in student satisfaction incurred after the Math Summer Bridge Program is not maintained over time for non-Native Hawaiian students) was rejected. For Native Hawaiian students, there was no statistically significant difference between means over time: Instructor Satisfaction, F (3, 96) = .072, p = .975; Support Services Satisfaction, F (3, 96) = .635, p = .594; Overall WCC Satisfaction, F (3, 96) = .423, p = .737; Total Satisfaction, F (3, 96) = .236, p = .871. Therefore, since the level of satisfaction attained after the intervention did not significantly change over time, the null hypothesis (the change in student satisfaction incurred after the Math Summer Bridge Program is not maintained over time for Native Hawaiian students) was rejected. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 167% Table 63 Ethnicity Grouping- One-Way Between Groups ANOVA Student Satisfaction Scores for 2015, 2014, 2013, and 2012 Cohorts Time period F Sig. Eta Squared Instructor Satisfaction Non-Native Hawaiian 1.404 .254 - Native Hawaiian .072 .975 - Support Services Satisfaction Non-Native Hawaiian 2.137 .108 - Native Hawaiian .635 .594 - Overall WCC Satisfaction Non-Native Hawaiian 3.458 .024 .184 Native Hawaiian .423 .737 - Total Satisfaction Non-Native Hawaiian 2.670 .059 - Native Hawaiian .236 .871 - Table 64 Ethnicity Grouping- One-Way Between Groups ANOVA Significant Differences-Student Satisfaction Scores for 2015, 2014, 2013, and 2012 Cohorts Time period Mean Difference Sig. Instructor Satisfaction - - Support Services Satisfaction - - Overall WCC Satisfaction Non-NH 2015 to 2014 .615 .020 Total Satisfaction - - Age Group Analysis: Traditional v. Nontraditional The third lens for analysis is by using age as a grouping factor (< 26 years old is traditional, 26+ years old is considered nontraditional). One-way ANOVAs were performed to check for a significant difference between means for each of the seven (7) questions (4 engagement questions and 3 satisfaction questions) with another measure for total engagement and one for total satisfaction. The first type of ANOVA was the one-way repeated measures ANOVA, to compute the differences between pre- and post-survey means and if they are MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 168% significant or not. One-way between-groups ANOVA were the second type, used to measure the differences between means over time (Cohort 2015 to Cohort 2012). The following sections further describe these analyses and how students’ scores related to one another when grouped by ethnicity (Traditional v. Nontraditional). Age group analysis: Student engagement pre/post MSB. There were five different one-way repeated measures tests performed to compare the means of Time 1 (pre-survey) and Time 2 (post-survey) answers for each individual student engagement question, as well as one with the total engagement score. Table 65 provides the means and standard deviations for each. Of the ten (10) pre/post measures analyzed [5 questions, 2 groups (traditional, nontraditional)], only one measure had statistically significant changes from pre- to post-survey (Nontraditional- Academic Peer Interaction). For traditional students, none of the five measures had significant changes between Time 1 and Time 2, however they all had small increases after the intervention (Instructor Interaction In Class, Wilks’ Lambda = .942, F (1, 33) = 2.019, p = .165; Instructor Interaction Outside of Class, Wilks’ Lambda = .990, F (1, 33) = .327, p = .571; Academic Peer Interaction, Wilks’ Lambda = .971, F (1, 33) = 1.000, p = .325; Social Peer Interaction, Wilks’ Lambda = .992, F (1, 33) = .251, p = .619; Total Engagement, Wilks’ Lambda = .959, F (1, 33) = 1.413, p = .243). Therefore, the null hypothesis (the Math Summer Bridge Program does not have a significantly positive effect on student engagement) was accepted. Nine measures had insignificant changes in means between Time 1 and Time 2 for nontraditional students. The one measure that did have a significant change (Academic Peer Interaction), had a very large effect size: Wilks’ Lambda = .811, F (1, 33) = 7.685, p = .009, multivariate partial eta squared = .189. The insignificant change in means between the pre- and MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 169% post-surveys for the other four measures for nontraditional students were as follows: (a) Instructor Interaction In Class, Wilks’ Lambda = 1.000, F (1, 33) = .000, p = 1.000, multivariate partial eta squared = .000; (b) Instructor Interaction Outside of Class, Wilks’ Lambda = .992, F (1, 33) = .280, p = .600, multivariate partial eta squared = .008; (c) Social Peer Interaction, Wilks’ Lambda = .947, F (1, 33) = 1.861, p = .182, multivariate partial eta squared = .053; and (d) Total Engagement was: Wilks’ Lambda = .895, F (1, 33) = 3.858, p = .058, multivariate partial eta squared = .105. Therefore, although there was only one question that had significant results, this means that the Math Summer Bridge Program did have a significant effect on student engagement for nontraditional students, and therefore, the null hypothesis is rejected. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 170% Table 65 Age Grouping- Descriptive Statistics Student Engagement Scores for Time 1 and Time 2-Total Time period N Mean Highest Possible Score Standard deviation Instructor Interaction In Class Traditional Time 1 (Pre-survey) 34 2.18 3 .716 Time 2 (Post-survey) 34 2.38 3 .551 Nontraditional Time 1 (Pre-survey) 34 2.32 3 .638 Time 2 (Post-survey) 34 2.32 3 .475 Outside of Class Traditional Time 1 (Pre-survey) 34 1.65 3 .544 Time 2 (Post-survey) 34 1.71 3 .629 Nontraditional Time 1 (Pre-survey) 34 1.76 3 .606 Time 2 (Post-survey) 34 1.71 3 .579 Peer Interaction Academic Traditional Time 1 (Pre-survey) 34 2.09 4 .866 Time 2 (Post-survey) 34 2.24 4 .654 Nontraditional** Time 1 (Pre-survey) 34 2.18 4 .834 Time 2 (Post-survey) 34 2.68 4 .806 Social Traditional Time 1 (Pre-survey) 34 2.15 4 .958 Time 2 (Post-survey) 34 2.24 4 .890 Nontraditional Time 1 (Pre-survey) 34 2.18 4 .936 Time 2 (Post-survey) 34 2.38 4 .888 Total Engagement Traditional Time 1 (Pre-survey) 34 8.06 14 2.24 Time 2 (Post-survey) 34 8.56 14 1.85 Nontraditional Time 1 (Pre-survey) 34 8.44 14 2.05 Time 2 (Post-survey) 34 9.09 14 1.86 * p < .05. ** p < .01. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 171% Age group analysis: Student engagement long-term effects for MSB. One-way between-groups ANOVA were performed to answer the second part of this question; whether or not these differences (or lack there of) were long-lasting for traditional or nontraditional students. A Levene’s test for homogeneity of variances was found to be significant for two of the ten measures (Traditional Instructor Interaction In Class; Nontraditional Instructor Interaction In Class). For Nontraditional Instructor Interaction In Class, Welch and Brown-Forsythe tests were not generated because there was no variance for Cohort 2012 means (100% of the responses were 3.00, the highest number possible). Responses were separated into four groups according to the year that they participated in the Math Summer Bridge Program (2012, 2013, 2014, and 2015). Three (3) measures were found to have statistically significant differences in means over time and seven (7) were not significant. Table 66 lists the F scores, significance, and effect size of each question; Table 67 presents the significant differences between groups for each question. Only one measure for traditional students had a significant increase in means over time: Instructor Interaction in Class, F (3, 63) = 2.704, p = .032, eta squared = .114. For the other four measures for traditional students, there were no statistically significant changes over time (Instructor Interaction Outside of Class, F (3, 63) = 1.082, p = .363; Academic Peer Interaction, F (3, 63) = .490, p = .690; Social Peer Interaction, F (3, 63) = .945, p = .424; Total Engagement, F (3, 63) = 2.487, p = .069). Therefore, for all five of these measures, since the level of interaction attained after the intervention was maintained and/or increased, the null hypothesis (the change in student engagement incurred after the Math Summer Bridge Program is not maintained over time for traditional students) was rejected. Two (2) of the five (5) measures for nontraditional students had a significant increase in means over time and they both had very large effect sizes a) Instructor Interaction in Class, F (3, MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 172% 79) = 6.622, p = .000, eta squared = .201, and b) Instructor Interaction Outside of Class, F (3, 79) = 2.790, p = .000, eta squared = .215. The other three measures did not have statistically significant changes over time: (a) Academic Peer Interaction, F (3, 79) = .814, p = .490; (b) Social Peer Interaction, F (3, 79) = .234, p = .872; and (c) Total Engagement, F (3, 79) = 1.210, p = .312. Since all five measures did not have means that significantly decreased from the levels attained after the Math Summer Bridge Program, the null hypothesis (the change in student engagement incurred after the Math Summer Bridge Program is not maintained over time for nontraditional students) was rejected. Table 66 Age Grouping- One-Way Between Groups ANOVA Student Engagement Scores for 2015, 2014, 2013, and 2012 Cohorts Time period F Sig. Eta Squared Instructor Interaction- In Class Traditional 2.704 .032 .114 Nontraditional 6.622 .000 .201 Instructor Interaction- Outside of Class Traditional 1.082 .363 - Nontraditional 2.790 .000 .215 Peer Interaction- School Traditional .490 .690 - Nontraditional .814 .490 - Peer Interaction- Social Traditional .945 .424 - Nontraditional .234 .872 - Total Engagement Traditional 2.487 .069 - Nontraditional 1.210 .312 - MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 173% Table 67 Age Grouping- One-Way Between Groups ANOVA Significant Differences-Student Engagement Scores for 2015, 2014, 2013, and 2012 Cohorts Time period Mean Difference Sig. Instructor Interaction- In Class Nontrad. 2015 to 2014 .394 .022 Nontrad. 2015 to 2013 .440 .004 Nontrad. 2015 to 2012 .657 .012 Instructor Interaction- Outside of Class Nontrad. 2015 to 2014 .549 .014 Nontrad. 2015 to 2012 1.119 .001 Nontrad. 2013 to 2012 .790 .035 Peer Interaction- School - - Peer Interaction- Social - - Total Engagement - - Age group analysis: Student satisfaction pre/post MSB. Student participants were separated by age (traditional v. nontraditional) and then four different one-way repeated measures tests were performed to compare the means of Time 1 (per-survey) and Time 2 (post- survey) answers for each individual student satisfaction question score, as well as a total satisfaction score. Table 68 displays the means and standard deviations for each. Satisfaction for three (3) out of the four (4) measures for traditional students had statistically significant increases and one (1) out of four (4) measures for nontraditional students had statistically significant increases after the intervention. The one measure that did not have statistically significant improvement post-intervention for traditional students was Overall WCC Satisfaction Wilks’ Lambda = .948, F (1, 33) = 1.821, p = .186. Although not significant, the scores did improve from Time 1 (pre-survey) (M = 4.32, SD = .912) and Time 2 (post-survey) (M = 4.56, SD = 5.61). For traditional students, the three measures that had significant tests had moderate to large effect sizes: (1) Instructor Satisfaction, Wilks’ Lambda = .856, F (1, 33) = 5.557, p = .024, multivariate partial eta squared = .144; MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 174% (2) Student Support Satisfaction, Wilks’ Lambda = .889, F (1, 33) = 4.139, p = .050, multivariate partial eta squared = .111; (3) Total Satisfaction, Wilks’ Lambda = .888, F (1, 33) = 4.156, p = .050, multivariate partial eta squared = .122. All four measures of satisfaction for nontraditional students increased from Time 1 (Pre- test) to Time 2 (Post-test). However only one was statistically significant: Student Support Satisfaction, Wilks’ Lambda = .986, F (1, 33) = 5.462, p = .026, multivariate partial eta squared = .142. The other three measures results are as follows: (1) Instructor Satisfaction, Wilks’ Lambda = .912, F (1, 33) = 3.194, p = .083; (2) Overall WCC Satisfaction, Wilks’ Lambda = .986, F (1, 33) = .463, p = .501; (3) Total Satisfaction, Wilks’ Lambda = .920, p = .099. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 175% Table 68 Ethnicity Grouping- Descriptive Statistics Student Satisfaction Scores for Time 1 and Time 2- Total Time period N Mean Highest Possible Score Standard Deviation Instructor Satisfaction Traditional* Time 1 (Pre-survey) 34 4.09 5 .900 Time 2 (Post-survey) 34 4.50 5 .564 Nontraditional Time 1 (Pre-survey) 34 4.44 5 .860 Time 2 (Post-survey) 34 4.71 5 .462 Support Services Satisfaction Traditional* Time 1 (Pre-survey) 34 4.18 5 .936 Time 2 (Post-survey) 34 4.53 5 .706 Nontraditional* Time 1 (Pre-survey) 34 4.41 5 .857 Time 2 (Post-survey) 34 4.79 5 .538 Overall WCC Satisfaction Traditional Time 1 (Pre-survey) 34 4.32 5 .912 Time 2 (Post-survey) 34 4.56 5 .561 Nontraditional Time 1 (Pre-survey) 34 4.56 5 .860 Time 2 (Post-survey) 34 4.68 5 .535 Total Satisfaction Traditional* Time 1 (Pre-survey) 34 12.59 15 2.57 Time 2 (Post-survey) 34 13.59 15 1.67 Nontraditional Time 1 (Pre-survey) 34 13.41 15 2.51 Time 2 (Post-survey) 34 14.18 15 1.24 * p < .05. ** p < .01. Age group analysis: Student satisfaction long-term effects for MSB. In order to address the second part of this question (whether or not the change from pre- to post-test was long-lasting), a one-way between-groups ANOVA with post-hoc tests was conducted to explore the impact of time on students’ satisfaction (instruction, support services, overall WCC, and MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 176% total) after participating in the Windward Community College Math Summer Bridge Program as measured by the paired (pre/post) satisfaction questions, when grouping students by age. A Levene’s test for homogeneity of variances was found to be significant for only one of eight tests (traditional-Overall WCC Satisfaction) and therefore, significance was determined by using Welch and Brown-Forsythe tests. Responses were separated into four groups according to the year that they participated in the Math Summer Bridge program (2012, 2013, 2014, and 2015). All four measures’ differences for both traditional and nontraditional students (eight total tests) were found to be not significant. Table 69 lists the F scores, significance, and effect size of each question. Although not statistically significant, all four measure for traditional students increased over time (2015 Cohort to 2012 Cohort): (a) Instructor Satisfaction, F (3, 63) = .127, p = .944; (b) Support Services Satisfaction, F (3, 63) = .300, p = .825; (c) Overall WCC Satisfaction, F (3, 63) = 1.655, p = .186; and (d) Total Satisfaction, F (3, 63) = .238, p = .870. Therefore, since the satisfaction attained directly after the Math Summer Bridge Program was sustained over time, the null hypothesis (the change in student satisfaction incurred after the Math Summer Bridge Program is not maintained over time for traditional students) was rejected. Similar to the traditional students, nontraditional students’ satisfaction was not statistically significant, but had small increases over time: (a) Instructor Satisfaction, F (3, 79) = .238, p = .908; (b) Support Services, F (3, 79) = .311, p = .817; (c) Overall WCC Satisfaction, F (3, 79) = .352, p = .788; and (d) Total Satisfaction, F (3, 79) = .146, p = .932. Therefore, since the level of satisfaction attained after the intervention was maintained over time, the null hypothesis (the change in student satisfaction incurred after the Math Summer Bridge program is not maintained over time for nontraditional students) was rejected. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 177% Table 69 One-Way Between Groups ANOVA Student Satisfaction Scores for 2015, 2014, 2013, and 2012 Cohorts Time period F Sig. Eta Squared Instructor Satisfaction Traditional .127 .944 - Nontraditional .238 .908 - Support Services Satisfaction Traditional .300 .825 - Nontraditional .311 .817 - Overall WCC Satisfaction Traditional 1.655 .186 - Nontraditional .352 .788 - Total Satisfaction Traditional .238 .870 - Nontraditional .146 .932 - Research Question 3 Summary When measuring student engagement and satisfaction (pre/post-survey) for all students in the 2015 cohort, the Math Summer Bridge Program had statistically significant effects on increasing academic peer interaction, total engagement, and instructor, support services, and total satisfaction. For all of these measures (looking at the four cohorts: 2015, 2014, 2013, 2012), the significant increase post-intervention was long lasting. When separating students by ethnicity, there were fewer significant increases in engagement and satisfaction. Only non-Native Hawaiian students had statistically significant increases in means for two measures (Academic Peer Interaction and Total Engagement). Student satisfaction only had significant increases for Native Hawaiian students, and all four were found statistically significant (Instructor Satisfaction, Support Services Satisfaction, Overall WCC Satisfaction, and Total Satisfaction). For all of these differences, they were also long lasting. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 178% Finally, when analyzing these questions by age group, there was only one significant increase in engagement: Academic Peer Interaction for nontraditional students. For satisfaction measures, there was a definite impact on traditional student satisfaction, improving three out of four measures (Instructor Satisfaction, Support Services Satisfaction, and Total Satisfaction), and also significant improvement for nontraditional support services satisfaction. For all significant measures for age groups, the effects were also long lasting. Table 70 displays all of the significant results for student engagement and satisfaction. Table 70 Research Question 3: Student Engagement and Satisfaction, Statistically Significant Results Measure F Sig. Partial Eta 2 Long- Lasting Total Sample Population Academic Peer Interaction 7.584 .008 .102 YES Total Engagement 6.674 .030 .065 YES Instructor Satisfaction 8.801 .004 .116 YES Support Services Satisfaction 9.651 .003 .126 YES Total Satisfaction 7.107 .010 .096 YES Ethnicity (Non-NH/NH) Non-NH Academic Peer Interaction 6.740 .016 .212 YES Non-NH Total Engagement 8.110 .009 .245 YES NH Instructor Satisfaction 7.142 .011 .148 YES NH Support Services Satisfaction 7.441 .009 .154 YES NH Overall WCC Satisfaction 4.095 .050 .091 YES NH Total Satisfaction 6.719 .013 .141 YES Age (Traditional/Nontraditional) Nontrad. Academic Peer Interaction 7.685 .009 .189 YES Trad. Instructor Satisfaction 5.557 .024 .144 YES Trad. Student Support Satisfaction 4.139 .050 .111 YES Nontrad. Student Support Satisfaction 5.462 .026 .142 YES Trad. Total Satisfaction 4.156 .050 .112 YES Research Question 4 Results: Persistence Research Question 4 measures if the WCC Math Summer Bridge Program has a significant impact on persistence, a student success measure, by comparing the MSB participants MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 179% (n = 294) to two different control groups. Control Group A consists of students who took the same developmental math course at Windward Community College over the same summer, but were not a part of the Math Summer Bridge Program (N=203). Control Group B (N = 460) consists of students who did not take a developmental math course over the summer, but instead were matched to individual MSB participants based on seven (7) characteristics: (1) First semester enrolled; (2) Current enrollment status (full- or part-time); (3) Age; (4) Ethnicity (Native Hawaiian or non-Native Hawaiian); (5) Sex; (6) Pell Grant recipient (or not); and (7) Compass placement test score. This research question is a two part, multi-level question asking: “Do the participants in the WCC Math Summer Bridge Program persist at a greater rate than a) non-MSB summer math students, or b) non-summer math students matched on seven (7) characteristics?” Also, “is there a significant difference between Native Hawaiian and non- Native Hawaiian students?” and “is there a difference between traditional and nontraditional students?” For the purpose of this study, persistence is considered as any student who is still currently enrolled in college (Fall 2015 semester). Whether students have earned a degree or not, if they are enrolled in at least one credit during the Fall 2015 semester, they are considered as persisting. If students have graduated already and not enrolled in coursework as of Fall 2015, they are not counted as persisting because they will be recorded in Research Question 5. In order to address this research question and determine whether or not the null hypothesis (that the WCC Math Summer Bridge Program does not have any impact on persistence rates) can be rejected, the researcher completed six (6) one-way between-measures analysis of variance tests (one for total sample population, one grouped by ethnicity, and one grouped by age; each of these performed to compare the experiment to the two control groups). The aim of these one-way ANOVA were to “compare the variance (variability in scores) MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 180% between the different groups [pre v. post 2015 surveys and post 2015 v. post 2014 v. post 2013 v. post 2012 surveys] (believed to be due to the intervention [WCC Math Summer Bridge Program]) with the variability within each of the groups (believed to be due to chance)” (Pallant, 2013, p. 258). One-way between-groups ANOVA were used to determine if there were any significant differences between the mean scores of the experiment group to the mean scores of each control group. For the purposes of this section, data will be presented in sections by grouping of student population; (1) total student population; (2) ethnicity (non-Native Hawaiian v. Native Hawaiian); and (3) age (traditional v. nontraditional). Total Group Analysis: Persistence First, to answer the question, “is there a difference in persistence rates for WCC Math Summer Bridge Participants compared to a) non-MSB summer math students, and b) non- summer math students matched on 7 characteristics?” a one-way between-measures ANOVA was conducted to compare persistence rates of the experiment group and each of the two control groups. Table 71 lists the means, F scores, significance, and effect size for the Experiment group and Control Group A. Table 72 does the same for the Experiment group and Control Group B. When comparing the experiment group to Control Group A (those who took the same summer math course but did not participate in the Math Summer Bridge Program), there was a significant difference in persistence rates: F (1, 495) = 15.287, p = .000, eta squared = .030. Then, when analyzing each cohort individually, all four cohorts, the WCC Math Summer Bridge students persisted at higher rates than the Control Group A students, however only the 2012 Cohort had statistically significant differences in means, F (1, 101) = 12.627, p = .001, eta squared = .111. Therefore, the null hypothesis (the Math Summer Bridge Program students do not persist at a greater rate than Control Group A) was rejected. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 181% Control Group B was matched on seven (7) different characteristics. Four hundred and sixty (460) non-summer math students enrolled at Windward Community College were matched to one hundred and twenty-six (126) students out of the 294 total WCC Math Summer Bridge participants. Matched Control Group B students were identified by the Institutional Research Office at Windward Community College and WCC Math Summer Bridge participants had anywhere from one to nineteen Control Group B matched students (matched on all 7 characteristics). In order to offset the skewed number of Control Group B students, experiment students’ data were duplicated to match the same amount of entries as their Control Group B counterparts. And because of this, data was not segregated by cohorts. When comparing means between the Experiment group and Control Group B, there was a statistically significant difference with a large effect size found between persistence rates: F (1, 1016) = 202.686, p = .000, eta squared = .166. Therefore, for the second part of this analysis, the null hypothesis (the Math Summer Bridge Program students do not persist at a greater rate than Control Group B) was rejected. Table 71 One-Way Between-Groups ANOVA Persistence Rates (Experiment v. Control A) Group Experiment M Control A M F Sig. Eta Squared Total .67 .50 15.287 .000 .030 2015 Cohort .93 .83 2.60 .155 - 2014 Cohort .62 .53 .821 .367 - 2013 Cohort .54 .49 .286 .593 - 2012 Cohort .56 .23 12.627 .001 .111 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 182% Table 72 One-Way Between-Groups ANOVA Persistence Rates (Experiment v. Control B) Group Experiment M Control B M F Sig. Eta Squared Exp. v. Control B .67 .26 202.686 .000 .166 Ethnicity Analysis (Non-Native Hawaiian v. Native Hawaiian): Persistence The second part of this question asks the same question, “is there a difference in persistence rates for WCC Math Summer Bridge Participants compared to a) non-MSB summer math students, and b) non-summer math students matched on 7 characteristics?” and adds, “is there a difference in significance based on ethnicity (non-Native Hawaiian v. Native Hawaiian)?” A one-way between-measures ANOVA was conducted to compare persistence rates of the experiment group and each of the two control groups. Table 73 lists the means, F scores, significance, and effect size for the Experiment group and Control Group A. Table 74 does the same for the Experiment group and Control Group B. Overall, there was a significant difference in persistence rates for both non-Native Hawaiians and Native Hawaiians when grouping students by ethnicity (Non-Native Hawaiian, F (1, 237) = 5.487, p = .020, eta squared = .023; Native Hawaiian, F (1, 256) = 7.447, p = .009, eta squared = .028). When separating data by cohort as well, only Native Hawaiian students in the 2012 Cohort had a significant difference, F (1, 48) = 7.507, p = .009, eta squared = .135. All other groupings did not have statistically significant differences in persistence rates (2015 Cohort Non-NH, F (1, 62) = .107, p = .745, eta squared = .002; 2015 Cohort NH, F (1, 58) = 1.645, p = .451, eta squared = .027; 2014 Cohort 2014 Non-NH, F (1, 48) = .254, p = .616, eta squared = .005; Cohort 2014 NH, F (1, 60) = 1.308, p = .257, eta squared = .021; Cohort 2013 Non-NH, F (1, 68) = .990, p = .323, eta squared = .014; Cohort 2013 NH, F (1, 82) = .232, p = .632, eta MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 183% squared = .003; Cohort 2012 Non-NH, F (1, 51) = 2.743, p = .154, eta squared = .051). Therefore, since there were significantly higher persistence rates for both non-Native Hawaiians and Native Hawaiians, the null hypothesis (non-Native Hawaiian Math Summer Bridge Program students do not persist at a greater rate than non-Native Hawaiian Control Group A students; Native Hawaiian Math Summer Bridge students do not persist at a greater rate than Native Hawaiian Control Group A students) was rejected for both. Comparing persistence rates between Experiment and Control Group B with ethnicity groupings also had statistically significant differences with moderate and very large effect sizes for non-Native Hawaiian and Native Hawaiian students, respectively (Non-NH, F = 43.667, p < .001, eta squared = .100; NH, F = 172.537, p < .001, eta squared = .218). For both comparisons to Control Group A and Control Group B, the Math Summer Bridge students persisted at higher rates than their non-MSB counterparts. Therefore, when comparing the Experiment group to Control Group B, the null hypothesis (non-Native Hawaiian Math Summer Bridge Program students do not persist at a greater rate than non-Native Hawaiian Control Group B students; Native Hawaiian Math Summer Bridge Program students do not persist at a greater rate than Native Hawaiian Control Group B students) was rejected. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 184% Table 73 Ethnicity Grouping- One-Way Between-Groups ANOVA Persistence Rates (Experiment v. Control A) Group Experiment M Control A M F Sig. Eta Squared Total Non-Native Hawaiian .64 .49 5.487 .020 .023 Native Hawaiian .69 .51 7.447 .009 .028 2015 Cohort Non-Native Hawaiian .84 .81 .107 .745 - Native Hawaiian .98 .90 1.645 .451 - 2014 Cohort Non-Native Hawaiian .65 .58 .254 .616 - Native Hawaiian .60 .42 1.308 .257 - 2013 Cohort Non-Native Hawaiian .53 .41 .990 .323 - Native Hawaiian .54 .60 .232 .632 - 2012 Cohort Non-Native Hawaiian .46 .23 2.743 .154 - Native Hawaiian .62 .25 7.507 .009 .135 Table 74 Ethnicity Grouping- One-Way Between-Groups ANOVA Persistence Rates (Experiment v. Control B) Group Experiment M Control B M F Sig. Eta Squared Exp. v. Control B Non-Native Hawaiian .59 .28 43.67 .000 .100 Native Hawaiian .71 .25 172.537 .000 .218 Age Group Analysis (Traditional v. Nontraditional): Persistence The third part of this question asks, yet again, the same question, “is there a difference in persistence rates for WCC Math Summer Bridge Participants compared to a) non-MSB summer math students, and b) non-summer math students matched on 7 characteristics?” and this time adds, “is there a difference in significance based on age (traditional v. nontraditional)?” A one- MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 185% way between-measures ANOVA was conducted to compare persistence rates of the experiment group and each of the two control groups. Table 75 lists the means, F scores, significance, and effect size for the Experiment group and Control Group A. Table 76 does the same for the Experiment group and Control Group B. Overall, when comparing Experiment to Control Group A traditional and nontraditional students, there was a statistically significant difference in persistence for nontraditional students (F (1, 217) = 13.654, p < .001, eta squared = .059) and the difference for traditional students was approaching significance (F (1, 276) = 3.883, p = .052, eta squared = .014). For traditional students in the 2012 Cohort, there was also a statistically significant difference between experiment and control groups and there was a large effect size (F = (1, 53) = 10.345, p = .002, eta squared = .163). All other comparisons did not have statistically significant differences in persistence rates, however, WCC Math Summer Bridge students did better than Experiment Group A students in every group except for the 2013 Cohort traditional students (2015 Traditional, F (1, 60) = .182, p = .671; 2015 Nontraditional, F (1, 60) = 3.698, p = .132; 2014 Traditional, F (1, 62) = 1.086, p = .301; 2014 Nontraditional, F (1, 46) = .044, p = .835; 2013 Traditional, F (1, 92) = .295, p = .588; 2013 Nontraditional, F (1, 58) = 2.361, p = .130; 2012 Nontraditional, F (1, 46) = 2.658, p = .138). Therefore, for both traditional and nontraditional students, the null hypothesis (the Math Summer Bridge Program students do not persist at a greater rate than Control Group A students) was rejected. When comparing the WCC Math Summer Bridge students (Experiment) to Control Group B, both traditional and nontraditional students had statistically significant differences in persistence with very large effect sizes: Traditional, F (1, 804) = 144.803, p < .001, eta squared = .153; Nontraditional, F (1, 210) = 60.540, p < .001, eta squared = .224. Therefore, since the MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 186% WCC Math Summer Bridge students persisted at a significantly higher rate than their Control Group B counterparts, the null hypothesis (the Math Summer Bridge Program students do not persist at a greater rate than Control Group B students) was also rejected. Table 75 Age Grouping- One-Way Between-Groups ANOVA Persistence Rates (Experiment v. Control A) Group Experiment M Control A M F Sig. Eta Squared Total Traditional .64 .53 3.883 .052 .014 Nontraditional .70 .46 13.654 .000 .059 2015 Cohort Traditional .90 .86 .182 .671 - Nontraditional .95 .80 3.698 .132 - 2014 Cohort Traditional .59 .45 1.086 .301 - Nontraditional .66 .63 .044 .835 - 2013 Cohort Traditional .50 .56 .295 .588 - Nontraditional .59 .39 2.361 .130 - 2012 Cohort Traditional .68 .27 10.345 .002 .163 Nontraditional .41 .19 2.685 .138 - Table 76 Age Grouping- One-Way Between-Groups ANOVA Persistence Rates (Experiment v. Control B) Group Experiment M Control B M F Sig. Eta Squared Exp. v. Control B Traditional .66 .27 144.803 .000 .153 Nontraditional .70 .23 60.540 .000 .224 Research Question 4 Summary Overall, when comparing WCC Math Summer Bridge students (Experiment) to either control group, they are persisting at significantly higher rates. When separating students by the MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 187% year they took a summer developmental math course, there were scores that were not statistically significant. However, the experiment group always had a higher persistence rate than the control groups (except for 2013 traditional students persisted at an insignificantly lower rate than their Control Group A counterparts). Table 77 and Table 78 displays all persistence rates’ means, F values, p values, and eta squared values for the experiment group versus Control Group A and Control Group B respectively. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 188% Table 77 Complete- One-Way Between-Groups ANOVA Persistence Rates (Experiment v. Control A) 2015 Cohort .93 .83 2.60 .155 - 2014 Cohort .62 .53 .821 .367 - 2013 Cohort .54 .49 .286 .593 - 2012 Cohort .56 .23 12.627 .001 .111 Ethnicity Total Non-Native Hawaiian .64 .49 5.487 .020 .023 Native Hawaiian .69 .51 7.447 .009 .028 Group Experiment M Control A M F Sig. Eta Squared Total Group Total .67 .50 15.287 .000 .030 2015 Cohort Non-Native Hawaiian .84 .81 .107 .745 - Native Hawaiian .98 .90 1.645 .451 - 2014 Cohort Non-Native Hawaiian .65 .58 .254 .616 - Native Hawaiian .60 .42 1.308 .257 - 2013 Cohort Non-Native Hawaiian .53 .41 .990 .323 - Native Hawaiian .54 .60 .232 .632 - 2012 Cohort Non-Native Hawaiian .46 .23 2.743 .154 - Native Hawaiian .62 .25 7.507 .009 .135 Age Total Traditional .64 .53 3.883 .052 .014 Nontraditional .70 .46 13.654 .000 .059 2015 Cohort Traditional .90 .86 .182 .671 - Nontraditional .95 .80 3.698 .132 - 2014 Cohort Traditional .59 .45 1.086 .301 - Nontraditional .66 .63 .044 .835 - 2013 Cohort Traditional .50 .56 .295 .588 - Nontraditional .59 .39 2.361 .130 - 2012 Cohort Traditional .68 .27 10.345 .002 .163 Nontraditional .41 .19 2.685 .138 - MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 189% Table 78 Complete - One-Way Between-Groups ANOVA Persistence Rates (Experiment v. Control B) Group Experiment M Control B M F Sig. Eta Squared Total Group .67 .26 202.686 .000 .166 Ethnicity Non-Native Hawaiian .59 .28 43.67 .000 .100 Native Hawaiian .71 .25 172.537 .000 .218 Age Traditional .66 .27 144.803 .000 .153 Nontraditional .70 .23 60.540 .000 .224 Research Question 5 Results: Graduation Research Question 5 measures the effects of the Math Summer Bridge Program on student graduation rates by comparing the MSB students (experiment group) to two different control groups. Control Group A consists of students who took the same developmental math course at Windward Community College over the same summer, but were not a part of the Math Summer Bridge Program (N=203). Control Group B (N = 460) consists of students who did not take a developmental math course over the summer, but instead were matched to individual MSB participants based on seven (7) characteristics: (1) First semester enrolled; (2) Current enrollment status (full- or part-time); (3) Age; (4) Ethnicity (Native Hawaiian or non-Native Hawaiian); (5) Sex; (6) Pell Grant recipient (or not); and (7) Compass placement test score. This research question is a two part, multi-level question asking: “Do the participants in the WCC Math Summer Bridge Program graduate at a greater rate than a) non-MSB summer math students, or b) non-summer math students matched on seven (7) characteristics?” Also, “is there a significant difference between Native Hawaiian and non-Native Hawaiian students?” and “is there a significant difference between traditional and nontraditional students?” For the purpose of this study, graduation is considered as any student who has earned a 2-year degree or higher before MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 190% the Fall 2015 semester. Many students have earned a 2-year degree (e.g. Associates of Arts in Liberal Arts, Associates of Arts in Hawaiian Studies, Associates in Science in Natural Science), and then transferred over to a 4-year institution. Even though they have not completed the current degree they are pursuing (e.g. Bachelor of Arts), since they earned a 2-year degree, they have been counted as a graduate for this study. Certificates or anything less than a 2-year degree have not been counted for this study. In order to address this research question and determine whether or not the null hypothesis can be rejected (that the WCC Math Summer Bridge Program does not have any impact on graduation rates), the researcher completed six (6) one-way between-measures analysis of variance tests (one for total sample population, one grouped by ethnicity, and one grouped by age; each of these performed to compare the experiment to the two control groups). The aim of these one-way ANOVA were to “compare the variance (variability in scores) between the different groups [pre v. post 2015 tests and post 2015 v. post 2014 v. post 2013 v. post 2012 tests] (believed to be due to the independent variable [WCC Math Summer Bridge Program]) with the variability within each of the groups (believed to be due to chance)” (Pallant, 2013, p. 258). One-way between-groups ANOVA were used to determine if there were any significant differences between the mean scores of the experiment group to the mean scores of each control group. For the purposes of this section, data will be presented in sections by grouping of student population; 1) total student population, 2) ethnicity (non-Native Hawaiian v. Native Hawaiian), and 3) age (traditional v. nontraditional). Total Group Analysis: Graduation First, to answer the question, “is there a difference in graduation rates for WCC Math Summer Bridge Participants compared to a) non-MSB summer math students, and b) non- MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 191% summer math students matched on 7 characteristics?” a one-way between-measures ANOVA was conducted to compare graduation rates of the experiment group and each of the two control groups. Table 79 lists the means, F scores, significance, and effect size for the Experiment group and Control Group A. Table 80 does the same for the Experiment group and Control Group B. When comparing the Math Summer Bridge students to Control Group A, they do not graduate at a higher rate than non-MSB summer math students although, these deficiencies are not statistically significant (Total, F (1, 495) = .880, p = .349, eta squared = .002; Cohort 2013, F (1, 152) = 1.887, p = .172, eta squared = .012; Cohort 2014, F (1, 110) = .011, p = .916, eta squared < .001; Cohort 2015, F (1, 122) = .086, p = .770, eta squared = .001). The only cohort that had statistically significant differences was Cohort 2012, and for this cohort, the experiment group has significantly higher graduation rates than the experiment group: F (1, 101) = 5.787, p = .022, eta squared = .054. Therefore for all but one comparison, statistically, there is no difference between the Math Summer Bridge students and Control Group A; the null hypothesis (the Math Summer Bridge Program students do not graduate at a significantly greater rate than Control Group A students) is accepted. Control Group B was matched on seven (7) different characteristics. Four hundred and sixty (460) non-summer math students enrolled at Windward Community College were matched to one hundred and twenty-six (126) students out of the 294 total WCC Math Summer Bridge participants. Matched Control Group B students were identified by the Institutional Research Office at Windward Community College and WCC Math Summer Bridge participants had anywhere from one to nineteen matches. In order to offset the skewed number of Control Group B students, experiment students’ data were duplicated to match the same amount of entries as their Control Group B counterparts. And because of this, data was not segregated by cohorts. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 192% When comparing means between the Experiment group and Control Group B, the MSB students had significantly higher graduation rates than their matched counterparts. This difference had a small effect size: F (1, 1016) = 68.895, p = .000, eta squared = .064. Therefore, for the second part of this analysis, the null hypothesis (Math Summer Bridge Program students do not graduate at a significantly greater rate than Control Group B students) was rejected. Table 79 One-Way Between-Groups ANOVA Graduation Rates (Experiment v. Control A) Group Experiment M Control A M F Sig. Eta Squared Total .17 .21 .880 .349 - 2015 Cohort .04 .05 .086 .770 - 2014 Cohort .13 .14 .011 .916 - 2013 Cohort .19 .28 1.887 .172 - 2012 Cohort .51 .28 5.787 .022 .054 Table 80 One-Way Between-Groups ANOVA Graduation Rates (Experiment v. Control B) Group Experiment M Control B M F Sig. Eta Squared Exp. v. Control B .26 .07 68.895 .000 .064 Ethnicity Analysis (Non-Native Hawaiian v. Native Hawaiian): Graduation The second part of this question asks the same question, “is there a difference in graduation rates for WCC Math Summer Bridge Participants compared to a) non-MSB summer math students, and b) non-summer math students matched on 7 characteristics?” and adds, “is there a difference in significance based on ethnicity (non-Native Hawaiian v. Native Hawaiian)?” A one-way between-measures ANOVA was conducted to compare graduation rates of the experiment group and each of the two control groups. Table 81 lists the means, F MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 193% scores, significance, and effect size for the Experiment group and Control Group A. Table 82 does the same for the Experiment group and Control Group B. When separating students by ethnicity, Native Hawaiian WCC Math Summer Bridge students had significantly lower graduation rates than their non-MSB counterparts, with a small effect size (Non-NH, F (1, 237) = .371, p = .543; NH, F (1, 256) = 5.826, p = .030, eta squared = .022). Out of the eight (8) experiment groups (4 cohorts x 2 ethnicity groups), only one (Cohort 2013 Native Hawaiian) did significantly worse than its Control Group A counterpart, however this was a large disparity (F (1, 82) = 12.162, p = .004, eta squared = .129). One experiment group had significantly better graduation rates: Cohort 2012 Native Hawaiian, F (1, 48) = 4.547, p = .035, eta squared = .087. The other six comparisons had no significant differences (Cohort 2012 Non-NH, F (1, 51) = 1.123, p = .294; Cohort 2013 Non-NH, F (1, 68) = 1.783, p = .174; Cohort 2014 Non-NH, F (1, 48) = .083, p = .775; Cohort 2014 NH, F (1, 60) = .182, p = .671; Cohort 2015 Non-NH, F (1, 62) = .000, p = 1.000; Cohort 2015 NH, F (1, 58) = .197, p = .659). Therefore, because the experiment students had significantly lower graduation rates for the total Native Hawaiian and 2013 Native Hawaiian groups, the null hypothesis (non-Native Hawaiian Math Summer Bridge Program students do not graduate at a significantly greater rate than non- Native Hawaiian Control Group A students; Native Hawaiian Math Summer Bridge Program students do not graduate at a significantly greater rate than Native Hawaiian Control Group A students) was accepted. Results when comparing the Experiment to Control Group B were much more definitive. For both non-Native Hawaiian and Native Hawaiian groups, the WCC Math Summer Bridge students have significantly higher graduation rates with small to moderate effect sizes (Non-NH, F (1, 394) = 18.411, p < .001, eta squared = .045; NH, F (1, 620) = 51.484, p < .001, eta squared MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 194% = .077). Therefore, since the experiment group has significantly higher graduation rates than Control Group B, the null hypothesis (non-Native Hawaiian Math Summer Bridge Program students do not graduate at a significantly greater rate than non-Native Hawaiian Control Group B students; Native Hawaiian Math Summer Bridge Program students do not graduate at a significantly greater rate than Native Hawaiian Control Group B students) was rejected for both non-Native Hawaiian and Native Hawaiian students. Table 81 Ethnicity Grouping- One-Way Between-Groups ANOVA Graduation Rates (Experiment v. Control A) Group Experiment M Control A M F Sig. Eta Squared Total Non-Native Hawaiian .18 .15 .371 .543 - Native Hawaiian .17 .30 5.826 .030 .022 2015 Cohort Non-Native Hawaiian .06 .06 .000 1.000 - Native Hawaiian .02 .00 .197 .659 - 2014 Cohort Non-Native Hawaiian .15 .13 .083 .775 - Native Hawaiian .12 .17 .182 .671 - 2013 Cohort Non-Native Hawaiian .21 .09 1.783 .174 - Native Hawaiian .17 .52 12.162 .004 .129 2012 Cohort Non-Native Hawaiian .46 .30 1.123 .294 - Native Hawaiian .54 .25 4.547 .037 .087 Table 82 Ethnicity Grouping- One-Way Between-Groups ANOVA Graduation Rates (Experiment v. Control B) Group Experiment M Control B M F Sig. Eta Squared Exp. v. Control B Non-Native Hawaiian .23 .08 18.411 .000 .045 Native Hawaiian .28 .07 51.484 .000 .077 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 195% Age Group Analysis (Traditional v. Nontraditional): Graduation The third part of this question asks, yet again, the same question, “is there a difference in graduation rates for WCC Math Summer Bridge Participants compared to a) non-MSB summer math students, and b) non-summer math students matched on 7 characteristics?” and this time adds, “is there a difference in significance based on age (traditional v. nontraditional)?” A one- way between-measures ANOVA was conducted to compare graduation rates of the experiment group and each of the two control groups. Table 83 lists the means, F scores, significance, and effect size for the Experiment group and Control Group A. Table 84 does the same for the Experiment group and Control Group B. Given the total experiment group had lower graduation rates (with a small effect size) than Control Group A, it is no surprise that the traditional students had similar results (F (1, 276) = 4.456, p = .045, eta squared = .015). The nontraditional experiment group (M = 24%), however, had higher graduation rates than Control Group A (M = 20%), but not significantly so (F (1, 217) = .559, p = .440, eta squared = .003). The only significant difference out of the eight comparisons was for nontraditional students in Cohort 2012 and the experiment group had much higher graduation rates (F (1, 46) = 7.784, p = .008, eta squared = .145). For all seven other comparison groups, there were no statistically significant differences: (a) Cohort 2012 traditional, F (1, 53) = .640, p = .427; Cohort 2013 traditional, F (1, 92) = 2.543, p = .144; (b) Cohort 2013 nontraditional, F (1, 58) = .079, p = .780; (c) Cohort 2014 traditional, F (1, 62) = 2.090, p = .245; (d) Cohort 2014 nontraditional, F (1, 46) = .989, p = .286; (e) Cohort 2015 traditional, F (1, 60) = .185, p = .669; and (f) Cohort 2015 nontraditional, F (1, 60) = .002, p = .968). Therefore, the null hypothesis (traditional Math Summer Bridge Program students do not graduate at a significantly greater rate than traditional Control Group A students) was accepted MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 196% for the total traditional student group because Control Group A has significantly higher graduation rates than the Experiment group. When comparing the experiment group to Control Group B, both traditional and nontraditional students had significantly higher graduation rates (traditional, F (1, 804) = 58.119, p < .001, eta squared = .067; nontraditional, F (1, 210) = 11.400, p = .001, eta squared = .051). Therefore, for these two analyses, the null hypothesis was (traditional Math Summer Bridge Program students do not graduate at significantly greater rate than traditional Control Group B students; nontraditional Math Summer Bridge Program students do not graduate at a significantly greater rate than nontraditional Control Group B students) rejected. Table 83 Age Grouping- One-Way Between-Groups ANOVA Graduation Rates (Experiment v. Control A) Group Experiment M Control A M F Sig. Eta Squared Total Traditional .12 .21 4.456 .045 .015 Nontraditional .24 .20 .599 .440 - 2015 Cohort Traditional .03 .05 .185 .669 - Nontraditional .05 .05 .002 .968 - 2014 Cohort Traditional .05 .15 2.090 .245 - Nontraditional .25 .13 .989 .286 - 2013 Cohort Traditional .13 .26 2.543 .144 - Nontraditional .27 .30 .079 .780 - 2012 Cohort Traditional .30 .35 .640 .427 - Nontraditional .65 .26 7.784 .008 .145 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 197% Table 84 Age Grouping- One-Way Between-Groups ANOVA Graduation Rates (Experiment v. Control B) Group Experiment M Control B M F Sig. Eta Squared Exp. v. Control B Traditional .25 .06 58.119 .000 .067 Nontraditional .28 .10 11.400 .001 .051 Research Question 5 Summary For Research Question 5, there were two very different findings when comparing the Math Summer Bridge students to Control Group A and to Control Group B. Overall, the experiment group had lower mean graduation rates than Control Group A, but only significantly lower for Native Hawaiian and traditional students, and with very low effect size. When further analyzing student graduation rates by cohort, there were three experiment cohorts that have higher graduation rates than their Control Group A counterparts (2012 Total, 2012 Native Hawaiian, and 2012 nontraditional). One Control Group A cohort did better than the experiment group (2013 Native Hawaiian). The rest of the groups had no statistically significant difference in graduation rates. Therefore, with respects to the hypothesis, the null hypothesis was accepted for most comparisons. When comparing the Math Summer Bridge students (experiment) to their non-MSB, matched characteristics counterparts (Control Group B), they have significantly higher graduation rates. The experiment group had medium effect sized significance in every way the data was analyzed (total, ethnicity, and age). Table 85 and Table 86 displays all graduation rates’ means, F values, p values, and eta squared values for the experiment group versus Control Group A and Control Group B respectively. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 198% Table 85 Complete- One-Way Between-Groups ANOVA Graduation Rates (Experiment v. Control A) Ethnicity Total Non-Native Hawaiian .18 .15 .371 .543 - Native Hawaiian .17 .30 5.826 .030 .022 2015 Cohort Non-Native Hawaiian .06 .06 .000 1.000 - Native Hawaiian .02 .00 .197 .659 - 2014 Cohort Non-Native Hawaiian .15 .13 .083 .775 - Native Hawaiian .12 .17 .182 .671 - 2013 Cohort Non-Native Hawaiian .21 .09 1.783 .174 - Native Hawaiian .17 .52 12.162 .004 .129 2012 Cohort Non-Native Hawaiian .46 .30 1.123 .294 - Native Hawaiian .54 .25 4.547 .037 .087 Group Experiment M Control A M F Sig. Eta Squared Total Group Total .17 .21 .880 .349 - 2015 Cohort .04 .05 .086 .770 - 2014 Cohort .13 .14 .011 .916 - 2013 Cohort .19 .28 1.887 .172 - 2012 Cohort .51 .28 5.787 .022 .054 Age Total Traditional .12 .21 4.456 .045 - Nontraditional .24 .20 .599 .440 - 2015 Cohort Traditional .03 .05 .185 .669 - Nontraditional .05 .05 .002 .968 - 2014 Cohort Traditional .05 .15 2.090 .245 - Nontraditional .25 .13 .989 .286 - 2013 Cohort Traditional .13 .26 2.543 .144 - Nontraditional .27 .30 .079 .780 - 2012 Cohort Traditional .30 .35 .640 .427 - Nontraditional .65 .26 7.784 .008 .145 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 199% Table 86 Complete - One-Way Between-Groups ANOVA Graduation Rates (Experiment v. Control B) Group Experiment M Control B M F Sig. Eta Squared Total Group .26 .07 68.895 .000 .064 Ethnicity Non-Native Hawaiian .23 .08 18.411 .000 .045 Native Hawaiian .28 .07 51.484 .000 .077 Age Traditional .25 .06 58.119 .000 .067 Nontraditional .28 .10 11.400 .001 .051 Research Question 6 Results: Cumulative GPA The final research question, Research Question 6, addresses the last student success measure, cumulative GPA. Research Question 6 aims to determine effects of the WCC Math Summer Bridge Program on student cumulative grade point averages (GPA) by comparing the MSB students (experiment group) to two different control groups. Control Group A consists of students who took the same developmental math course at Windward Community College over the same summer, but were not a part of the Math Summer Bridge Program (N=203). Control Group B (N = 460) consists of students who did not take a developmental math course over the summer, but instead were matched to individual MSB participants based on seven (7) characteristics: (1) First semester enrolled; (2) Current enrollment status (full- or part-time); (3) Age; (4) Ethnicity (Native Hawaiian or non-Native Hawaiian); (5) Sex; (6) Pell Grant recipient (or not); and (7) Compass placement test score. This research question is a three part, multi-level question asking: “Do the participants in the WCC Math Summer Bridge Program have higher cumulative GPAs than a) non-MSB summer math students, or b) non-summer math students matched on seven (7) characteristics?” Also, “is there a significant difference between Native Hawaiian and non-Native Hawaiian students?” and “is there a difference between traditional and MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 200% nontraditional students?” For the purpose of this study, cumulative GPA was computed using all courses taken at Windward Community College, including developmental or remedial classes. In order to address this research question and determine whether or not the null hypothesis can be rejected (that the WCC Math Summer Bridge Program does not have any impact on cumulative GPA), the researcher completed six (6) one-way between-measures analysis of variance tests (one for total sample population, one grouped by ethnicity, and one grouped by age; each of these performed to compare the experiment to the two control groups). The aim of these one-way ANOVA were to “compare the variance (variability in scores) between the different groups [pre v. post 2015 tests and post 2015 v. post 2014 v. post 2013 v. post 2012 tests] (believed to be due to the independent variable [WCC Math Summer Bridge Program]) with the variability within each of the groups (believed to be due to chance)” (Pallant, 2013, p. 258). One-way between-groups ANOVAs were used to determine if there were any significant differences between the mean scores of the experiment group to the mean scores of each control group. For the purposes of this section, data will be presented in sections by grouping of student population; 1) total student population, 2) ethnicity (non-Native Hawaiian v. Native Hawaiian), and 3) age (traditional v. nontraditional). Total Group Analysis: Cumulative GPA First, to answer the question, “is there a difference in cumulative GPAs for WCC Math Summer Bridge Participants compared to a) non-MSB summer math students, and b) non- summer math students matched on 7 characteristics?” a one-way between-measures ANOVA was conducted to compare cumulative GPAs of the experiment group and each of the two control groups. Table 87 lists the means, F scores, significance, and effect size for the Experiment group and Control Group A. Table 88 does the same for the Experiment group and MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 201% Control Group B. When comparing the total Math Summer Bridge student group (experiment) to Control Group A, there was no significant difference between cumulative GPAs (F (1, 495) = .051, p = .822). Even when separating students by cohort, the experiment group generally has higher cumulative GPAs, but they aren’t statistically significant: (a) Cohort 2012, F (1, 101) = 1.168, p = .240; (b) Cohort 2013, F (1, 152) = 2.816, p = .095; (c) Cohort 2014, F (1, 110) = .006, p = .936; and (d) Cohort 2015, F (1, 122) = 1.637, p = .203. Therefore, the null hypothesis (the Math Summer Bridge Program students do not have significantly higher cumulative GPAs than Control Group A students) was accepted. The Math Summer Bridge students (experiment) had significantly higher cumulative GPAs (M = 2.91) than their Control Group B counterparts (M = 1.90) with a very large effect size (F (1, 1016) = 230.487, p = .000, eta squared = .185). Therefore, the null hypothesis (the Math Summer Bridge Program students do not have significantly higher cumulative GPAs than Control Group B students) was rejected; the experiment group has significantly higher cumulative GPAs than Control Group B students. Table 87 One-Way Between-Groups ANOVA Cumulative GPA (Experiment v. Control A) Group Experiment M Control A M F Sig. Eta Squared Total 2.73 2.71 .051 .822 - 2015 Cohort 2.92 2.66 1.637 .203 - 2014 Cohort 2.64 2.62 .006 .936 - 2013 Cohort 2.60 2.84 2.816 .095 - 2012 Cohort 2.81 2.62 1.168 .240 - MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 202% Table 88 One-Way Between-Groups ANOVA Cumulative GPA (Experiment v. Control B) Group Experiment M Control B M F Sig. Eta Squared Exp. v. Control B 2.91 1.90 230.487 .000 .185 Ethnicity Analysis (Non-Native Hawaiian v. Native Hawaiian): Cumulative GPA The second part of this question looks at differences when grouping the students based on ethnicity. It asks the same question, “is there a difference in cumulative GPAs for WCC Math Summer Bridge Participants compared to a) non-MSB summer math students, and b) non- summer math students matched on 7 characteristics?” and adds, “is there a difference in significance based on ethnicity (non-Native Hawaiian v. Native Hawaiian)?” A one-way between-measures ANOVA was conducted to compare cumulative GPAs of the experiment group and each of the two control groups. Table 89 lists the means, F scores, significance, and effect size for the Experiment group and Control Group A. Table 90 does the same for the Experiment group and Control Group B. When separating students by ethnicity, there were no significant differences in cumulative GPA for either non-Native Hawaiian (F (1, 237) = .271, p = .603) or Native Hawaiian students (F (1, 256) = .408, p = .524). The same is true when dividing students by years/cohorts and by ethnicity; there are no significant differences in cumulative GPA between the experiment group and Control Group A: (a) Cohort 2012 non-NH, F (1, 51) = .651, p = .288; (b) Cohort 2012 NH, F (1, 48) = 1.146, p = .290; (c) Cohort 2013 non-NH, F (1, 68) = .852, p = .359; (d) Cohort 2013 NH, F (1, 82) = 2.360, p = .128; (e) Cohort 2014 non-NH, F (1, 48) = .059, p = .810; (f) Cohort 2014 NH, F (1, 60) = .654, p = .422; (g) Cohort 2015 non-NH, F (1, 62) = 1.805, p = .184; and (h) Cohort 2015 NH, F (1, 58) = 1.868, p = .177). Therefore, when MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 203% comparing the cumulative GPAs of the experiment group and Control Group A, with specific ethnicity grouping, there was no significant difference between the groups, and the null hypothesis (non-Native Hawaiian Math Summer Bridge Program students do not have significantly higher cumulative GPAs than non-Native Hawaiian Control Group A students; Native Hawaiian Math Summer Bridge Program students do not have higher cumulative GPAs than Native Hawaiian Control Group A students) was accepted. When comparing the Math Summer Bridge students (experiment group) to Control Group B, there were statistically significant differences for both non-Native Hawaiian (F (1, 394) = 96.046, p < .001, eta squared = .196) and Native Hawaiian students (F (1, 620) = 139.240, p < .001, eta squared = .183). The MSB students had much higher cumulative GPAs than their Control Group B counterparts, and therefore, the null hypothesis (non-Native Hawaiian Math Summer Bridge Program students do not have significantly higher cumulative GPAs than non-Native Hawaiian Control Group B students; Native Hawaiian Math Summer Bridge Program students do not have significantly higher cumulative GPAs than Native Hawaiian Control Group B students) was rejected. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 204% Table 89 Ethnicity Grouping- One-Way Between-Groups ANOVA Cumulative GPA (Experiment v. Control A) Group Experiment M Control A M F Sig. Eta Squared Total Non-Native Hawaiian 2.85 2.78 .271 .596 - Native Hawaiian 2.65 2.57 .408 .524 - 2015 Cohort Non-Native Hawaiian 3.12 2.79 1.805 .184 - Native Hawaiian 2.79 2.26 1.868 .177 - 2014 Cohort Non-Native Hawaiian 2.86 2.80 .059 .810 - Native Hawaiian 2.52 2.27 .654 .422 - 2013 Cohort Non-Native Hawaiian 2.58 2.80 .852 .359 - Native Hawaiian 2.61 2.91 2.360 .128 - 2012 Cohort Non-Native Hawaiian 2.92 2.67 .651 .288 - Native Hawaiian 2.75 2.55 1.146 .290 - Table 90 Ethnicity Grouping- One-Way Between-Groups ANOVA Cumulative GPA (Experiment v. Control B) Group Experiment M Control B M F Sig. Eta Squared Exp. v. Control B Non-Native Hawaiian 3.09 2.09 96.046 .000 .196 Native Hawaiian 2.79 1.78 139.240 .000 .183 Age Group Analysis (Traditional v. Nontraditional): Cumulative GPA Finally, the third part of this question asks, yet again, the same question, “is there a difference in cumulative GPAs for WCC Math Summer Bridge Participants compared to a) non- MSB summer math students, and b) non-summer math students matched on 7 characteristics?” and this time adds, “is there a difference in significance based on age (traditional v. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 205% nontraditional)?” A one-way between-measures ANOVA was conducted to compare cumulative GPAs of the experiment group and each of the two control groups. Table 91 lists the means, F scores, significance, and effect size for the Experiment group and Control Group A. Table 92 does the same for the Experiment Group and Control Group B. Overall, there were no significant differences between the experiment group and Control Group A for traditional (F (1, 276) = .194, p = .660) or nontraditional students (F (1, 217) = .006, p = .940). When further analyzing the data by cohort, six (6) out of eight (8) comparisons did not have any significant differences in cumulative GPA: (a) Cohort 2012 traditional, F (1, 53) = .024, p = .877; (b) Cohort 2013 traditional, F (1, 92) = 2.855, p = .094; (c) Cohort 2013, nontraditional, F (1, 58) = .272, p = .604; (d) Cohort 2014 traditional, F (1, 62) = .042, p = .861; (e) Cohort 2014 nontraditional, F (1, 46) = .000, p = .995; and (f) Cohort 2015 nontraditional, F (1, 60) = .547, p = .462). The last two measures, however did have significantly different cumulative GPAs, in which the experiment group had higher cumulative GPAs than the Control Group A students: Cohort 2015 traditional students, F (1, 60) = 9.024, p = .004, eta squared = .131; Cohort 2012 nontraditional students, F (1, 46) = 3.414, p = .033, eta squared = .069. Since there were two cohorts that had significantly higher cumulative GPAs, the null hypotheses (traditional Math Summer Bridge Program students do not have significantly higher cumulative GPAs than traditional Control Group A students; nontraditional Math Summer Bridge Program students do not have significantly higher cumulative GPAs than nontraditional Control Group A students) were rejected. When grouping students by age, both traditional and nontraditional students who participated in the WCC Math Summer Bridge Program had significantly higher cumulative GPAs compared to their counterparts from Control Group B (Traditional, F (1, 804) = 209.339, p MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 206% < .001, eta squared = .207; nontraditional, F (1, 210) = 29.200, p < .001, eta squared = .122). Therefore, the null hypothesis (traditional Math Summer Bridge Program student do not have significantly higher cumulative GPAs than traditional Control Group B students; nontraditional math Summer Bridge Program students do not have significantly higher cumulative GPAs than nontraditional Control Group B students) was rejected when comparing the experiment group to Control Group B. Table 91 Age Grouping- One-Way Between-Groups ANOVA Cumulative GPA (Experiment v. Control A) Group Experiment M Control A M F Sig. Eta Squared Total Traditional 2.60 2.55 .194 .660 - Nontraditional 2.88 2.89 .006 .940 - 2015 Cohort Traditional 3.12 2.38 9.024 .004 .131 Nontraditional 2.73 2.97 .547 .462 - 2014 Cohort Traditional 2.38 2.32 .042 .861 - Nontraditional 2.99 2.99 .000 .995 - 2013 Cohort Traditional 2.45 2.76 2.855 .094 - Nontraditional 2.85 2.97 .272 .604 - 2012 Cohort Traditional 2.56 2.59 .024 .877 - Nontraditional 3.13 2.66 3.414 .033 .069 Table 92 Age Grouping- One-Way Between-Groups ANOVA Cumulative GPA (Experiment v. Control B) Group Experiment M Control B M F Sig. Eta Squared Exp. v. Control B Traditional 2.86 1.78 209.339 .000 .207 Nontraditional 3.09 2.37 29.200 .000 .122 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 207% Research Question 6 Summary The means of cumulative GPAs of Math Summer Bridge Program students were generally higher than those of Control Group A for all three ways of analysis (total, ethnicity, and age), however generally these differences weren’t statistically significant. There were only two (2) groups out of the twenty-five (25) analyzed that had statistically significant differences between cumulative GPAs. In both Cohort 2015 traditional students (F (1, 60) = 9.024, p = .004, eta squared = .131) and Cohort 2012 nontraditional students (F (1, 46) = 3.414, p = .033, eta squared = .069), it was the experiment group that had higher cumulative GPAs than the Control Group A students. The Math Summer Bridge students’ (experiment group) cumulative GPAs were significantly higher than those matched based on 7 characteristics (Control Group B). Every way of comparison (total, ethnicity, and age) had large effect sizes. Therefore, it is concluded that the null hypothesis be rejected. Table 93 and Table 94 displays all graduation rates’ means, F values, p values, and eta squared values for the experiment group versus Control Group A and Control Group B respectively. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 208% Table 93 Complete- One-Way Between-Groups ANOVA Cumulative GPA (Experiment v. Control A) Ethnicity Total Non-Native Hawaiian 2.85 2.78 .271 .596 - Native Hawaiian 2.65 2.57 .408 .524 - 2015 Cohort Non-Native Hawaiian 3.12 2.79 1.805 .184 - Native Hawaiian 2.79 2.26 1.868 .177 - 2014 Cohort Non-Native Hawaiian 2.86 2.80 .059 .810 - Native Hawaiian 2.52 2.27 .654 .422 - 2013 Cohort Non-Native Hawaiian 2.58 2.80 .852 .359 - Native Hawaiian 2.61 2.91 2.360 .128 - 2012 Cohort Non-Native Hawaiian 2.92 2.67 .651 .288 - Native Hawaiian 2.75 2.55 1.146 .290 - Group Experiment M Control A M F Sig. Eta Squared Total Group Total 2.73 2.71 .051 .822 - 2015 Cohort 2.92 2.66 1.637 .203 - 2014 Cohort 2.64 2.62 .006 .936 - 2013 Cohort 2.60 2.84 2.816 .095 - 2012 Cohort 2.81 2.62 1.168 .240 - Age Total Traditional 2.60 2.55 .194 .660 - Nontraditional 2.88 2.89 .006 .940 - 2015 Cohort Traditional 3.12 2.38 9.024 .004 .131 Nontraditional 2.73 2.97 .547 .462 - 2014 Cohort Traditional 2.38 2.32 .042 .861 - Nontraditional 2.99 2.99 .000 .995 - 2013 Cohort Traditional 2.45 2.76 2.855 .094 - Nontraditional 2.85 2.97 .272 .604 - 2012 Cohort Traditional 2.56 2.59 .024 .877 - Nontraditional 3.13 2.66 3.414 .033 .069 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 209% Table 94 Complete - One-Way Between-Groups ANOVA Cumulative GPA (Experiment v. Control B) Group Experiment M Control B M F Sig. Eta Squared Total Group 2.91 1.90 230.487 .000 .185 Ethnicity Non-Native Hawaiian 3.09 2.09 96.046 .000 .196 Native Hawaiian 2.79 1.78 139.240 .000 .183 Age Traditional 2.86 1.78 209.339 .000 .207 Nontraditional 3.09 2.37 29.200 .000 .122 Summary of Results Chapter 4 consisted of two parts. The first part provided descriptive statistics of the experiment and control groups’ demographics as well as input characteristics by a) ethnicity (non-Native Hawaiian v. Native Hawaiian) and b) age (traditional v. nontraditional). For both the experiment and control groups, proportions of males and females were similar to the total Windward Community College student population as well as national averages for community colleges. They had a significantly higher proportion of Native Hawaiian students in the experiment group and Control Group B, and larger proportions of full-time students than WCC and national averages. Of the six scales used in this study (pre- and post-surveys for College Self-Efficacy Scale, pre- and post-surveys for Sense of Belonging Inventory-Antecedents, and pre- and post-surveys for Sense of Belonging Inventory-Psychology), four of them had strong internal consistency and the other two had moderate internal consistency, reinforcing the researcher’s decision to use the data collected. The second part of Chapter 4 addressed each of the six (6) research questions. The first three research questions measured the potential direct impact of the Math Summer Bridge Program on three types of indicators for student success (1) college self-efficacy, 2) sense of MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 210% belonging; 3) engagement and satisfaction). The last three research questions compared Math Summer Bridge Program participants to two different control groups on three different student success measures (persistence, graduation, and cumulative GPA). For Research Question 1 (Is there a significant change in college self-efficacy for students who participated in the Math Summer Bridge Program? Is this change maintained over time?), it was found that the Windward Community College Math Summer Bridge program has a statistically significant impact on students’ college self-efficacy, especially for traditional students, and that this increase is long lasting. Although mean scores did increase between Time 1 (pre-survey) and Time 2 (post-survey), the data did not show any significant impact on sense of belonging (SOBI-A or SOBI-P) for Research Question 2 (Is there a significant change in sense of belonging for students who participated in the Math Summer Bridge Program? Is this change maintained over time?). In answering Research Question 3 (Is there a significant change in engagement and satisfaction for students who participated in the Math Summer Bridge Program? Is this change maintained over time?), all scores went up for all five (5) engagement questions, however they only significantly increased for academic peer interaction, and total engagement. Non-Native Hawaiian and nontraditional students especially had significant increases in academic peer interaction and were long-lasting. All four satisfaction questions (Instructor, Support Services, Overall WCC, and Total) increased between the pre- and post-surveys and were found to be especially significant for Native Hawaiian students and both traditional and nontraditional students. For all of these measures, the increases that occurred because of the intervention were maintained over the following three years that were measured for this study. The last three questions, measuring student success based on persistence, graduation, and cumulative GPA, found two major themes. When comparing the MSB students (experiment) to MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 211% students who took summer math during the same year, but were not a part of the MSB Program (Control Group A), the Experiment Group persisted at significantly higher rates, graduated at comparable rates, and had comparable cumulative GPAs. The experiment group persisted at significantly higher rates than Control Group A for all identified groups (total, non-Native Hawaiian, traditional, and nontraditional). It is also noted that the 2012 cohort experiment group (3 years post intervention) had significantly higher graduation rates than their Control Group A counterparts; the only significantly different statistic for Research Question 5 (Do the Math Summer Bridge Program students graduate at a greater rate than non-Math Summer Bridge Program students?). The second control group used in answering Research Questions 4-6 consisted of 460 students who matched 126 of the 294 Math Summer Bridge Program students on seven (7) characteristics: (1) First semester enrolled; (2) Current enrollment status (full- or part-time); (3) Age; (4) Ethnicity (Native Hawaiian or non-Native Hawaiian); (5) Sex; (6) Pell Grant recipient (or not); and (7) Compass placement test score. It is hypothesized that Control Group B is a closer match to experiment students and there were very promising results. For all three questions, the experiment group had significantly better persistence rates, graduation rates, and cumulative GPAs for all groups analyzed (total, non-Native Hawaiian, Native Hawaiian, traditional, and nontraditional). In conclusion, a fairly large sample (N = 957) was used in this study to determine the effects of a math summer bridge program on community college student success. Overall, it has shown to have positive significant effects on college self-efficacy, engagement, and satisfaction. Math Summer Bridge students do just as well, if not significantly better on all three student success measures identified in this study (persistence rates, graduation rates, and cumulative MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 212% GPA) compared to two different control groups. Chapter 5 will further discuss the results of this study and provide hypotheses for any shortcomings. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 213% CHAPTER 5 DISCUSSION As mentioned in the beginning of this study, there is a critical need for college-educated individuals to sustain not only their own personal wealth, but also to perpetuate the growth of our economy (Baum, et al., 2013). However, especially for individuals coming from lower socioeconomic backgrounds, there is a discrepancy in the quality of education earned in K-12 and the level needed in order to succeed in college (Darling-Hammond, 2007; Eger & McDonald, 2012). In order to help students prepare for college-level math, community colleges (and some 4-year institutions) provide developmental/remedial coursework. Unfortunately, across the nation, the success rates for these individual courses (passing the individual course and completing remediation), as well as the success rates of the student as a whole (e.g. persistence or graduation), are dismal (Bailey, et al., 2009; Bahr, 2012). With nationwide state budget cuts to public community college funding, it is not only an ethical obligation, but also a fiscal responsibility for institutions to create innovative ways in helping students complete remediation, and ultimately earn their college degree. The purpose of this study was to investigate the effectiveness of a math summer bridge program at Windward Community College, more specifically, if it has positive effects on six (6) key measures for student success: (1) college self-efficacy; (2) sense of belonging; (3) engagement and satisfaction; (4) persistence; (5) graduation; and (6) cumulative GPA. This study also wants to go further in investigating if these effects are different for Native Hawaiian or nontraditional students. The first three items were chosen because they have been proven to be psychological, cognitive, and behavioral characteristics that lead to college success (Astin, 1993, 1999; Bandura, 1986, 2001; DeWitz, et al., 2009; Gore, 2006; Hall & Ponton, 2005; MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 214% Jameson & Fusco, 2014; Majer, 2009; Martin & Dowson, 2009; Multon, et al., 1991; Wright, et al., 2012). The last three are outcomes used to measure college success. In addition, special consideration was given to a) Native Hawaiian students and b) nontraditional students because both of these populations are more likely to place into developmental/remedial math courses and less likely to be successful in completing remediation and earning a college degree (Bahr, 2010; OHA 2014). Also, at Windward Community College, each group makes up the majority of the population (ethnicity and age), and nationally, nontraditional students make up the majority of the community college student body (AACC, 2015). This two-part study consists of a pretest/posttest and cross-sectional design with comparisons to two different control groups. There were 957 students involved; 294 students participated in the Math Summer Bridge (MSB) Program (Experiment Group), 203 students in Control Group A (non-MSB summer math students), and 460 students included in Control Group B (non-summer math students matched on 7 characteristics). Using a researcher-created survey, including items from the College Self-Efficacy Inventory (Solberg, et al., 1993) and Sense of Belonging Instrument (Hagerty & Patusky, 1995), the six (6) different student success indicators were measured pre/post intervention for the 2015 MSB cohort, and then post-survey answers were compared for all students in the four cohorts (2012, 2013, 2014, and 2015) as a cross- sectional study to explore the longevity of the changes acquired through participating in the WCC Math Summer Bridge Program. Multilinear regressions and one-way between groups ANOVA statistical techniques were used to answer the six research questions. The first three measure mindset and/or behaviors that lead to student success (college self-efficacy, sense of belonging, and engagement and satisfaction). Data was collected from the experiment group (n = 165) to address these three questions: MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 215% •! RQ1. Is there a significant change in college self-efficacy for students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 1A. Is there a significant change in college self-efficacy for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 1B. Is there a significant change in college self-efficacy for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? •! RQ 2. Is there a significant change in sense of belonging for students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 2A. Is there a significant change in sense of belonging for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 2B. Is there a significant change in sense of belonging for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? •! RQ 3. Is there a significant change in engagement and satisfaction for students who participated in the Math Summer Bridge Program? Is this change maintained over time? o! RQ 3A. Is there a significant change in student engagement and satisfaction for Native Hawaiian or non-Native Hawaiian students who participated in the Math Summer Bridge Program? Is this change maintained over time? MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 216% o! RQ 3B. Is there a significant change in student engagement and satisfaction for nontraditional or traditional students who participated in the Math Summer Bridge Program? Is this change maintained over time? The following three questions measure student success through persistence, graduation, and cumulative grade point average (GPA). Data was collected from the experimental group (n = 294), Control Group A (n = 203), and Control Group B (n = 460) to address these questions: •! RQ 4. Do the Math Summer Bridge Program students persist at a greater rate than non-Math Summer Bridge Program students? o! RQ 4A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students persist at a greater rate than their non-Math Summer Bridge Program counterparts? o! RQ 4B. Do the nontraditional or traditional Math Summer Bridge Program students persist at a greater rate than their non-Math Summer Bridge Program counterparts? •! RQ 5. Do the Math Summer Bridge Program students graduate at a greater rate than non- Math Summer Bridge Program students? o! RQ 5A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students graduate at a greater rate than their non-Math Summer Bridge Program counterparts? o! RQ 5B. Do the nontraditional or traditional Math Summer Bridge Program students graduate at a higher rate than their non-Math Summer Bridge Program counterparts? •! RQ 6. Do the Math Summer Bridge Program students have a higher cumulative grade point average (GPA) than non-Math Summer Bridge Program students? MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 217% o! RQ 6A. Do the Native Hawaiian or non-Native Hawaiian Math Summer Bridge Program students have higher cumulative GPAs than their non-Math Summer Bridge Program counterparts? o! RQ 6B. Do the nontraditional or traditional Math Summer Bridge Program students have higher cumulative GPAs than their non-Math Summer Bridge Program counterparts? Discussion of Findings The major findings identified in this study (presented in Table 95) are grouped into two main categories: 1) direct effects that the WCC Math Summer Bridge has on student attitudes, beliefs, and behaviors, and 2) WCC Math Summer Bridge students’ performance compared to two different control groups. Overall, there were significant changes in college self-efficacy, and engagement and satisfaction. The WCC Math Summer Bridge students (experiment group) also had significantly higher persistence rates than Control Group A students, and had similar graduation rates and cumulative GPAs. However, when comparing the experiment group to Control Group B, they had significantly better scores on all three success measures (persistence, graduation, and cumulative GPA). The following section will further discuss the findings on each of the six research questions. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 218% Table 95 Total Study Results Note. “X” the experiment group has statistically significantly higher scores than the control group (p < .05). “-X” the experiment group has statistically significantly lower scores than the control group (p < .05). Attitudes/Beliefs/Behaviors Performance CSE SOBI Engagement and Satisfaction Performance Pre/Post CSE Maintained Pre/Post SOBI-A Maintained Pre/Post SOBI-P Maintained Pre/Post Instructor Interaction In Class Maintained/Increased over time Pre/Post Instructor Interaction Outside Class Maintained/Increased over time Pre/Post Academic Peer Interaction Maintained Pre/Post Social Peer Interaction Maintained Pre/Post Total Engagement Maintained Pre/Post Instructor Satisfaction Maintained Pre/Post Support Services Satisfaction Maintained Pre/Post Overall WCC Satisfaction Maintained/Increased over time Pre/Post Total Satisfaction Maintained Persistence v. Control A Persistence v. Control B Graduation v. Control A Graduation v. Control B Cumulative GPA v. Control A Cumulative GPA v. Control B Total Cohort X X X X X X X X X X X X X X X X X X Non-Native Hawaiian X X X X X X X X X Native Hawaiian X X X X X X X X X X X X - X X X Traditional X X X X X X X X X X - X X X Nontraditional X X X X X X X X X X Running&head:&MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF<EFFICACY&&&&&&&&&&&&&&&&&& & & Research Question 1 Findings: Impact on College Self-Efficacy After completing one-way ANOVAs, it was found that the Windward Community College Math Summer Bridge Program had a significant positive effect on students’ college self- efficacy and these changes were long lasting. Significant changes were found for the total experiment population (Time 1, M = 153; Time 2, M = 161.13) as well as for traditional students (Time 1, M = 143.50; Time 2, M = 156.59). Although pre/post-survey scores were not found to be significant for non-Native Hawaiian, Native Hawaiian, and nontraditional students, all College Self-Efficacy Inventory scores went up from pre- to post-survey and did not significantly drop over time. The researcher concludes that this intervention has an equally significant effect on all populations, no matter their ethnicity (non-Native Hawaiian or Native Hawaiian) or age (traditional or nontraditional). It is also important to note, when looking at the individual means for CSE scores, non-Native Hawaiian students (Time 1, M = 154.04; Time 2, M = 162.23) and nontraditional students (Time 1, M = 163.00; Time 2, M = 165.68) had higher overall means than the total group. Therefore, the intervention had the largest effects on students with the lowest college self-efficacy scores, which in the opinion of the researcher, is a key component of an effective intervention; being able to help the students most in need. Research Question 2 Findings: Impact on Sense of Belonging According to the pre- and post-survey responses, the WCC Math Summer Bridge Program had no effect on sense of belonging (Affective or Psychological components). However, the researcher questions whether or not this tool or all of the questions in this inventory were appropriate to use for this study/population because the lack of results conflicts with those found in Research Question 3 (student engagement and satisfaction), where there were significant positive increases. Instead of accepting the null hypothesis (the intervention has MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY& & & 220& no effect on sense of belonging), the researcher believes that some of the questions asked in this instrument were not relevant to what the students’ perceived to be important. For example, after the intervention, the SOBI-A question, “It is important that I fit in at school” decreased and did not have many students strongly agree with this statement (strongly agree = 4, M = 2.43). Especially at a commuter campus, where students spend limited amounts of time there (compared to campuses where students live on campus), these students might not particularly care if they “fit in” and instead care more about understanding course content and working towards earning a degree. Also, many of these students are nontraditional and are coming to college for the sole purpose in earning a degree, so it doesn’t matter to them if they “fit” in college, as long as they’re getting a quality education and working towards earning their degree. Previous research links sense of belonging to motivation, achievement, and influence on behavior (e.g. engagement with peers and/or instructors) (Bozak, 2013; Hagerty, et al., 1992; Hagerty & Patusky, 1995; Strayhorn, 2012), and since these same students who have low and unchanging Sense of Belonging Inventory scores also have high achievement (compared to their peers in two control groups) and increasing positive college behaviors (e.g. engagement with instructors and peers), there is a disconnect between the measure used (Sense of Belonging Instrument) and what is trying to be measured (students’ sense of belonging). One third of the questions (6 out of 18) included in the SOBI pertain to “fitting in” at college. Instead, the researcher hypothesizes that using questions focused on feeling welcome and/or comfortable on campus would render different responses. Looking back to the original definition of sense of belonging, the focus is on feeling “cared about, accepted, respected, valued by, and important to the group (e.g., campus community) or others on campus (e.g. faculty, peers)” (Strayhorn, 2012, p. 3). Therefore, even though the Sense of Belonging Instrument has been a proven tool for MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY& & & 221& community college students, the verbiage used might not be the most appropriate when trying to link the underlying thoughts and perceptions with the true concept of sense of belonging. Research Question 3 Findings: Impact on Student Engagement and Satisfaction When analyzing the change in student engagement and satisfaction, there were positive, statistically significant increases for the experiment group as a whole, which resulted in rejecting the null hypothesis (the WCC Math Summer Bridge Program had no effect on student engagement and satisfaction). All scores for student engagement (5 questions 5 different ways of analysis-total, ethnicity, and age), had increases from pre- to post-survey responses, and positive increases years after the intervention. There were statistically significant increases for total engagement for the entire 2015 cohort, as well as for non-Native Hawaiian students. There were also statistically significant increases in academic peer interaction for the entire 2015 Cohort, as well as for non-Native Hawaiian and nontraditional students. Also, although immediately after the intervention, instructor interaction (in and outside of class) did not have statistically significant increases, it did significantly increase over the next three years. Three years post intervention (2012 Cohort), 100% of the students who responded to the survey reported that they interact with their instructors 4+ times/week in class (highest option available on the survey). This study shows that the WCC Math Summer Bridge Program creates an environment where students get used to and feel comfortable with interacting with their instructors and peers, including asking questions, and seeking help when needed, a skillset shown to increase the likelihood of not only college success, but also in life in general (Dweck, 2008). Similar to student engagement, the WCC Math Summer Bridge Program had significant positive effects on student satisfaction, especially for Native Hawaiian and traditional students. MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY& & & 222& This satisfaction persisted over time; with non-Native Hawaiian and nontraditional students having 100% of the respondents “agree” or “strongly agree” with support services satisfaction and overall WCC satisfaction. The researcher believes that strong satisfaction is closely related to the high levels of college self-efficacy and engagement. When students feel comfortable asking questions, seeking help when needed, and overall have confidence in their abilities as a college student, they have positive interactions with their instructors, support services (e.g. counselors or financial aid offices), and the campus in general. These interactions and experiences are dynamic and have a cyclical relationship: the more students have positive interactions with their instructors/peers/supporting faculty, the more comfortable they feel asking for help when needed, the more likely they are to be successful in their classes, which in turn makes them more willing to interact with an increased number of faculty and peers. Therefore, the WCC Math Summer Bridge Program had a positive effect on student engagement, but it doesn’t seem to affect one specific population (ethnicity or age group) more than another. It is concluded that the MSB students acquired or began to acquire the skill sets, habits, and mindsets of positive engagement, satisfaction, and college self-efficacy that are necessary to succeed in school, which is reflected in the experiment groups’ performance on three different student success measures (persistence, graduation, and cumulative GPA) when comparing them to two different control groups. Research Question 4 Findings: Impact on Persistence Research Question 4 provided concrete evidence that students in the Windward Community College Math Summer Bridge Program persist at greater rates than their peers. Control Group A participants were chosen because these students were in similar points in their academic careers, and were taking the same level of developmental math as their MSB MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY& & & 223& counterparts. However, this is the only matched similarity. It is important to note that students usually do not receive financial aid over the summer because they used it all during the fall and spring semesters. Therefore, students who enroll in summer school are more likely to have the financial means to pay for school on their own and be at a higher socioeconomic status than those in the experiment group. For each of the four cohorts (2012, 2013, 2014, 2015), as well as looking at the overall experiment v. control group, the MSB students persisted at higher rates. This was significant for the total experiment group (67% persistence rate) v. Control Group A (50% persistence rate) and the 2012 Cohort (56% persistence rate) v. Control Group A (23% persistence rate). It is believed that although this is a positive measure, the higher percentage of MSB students still enrolled in school attributes to the lack of significant differences in graduation rates found in Research Question 5’s Control Group A comparison. Control Group B is a more reliable comparison group to the experiment group because not only are they matched on the course level they placed into (Compass placement score), they are also matched on six other characteristics that research has shown to be linked to student success: (1) first semester enrolled; (2) current enrollment status (full- or part-time); (3) age; (4) ethnicity (Native Hawaiian or non-Native Hawaiian); (5) sex; and (6) Pell Grant recipient (reflecting economic status). The results of this analysis reveled that the experiment group have significantly higher persistence rates (67%) than their Control Group B counterparts (26%). When comparing the WCC Math Summer Bridge students to both control groups, every MSB population had higher persistence rates and all but one (traditional students to Control Group A) were statistically significant. Therefore, these findings reiterate that the WCC Math Summer Bridge Program intervention is a program that equally benefits all students, no matter their age or if they are Native Hawaiian or not. The majority of the WCC Math Summer Bridge MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY& & & 224& students placed into the lowest level of remedial math (3 levels below college level) (62.9%), making them the least likely to persist (Bahr, 2012), however they (one year post-intervention, MSB 2014 Cohort persistence rate of 62%) are persisting at higher rates than the total student body at Windward Community College (Fall to Fall persistence rate of 45.38%) (Windward Community College, 2014). Therefore, the WCC Math Summer Bridge students are gaining both academic and soft skills (e.g. study habits, interpersonal skills) that are supporting their successful persistence in school. Research Question 5 Findings: Impact on Graduation Research Question 5 sought to find out if the Windward Community College Math Summer Bridge Program had a positive effect on graduation rates compared to two different control groups. First, when comparing the experiment group to Control Group A (students who took the same developmental math course in the same summer, but not in the MSB Program), three out of the five groups (Total, non-Native Hawaiian, and nontraditional), did not have any significant differences in graduation rates. Although Native Hawaiian and traditional students in the experiment group had significantly lower graduation rates than their Control Group A counterparts, when comparing each of these groups by cohort, for Native Hawaiian students, there was only one year (2013) with a significantly lower graduation rate, and there were no significantly lower graduation rates for any cohort for traditional students. The 2012 cohort as a whole had significantly higher graduation rates (51%) than their Control Group A counterparts (28%). The WCC Math Summer Bridge students had higher graduation rates than their Control Group B counterparts. For every single group (Total, Non-Native Hawaiian, Native Hawaiian, Traditional, and Nontraditional), the experiment group had statistically significant graduation MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY& & & 225& rates (p ≤ .001) with averages ranging from 23% to 28%, compared to Control Group B’s 6% to 10%, which is similar to Windward Community College’s graduation rate of 12% (Windward Community College, 2014). In comparing the 2012 Cohort’s graduation rate of 51% to Windward Community College’s 3-year graduation rate of 12%, there’s a clear difference in the success of the experiment group, not only to it’s more reliable control group (Control Group B), but also to the general campus’ graduation rates. Therefore, the WCC Math Summer Bridge Program has a significant impact on graduation rates as well. Research Question 6 Findings: Impact on Cumulative GPA Finally the findings from Research Question 6, looking at cumulative GPA, were very similar to those from Research Question 5 (graduation rates). When comparing the experiment group to each control group, there were two distinct patterns. The first pattern was found when comparing the WCC Math Summer Bridge students to Control Group A. Although none of the groups had statistically significant differences in cumulative GPA, the experiment group generally had higher means than Control A. It is important to take into consideration the experiment group’ large population placing into the lowest level of remedial math. And although Control Group A students were also taking remedial/developmental courses, it is unclear as to which level they started at. The second distinct pattern that emerged from cumulative GPA analysis, is that the experiment group (M = 2.91) had significantly higher cumulative GPAs than their Control Group B counterparts (M = 1.90), and for every group (Total, non-Native Hawaiian, Native Hawaiian, traditional, and nontraditional), its effect size was very large (p < .001, eta squared ranging from .122 to .207). The average cumulative GPA of the WCC Math Summer Bridge students (2.91) is also higher than the average cumulative GPA for the total Windward Community College MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY& & & 226& student body (2.47). Therefore, in trend with the other two student success measures, the experiment group had just as good, if not significantly higher cumulative GPAs than their two control group comparisons and the general Windward Community College student body. Limitations As a single-institution, quantitative study, the researcher identified major limitations. These limitations include threats to internal & external validity, as well as bias. First, with regards to external validity, it is important to note the sample population’s distribution with key demographics. There were significantly larger amounts of Native Hawaiian students included in the experiment group compared to the Windward Community College total student population (χ 2 (1, n=294) = 52.792, p = .000). Windward Community College has the largest proportion of Native Hawaiian students in the University of Hawaii System. Although studies have shown that many community colleges with higher minority populations face similar problems, the researcher cautions that different cultures might take to some of these key practices more than others (e.g. free breakfast on test days, intrusive counseling). As previously mentioned, the post-survey was administered right before the students took their final exam. Although this was a practical way to ensure the post-survey was taken, it probably did not provide the most realistic scores because students were anxious and under higher amounts of stress in anticipation of their final exam. The researcher relied on the samples for each group (experiment, Control Group A, Control Group B) based on what was available, not what is/would be ideal. For the experiment group’s 2012, 2013, and 2014 cohort respondents, the researcher relied on individuals that responded to the email request. Therefore, it is assumed that these students possessed the means (e.g. computer/email access) and the skill sets to reply back. The researcher predicts that there would be different findings in terms of college self-efficacy for the smaller proportion of students who MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY& & & 227& stopped out of college. Also, for the two control groups, the researcher relied on the entire developmental math student population for each summer, however summer school enrollments at the institution are much smaller than the fall/spring semesters and probably less diverse. As previously mentioned, these students generally are not able to use financial aid because they already extinguished those resources for fall/spring semesters, making the summer school population skewed towards individuals with more financial resources. Therefore, these findings are cautioned when making assumptions for generalizability to other community colleges and higher education institutions, as a whole. Internal validity threats are also present in this study. Unfortunately, when conducting research in the field, this is inevitable but it is up to the researcher to identify and minimalize them if at all possible (Creswell, 2008). First, the researcher relied on the assumption that the Math Summer Bridge students were responding with the most accurate and truthful responses possible. Then, even with them responding as honestly as possible, their perception of what is happening (e.g. the helpfulness of the faculty/staff) might not be what is actually happening or what the majority of others perceive as what is happening. Thirdly, the researcher is assuming that the students fully understood the questions on the survey and how to answer them. And finally, this being what the researcher views as most problematic, in terms of the post-survey, it was given to students right before they took their final exam. This was to ensure that they complete the survey (and give the researcher a complete pre/post pairing), but it was not viewed as the most ideal time. Students were visibly and verbally showing heightened levels of stress associated with taking final exams and therefore, the researcher believes that their survey scores, especially for college self-efficacy, were skewed to be lower than what they would be normally. MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY& & & 228& As stated by McEwan & McEwan (2003), there is always an “omnipresent possibility that the comparison… is different in some unknown way” (p. 37). In order to protect the study from this type of bias, the researcher included a large sample population (N = 957), which would naturally control for some of these unknown factors. Ensuring the researcher was not in the room when students took their surveys curbed researcher bias. Instead there were student aids to assist when a student had issues, such as inputting the correct url, how to select an answer, or other computer-related issues. Student aids did not assist with interpreting the question or watch over those completing the survey. The study attempted to minimize all threats to external and internal validity as well as bias through research design. Alternative Framework Using the data collected from this study, a new framework was created to help not only practitioners (e.g. instructional faculty or academic advisors), but also administrators and students understand the important dynamics between: (a) thoughts, attitudes, and beliefs (e.g. college self-efficacy); (b) behaviors (e.g. interaction with instructors); and (c) student success outcomes (e.g. persistence and graduation). This framework moves away from Astin’s IEO Framework, instead of looking at each component in a linear progression, this framework focuses on the multi-directional dynamics occurring between key facets of an individual’s life (college self-efficacy, engagement, satisfaction, sense of belonging, and academic achievement) and the impact of a multifaceted intervention’s effect on a) each of these key facets individually, and b) the larger student success system as a whole. Because each of the key facets depend on and effect one another, a multi-faceted intervention that improves each exponentially grows the system, setting the student on a new higher trajectory for success, which is in agreement with Gore (2006), stating that an intervention that abruptly increases college self-efficacy could put a MATH&SUMMER&BRIDGE&EFFECTS&ON&SELF&EFFICACY& & & 229& student on a higher trajectory for success. When students experience interventions that target improvement on multiple aspects of an individual’s life (e.g. mindset and academic subject- math), it will have a larger, long-lasting effect on their successes. Figure 4 illustrates this framework. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 230% Figure 4. Alternative Framework MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 231% Implications for Practice Findings from this study provided two major implications for practice. Although this study was conducted at one institution, it is believed that these suggestions are a template that could be applied at any institution with special modifications to the populations they serve. The two recommendations are creating 1) an academic community, and 2) multi-faceted interventions. Academic Community This study validated the need for academic communities. Especially with community college students/students who commute to campus, it is harder for them to feel connected to the campus (Astin, 1999). Academic communities are not new to this study. They are a concept that have been used before and have especially been shown to be effective with Native Hawaiian students in STEM (science, technology, engineering, math) majors (Kaakua, 2014). Academic communities provide a venue for students of similar characteristics (e.g. placement into developmental math courses) come together to associate with one another. Through not only creating an environment, but also ensuring the opportunities for students to engage with one another, they not only make social connections, but also academic ones, which research, including this study, has shown to be beneficial for student success (Di Tommaso, 2012). Beyond engaging with one another, academic communities include discussions and mentoring between students and peer mentors, instructors, other faculty, and administration. It is through these relationships that leaders (peer or faculty) can help students navigate through difficult periods of time (e.g. failing an exam). In academic communities, leaders can share personal experiences (including failures) and anecdotes on how they came to be where they are today. However, what the researcher believes to be the most important component of the MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 232% academic community is the positive pressure from peers (positive peer pressure) and mentors to have students set high standards for themselves and the support to allow them the opportunities to grow from setbacks. Openly acknowledging and embracing “failures” as just a step towards success is a key component of having a “growth mindset” (a concept that has been linked not only to academic success, but also successful people in general) (Dweck, 2008). Academic communities are a tool in which growth mindset practices can be implemented. This practices follows the Alternative Framework emphasizing the importance of a centralized intervention that addresses multiple facets of the individual student. Multi-facet Interventions The second implication concluded from this study is the importance that multi-faceted interventions have in student success, consisting of two tiers. The first tier is that the Windward Community College Math Summer Bridge Program targeted not just the subject matter (developmental mathematics), but it also integrated the practice of (1) necessary soft-skills (e.g. study skills/habits); (2) opportunities for students to engage with instructors and other key faculty/administration through the academic community setting; and (3) addressing other barriers that students might be facing (e.g. hunger, homelessness, court hearings) and using them as an opportunity to build relationships. For each of these facets of the WCC Math Summer Bridge intervention, high standards and expectations were set, along with the tools necessary to empower the students. Although the data did not support the idea of the MSB Program improving sense of belonging, it did significantly increase student engagement with both instructors and peers, as well as increase student satisfaction. The MSB Program did so by incorporating different practices that made the students feel welcome and a part of the WCC ‘ohana (family). The first MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 233% practice that contributed was providing free lunch everyday and free breakfasts on test days. By providing free meals, it (1) made sure that hunger was not a distraction from learning; and (2) created daily avenues for students to interact with one another, and with instructors, faculty, and administration. When students interact with faculty and administration on a less formal level, they become more confident in engaging in academic conversation as well; contributing to class discussions and asking for help when needed, thus boosting college self-efficacy, sense of belonging, and engagement and satisfaction. This cycle intensifies and grows: students are more likely to take new risks because they are confident in their knowledge and support system, and when they make new strides, the cycle repeats itself. Students’ thoughts, attitudes, and behaviors are continually feeding off one another and contributing to the individual’s success. The second tier of student success multi-faceted interventions are focused on is putting their thoughts, attitudes, and behaviors into action. Human beings do not naturally learn one specific skill in isolation. As babies, we learn how to socialize and communicate while developing our motor skills. We better understand concepts when also learning their relationships to others and/or within a larger context. Therefore, it follows the same logic that when wanting to help students become successful college students, an intervention that helps them improve upon individual skill sets (e.g. mathematics) while also applying them in other and/or larger contexts (e.g. study habits, communication/connections with faculty and administration), will be the most impactful. The most important component to this philosophy is making sure that the students practice these concepts, not just learn about them. Instead of students learning about how much time should be allotted to studying, the WCC Math Summer Bridge Program mandated that they completed 2 hours of study hall per day in school with tutors readily at hand. The WCC Math Summer Bridge Program faculty (instructors and advisors), did MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 234% not just tell students to set high goals for themselves, they required students to retake their chapter exams if they did not earn a 95% or higher and made sure they had a variety of studying aids (practice tests, tutors, study rooms, etc.) to help the students improve their scores on the retake. The final benefit of multi-faceted programs such as the Windward Community College Math Summer Bridge Program, is that there does not seem to be any specific time when these programs should be implemented. Students who participated in, and benefitted from, the WCC MSB were in various stages of their academic career. The researcher believes that this was a beneficial component of the program. Students who were in their second year of college provided mutually beneficial mentoring to those who were just beginning their academic careers. What does seem to be important is that there is a strong/important commonality (usually with a high level of difficulty) uniting these students, in this example, developmental mathematics. The power of students making connections to their peers, faculty, and administrators, while also acquiring skill sets needed to succeed in college (academic and non-academic), transcends their immediate goal of passing a math class and instead propels the individual on a new trajectory for success. Future Research This study aimed to investigate the effects of the WCC Math Summer Bridge Program on six different measures linked to or indicating student success. It also aimed to investigate the different effects of this intervention on two specific groups (Native Hawaiians and nontraditional students). Because of the limitations of this study, it is believed that further research in different forms will give a better understanding of the reasons for the success of the WCC Math Summer MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 235% Bridge students and the effects that this intervention has on college self-efficacy, sense of belonging, engagement, and satisfaction. New Instrument to Measure Sense of Belonging As shown through the contradictions in results for the Sense of Belonging Inventory (SOBI) and Engagement and Satisfaction questions, there is a need to create a new measure for sense of belonging. It is hypothesized that one of the main reasons why the SOBI didn’t properly measure sense of belonging is because of the language used throughout the questionnaire. Many of the questions pertained to “fit” (e.g. “I wonder if I really fit in college” or “It is important that I fit in at school”), where especially for nontraditional students who perceive “normal/fit” college students as being 18-22 years old, they would not believe they “fit” into that category. And it wouldn’t matter if they “fit in” as long as they feel comfortable asking questions and talking with their instructors and peers (engagement) and were satisfied with their instructors, support services, and the overall campus. Instead, the researcher proposes creating a Sense of ʻOhana Inventory, where the emphasis is on connectedness on campus using appropriate language. Qualitative Focus Groups Although the quantitative survey collected in this survey showed an increase in college self-efficacy, engagement, and satisfaction, the researcher believes that conducting qualitative focus groups will better highlight the specific components of the WCC MSB Program that contributed to this success. At each point in a student’s career (e.g. immediately after intervention, 2 years post-intervention, 3-years post intervention), what stands out to them as important parts of the MSB Program? What, if anything, do they still practice/incorporate in their academic and/or professional lives? How do their answers compare to other students who MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 236% took a summer math course but were not a part of the MSB Program (Control Group A)? How do the answers of MSB students with higher success marks (e.g. persistence/graduation/cumulative GPA) compare to MSB students who stopped out of school without earning a degree? By understanding what the students perceive as important and what was successfully implemented, we could possibly integrate these practices into different general veins of the community college system (e.g. academic advising sessions, financial aid/admissions application processes, each branch of academic courses). Mixed-Methods Longitudinal Design For the purposes of this study, cross-sectional data was used in order to determine if there were long-term effects from the WCC Math Summer Bridge Program. However, to get a more accurate understanding of how this intervention affects the individual person throughout their life, it is recommended that a longitudinal design be done. Implementing quantitative (e.g. researcher-created survey) and qualitative (e.g. interviews) measures at the same time, answers from both can be compared to one another to better understand the dynamics of specific components of the intervention (e.g. mandatory test retake policy, free lunch, quality/number of tutors) on key indicators of student success (e.g. college self-efficacy). Repeating these procedures at multiple points in time for the same individual will thoroughly describe the short and long-term effects of the intervention as well as help account for unforeseen barriers (e.g. job loss/gain, family responsibilities). The Pipeline to and through Community College As previously stated, the community college campus is much more diverse than its four- year campus counterparts. Some students enroll right after high school, but the majority of students make this transition many years after their high school graduation. And quite a few of MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 237% these students are a combination of the two, starting college and then stopping out, numerous times. With ever changing guidelines and standards for high school mathematics, as well as disparities between those standards and actual achievement, another suggestion for research is identifying the reasons for students’ math deficiencies as well as supplemental interventions before entering college to ensure student success in college level math courses. Beyond identifying obstacles/barriers before entering college, it is important for institutions to understand the patterns of enrollment for community college students. Many of these students do not complete their degree in one continuous enrollment period; instead they stop out and return (sometimes multiple times). If practitioners are able to identify critical periods in a student’s college experience and implement multi-faceted interventions, we will be able to help students earn their degree more efficiently. Conclusion The value of a college degree is at an all-time high, essential for earning a living wage in the United States. With disparities in K-12 education, and adults entering college from the workforce, community colleges provide access to those who would not be accepted to the traditional 4-year institution. Unfortunately, after gaining access, almost half (43.6%) of all community college students stop out without earning any degree or transferring to another institution (Shapiro & Dundar, 2012). Remediation plays a major role in this issue. Students who place into developmental/remedial courses not only have to take extra courses to get to college level work, and ultimately earning a degree, they are also more likely to stop out of school due to low self-esteem or weak motivation (Bahr, 2012). This study attempted to analyze an innovative intervention aimed to address not only mathematic deficiencies, but also college self-efficacy, sense of belonging, engagement and MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 238% satisfaction. It was found that not only does the Windward Community College Math Summer Bridge Program have positive effects on students’ math grades (89.19% pass rates) compared to the total Windward Community College developmental pass rates (22%-70%), but they are also more likely to persist, graduate, and have higher cumulative GPAs than their peers who did not participate in the MSB Program. It is postulated that these higher success measures are due to the intervention’s positive and long-lasting impact on college self-efficacy, and student engagement and satisfaction. From this study, a new framework was created to understand the working dynamics of student success indicators (e.g. attitudes, beliefs, and behaviors) and the cyclical growth relationship they have with student success measures (e.g. persistence, graduation, and cumulative GPA). It is hoped that from this study, practitioners can create multi-faceted interventions modeled after the Windward Community College Math Summer Bridge Program to increase developmental math remediation as well as overall student success. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 239% References Achieving the Dream (2014). About Us. Retrieved from the Achieving the Dream website: achievingthedream.org/about-us ACT. (2012). National condition of college & career readiness 2012. Retrieved from: media.act.org/documents/CCCR12-NationalReadinessRpt.pdf. Adams, P., Adams, W., Franklin, D., Gulick, D., Gulick, F., Shearn, E., & Mireles, S. (2012). Remediation: Higher education’s bridge to nowhere. Retrieved from Complete College America website: http://www.completecollege.org/docs/CCA-Remediation-final.pdf Adelman, C. (2006). The toolbox revisited: Paths to degree completion from high school through college. U.S. Department of Education. Washington, DC: Office of vocational and Adult Education. Almy, S. & Theokas, C. (2010). Not prepared for class: high-poverty schools continue to have fewer in-field teachers. (November 18). Washington DC: The Education Trust. Retrieved from: http://www.edtrust.org/dc/publication/not-prepared-for-class-high-poverty-schools- continue-to-have-fewer-in-field-teachers American Association of Community Colleges. (2013). Community College Enrollment. Retrieved from the American Association of Community Colleges website: http://www.aacc.nche.edu/AboutCC/Trends/Pages/enrollment.aspx American Association of Community Colleges. (2014). 2014 Fact Sheet. Retrieved from the American Association of Community Colleges website: www.aacc.nche.edu/AboutCC/Documents/Facts14_Data_R3.pdf MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 240% American Association of Community Colleges. (2015). 2015 Fact Sheet. Retrieved from the American Association of Community Colleges website: http://www.aacc.nche.edu/AboutCC/Documents/FactSheet2015.pdf American Association of Community Colleges. (2016). 2016 Fact Sheet. Retrieved from the American Association of Community Colleges website: www.aacc.nhce.edu/AboutCC/Documents/AACCFactSheet2016.pdf American Association of University Women. (2013). Women in Community Colleges: Access to Success. Washington, DC: St. Rose, A., & Hill, C. 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. W. (1975). Preventing students from dropping out. San Francisco: Jossey-Bass. Astin, A. W. (1977). Four Critical Years. San Francisco: Jossey-Bass. Astin, A. W. (1985). Achieving educational excellence: A critical assessment of priorities and practices in higher education. San Francisco, CA: Jossey-Bass. Astin, A. W. (1993). What Matters in College? Four Critical Years Revisited. San Francisco, CA: Jossey-Bass. Astin, A. W. (1999). Student involvement: A developmental theory for higher education. Journal of College Student Development, 40, 518-529. Astin, A. W., & Antonio, A. L. (2012). Assessment for Excellence: The Philosophy and Practice of Assessment and Evaluation in Higher Education. Lanham, MD: Rowman & Littlefield Publishers, Inc. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 241% Atherton, M. C. (2014). Academic preparedness of first-generation college students: Different perspectives. Journal of College Student Development, 55(8), 824-829. Bahr, P. (2010). Preparing the underprepared: An analysis of racial disparities in post secondary mathematics remediation. The Journal of Higher Education, 81(2), 209-237. muse.jhu.edu.libproxy.usc.edu/journals/journal_of_higher_education/v081/81.2.bahr.html Bahr, P. R. (2012). Deconstructing remediation in community colleges: Exploring associations between course-taking patterns, course outcomes, and attrition from the remedial math and remedial writing sequences. Research in Higher Education, 53, 661-693. Bailey, T., Jeong, D., & Cho, S. (2009). Referral, enrollment, and completion in developmental education sequences in community colleges. Economics of Education Review, 29, 255- 270. Bailey, T., Jeong, D. W., & Cho, S. (2010). Referral, enrollment, and completion in developmental education sequences in community colleges. Economics of Education Review, 29, 255-270. Bandura, A. (1986). Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice Hall. Bandura, A. (1997). Self-Efficacy: The Exercise of Control. New York: Freeman. Bandura, A. (2001). Social Cognitive Theory: An agentic perspective. Annual Review of Psychology, 52, 1-26. Barbatis, P. (2010). Underprepared, ethnically diverse community college students: Factors contributing to persistence. Journal of Developmental Education, 33(3), 14-18, 20, 22, 24. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 242% Barry, C. L., & Finney, S. J. (2009). Can we feel confident in how we measure college confidence? A psychometric investigation of the College Self-Efficacy Inventory. Measurement and Evaluation in Counseling and Development, 42(30), 197-222. Baum, S., Ma, J., & Payea, K. (2013). Education pays 2013: The benefits of higher education for individuals and society. (Trends in Higher Education Series). Retrieved from: http://trends.collegeboard.org/education-pays 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, & Education, 9(1), 29- 50. Berger, J. B. (2000). Optimizing capital, social reproduction, and undergraduate persistence: A sociological perspective. In J. M. Braxton (Ed.), Reworking the student departure puzzle (pp. 95-126). Nashville, TN: Vanderbilt University Press. Bong, M., & Skaalvik, E. M. (2003). Academic self-concept and self-efficacy: How different are they really? Educational Psychology Review, 15, 1-40. Bordes-Edgar, V., Arredondo, P., Kurpius, S. R., & Rund, J. (2011). A longitudinal analysis of Latina/o students’ academic persistence. Journal of Hispanic Higher Education, 10(4), 358-368. Bozak, S. (2013). College Students’ Sense of Belonging and Social Support: Potential Factors in Resilience (Doctoral dissertation). Retrieved from ProQuest LLC. (3560392) Brayboy, B., Castagno, A., & Maughan, E. (2007). Equity and justice for all? Examining race in educational scholarship. Review of Research in Education, 31, 159-194. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 243% Brewer, M. B. (2000). Research design and issues of validity. In H. T. Reis & C. M. Judd (Eds), Handbook of research methods in social psychology (pp. 3-16). New York, NY: Cambridge University Press. Brown, S. D., Tramayne, S., Hoxha, D., Telander, K., Fan, X., & Lent, R. W. (2008). Social cognitive predictors of college students’ academic performance and persistence: A meta- analytic path analysis. Journal of Vocational Behavior, 72, 298-308. Boyd, D. (2009). Projected state and local budget surplus as a percent of revenues, 2016. Rockefeller Institute of Government, Unpublished manuscript. Buchman, C., & DiPrete, T. (2006). The growing female advantage in college completion: The role of family background and academic achievement. American Sociological Review, 71, 143-164. Carnevale, A., Smith, N., & Strohl, J (2013). Recovery: Projections of jobs and education requirements through 2020. Retrieved from Georgetown University, Georgetown Public Policy Institute Center on Education and the Workforce website: http://www9.georgetown.edu/grad/gppi/hpi/cew/pdfs/recovery2020.fr.web.pdf Calcagno, J. C., Bailey, T., Jenkins, D., Kienzl, G., & Leinbach, T. (2008). Community college student success: What institutional characteristics make a difference? Economics of Educational Review. 27(2008), 632-645. Castellanos, J., & Jones, L. (2004). Latino/a undergraduate experiences in American higher education. In J. Castellanos & L. Jones (Eds.), The majority in the minority. Sterling, VA: Stylus. Chemers, M. M., Hu, L., & Garcia, B. F. (2001). Academic self-efficacy and first-year college student performance adjustment. Journal of Educational Psychology, 93, 55-64. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 244% Civil Rights Project (2012). Building Pathways to Transfer: Community Colleges that Break the Chain of Failure for Students of Color. Los Angeles, CA: Ganadara, P., Alvarado, E., Driscoll, A., & Orfield, G. Clark, L. (2012). When nontraditional is traditional: A faculty dialogue with graduating community college students about persistence. Community College Journal of Research and Practice, 36(7), 511-519. Condron, D., & Roscigno, V. (2003). Disparities within: Unequal spending and achievement in an urban school district. Sociology of Education, 76(1), 18-36. Creswell, J. W. (2014). Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. Thousand Oaks, CA: SAGE Publications, Inc. Crisp, G., & Nora, A. (2010). Hispanic student success: Factors influencing the persistence and transfer decisions of Latino community college students enrolled in developmental education. Research in Higher Education, 51(2), 175-194. Cushman, K. (2007). Facing the culture: First-generation college students talk about identity, class, and what helps them succeed. Educational Leadership, 44-47. Darling-Hammond, L. (2007). The flat earth and education: How America’s commitment to equity will determine our future. Educational Researcher, 36(6), 318-334. De Paola, M., Ponzo, M., & Scoppa, V. (2013). Class size effects on student achievement: Heterogeneity across abilities and fields. Education Economics, 21(2), 135-153. DeWitz, S. J., Woolsey, M. L., & Walsh, W. B. (2009). College student retention: An exploration of the relationship between self-efficacy beliefs and purpose in life among college students. Journal of College Student Development, 50(1), 19-34. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 245% Di Tommaso, K. (2012). Developmental students: The challenge of peer relationships. Community College Journal of Research and Practice, 36(12), 941-954. Dweck, C. S. (2008). Mindset: The New Psychology of Success. New York, NY: Ballantine Books. Eger, R., McDonald, B. (2012). Moving toward comparability: Assessing per student costs in K- 12. Journal of Public Budgeting, Accounting, and Management, 24(4), 609-638. Flores, S. M., & Park, T. J. (2013). Race, ethnicity, and college success: Examining the continued significance of the minority-serving institution. Educational Researcher, 42(3), 115-128. Gilroy, M. (2010). Studies show remedial courses in community colleges yield mixed results. The Hispanic Outlook in Higher Education, 20(12). Retrieved from http://search.proquest.com.libroxy.usc.edu/docview/219196246. Goldin, C., Katz, L., Kuziemko, I. (2006). The homecoming of American college women: The reversal of the college gender gap. Journal of Economic Perspectives, 20, 133-156. Goldrick-Rab, S. (2010). Challenges and opportunities for improving community college student success. American Educational Research Association. 80(3), 437-469. Gore, P. A., JR. (2006). Academic self-efficacy as a predictor of college outcomes: Two incremental validity studies. Journal of Career Assessment, 14, 92-115. Gore, P. A., Jr., Leuwerke, W. C., & Turley, S. E. (2005). A psychometric study of the college self-efficacy inventory. Journal of College Student Retention: Research, Theory and Practice, 7, 227-244. Hagerty, B. M. K., Lynch-Sauer, J., Patusky, K. L., Bouwsema, M., & Collier, P. (1992). Sense of belonging: A vital mental health concept. Archives of Psychiatric Nursing, 6, 172-177. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 246% Hagerty, B. M. K., & Patusky, K. (1995). Developing a measure of Sense of Belonging. Nursing Research, 44(1), 9-13. Hall, J. M., & Ponton, M. K. (2005). Mathematics self-efficacy of college freshman. Journal of Developmental Education, 28(3), 26-32. Handel, S. & Williams, R. (2012). Reimagining remediation. Change: The Magazine of Higher Learning, 43(2), 28-33. Harris, B. S. (2006-2007). The importance of creating a “sense of community”. Journal of College Student Retention, 8(1), 83-105. Hekimoglu, S., & Kittrell, E. (2010). Challenging students’ beliefs about mathematics: The use of documentary to alter perceptions of efficacy. Primus, 20(4), 1-33. Hendrick, R. Z., Hightower, W. H., & Gregory, D. E. (2006). State funding limitations and community college open door policy: confliction priorities?. Community College Journal of Research and Practice, 30(8). 627-640. Retrieved from: http://www.tandfonline.com.libproxy.usc.edu/doi/abs/10.1080%2F10668920600746078# .UhH7AhaT-5d Hollis-Sawyer, L. (2011). A math-related decrement stereotype threat reaction among older nontraditional college learners. Educational Gerontology, 37, 292-306. Ishitani, T. T. (2003). A longitudinal approach to assessing attrition behavior among first generation students: Time-varying effects on pre-college characteristics. Research in Higher Education, 44, 433-449. Johnson, I. (2008). Enrollment, persistence and graduation of in-state students at a public research university: Does high school matter? Research in Higher Education, 49, 776- 793. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 247% Jones, S. J., & Hansen, K. (2014). Technology review: Virtual intrusive advising-supporting community college students through web-based synchronous technologies. Community College Enterprise, 20(1), 89-93. Jurgens, J. (2010). The evolution of community colleges. College Student Affairs Journal, 28(2), 251-261. search.proquest.com.libproxy.usc.edu/docview/737539278/fulltextPDF?accountid=14749 Kaakua, J. K. (2014). Self-efficacy beliefs and intentions to persist of Native Hawaiian and non- Native Hawaiian science, technology, engineering, and mathematics majors (Doctoral dissertation). Retrieved from the University of Southern California Digital Library. Kahn, J. H. & Nauta, M. M. (2001). Social-cognitive predictors of first-year college persistence: The importance of proximal assessment. Research in Higher Education, 42,633-652. Keith, P. M. (2007). Barriers and nontraditional students’ use of academic and social services. College Student Journal, 41(4), 1123-1127. Kemp, W. C. (2002). Persistence of adult learners in distance education. American Journal of Distance Education, 16, 65-81. Kerrigan, M. R., & Slater, D. (2010). Collaborating to create change: How El Paso community college improved the readiness of its incoming students through Achieving the Dream (Culture of Evidence Series, Report No. 4). New York, NY: Columbia University, Teachers College, Community College Research Center. Korbin, J. L., Patterson, B. F., Shaw, E. J., Mattern, K. D., & Barbuti, S. M. (2008). Validity of the SAT for predicting first-year college grade point average. New York, NY: College Board. Retrieved from http://professionals.collegeboard.com/data-reports- research/cb/validity-of-sat-predicting-fycgpa MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 248% Lanni, J. C. (1997, May). Modeling student outcomes: A longitudinal study. Paper presented at the 37 th Annual Forum of the Association for Institutional Research, Orlando, FL. (ERIC Document Reproduction Service No. ED410670) Lent, R. W., Brown, S. D., & Hackett, G. (1994). Toward a unifying social cognitive theory of career and academic interest, choice and performance [Monograph]. Journal of Vocational Behavior, 45, 79-122. Lim, C. K. (2001). Computer self-efficacy, academic self-concept, and other factors as predictors of satisfaction and future participation of adult learners in web-based distance education. American Journal of Distance Education, 15, 41-51. Lopez, F. G., Lent, R. W., Brown, S. D., & Gore, P. A., Jr. (1997). Role of social-cognitive expectations in high school students” mathematics-related interest and performance. Journal of Counseling Psychology, 44, 44-52. Long, M., Latarola, P., & Conger, D. (2009). Explaining gaps in readiness for college-level math: The role of high school courses. Retrieved from www.mitpressjournals.org/doi/pdf/10.1162/edfp.2009.4.1.1 Low, L. (2000). Are College Students Satisfied? A National Analysis of Changing Expectations. Indianapolis, IN: USA Group, Inc. Lucas, C. J. (2006). American Higher Education. New York, NY: Palgrave Macmillan. Lundberg, C. A. (2014). Peers and faculty as predictors of learning for community college students. Community College Review, 42(2), 79-98. Lundberg, C. A., McIntire, D. D., & Creasman, C. T. (2008). Sources of support and self- efficacy for adult students. Journal of College Counseling, 11(1), 58-72. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 249% MacCann, C., Fogarty, G. J., & Roberts, R. D. (2012). Strategies for success in education: Time management is more important for part-time than full-time community college students. Learning and Individual Differences, 22, 618-623. Majer, J. M. (2009). Self-efficacy and academic success among ethnically diverse first- generation community college students. Journal of Diversity in Higher Education, 2(4), 243-250. Martin, A. J., & Dowson, M. (2009). Interpersonal relationships, motivation, engagement, and achievement: Yields for theory, current issues, and educational practices. Review of Educational Research, 79(1), 327-365. McCarron, G. P. & Inkelas, K. K. (2006). The gap between educational aspirations and attainment for first-generation college students and the role of parental involvement. Journal of College Student Development, 47, 534-549. McEwan, E. K., & McEwan, P. J. (2003). Making sense of research: What’s good, what’s not, and how to tell the difference. Thousand Oaks, CA: Corwin Press, Inc. Melguizo, T. (2010). Are students of color more likely to graduate from college if they attend more selective institutions? Evidence from a cohort of recipients and non-recipients of the Gates Millennium Scholarship Program. Educational Evaluation and Policy Analysis, 32, 230-248. Mercer, D. (1993). Older coeds: Predicting who will stay this time. Journal of Research and Development in Education, 26, 153-163. Mullin, C. (2012). Why access matters: The community college student body. Retrieved from the American Association of Community Colleges website: http://www.aacc.nche.edu/Publications/Briefs/Documents/PB_AccessMatters.pdf MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 250% 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, 28, 30-38. Nakajima, M. A., Dembo, M. H., & Mossler, R. (2012). Student persistence in community colleges. Community College Journal of Research and Practice, 36(8), 591-613. National Center for Education Statistics. (2012). Digest of education statistics: Table 384. Retrieved from: http://nces.ed.gov/programs/digest/d12/tables/dt12_384.asp National Center for Education Statistics (2014). The Condition of Education. (U.S. Department of Education Publication no. NCES 2014-083). Washington, DC: U.S. Government Printing Office. NCES. (2015). 2011-12 National Postsecondary Student Aid Study (NPSAS:12) [AACC analysis]. Office of Hawaiian Affairs, Research Division, Special Projects. (2014). A Native Hawaiian Focus on Post Secondary Education within the University of HawaiiʻI System. Honolulu, HI. Oja, M. (2011). Student satisfaction and student performance. Journal of Applied Research in the Community College, 19(1), 50-56. Orozco, G. L., Alvarez, A. N., & Gutkin, T. (2010). Effective advising of diverse students in community colleges. Community College Journal of Research and Practice, 34(9), 717- 737. Pajares, F., & Miller, M. D. (1995). Mathematics self-efficacy and mathematics performances: The need for specificity of assessment. Journal of Counseling Psychology, 42, 190-198. Pallant, J. (2013). SPSS survival manual: A step by step guide to data analysis using IBM SPSS (5 th ed.). New York, NY: McGraw Hill Education. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 251% Pascarella, E. T. (2006). How college affects students: Ten directions for future research. Journal of College Student Development, 47, 508-520. Perna, L. W., Klein, M. W., & McLendon, M. K. (2014). Insights and implications for state policy-makers. The ANNALS of the American Academy of Political and Social Science, 655(1), 209-230. Philibert, N. Allen, J., & Elleven, R. (2008). Nontraditional students in community colleges and the model of college outcomes for adults. Community College Journal of Research and Practice, 32(8), 582-596. Prado, R. E. (2012). Peers as institutional agents: Acquiring social capital through peer interactions. Retrieved from the University of Southern California Digital Library Database. Rendon, L. I., Jalomo, R., & Nora, A. (2001). Minority student persistence. In Braxton’s (Ed.), Rethinking the departure puzzle: New theory and research on college student retention. Nashville: Vanderbilt University Press. Ritt, E. (2008). Redefining tradition: Adult learners and higher education. Adult Learning, 19, 12-16. Robbins, S. B., Allen, J., Casillas, A., Peterson, C. H., & Le, H. (2006). Unraveling the differential effects of motivational and skills, social, and self-management measures from traditional predictors of college outcomes. Journal of Educational Psychology, 98(3), 598-616. 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, 261-288. doi: 10.1037/0033-2909.130.2.261 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 252% Robinson Kurpius, S. E., & Stafford, M. E. (2006). Testing and Measurement: A User-Friendly Guide. Thousand Oaks, CA: Sage Publications, Inc. Rubin, M. (2012). Social class differences in social integration among students in higher education: A meta-analysis and recommendations for future research. Journal of Diversity in Higher Education, 5, 22-38. Rubin, M., Denson, N., Kilpatrick, S., Matthews, K. E., Stehlik, T. & Zyngier, D. (2014). “I am working-class”: Subjective self-definition as a missing measure of social class and socioeconomic status in higher education. Educational Researcher, 43(4), 196-200. Ryan, M. G. (2013). Improving retention and academic achievement for first-time students at a two-year college. Community College Journal of Research and Practice, 37, 131-134. Sablan, J. R. (2014). The challenge of summer bridge programs. American Behavioral Scientist, 58(8), 1035-1050. Sawyer, R. (2013). Beyond correlations: usefulness of high school GPA and test scores in making college admissions decisions. Applied Measurement in Education, 26(2), 89-112. Shapiro, D., & Dundar, A. (2012). Signature Report: Completing College: A national view of student attainment rates. Retrieved from National Student Clearinghouse Research Center website: content.nwacc.edu/publications/CostContainment/NSC_Signature_Report_4.pdf Schlossberg, N. K., Lynch, A. Q., & Chickering, A. W. (1989). Improving Higher Education Environments for Adults. San Francisco, CA: Jossey-Bass. Silver-Pacuilla, H., Perin, D., Miller, B. (2013). Introduction to special issue of community college review: skills and trajectories of developmental education learners. Community College Review, 41(2), 111-117. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 253% Smith, J. S. (2007). Using data to inform decisions: Intrusive faculty advising at a community college. Community College Journal of Research and Practice, 31(10), 813-831. Solberg, V. S., O’Brien, K., Villareal, P., Kennel, R., & Davis, B. (1993). Self-efficacy and Hispanic college students: Validation of the college self-efficacy instrument. Hispanic Journal of Behavioral Sciences, 15, 80-95. doi: 10.1177/07399863930151004 Starr, A., Betz, E. L., & Menne, J. (1972). Differences in college student satisfaction: Academic dropouts, nonacademic dropouts, and nondropouts. Journal of Counseling Psychology, 19, 318-322. State of the Union Address. (2012). Retrieved November 6, 2013, from http://www.whitehouse.gov/state-of-the-union-2012 Stevens, J. (1996). Applied Multivariate Statistics for the Social Sciences (3 rd edition). Mahwah, NJ: Lawrence Erlbaum. Strayhorn, T. L. (2012). College Students’ Sense of Belonging: A Key to Educational Success for All students. New York, NY: Routledge Publishing. Taniguchi, H. & Kaufman, G. (2005). Degree completion among nontraditional college students. Social Science Quarterly, 86(14), 912-927. The College Board. (2014). Trends in College Pricing. Collegeboard.org: Sandy Baum & Jennifer Ma. Retrieved from: https://secure- media.collegeboard.org/digitalServices/misc/trends/2014-trends-college-pricing-report- final.pdf Thompson, M. D. (2001). Informal student-faculty interaction: Its relationship to educational gains in science and mathematics among community college students. Community College Review, 29(1), 35-57. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 254% Tierney, W. (1992). An anthropological analysis of student participation in college. Journal of Higher Education, 63(6), 603-618. Tinto, V. (1993). Leaving college: Rethinking the causes and cures of student attrition (2 nd ed.). Chicago: University of Chicago Press. Tyler-Smith, K. (2006). Early attrition among first time eLearners: A review of factors that continue to drop-out, withdrawal and non-completion rates of adult learners undertaking eLearning programmes. Journal of Online Teaching and Learning, 2, 73-85. U.S. Bureau of the Census (2016). Census Glossary. Retrieved from www.census.gov/glossary/ U.S. Department of Education, National Center for Education Statistics. (2013). Digest of Education Statistics, 2012 (NCES 2014-015), Chapter 3. Retrieved from: nces.ed.gov/fastfacts/display.asp?id=76 Walker, T. R. (2008). Institutional factors impacting student satisfaction and persistence. (Doctoral dissertation). Retrieved from ProQuest Dissertations and Theses Database. (Publication no. AAT 3310714). Wassmer, R., Moore, C., & Shulock, N. (2004). The effect of racial/ethnic composition on transfer rates in community colleges: Implications for policy and practice. Research in Higher Education, 45(6), 651-672. White, E., & Schulenberg, J. (2012). Academic advising-A focus on learning. About Campus, 11-17. doi: 10.1002/abc.20082 White, J. W., & Lowenthal, P. R. (2011). Minority college students and tacit “codes of power”. The Review of Higher Education, 34(2), 283-318. Windward Community College (2014). Quick Facts for 2013. Retrieved from http://windward.hawaii.edu/ir/QuickFacts/Quick%20Facts%202014_5-29-14.pdf MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 255% Windward Community College. (2015). Annual Compass Placement Test Results. Internal data report. Wirt, L. G. & Jaeger, A. J. (2014). Seeking to understand faculty-student interaction at community colleges. Community College Journal of Research and Practice, 38(11), 980- 994. Wright, S. L., Jenkins-Guarnieri, M. A., & Murdock, J. L. (2012). Career development among first-year college students: College self-efficacy, student persistence, and academic success. Journal of Career Development, 40(4), 292-310. Zajacova, A., Lynch, S. M., & Espenshade, T. J. (2005). Self-efficacy, stress and academic success in college. Research in Higher Education, 46, 677-706. Doi: 10.1007/s11162- 004-4139-z Zimmerman, B. J., Bandura, A., & Martinez-Pons, M. (1992). Self-motivation for academic attainment: The role of self-efficacy beliefs and personal goal setting. American Educational Research Journal, 29, 663-676. Doi: 10.3102/00028312029003663 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 256% APPENDICES Appendix A: Student Survey Note: Survey has multiple paths, which are all listed below. Students would not see all of the questions based on their answers to specific dichotomous questions. 2012-2014 Math and College Self-Efficacy Survey 1. Aloha! My name is Sarah Akina and I am a doctoral candidate at the University of Southern California Rossier School of Education. I am collecting data for my dissertation entitled The Effects of a Math Summer Bridge Program on College Self-Efficacy and other Student Success Measures in Community College Students this summer and would greatly appreciate your help. The purpose of this survey is to identify how summer math courses and the WCC Math Summer Bridge Program impact students' college self-efficacy, persistence/graduation, and cumulative GPA. You have been asked to take part in this study because you have participated in the TRiO SSS Math Summer Bridge Program between 2012 and 2014. We want to identify factors that compose a successful math course/experience. Your input will be greatly valued and your insight will be used to improve academic and student services at TRiO SSS and Windward Community College, specifically in regards to how we can support developmental math courses. You will be asked to please provide your first and last name or UH Username or UH Student ID number here to track your coursework progress through the Fall 2015 semester. Your identifying information will not be published in the survey and will be kept strictly confidential. Participation is voluntary. You do not need to participate if you do not want to. If you do not take part in this survey, there will be no effects on your academic or student services at Windward Community College or any other repercussions. All of your identifying information (Name, Student ID number, etc.) will be kept confidential and will not be published anywhere. This survey will be kept strictly confidential. If you have any questions, please feel free to contact Sarah Akina at inouyes@hawaii.edu or (808) 235- 7326. Mahalo in advance for your participation! Please input your First and Last Name or UH Username or UH Student ID Number in the box below: * MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 257% 2012-2014 Math and College Self-Efficacy Survey 2. Please choose the ethnicity you most identify with (select only one). * American Indian or Alaska Native Asian Indian Chinese Filipino Japanese Korean Laotian Thai Vietnamese Other Asian African American or Black Guamanian or Chamorro Micronesian Native Hawaiian or Part Native Hawaiian Samoan Tongan Other Pacific Islander Caucasian or White Latino/a 3. Were any of your ancestors Native Hawaiian? * Yes No 4. What is your sex? * Female Male MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 258% 5. What is the highest level of education your mother completed? * Intermediate school or below High school Some college or two-year degree Undergraduate (4-year) degree Graduate/professional degree Do not know 6. What is the highest level of education your father completed? * Intermediate school or below High school Some college or two-year degree Undergraduate (4-year) degree Graduate/professional degree Do not know 7. How do you identify you (and your family's) economic status? * I/we have more than enough money to live comfortably. I/we have just enough money to live comfortably. I/we do not have enough money to live comfortably. 8. Will you receive the Federal Pell Grant during the 2015-2016 school year? * Yes No Do not know Not applicable (will not be enrolled in school) Not eligible (financial aid probation or suspension) MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 259% 9. What was your cumulative high school GPA? * 4.00 3.50-3.99 3.00-3.49 2.50-2.99 2.00-2.49 1.50-1.99 below 1.5 Do not know 10. Did you attend college during the fall after graduating from high school? * Yes No MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 260% 2012-2014 Math and College Self-Efficacy Survey 11. If yes, did you stop out of college at any time? * Yes No MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 261% 2012-2014 Math and College Self-Efficacy Survey 12. Are you responsible for any dependents? * Yes No MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 262% 2012-2014 Math and College Self-Efficacy Survey 13. If yes, how many? 1 2 3 4 5 6 7+ <1 year old 1-2 years old 3-4 years old Elementary school age Intermediate school age High school age Dependent 1 Dependent 2 Dependent 3 Dependent 4 Dependent 5 Dependent 6 Dependent 7 14. How old is/are your dependent(s)? MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 263% 2012-2014 Math and College Self-Efficacy Survey 15. For the Fall 2015 semester, will you be enrolled in college? * Yes No MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 264% 2012-2014 Math and College Self-Efficacy Survey 16. If no, why? * Stopped out without earning a degree due to financial reasons Stopped out without earning a degree due to family responsibilities Graduated with a degree Other (please specify) 17. During the WCC Math Summer Bridge Program, how many times did you talk to/interact with your math instructor during class? * 0 times/week 1-3 times/week 4+ times/week 18. During the WCC Math Summer Bridge Program, how many times did you talk to/interact with your math instructor outside of class? * 0 times/week 1-3 times/week 4+ times/week 19. While participating in the WCC Math Summer Bridge Program, I felt comfortable talking with my Windward Community College math instructor about (select ALL that apply) * Math curriculum (i.e., math major, careers in math, etc.) Non-math related school topics (i.e., recommendations for science or English courses, academic advising, etc.) Non-academic topics (i.e., hobbies, movies, etc.) None of the above MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 265% 20. For your most recent classes, how many times did you talk to/interact with your instructors during class? * 0 times/week 1-3 times/week 4+ times/week 21. For your most recent classes, how many times did you talk to/interact with your instructors outside of class? * 0 times/week 1-3 times/week 4+ times/week 22. On average, how many times did you interact with your peers at school for social purposes (e.g. eat lunch, drink coffee, socialize)? * 0 times/week 1-3 times/week 4-6 times/week 7+ times/week 23. On average, how many times did you interact with your peers at school for academic purposes (e.g. study groups, work on projects)? * 0 times/week 1-3 times/week 4-6 times/week 7+ times/week Strongly disagree Disagree Neutral Agree Strongly agree 24. I am very satisfied with the academic instruction I've received at Windward Community College. * Strongly disagree Disagree Neutral Agree Strongly Agree 25. I am very satisfied with the student support services (e.g. academic advising, financial aid office) I've received at Windward Community College. * MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 266% Strongly Disagree Disagree Neutral Agree Strongly Agree 26. Overall, I am very satisfied with Windward Community College as a whole. * MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 267% 2012-2014 Math and College Self-Efficacy Survey 27. If yes, where are you currently enrolled in college? (Fall 2015 semester home institution) * Windward Community College Other UH Community College University of Hawaii at Manoa University of Hawaii at West Oahu Other (please specify) 28. For the Fall 2015 semester, what is your enrollment status? * Full-time (12+ credits/semester) Part-time (6-11 credits/semester) Below part-time (1-5 credits/semester) 29. During the WCC Math Summer Bridge Program, how many times did you talk to/interact with your math instructor during class? * 0 times/week 1-3 times/week 4+ times/week 30. During the WCC Math Summer Bridge Program, how many times did you talk to/interact with your math instructor outside of class? * 0 times/week 1-3 times/week 4+ times/week MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 268% 31. While participating in the WCC Math Summer Bridge Program, I felt comfortable talking with my Windward Community College math instructor about (select ALL that apply) * Math curriculum (i.e., math major, careers in math, etc.) Non-math related school topics (i.e., recommendations for science or English courses, academic advising, etc.) Non-academic topics (i.e., hobbies, movies, etc.) None of the above 32. For your most recent classes, how many times did you talk to/interact with your instructors during class? * 0 times/week 1-3 times/week 4+ times/week 33. For your most recent classes, how many times did you talk to/interact with your instructors outside of class? * 0 times/week 1-3 times/week 4+ times/week 34. On average, how many times do you interact with your peers at school for social purposes (e.g. eat lunch, drink coffee, socialize)? * 0 times/week 1-3 times/week 4-6 times/week 7+ times/week 35. On average, how many times do you interact with your peers at school for academic purposes (e.g. study groups, work on projects)? * 0 times/week 1-3 times/week 4-6 times/week 7+ times/week MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 269% Strongly disagree Disagree Neutral Agree Strongly agree 36. I am very satisfied with the academic instruction I've received at Windward Community College so far. * Strongly disagree Disagree Neutral Agree Strongly Agree 37. I am very satisfied with the student support services (e.g. academic advising, financial aid office) I've received at Windward Community College so far. * Strongly Disagree Disagree Neutral Agree Strongly Agree 38. Overall, I am very satisfied with Windward Community College as a whole. * MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 270% 2012-2014 Math and College Self-Efficacy Survey 39. Are you currently employed? * Yes No MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 271% 2012-2014 Math and College Self-Efficacy Survey 40. If yes, how many hours per week on average do you work? * 0-10 11-20 21-30 31-40 41+ 41. Where do you work? * On campus Off campus Both on and off campus MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 272% 2012-2014 Math and College Self-Efficacy Survey 42. What is your long term career goal? * Pursue a career in a STEM (Science, Technology, Engineering, or Math) field Pursue a career in a Business field Pursue a career in something other than STEM or Business fields MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 273% 2012-2014 Math and College Self-Efficacy Survey Not at all confident 1 2 3 4 5 6 7 8 9 Extremely confident 10 Make new friends at college Successfully complete my current math course Talk to college staff/faculty Talk to college administration Manage time effectively Ask a question in class Participate in class discussions Research a term paper Do well on my exams Join a student organization Talk to my professors outside of class Ask a professor a question about math coursework Take good class notes Complete the Math Requirement (pass Math 25 or Math 28) 43. The following items from the Solberg, O'Brien, Villareal, Kennel, and Davis College Self-Efficacy Inventory concern your confidence in various aspects of college. Using the scale below, please indicate how confident you are as a student that you could successfully complete the following tasks. If you are extremely confident, mark at 10. If you are not at all confident, mark 1. If you are more or less confident, find the number between 10 and 1 that best describes you. Item responses are aggregated across all student respondents in order to better understand how confident the "average" student feels. Levels of confidence vary from person to person, and there are no right or wrong answers; just answer honestly . * MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 274% Seek help from my tutors and instructors to better understand math concepts Understand my textbook Keep up to date with my schoolwork Write course papers Complete my math homework Not at all confident 1 2 3 4 5 6 7 8 9 Extremely confident 10 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 275% 2012-2014 Math and College Self-Efficacy Survey Strongly disagree 1 Disagree 2 Agree 3 Strongly agree 4 I wonder if I really fit in college. It is important to be valued by others, especially at school. I am not sure if I fit with my student peers. I have felt valued in the past. I describe myself as a misfit. It is important that I fit in at school. People accept me. I have qualities. What I offer is valued by others at school (peers, faculty, instructors). I have no place in this world. I want to be a part of things. It is important that my opinions are valued. I'd rather observe college life than participate in it. Others recognize my strengths. I don't really fit in at school. 44. The following 18 items from the Hagerty and Patusky Sense of Belonging Inventory concern your perception of sense of belonging. Using the scale below, please indicate how true each statement is for you. If you strongly disagree, mark at 1. Conversely, if you strongly agree, mark at 4. If you agree more or less, find the number between 1 and 4 that best describes you. There are no right or wrong answers; just answer honestly. * MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 276% I feel left out at school. I make myself fit in at school. I do not feel valued or important at school. Strongly disagree 1 Disagree 2 Agree 3 Strongly agree 4 MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 277% 2012-2014 Math and College Self-Efficacy Survey Mahalo for your participation! Your insight is greatly appreciated and will be used to improve TRiO Student Support Services at Windward Community. If you have any questions, please feel free to contact Sarah Akina at inouyes@hawaii.edu or (808) 253-7326. MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 278% Appendix B: College Self-Efficacy Inventory The following 19 items concern your confidence in various aspects of college. Using the scale below, please indicate how confident you are as a student that you could successfully complete the following tasks. If you are extremely confident, mark at 10. If you are not at all confident, mark a 1. If you are more or less confident, find the number between 10 and 1 that best describes you. Item responses are aggregated across all student respondents in order to better understand how confident the “average” student feels. Levels of confidence vary from person to person, and there are no right or wrong answers; just answer honestly. 1 2 3 4 5 6 7 8 9 10 Not at all Extremely confident confident 1.! Make new friends at college 2.! Successfully complete my current math course 3.! Talk to college staff/faculty 4.! Talk to college administration 5.! Manage time effectively 6.! Ask a question in class 7.! Participate in class discussions 8.! Research a term paper 9.! Do well on my exams 10.! Join a student organization 11.! Talk to my professors outside of class 12.! Ask a professor a question about math coursework 13.! Take good class notes 14.! Complete the Math Requirement (pass Math 28 or 25) 15.! Seek help from my tutors and instructors to better understand math concepts 16.! Understand my textbook 17.! Keep up to date with my schoolwork 18.! Write course papers 19.! Complete my math homework MATH%SUMMER%BRIDGE%EFFECTS%ON%SELF%EFFICACY% % % 279% Appendix C: Sense Of Belonging Inventory Questions The following 18 items from the Hagerty and Patusky Sense of Belonging Inventory concern your perception of sense of belonging. Using the scale below, please indicate how true each statement is for you. If you strongly disagree, mark at 1. Conversely, if you strongly agree, mark at 4. If you agree more or less, find the number between 1 and 4 that best describes you. There are no right or wrong answers; just answer honestly. 1--------------------------2-------------------------3---------------------------4 Highly Disagree Agree Highly Disagree Agree 1.! I wonder if I really fit in college 2.! It is important to be valued by others, especially at school 3.! I am not sure if I fit with my student peers 4.! I have felt valued in the past 5.! I describe myself as a misfit 6.! It is important that I fit in at school 7.! People accept me 8.! I have qualities 9.! What I offer is valued by others at school (peers, faculty, instructors) 10.!I have no place in this world 11.!I want to be a part of things 12.!It is important that my opinions are valued 13.!I’d rather observe college life than participate in it 14.!Others recognize my strengths 15.!I don’t really fit in at school 16.!I feel left out at school 17.!I make myself fit in at school 18.!I do not feel valued or important at school Questions 2, 4, 6, 8, 11, 12, 14, and 17 address the SOBI-A (Antecedents or precursors of sense of belonging). Questions 1, 3, 5, 7, 9, 10, 13, 15, 16, and 18 address the SOBI-P (psychological state of sense of belonging). Students do not see the SOBI as shown in Appendix C. This is only for reference.

Abstract (if available)

Abstract This study uses Astin’s (1999) Inputs-Environment-Outcomes (I-E-O) framework to investigate if a developmental math summer bridge program had a significant effect on six (6) student success measures

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Developmental math in California community colleges and the delay to academic success

PDF

The effects of math anxiety and low self-efficacy on students’ attitudes and interest in STEM

PDF

A longitudinal study on the opportunities to learn science and success in science in the California community college system

PDF

How extending time in developmental math impacts persistence and success: evidence from a regression discontinuity in community colleges

PDF

Examining mathematics anxiety and mathematics teaching self-efficacy of elementary school teachers in a Hawaii school district

PDF

Accountability models in remedial community college mathematics education

PDF

Mathematics identity and sense of belonging in mathematics of successful African-American students in community college developmental mathematics courses

PDF

The perceptions and attitudes of “low-router” students in developmental math

PDF

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

PDF

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

PDF

Math achievement and self-efficacy of linguistically and ethnically diverse high school students: their relationships with English reading and native language proficiency

PDF

The transfer of online instruction to TESOL candidates' perceived self-efficacy of teaching English language learners

PDF

Input-adjusted transfer scores as an accountability model for California community colleges

PDF

Stigma, self-efficacy, and spheres of influence: factors that impact college success for formerly incarcerated college students

PDF

A case study: how a high performing school in a predominantly Hispanic community is curtailing the dropout rate through principal leadership and school organization

PDF

Perspectives of Native American community college students

PDF

A formative evaluation of the student support services TRIO program for low income and first generation college bound students self-efficacy at Butte-Glenn Community College District

PDF

Examining the impact of continuation high schools on students' self-efficacy

PDF

A case study: what factors influenced classical music professors' self-efficacy in teaching and their self perceived self-efficacy on how to teach/mentor their classical music students to degree ...

PDF

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

Asset Metadata

Creator Akina, Sarah Emiko (author)

Core Title The effects of a math summer bridge program on college self-efficacy and other student success measures in community college students

School Rossier School of Education

Degree Doctor of Education

Degree Program Education (Leadership)

Publication Date 04/20/2016

Defense Date 03/19/2016

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

Tag college self-efficacy,community college,developmental math,math summer bridge program,OAI-PMH Harvest

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Keim, Robert G. (committee chair), Eschenberg, Ardis (committee member), Hocevar, Dennis (committee member), Oliveira, Judy Ann (committee member)

Creator Email inouyes@hawaii.edu,seinouye@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-236800

Unique identifier UC11276423

Identifier etd-AkinaSarah-4321.pdf (filename),usctheses-c40-236800 (legacy record id)

Legacy Identifier etd-AkinaSarah-4321.pdf

Dmrecord 236800

Document Type Dissertation

Format application/pdf (imt)

Rights Akina, Sarah Emiko

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