The influence of emotion regulation, self-efficacy, and PTSD symptom severity on College student veterans: a social cognitive model predicting college GPA

The purpose of this research study was to investigate observed variables predicting academic outcomes using structural equation modeling. The results were based on an analysis of variables using a national sample of college student servicemembers and veterans (n = 412). Study participants were recruited with the collaboration of veteran support center administrators and academic administrators at various nationwide colleges and universities. Observations were made from survey responses. Cognitive reappraisal and expression suppression were measured using the Emotion Regulation Questionnaire (ERQ). Additionally, self-efficacies were measured using the Regulatory Emotional Self-Efficacy Scale (RESES) and the College Self-Efficacy Inventory (CSEI). Posttraumatic symptom severity was measured using the Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5). All were predictors for self-reported cumulative grade point average. Structural path analyses were conducted using Mplus Version 7 to infer causal versus noncausal correlations among variables. Model fit indices suggested a college self-efficacy model is indeed the best fitting model as compared to other structural models in this study. Furthermore, an alternative latent model analysis suggested significant predictive strength from cognitive reappraisal as well as expression suppression when predicting college self-efficacy. This is supported by significant correlations that expression suppression and cognitive reappraisal have with college self-efficacy. Moreover, cognitive reappraisal has a significant correlation with cumulative grade point average suggesting that cognitive reappraisal has a unique role in generating positive academic outcomes. Further, college self-efficacy and regulatory emotional self-efficacy partially mediated the significant correlation between cognitive reappraisal and PTSD severity. Additionally, a series of one-way ANOVAs compared participants based upon race, ethnicity, gender, and military branch. Hochberg posthoc analyses suggested that the military reserves had significantly lower rates of PTSD severity as compared to other military branches. Finally, Hochberg posthoc analyses indicated nonsignificant differences in cumulative GPA across gender, military branch, and race. In summary, study results supported and confirmed that cognitive reappraisal and expression both have predictive strength in relationship to college self-efficacy and that regulatory emotional self-efficacy has strength in predicting PTSD severity. It is suggested future qualitative and quantitative studies focus on the predictive relationships that emotion regulation and self-efficacy beliefs have with PTSD severity and academic outcomes among college student servicemembers and veterans. iv

THE INFLUENCE OF EMOTION REGULATION, SELF-EFFICACY, AND PTSD SYMPTOM SEVERITY ON COLLEGE STUDENT VETERANS: A SOCIAL COGNITIVE MODEL PREDICTING COLLEGE GPA by David Matthew Shirley A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Educational Psychology The University of Utah August 2019 Copyright © David Matthew Shirley 2019 All Rights Reserved The University of Utah Graduate School STATEMENT OF DISSERTATION APPROVAL The dissertation of David Matthew Shirley has been approved by the following supervisory committee members: Amy Jo Metz , Chair May 2, 2019 Date Approved Craig Bryan , Member May 2, 2019 Date Approved David Robert Davies , Member May 2, 2019 Date Approved Jennifer Marie Taylor , Member May 2, 2019 Date Approved Dan J. Woltz , Member May 2, 2019 Date Approved and by Chair of the Department of Anne E. Cook Educational Psychology and by David B. Kieda, Dean of The Graduate School. ABSTRACT The purpose of this research study was to investigate observed variables predicting academic outcomes using structural equation modeling. The results were based on an analysis of variables using a national sample of college student servicemembers and veterans (n = 412). Study participants were recruited with the collaboration of veteran support center administrators and academic administrators at various nationwide colleges and universities. Observations were made from survey responses. Cognitive reappraisal and expression suppression were measured using the Emotion Regulation Questionnaire (ERQ). Additionally, self-efficacies were measured using the Regulatory Emotional Self-Efficacy Scale (RESES) and the College Self-Efficacy Inventory (CSEI). Posttraumatic symptom severity was measured using the Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5). All were predictors for self-reported cumulative grade point average. Structural path analyses were conducted using Mplus Version 7 to infer causal versus noncausal correlations among variables. Model fit indices suggested a college self-efficacy model is indeed the best fitting model as compared to other structural models in this study. Furthermore, an alternative latent model analysis suggested significant predictive strength from cognitive reappraisal as well as expression suppression when predicting college self-efficacy. This is supported by significant correlations that expression suppression and cognitive reappraisal have with college selfefficacy. Moreover, cognitive reappraisal has a significant correlation with cumulative grade point average suggesting that cognitive reappraisal has a unique role in generating positive academic outcomes. Further, college self-efficacy and regulatory emotional self-efficacy partially mediated the significant correlation between cognitive reappraisal and PTSD severity. Additionally, a series of one-way ANOVAs compared participants based upon race, ethnicity, gender, and military branch. Hochberg posthoc analyses suggested that the military reserves had significantly lower rates of PTSD severity as compared to other military branches. Finally, Hochberg posthoc analyses indicated nonsignificant differences in cumulative GPA across gender, military branch, and race. In summary, study results supported and confirmed that cognitive reappraisal and expression both have predictive strength in relationship to college self-efficacy and that regulatory emotional self-efficacy has strength in predicting PTSD severity. It is suggested future qualitative and quantitative studies focus on the predictive relationships that emotion regulation and self-efficacy beliefs have with PTSD severity and academic outcomes among college student servicemembers and veterans. iv To my parents, brother, friends, and all those who deserve a warm welcome home. In order to succeed, people need a sense of self-efficacy, to struggle together with resilience to meet the inevitable obstacles and inequities of life. -Bandura, A. (1977) Social Learning Theory TABLE OF CONTENTS ABSTRACT ....................................................................................................................... iii LIST OF TABLES ............................................................................................................. ix LIST OF FIGURES ........................................................................................................... xi ACKNOWLEDGEMENTS .............................................................................................. xii Chapters 1. INTRODUCTION ...........................................................................................................1 2. REVIEW OF THE LITERATURE .................................................................................5 Attending College as a Servicemember or Veteran ........................................................5 Posttraumatic Stress Disorder .......................................................................................10 PTSD, Self-Efficacy, and Academic Outcomes ....................................................14 Emotion Regulation .......................................................................................................23 Emotion Regulation, PTSD, and Academic Outcomes .........................................28 Self-Efficacy ..................................................................................................................37 Regulatory Emotional Self-Efficacy ......................................................................43 College Self-Efficacy .............................................................................................45 Purpose of the Present Study ..................................................................................48 Research Questions .......................................................................................................50 Hypotheses ....................................................................................................................51 3. METHODS ...................................................................................................................56 Participants ....................................................................................................................56 Sampling Procedures .....................................................................................................58 Sampling Size and Power ..............................................................................................59 Measures ........................................................................................................................59 Emotion Regulation Questionnaire (ERQ) ............................................................60 Regulatory Emotional Self-Efficacy Scale (RESES) ............................................64 College Self-Efficacy Inventory (CSEI) ................................................................67 Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5) ...............................71 Data Analysis ................................................................................................................72 Path Analysis .................................................................................................................73 Path Model ....................................................................................................................74 Model Fit .......................................................................................................................75 4. RESULTS .....................................................................................................................76 Statistical Data Analysis ................................................................................................77 Structural Equation Model Path Analysis .....................................................................85 Latent Model Analysis ..................................................................................................95 Mediational Model Analysis .........................................................................................98 5. DISCUSSION .............................................................................................................116 Research Questions .....................................................................................................116 Hypotheses ..................................................................................................................117 Model Comparisons ....................................................................................................118 Original Structural Model ....................................................................................119 College Self-Efficacy Model ................................................................................122 Regulatory Emotional Self-Efficacy Model ........................................................124 Alternate Latent Model ........................................................................................125 Significant Individual Relationships ...........................................................................126 Regulatory Emotional Self-Efficacy ....................................................................126 College Self-Efficacy ...........................................................................................129 Emotion Regulation .............................................................................................130 Limitations of the Present Study .................................................................................133 Implications From the Present Study ..........................................................................136 Future Research Directions .........................................................................................140 Conclusion ...................................................................................................................143 Appendices A: EMOTION REGULATION QUESTIONNAIRE ......................................................144 B: REGULATORY EMOTIONAL SELF-EFFICACY SCALE ....................................145 C: COLLEGE SELF-EFFICACY INVENTORY ...........................................................146 D: PTSD CHECK LIST-5 (PCL-5) ................................................................................147 REFERENCES ................................................................................................................149 viii LIST OF TABLES Tables 2.1 DSM-5 Research Criteria for PTSD ............................................................................52 4.1 Frequencies and Percentages for Sample Demographic Variables............................101 4.2 Means and Standard Deviations for Measurement Sum Scores ................................103 4.3 Correlations Between Measurement Sum Scores ......................................................104 4.4 One-Way Analysis of Variance by Hispanic/Latino Ethnicity ..................................104 4.5 PTSD Severity Means and Standard Deviations by Hispanic/Latino Ethnicity ........104 4.6 Cumulative GPA Means and Standard Deviations by Hispanic/Latino Ethnicity ....105 4.7 One-Way Analysis of Variance by Gender ...............................................................105 4.8 PTSD Severity Means and Standard Deviations .......................................................106 4.9 GPA Means and Standard Deviations........................................................................107 4.10 One-Way Analysis of Variance by Military Branch................................................107 4.11 Posthoc Analysis of PTSD Severity by Military Branch and Race .........................108 4.12 One-Way Analysis of Variance by Race .................................................................108 4.13 Goodness of Fit Comparison: Path Models Predicting Academic Outcomes .........109 4.14 Analysis of Original Recursive Path Model With Multiple Regression ..................109 4.15 Analysis of RESES Recursive Path Model With Multiple Regression ...................110 4.16 Analysis of CSEI Recursive Path Model With Multiple Regression ......................111 4.17 Goodness of Fit: Path Models Predicting Academic Outcomes ..............................111 4.18 Regressions to Generate Path Coefficients and Disturbance Variances ..................112 4.19 Regression to Establish Causal and Outcome Variable Correlation ........................112 4.20 Mediational Model Analysis Results .......................................................................112 x LIST OF FIGURES Figures 2.1 Examples of Hedonic and Counterhedonic Emotion Regulation ................................54 2.2 Proposed Original Path Model Correlates ...................................................................54 2.3 Proposed College Self-Efficacy Inventory (CSEI) Path Model .................................55 2.4 Proposed Regulatory Emotional Self-Efficacy Scale (RESES) Path Model ...............55 4.1 Original Path Analysis Model ....................................................................................112 4.2 Regulation Emotional Self-Efficacy Scale (RESES) Path Analysis Model ..............113 4.3 College Self-Efficacy Inventory (CSEI) Path Analysis Model .................................113 4.4 Latent Model Predicting Academic Outcome............................................................113 4.5 Direct Effect Between Cumulative GPA and Cognitive Reappraisal ....................... 114 4.6 College Self-Efficacy Mediation Model Predicting GPA .........................................114 4.7 Regulatory Emotional Self-Efficacy Mediation Model Predicting GPA ..................114 4.8 Direct Effect Between PTSD Severity and Cognitive Reappraisal ........................... 115 4.9 College Self-Efficacy Mediation Model Predicting PTSD Severity .........................115 4.10 Regulatory Self-Efficacy Mediation Model Predicting PTSD Severity ..................115 ACKNOWLEDGEMENTS I want to recognize U.S. Servicemembers and Veterans for their dedication and sacrifice to their country. I honor and respect your courage to serve and protect your fellow citizens and defend the familiar ground at home. I am inspired by the ways in which you persevere and overcome significant life adversities due to the inherent challenges of military service and the all too common tragedies of war. I also want to thank all the people who contribute to the University of Utah Counseling Center. I admire and emulate your steadfast dedication to social justice and your unrelenting dedication to provide mental health services to a culturally diverse population of college students. Thank you for your tireless support and commitment to clinical trainees and offering your clinical expertise in the field of counseling and clinical psychology. I am particularly thankful to Francis N. Harris, PhD for her commitment to supervising and guiding counseling psychology practicum and intern students. You are an incredibly generous and caring individual who holds genuine interest in the professional growth of your supervisees. I am forever grateful to my advisor, A.J. Metz, PhD, whose reservoir of acceptance and calm resolve fostered my capacity to persevere and overcome the challenges of graduate school. I further want to extend gratitude to all my committee members for supporting my professional development as a researcher and scientistpractitioner. I also want to express my appreciation to individuals who significantly influenced my development as a scientist-practitioner in the field of psychology. Special thanks to Danial Olympia, PhD, Dorlene Walker, PhD, John J, Peregoy, PhD, and Sue Morrow, PhD, for their mentorship and support in preparing me to become an academic researcher and clinician in the field of educational and counseling psychology. Finally, I want to thank Dr. Cecelia H. Foxley for awarding me a scholarship which supported my academic efforts to complete my dissertation research and meet program requirements for graduation from the University of Utah Counseling Psychology Doctoral Program. xiii CHAPTER 1 INTRODUCTION Ninety-six percent of all higher education institutions enroll military servicemembers and veterans (Queen & Lewis, 2014). In 2009, there were almost 500,000 college student veterans; while in 2013, this number rose to over 1,000,000 (Queen & Lewis, 2014). This increase in enrollment may be due to the passage of the Post-9/11 Veterans Education Assistance Act of 2008, otherwise known as the Post-9/11 GI Bill. According to the National Center for Veterans Analysis and Statistics (NCVAS; 2014), the Post-9/11 GI Bill provides full funding for a public, 4-year undergraduate degree to any veteran who has served 3 or more years of active military duty since September 11, 2001. The increase in college student enrollment may also reflect the fact that more veterans are taking advantage of programming and services through the Veteran’s Administration. For example, the Veterans Benefits Administration (VBA) reported that over 900,000 veterans are currently receiving benefits compared to approximately 560,000 in 2006; a 42% increase in service utilization (NCVAS, 2014). Colleges and universities across the nation are directing efforts to expand and strengthen their support services for college student veterans to meet the increasing enrollment rate of this unique population of students. According to a national assessment of institutions (n = 690) conducted by the American Council on Education (ACE), there 2 has been a 12% increase in the overall presence of campus programs and services for college student servicemembers and veterans from 2009 to 2012 (McBain, Kim, Cook, & Snead, 2012). Additionally, a survey conducted by the Student Affairs Administrators in Higher Education (2013) reported that nearly 75% of the total number of responding institutions (n = 239) indicated having a specific staff or office dedicated to meeting the needs of student veterans and active duty military students. For example, the Student Veterans of America (SVA) is a nonprofit coalition of student veteran groups appearing on college campuses nationwide. The SVA supports local student-led chapters by providing money for student scholarships and the building of campus Vet Centers. Additionally, the SVA provides a support network, leadership training, a national conference, and conducts research to inform policymakers, service providers, colleges, and the public in order to advocate for college student servicemembers and veterans. Despite the increase in fiscal and human resources being dedicated to support college student servicemembers and veterans, retention rates continue to remain low; in fact, only 50% of college student veterans complete a degree (Cate, 2014). This low retention rate is often attributed to the difficulties associated with transitioning from overstimulating environments associated with the arousal and stress of war to civilian life (Lighthall, 2012). The transition from active duty military life to civilian life not only represents a greater degree of isolation from the camaraderie and structure experienced while deployed on active duty, but also represents a continuation and a plausible exacerbation of psychological symptoms and severity associated with the recovery from the psychological wounds of being exposed to military combat-related trauma. According to a national survey exploring the psychological symptoms of veterans (n = 434 - 439) 3 returning from Operation Iraqi Freedom (OIF) and Operation Enduring Freedom (OEF), 58% reported exposure to combat and 44.6% of those with combat exposure reported being exposed to moderate or higher (heavy) combat (Rudd, Goulding, & Bryan, 2011). Further, according to Rudd et al. (2011), 34.6% reported experiencing “severe anxiety” based upon measurements of anxiety using the General Anxiety Disorder-7 (GAD-7), 23.7% reported experiencing “severe depression” based upon measurements of depression using the Patient Health Questionairre-9 (PHQ-9) and 45.6% reported significant symptoms of PTSD based upon measurements of PTSD using the PCL-M (PTSD Checklist-Military version). College student veterans face similar challenges as other nontraditional student populations as they tend to be older, have families to support, and juggle work and school. However, unlike their nontraditional peers, college student veterans may need to interrupt their education due to military obligations and may struggle with physical disabilities and mental health issues associated with their military service. Thus, college student veterans may have some of the same struggles as other nontraditional students and may also face unique challenges. At this time, there is limited research on the factors associated with college student success for college student servicemembers and veterans (Student Affairs Administrators in Higher Education, 2013). It is therefore a challenge to draw conclusions about the issues which may impede their successful academic outcomes and cause barriers to persistence towards earning a college degree. The purpose of the current research study is to better understand the potential factors related to the academic success of college student servicemembers and veterans. Prior research supports the role of college self-efficacy in predicting academic success of 4 a general student population, while research on college student servicemembers and veterans specifically describe the role of PTSD symptom severity. Emotional regulation strategies and regulatory emotional self-efficacy have been shown to impact affect and could therefore impact PTSD symptomatology. Thus, the current research study will explicate the influence of emotion regulation strategies, regulatory emotional selfefficacy, college self-efficacy, and PTSD symptom severity on academic success as measured by college cumulative GPA. Because specific linear relationships are hypothesized, a path model will be utilized. The following literature review will begin with a brief discussion about the upward trend of servicemembers and veterans becoming the beneficiaries of education programming. The literature review will then continue with an overview of PTSD and the influence of PTSD symptom severity upon self-efficacy and academic outcomes. Next, emotion regulation, cognitive appraisal, and expression suppression will be addressed, highlighting research demonstrating that emotion regulation strategies are significant contributing factors related to the etiology and maintenance of PTSD symptoms and severity among college students. Additionally, self-efficacy as a broad construct will be examined as a social cognitive factor influencing the academic outcomes of college student servicemembers and veterans. Finally, research literature associated with perceived regulatory emotional self-efficacy and college self-efficacy will be introduced and discussed as potential significant factors influencing and predicting the academic outcome of college student servicemembers and veterans. CHAPTER 2 REVIEW OF THE LITERATURE Attending College as a Servicemember or Veteran The Post-9/11 Veterans Educational Assistance Act of 2008 expanded higher education benefits to military servicemembers and veterans who have served since September 11, 2001. Specifically, the “new G.I. Bill” provides 100% funding of a public, 4-year undergraduate degree for any veteran who has served 3 or more years on active duty since 9/11. These benefits can be transferred to a spouse or child if the veteran has served (or agrees to serve) for 10 years. According to the Department of Veterans Affairs Annual Benefits Report 2015, there has been a 95% increase in the number of beneficiaries from the Post-9/11 Veterans Educational Assistance Program since the new education benefit became effective on August 1, 2009. This trend reflects an everincreasing number of servicemembers and veterans who are returning from active duty and utilizing the Post-9/11 GI Bill to further their education. From 2009 to 2012, there were nearly 500,000 veteran beneficiaries (Sander, 2012). In the 2015 fiscal year, an additional 230,000 veterans across all seven active education programs utilized this educational assistance program. It is clear there is a significant ongoing upward trend in the number of veterans enrolling in colleges and universities as undergraduate and graduate students (National Center for Veterans Analysis and Statistics, 2014). 6 According to a recent survey conducted by Cate and Davis (2016) consisting of 1,352 participants, the college student veteran population is largely male (73%) with the majority of this entire sample identifying as nontraditional college students over the age of 25 (80.4%) and large portions this sample of veterans identifying themselves as being between the ages of 25-30 and 40-50 years-of-age. Additionally, the ethnicity of veterans is represented as being White/Caucasian (70.86%) and Hispanic/Latino (8.58%) with further demographic representations of ethnicity from this sample as being AfricanAmerican (8.06%), Asian (2.48%), Middle Eastern (0.23%), Multi or Biracial (6.78%), Native American/American Indian/Alaskan Native (1.13%), Native Hawaiian/Pacific Islander (0.60%), and those who identified as other or not listed (1.28%). The majority of respondents identified their military rank as being enlisted veterans (90.35) versus officer/warrant officer (9.65%) and a large percentage identified as Army veterans (43.65%), while all other branches were equally represented (18.4% to 18.49%) with the exception of the Coast Guard (1.78%). Furthermore, the vast majority represented veterans during the OEF era (46.30%) and OIF era (37.72%) largely attending 2-year (24.67%) and 4-year (58.59%) institutions. Cate (2014) also described descriptive statistics in which college student servicemembers and veterans reported being first generation college students (62%) with nearly half represented by married (47.3%) and divorced (12.12%) couples and a large number being responsible for the parenting of their own children (47%). Although the Federal Government has provided a strong incentive for veterans to attend college, this population of students has low retention and graduation rates. For example, a recent study conducted by the Veteran Students of America, the National 7 Student Clearinghouse, and the U.S. Department of Veterans Affairs (2014) reviewed the academic records of 859,297 veterans of the Iraq and Afghanistan wars who received Post-9/11 GI Bill benefits. They found that only 51.7% (n = 407,483) completed programs ranging from vocational training to advanced degree programs. Between 2002 and 2013 small percentages of veterans reported earning a certificate (10.3%), associate’s (48.5%), or baccalaureate-level degree (38.6%) as their first degree earned, while an additional percentage of veterans reported earning a baccalaureate-level degree as their highest level of education attained (47.3%) indicating that less than half of veterans continued their education and earned an additional higher education degree beyond a certificate or associate’s degree (Cate, 2014). In addition to low retention rates, college student servicemembers and veterans are more likely to report a perception of lower gains on academic and work-related skills, knowledge, and personal development in comparison to their nonveteran and civilian counterparts. Student servicemember and veterans were also less likely to report gains in academic areas such as acquiring a broad general education, quantitative skills, and learning on their own. Further, their civilian peers were more likely to report high gains in speaking skills, working with others, and solving complex real-world problems. However, student servicemembers and veterans start from a different developmental baseline in comparison to civilian students due to their advanced age and accumulated life experiences, which could potentially moderate the possible perceived gains during their postsecondary education (Kim & Cole, 2013). Notwithstanding the evidence associated with low retention and graduate rates, 8 our understanding of college student veteran retention and graduation rates is limited. The National Association of Student Personnel Administrators (NASPA, 2013) suggested that it’s difficult to make any generalizations about college student veterans when only 26% of colleges and universities disaggregate their data to specifically look at these populations of students independently (NASPA, 2013). According to NASPA’s report, postsecondary institutions appear to have access to unreliable data and utilize insufficient measurement tools to analyze enrollment, retention, and completion rates of college student veterans and servicemembers. Further, the report indicates that only 22% of institutions are disaggregating data related to retention and completion rates for active duty military and veteran students. This leads to a limited understanding of the underlying cause for undergraduate college student servicemembers and veterans who seem to demonstrate a pattern of stopping out or dropping out of college and suggests a need for insight related to the factors affecting the academic success and persistence of student servicemembers and veterans across U.S. American colleges and universities (NASPA, 2013). The low retention and graduation rates of college student veterans may be attributed to demographic characteristics. For example, most college student veterans are male. Female college students tend to have higher GPAs and better retention rates than male college students. In addition, 62% percent of veteran college students are first generation college students, meaning they are the first person in their family to attend college. In the general college student population, those who are first generation are more likely to be racial/ethnic minorities and far more likely to leave school before graduation. In six years, only 40% of first generation college students earn an associate’s or a 9 bachelor’s degree. Moreover, 85% of college student veterans are nontraditional college students. Nontraditional college students are typically defined as a student over the age of 24 years old, often with family and work responsibilities. Nearly half of all college student servicemembers and veterans have families with 47.3% being married and 47% who report being responsible for the care of their own children. College student veterans may face unique challenges as students. The first significant challenge for college student servicemembers and veterans is associated with transitioning from combat to civilian life, and one component of the civilian life may perhaps include the transition into college environments (Romero, Riggs, & Ruggero, 2015). During such transitions college student servicemembers and veterans often feel a sense of detachment from education related settings and can frequently encounter systemic and/or personal stressors related to college environments (Elliot, 2015). Systemic stressors could perhaps include school policies and infrastructure being perceived as being unaccommodating to college student veterans and servicemembers, while personal stressors might include tenuous interactions with faculty members, instructors, and civilian students. For example, potential personal stressors could be attributed to military values among college student veterans and servicemembers being perceived as notably and broadly divergent from civilian values and/or the existence of appreciably disparate political views associated with the United States involvement in foreign wars and conflicts (Livingston, Havice, Cawthon, & Fleming, 2011). 10 Posttraumatic Stress Disorder In addition to the unique challenges described above, college student servicemembers and veterans may struggle with posttraumatic stress disorder (PTSD). Posttraumatic stress disorder has been in the past and continues to be a concerning and widespread problem among military personnel. The estimated prevalence rate of combat related PTSD among military personnel returning from deployment to Iraq and Afghanistan ranges from 5% (Hoge, Auchterlonie, & Milliken, 2006) to 45% (Helmer et al., 2007) based upon assessment approach, sample, time frame, and location (Peterson, Wong, Haynes, Bush, & Schillerstrom, 2010). Another recent historical study of over 18,000 United States Army soldiers from four Active Component and two National Guard infantry brigade combat teams that had been deployed to Iraq or Afghanistan found similar rates of combat related PTSD (Thomas et al., 2010). Specifically, Thomas et al. (2010) found a 9-31% prevalence rate for PTSD depending upon the level of functional impairment self-reported by military personnel. The prevalence rates for PTSD reported above do not disaggregate the rates of PTSD among college student servicemembers and veterans; however, the prevalence of PTSD among military personnel as an entire population reflects the distinct possibility and likelihood that PTSD is a diagnostic disorder experienced by a population of college students. For the purpose of defining PTSD, the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) is the most widely accepted nomenclature used by clinicians and/or researchers for the classification of mental disorders (Table 2.1). According to the DSM-5, posttraumatic stress disorder (PTSD) is a mental health problem in which some individuals experience a combination of 11 characteristic symptoms following exposure to one or more traumatic events (DSM–5; American Psychiatric Association, 2013). The PTSD diagnostic criteria published in the DSM-5 and described below is empirically based on a rigorous 5-year process involving a sub-group of members and advisors focusing on trauma/stress related and dissociative disorders and reviewed by the Scientific Review Committee, Clinical and Public Health Committee, Forensic Committee, DSM-5 Task Force, and a Summit Committee under the aegis of the APA Board of Trustees (Friedman, 2013). The first essential criterion for PTSD is exposure to a traumatic event(s) which is specifically defined being exposed to actual or threatened death, serious injury or sexual violence in such a manner that an individual directly experienced the traumatic event(s), personally witnessed the event(s) occurring to others, learned that a traumatic event(s) occurred to a close family member or close friend, or experienced repeated or extreme exposure to the aversive details of a traumatic event(s) (May & Wisco, 2016). The essential features of PTSD are recognized as a constellation of characteristic symptoms, which chronologically follow the initial exposure to a traumatic event(s). The primary features of PTSD symptomatology are recognized within four broad categories or clusters and include intrusion symptoms, avoidance of stimuli, negative alterations in cognition and mood, and/or marked alterations in arousal and reactivity (DSM–5; American Psychiatric Association, 2013). The clinical presentation of PTSD varies among individuals across the four categories of symptoms described above. Some individuals express the disorder through various fear based reexperiencing, emotional, and behavioral symptoms, while others experience dysphoric mood states, negative cognitions, or dissociative states, or 12 demonstrate arousal and reactive externalization. This type of broad approach to defining PTSD provides an adequate listing of symptoms and symptom clusters to identify the most typical clinical presentations (Friedman, 2013). Below are brief and concise descriptions of the four categories or clusters of symptoms in which a diagnosis of PTSD is based upon: intrusions, avoidance, negative alterations in cognition and mood, and marked alterations in arousal and reactivity. First, intrusion symptoms are the recurrent, involuntary, intrusive recollections of traumatic event(s) or distressing memories related to the traumatic event(s), which are separate and distinguished from depressive rumination when the recollections and memories are directly associated with involuntary and intrusive distressing memories. The recurrent memories tend to evoke sensory, emotional, or physiological behavioral aspects of PTSD and are commonly expressed through distressing dreams or nightmares in which the traumatic event(s) or some representation of the threat associated with the traumatic event(s) is reexperienced. Additionally, dissociative states are also common and potentially experienced as a visual or other sensory intrusion lasting anywhere from a few seconds to a few hours or even a few days in which an individual will relive components of a specific traumatic event(s) and possibly behave as if the traumatic event were occurring in the present moment. These types of episodes are frequently referred to as “flashbacks” in which the episodic occurrence is potentially triggered by external and internal events evoking intense prolonged psychological distress and/or a heightened arousal associated with a physiological reactivity. The “flashback” has the potential to be experienced on a continuum from a partial to total loss of awareness of current surroundings or orientation to reality (Friedman, 2013). 13 Secondly, the avoidance of stimuli associated with the traumatic event(s) is consistent and common. Avoidance symptoms include persistent and intentional efforts to avoid thoughts, memories, and emotions related to the traumatic event(s). However, the behavioral expression of symptomatology associated with the avoidance of stimuli is often manifested in efforts to avoid activities, objects, situations, or people associated with the traumatic event(s) that tend to trigger an arousal of recollected thoughts and/or emotions (Friedman, 2013). Third, persistent erroneous negative alterations in cognition and/or persistent mood state tend to begin or worsen over time after the initial exposure to the traumatic event(s) and are expressed in various forms, such as an inability to recall important aspects associated with the traumatic event(s) and/or persistent and exaggerated negative expectations toward important aspects of life as applied to oneself, others, or the future. Furthermore, the behavioral expression of symptoms is often recognized in the potential for marked diminishment in participation in previously enjoyed activities and/or efforts to become detached and/or estranged from others (Friedman, 2013). Lastly, the potential for marked alterations in arousal and reactivity are recognized as a combination of various behavioral expressions associated with PTSD. The various behavioral expressions associated with marked alterations in arousal and reactivity include the potential for verbal and/or physical aggression toward people or objects, reckless or self-destructive behavior, and/or exaggerated startle responses to external stimuli. Individuals experiencing the symptoms of PTSD are also prone to express irritability and anger with behaviors such as yelling, physical fighting, and damaging or destroying various objects. Additionally, behavior related to the experience 14 of PTSD symptoms potentially includes reckless, self-injurious, or even life-threatening behaviors, such as thrill-seeking activities, excessive substance use, and/or actions associated with suicidal ideation. Further, individuals experiencing PTSD tend to exhibit sensitive reactivity toward unexpected external stimuli and/or express sensitivity toward potential threats in the form of hypervigilance, including perceived external threats that are distinctly associated with the traumatic event(s) as well as those that are seemingly unrelated. Finally, difficulties associated with concentration are commonly reported and might include difficulty remembering daily life events or a diminished capacity to focus on immediate tasks. Problems associated with sleep onset and maintenance can also contribute to issues related to concentration in which elevated arousal due to nightmares and concern for immediate safety can interfere with adequate quality and quantity of sleep (Friedman, 2013). PTSD, Self-Efficacy, and Academic Outcomes Rumann and Hamrick (2010) conducted a qualitative study in which semistructured interviews were completed with five men and one woman who had reenrolled at a university as full time students following an 11-16 month deployment in Iraq, Afghanistan, or Kuwait. Prior to this qualitative study, research tended to focus on the impact of federal assistance programs and academic achievement among college student veterans, largely neglecting the issues related to the transition from war zone deployments. The majority of respondents in this study reported challenges associated with resuming or developing relationships, while also reporting lingering stress and anxiety associated with returning to college. Furthermore, the majority of respondents 15 reported mild symptoms of PTSD and experiencing lingering remnants of high stress levels associated with deployment and combat within the context of being a college student (Rumann & Hamrick, 2010). DiRamio, Ackerman, and Mitchell (2008), also conducted a qualitative study examining the needs of college student veterans transitioning from military service to full time college enrollment at a 4-year institution. DiRamio et al. (2008) interviewed 25 college student veterans who reported serving in the Iraq and Afghan conflicts between 2003 and 2007. All 25 college student veterans were enrolled full time at one of the three research universities representing the three following U.S. geographical regions: northern, southern, and western. The data from this qualitative study revealed that the transition from the military to college civilian life included academic preparation issues that were compounded by symptoms of PTSD, which were cited numerous times throughout the transcripts of interviews with the college student veterans. Furthermore, health issues emerged as important themes in which mental health problems associated with anger and PTSD were expressed along with the implication that current college campus resources are inadequate in meeting the mental health needs of college student veterans (DiRamio et al. 2008). The two comparative qualitative studies conducted by DiRamio et al. (2008) and Rumann and Hamrick (2010) both describe PTSD symptoms as being factors potentially interfering with the academic efforts among those who were interviewed. However, the qualitative research of DiRamio et al. (2008) and Rumann and Hamerick (2010) does not describe the prevalence of PSTD symptoms among college student servicemembers and veterans, nor are the academic experiences associated with the PTSD symptoms and the 16 academic outcomes associated with PTSD explored in detail. The authors only indicate that PTSD symptomatology is a potential common factor related to the experience of college student veterans enrolled with full time status at 4-year institutions (Ackerman, DiRamio, & Mitchell, 2008). Additional research points to PTSD symptoms and symptom severity as factors influencing academic success and persistence among college student servicemembers and veterans (Nyaronga & Toma, 2015). According to Rudd et al. (2011), rates, types, and severity of PTSD symptoms experienced by college student veterans are parallel to those reported by active servicemembers. Their study explored various psychological symptoms and severity related to suicide risk in a national population of student veterans (n = 425) of which almost 46% of the study participants reported experiencing significant symptoms exceeding the cut off score for PTSD. Furthermore, relevant research reports the prevalence of PTSD symptoms among approximately 304,000 Army and Marine servicemembers returning from deployments to Operation Enduring Freedom (OEF, Afghanistan), Operation Iraqi Freedom (OIF, Iraq, Kuwait, Qatar), and other locations (e.g., Bosnia, and Kosovo). Hoge et al. (2006) confirmed previous research showing that military deployment and exposure to combat increases the risk for PTSD based upon data reported by returning service-members who completed a Post Deployment Health Assessment (PDHA) between 2003 and 2004 immediately upon return from deployment. The PDHA includes the Primary Care-PTSD screen (PC-PTSD), which is a 4-item screen for PTSD in which an endorsement of 2 out of the 4 items is considered to be indicative of risk for PTSD. The prevalence of endorsing 2 or more of the PC-PTSD items within the PDHA was 9.8% for OIF servicemembers and 4.7% for OEF servicemembers and 17 2.1% for other locations (Hoge et al. 2006). The prevalence and severity of experiencing PTSD symptoms described above within a large population of servicemembers returning from war zone deployments represents a distinct possibility for a constellation of unique academic challenges among student servicemembers and veterans attending U.S. American college and universities (Nyronga & Toma, 2015). Bryan, Bryan, Hinkson, Bichrest, and Ahern (2014) examined the relationship among self-reported depression symptom severity, PTSD symptom severity, and grade point average in a sample of 276 student servicemembers and veterans ranging in age from 19 to 78 years old. The study consisted of an anonymous online survey in which college student servicemembers and veterans provided self-reported responses to survey items related to GPA, depression severity, PTSD severity, and frequency of academic problems such as late assignments, low grades, failed exams, and skipped classes. The researchers reported small beta weights for self-reported PTSD symptom severity and did not consider the variable as a significant predictor for any of the four academic problems considered within the context of the study (late assignments:  = 0.014, p = 0.84; low grades:  = 0.062, p = 0.43; failed exams:  = 0.030, p = 0.68; and skipped classes:  = -0.049, p = 0.50). Furthermore, PTSD was not found to be a significant predictor of GPA ( = -0.040, p = 0.62). However, the intercorrelations reported in the study indicated a significant correlation between self-reported depression symptom severity and PTSD symptom severity (r = .68, p = 0.01) as well as a proportion of participants with a combination of being screened positive for depression and PTSD (26.0%, CI = 21.9 30.5). Additionally, when PTSD and depression symptoms were simultaneously included in the regression equation, the model was found to be statistically significant (F(4,254) = 18 5.729, p < 0.001) and accounted for 8.3% of the total variance in GPA. Furthermore, an interaction between PTSD and depression added to the regression model in the study indicating what the researchers referred to as a “trend toward significance” ( = -0.378, p = 0.08), which explained an additional small percentage of the variance in GPA (0.01%). Finally, PTSD symptom severity was independently associated with significantly lower GPA ( = -0.151, p = 0.01) as was depression ( = -0.209, p = 0.001), while the proportion of the entire sample who were screened positive for PTSD and positive for depression were also found, on average, this group self-reported a lower GPA (Mean GPA = 3.27, SE =  0.0, p = 0.05) in comparison to those in all other groups. In other words, groups who screened negative for depression and PTSD, screened negative for depression and positive for PTSD, and screened positive for depression and negative for PTSD. Similar results were also reported in a study examining the degree to which posttraumatic stress (PTS) symptoms were associated with academic performance (GPA) and educational self-efficacy (Barry, Whiteman, & MacDermid-Wadsworth, 2012). This particular study included a combined sample of 78 combat exposed student servicemembers and veterans, 53 noncombat exposed student servicemembers and veterans, 38 Reserve Officers’ Training Corps (ROTC) students, and 79 civilian students (n = 248) from various private and public institutions, with varying sizes of enrollment ranging from 4,000 to 40,000 within a Midwestern state. The study reported a significant difference in the gradation of PTSD symptoms being endorsed by college student servicemembers and veterans exposed to combat related trauma in comparison to student veterans without combat exposure, including ROTC students and civilian students 19 (F(3, 249) = 5.13, p < 0.001). In other words, student servicemembers and veterans who reported being exposed to combat related trauma endorsed greater PTSD severity (M =32.95, SD = 13.95). Furthermore, in regards to academic correlates, the results of this particular study indicate that PTS symptoms were negatively associated with and significantly predicted GPA ( = -.18, p < 0.01) and that PTSD symptoms were also negatively associated with and significantly predicted education self-efficacy ( = -.22, p < 0.01) such that as PTSD symptoms increased, GPA and educational self-efficacy decreased. Finally, PTSD symptoms were positively associated with and significantly predicted extrinsic motivation ( = .28, p < 0.001) among student servicemembers and veterans who identified as being exposed to combat related trauma (Barry et al., 2012). The results of the preceding studies conducted by both Bryan et al. (2014) and Barry et al. (2012) appear to indicate that PTSD may be related to lower GPA and an increased likelihood of stop out or drop out among college student servicemembers and veterans. However, there is a possibility that the particular populations of college student servicemembers and veterans in both studies included individuals who persisted beyond the first semester or first year in college. As a result, an entire population of college student servicemembers and veterans experiencing symptoms of PTSD/PTS who enrolled in college and stopped during or after the first semester or first year were perhaps overlooked and not included in the studies. For example, Boyraz, Granda, Baker, Tidwell, and Waits (2016) examined GPA and effort regulation as potential mediators between PTSD symptomatology and college persistence among first year college students (n = 484; mean age = 18.23, SD = .61) who reported a life time exposure to traumatic events (12.4% of the sample population met the full criteria for PTSD). The results of the 20 study indicate that indirect effects of PTSD symptomatology on first year GPA through effort regulation were significant ( = -.12, p < .001) and that an indirect two-step path from PTSD symptomatology to second year enrollment through effort regulation and first year GPA was also significant ( = -.08, p < .001). The results suggest that the experience of PTSD symptoms among first year students who entered college reporting high PTSD symptomatology were more likely to also report lower levels of effort regulation which in turn plausibly predicted their academic outcomes related to first year GPA and persistence for second year enrollment. In essence, the study targeted a population of students early in their academic college career before dropping or stopping out before the second year of enrollment and discovered that those students with high PTSD symptomatology are at the greatest risk for poor academic outcomes due to difficulties with effort regulation (Boyraz et al., 2016). An additional study of the trajectory of PTSD symptom severity during the first year of college further supports the results reported by Boyraz et al. (2016) in which patterns of PTSD symptoms among incoming freshman (n = 944; mean age = 18.11, SD =0.44) were measured using the PTSD Checklist Civilian version (PCL-C; Read, Bachrach, Wright & Colder, 2016). Three measurements of PTSD symptom severity were examined longitudinally in the form of a latent transition analysis over the course of the first year of college enrollment. The three waves of PTSD assessment that were conducted throughout the academic year (T1, T2, T3) using the PCL-C, which yielded a three class solution for PTSD symptom severity: 1) severe symptoms, 2) moderate symptoms, and 3) no symptoms. The findings from this latent transition analysis indicated significant variability in the observed PTSD symptom severity change over the 21 trajectory of being enrolled during the first year of college. The results of the study identified the most typical pattern of observed change in PTSD symptom severity and described the pattern as being inclined toward a natural resolution of PTSD symptoms, which is consistent with the research literature providing evidence for a natural recovery process in those experiencing PTSD symptomatology (Blasco-Ros, Sánchez-Lorente, & Martinez, 2010). However, patterns of a severe class of PTSD symptoms were found among portions of the sample in which entrenched PTSD symptoms showed very little amelioration and indeed presented evidence for symptomatic worsening during the first year of college (Read et al., 2016). The study further reported patterns of stability and change observed within the latent transition analysis in which particular periods of risk were also identified due to symptomatic severity related to PTSD. The populations at greatest risk were those considered to be moderately or severely symptomatic as such that the progression of the academic year showed an increase in the probability of individuals in the moderate class of symptom severity moving into the severe class of symptoms during the second semester. Furthermore, the individuals already classified as being in the severe class of symptoms remained stable throughout the academic year without any change in symptom severity (Read et al., 2016). Finally, Elliot (2015) analyzed a conceptual model in which PTSD and depression were both considered to be mental health issues and potential mediators predicted by college student veterans’ military background, social support, social stressors. The mental health issues reported by college student servicemembers and veterans then predicted aspects of their college experience in terms of being unfairly judged, not fitting in, and feeling uncomfortable on a university campus. Essentially, the conceptual model 22 predicted the college campus experiences of college student veterans and servicemembers based upon previous military deployment, various military related stressors, and the symptoms of mental illness within various levels of social support during and after military deployment. The mental health issues associated with PTSD symptoms and the number of symptomatic days related to depression in the conceptual model functioned as mediating variables to the challenges associated with college student veterans experience of pursuing postsecondary education on a college campus (Elliot, 2015). Elliot’s (2015) analysis of college student veterans’ transition from military service to a postsecondary education environment included respondents (n = 626) attending one out of two research universities (36.6%), a 4-year college (47.9%), or a community college (15.5%). The mean age of respondents in the study was 34 (SD = 8.57) and the average year in school was 2.4 or anywhere between sophomore and junior (SD = 1.02). The conceptual model was assessed by the root mean square error of approximation (RMSEA = .028), the standardized root mean square residual (SMSR = .012), and the comparative fit index (CFI = .99), which are considered to be indicative of good overall model fit. Furthermore, results of the study indicated that the mental health issues being considered in the study were both significant predictors of negative experiences on campus. The statistical outcome associated with PTSD (b = .231, SE = .038, p = .001) and depression (b = .156, SE = .029, p = .001) suggest that symptoms of PTSD and depression are strongly associated with and plausibly predict feeling unfairly judged, feeling like one does not fit in, and feeling uncomfortable on campus (Elliot, 2015). The results of the Elliot (2015) study are meaningful in the sense that respondents 23 are established students who are well beyond their first year of college and continue to experience socioemotional challenges on campus that are plausibly predicted by PTSD symptomatology and previous military experiences. The respondents reported successfully progressing into their second or third year of their college experience and yet also reported ongoing socioemotional challenges within their pursuit of a postsecondary education. This study perhaps identified a population of college student veterans who persevered and succeeded within an academic environment despite the challenges associated with various experiences stemming from military service and PSTD symptomatology. However, the outcome of the study suggests college student veterans experiencing PTSD symptoms perhaps remain at risk of experiencing poor academic outcomes ultimately leading to the potential for stopping or dropping out of college due to their persistent experiences on campus associated with feeling unfairly judged, feeling like one does not fit in, and feeling uncomfortable on campus (Elliot, 2015). Emotion Regulation PTSD symptomatology and severity are plausibly regulated by one’s capability to regulate affective states through the utilization of various components associated with emotion regulation. Emotions are dynamic multi-componential processes occurring and changing over time, which broadly involve periods of interrelated response components consisting of latency, rise time, magnitude, duration, and the onset of responses across behavioral, experiential, and/or physiological domains. However, intrinsic emotion regulation as a singular response component will be focused upon here for the purpose of this study and refers to the automatic or controlled emotion regulatory processes by 24 which individuals consciously or unconsciously influence the particular and specific emotions to be experienced, when to experience the emotions, and how to experience and express the emotions (Gross, 1998). Emotion regulation can be considered as a single overlapping response component within an overarching concept understood to be affect regulation, which also includes coping and mood regulation as two additional response components. Emotion regulation specifically refers to an individual making attempts to influence which emotion to experience, when to experience the emotion, and how to experience or express the emotion (Gross, 1998). Whereas mood regulation is distinguished from emotion regulation in that moods generally include less well defined behavioral response tendencies in comparison to the explicit focus on altering subjective feeling states related to emotions (Larsen, 2009). Further, the defining feature associated with emotion regulation and its focus on altering feeling states includes the activation of a goal so as to influence the trajectory of an experienced emotion (Gross & John, 2003). For example, one may intrinsically regulate a feeling of sadness for the purpose of simply feeling less sad; however, the emotional regulation goal may also be a means to achieving a particular outcome, such as mitigating the feeling of anger towards a coworker or supervisor in order to maintain employment. There are five defining aspects to consider in relationship to Gross’ (1998) response tendency perspective of emotion regulation which are relevant to the current literature review. First, individuals tend to increase, maintain, and/or decrease negative and positive aspects of experienced emotion (Parrott & Smith, 1993). Second, neural emotion circuits do not appear to overlap completely with one another, which suggests 25 that circuits involved in regulating emotions also may not overlap completely with each other, and that there may perhaps be important differences in emotion regulatory processes across emotions (LeDoux, 1994; Panksepp, 1998). Third, Gross’ (1998) definition of emotion regulation places an emphasis upon the intrinsic emotion regulation of the self as compared to the potential attempts to utilize extrinsic emotion regulation to influence the emotions of others as an emotion regulatory process (Gross & John, 2003). Fourth, emotion regulation exists on a continuum from being a conscious, effortful, and controlled regulatory process to being an unconscious, effortless, and automatic regulatory process (Shiffrin & Schneider, 1977), Fifth, there is no a priori assumptions as to whether emotion regulation is good or bad, which essentially circumvents any confusion as to whether emotion regulation is predefined as being a maladaptive defense or an adaptive coping mechanism (Parker & Endler, 1996). Emotion regulation is thought to include motivation and effort to down regulate or decrease negative emotions such as sadness, anger, and anxiety and up regulate or increase positive emotions such as love, happiness, and joy, which is consistent with traditional hedonic perspectives of emotion regulation. (Gross & John, 2003; Quoidbach, Berry, Hansenne, & Mikolajczac, 2010). However, the potential for individuals to engage in the intrinsic counterhedonic regulation of emotion also exists in which negative emotions are increased or positive emotions are decreased based upon the motivation to achieve instrumental goals and/or are influenced by broader cultural expectations of how emotions should be expressed in particular contexts (Szczurek, Monin, & Gross, 2012). The dualistic nature of emotion regulation is depicted (see Figure 2.1) in which counterhedonic and hedonic regulation of emotions are separated depending upon the 26 context of the emotion being experienced by an individual and their motivation for expressing a particular emotion (Gross, 2014). For the purpose of achieving various goals that are associated with emotion regulation, individuals will employ emotion regulation strategies to change the intensity, duration, and/or quality of an experienced emotion. The intensity of an emotion is often controlled by increasing or decreasing the emotional experience or altering specific behaviors associated with the emotion. For example, a parent who changes or hides their feelings of distress from their child (Smith & Klienman, 1989). Additionally, there are times when individuals attempt to change the duration of experiencing an emotion by increasing or decreasing the amount of time in which an emotion is experienced. For example, attempting to extend the feeling of happiness or joy when sharing good news with friends and/or family members (Gable, Reis, Impett, & Asher, 2004). Lastly, there are also additional times in which a change in the quality of an emotional response is sought out by individuals to alter the experienced emotion. For example, an individual who describes a “silver lining” in an otherwise miserable and uncomfortable situation so as to change the emotional experience or an alternate example is an individual who attempts to recognize the serious nature of a humorous situation to decrease any positive feelings associated with the serious/humorous event (Samson & Gross, 2012). Emotions are typically seen as unfolding over a variable period lasting anywhere from a few seconds to a few minutes, which is generally conceptualized within the modal model of emotion (Barrett, Ochsner, & Gross, 2007). However, Gross (2014) further describes the process model of emotion based upon the modal model, which suggests that an emotion generation process follows a situation attention appraisal response sequence. 27 The emotion generation process sequence essentially begins with a psychologically and emotionally relevant situation associated with an aspect of the external environment or is alternatively an activation of an internal representation of the same event. For example, a college student may be called upon in a class to respond to a question posed by an instructor or professor, which is a psychologically relevant situation occurring in the immediate environment that potentially evokes a range of emotions such as anger, anxiety, joy, or happiness. However, the same exact event could be internally represented or imagined, which plausibly evokes the same range of emotions. The situation is then followed by an individual’s attention being directed toward or away from the emotional situation, which is also appraised in such a way that the situation is evaluated or interpreted so as to generate a response giving rise to changes across behavioral, experiential, and/or physiological domains (Gross, 2014). The sequential format of the emotion is thought to continue in a circular, recursive, and dynamic format in which the emotion generation process is potentially helpful or harmful depending upon the intensity, duration, frequency, or type of emotion experienced by a given individual (Gross, 2014). The process model of emotion presents five temporal points in the emotion generative process at which emotions are potentially regulated with emotion regulation processes that are typically categorized into families. There are five families of emotion regulation processes, which consist of situation selection, situation modification, attentional deployment, cognitive change, and response modulation (Scheppes, Suri, & Gross, 2015). Cognitive change and response modulation are the two families of emotion regulation processes that will be described for the purpose of this study. First, cognitive 28 change modulation is a family of strategies which are employed as a means of modifying the late semantic meaning processing of an emotion such as cognitively reappraising the emotional meaning of a situation so as to construe the situation in unemotional terms. Second, response modulation refers to a family of strategies that essentially targets the final stage of the generative-emotion process with a distinct modification of the behavioral, experiential, and/or physiological components of an activated response system, such as expression suppression in which an individual tends to inhibit a behavioral manifestation of an emotion (Scheppes et al. 2015). Emotion Regulation, PTSD, and Academic Outcomes There is evidence demonstrating that the emotion regulation strategies are significant contributing factors related to the etiology and maintenance of PTSD symptoms and severity among college students. For example, Tull, Barrett, McMillan, and Roemer (2007) explored the relationship between posttraumatic stress (PTS) symptom severity and difficulties with effective emotion regulation strategies among college students (n = 194). PTS symptom severity was found to be associated with lower levels of emotional acceptance, difficulty engaging in goal directed behavior when distressed, challenges associated with impulse control, restricted degree of emotional clarity, and limited access to effective emotion regulation strategies (Tull et al. 2007). The study reported multiple regression coefficients indicating significant positive associations between PTS symptoms severity based upon the PTSD Checklist (PCL) total score and over all scores of the Difficulties in Emotion Regulation Scale (DERS; r = .59, p < .01). Additionally, the relationships between total PTS symptom severity and three of 29 the six DERS subscales; lack of emotional acceptance (r = .49), impulse control difficulties (r = .59), and lack of access to effective strategies (r = .61), all remained significant (p < .01). The study used a hierarchical regression analysis to examine difficulties in emotion regulation as a predictor of PTS symptom severity, controlling for income level and negative affect. The results indicated that the DERS total score remained as a significant predictor of PTS symptom severity (B = .30, p < .01) while negative affect accounted for a significant amount of the variance in the model (R2 = .35, p < .001). Furthermore, there were significant differences between mean DERS total scores when comparing participants PCL scores above and below the cutoff scores for a clinical diagnosis of PTSD; F(1, 106) = 42.48, p = .001. Significant differences were also associated with the mean of all subscales scores apart from a lack of emotional awareness subscale. The results of posthoc ANCOVAs indicated that participants above the PCL cutoff score reported significantly greater scores associated with impulse control difficulties, F(1, 104) = 13.33, p < .001; limited or lack of access to effective emotion regulation strategies F(1, 104) = 8.82, p < .01; and limited or lack of emotional clarity, F(1, 104) = 5.79, p < .05 (Tull et al., 2007). In addition, O’Bryan, McLeish, Kraemer, and Fleming (2015) confirmed aspects of the previous findings with evidence in a study examining the relationship between PTS symptoms and aspects of emotion regulation in a sample of undergraduate students (n = 297) with a mean age of 20.46 (SD = 4.64). The DERS was used to assess difficulties with emotion regulation and the Posttraumatic Diagnostic Scale (PDS) was used to assess PTS symptoms based on DSM-IV criteria for PTSD. The results of the study indicated that greater difficulties with aspects of emotion regulation, such as 30 emotional acceptance and accessing emotion regulation strategies when distressed, indeed predicted the cluster of PTS symptoms associated with negative alterations and mood while greater difficulties with emotional acceptance alone predicted avoidance and hyperarousal symptom severity above and beyond the number of trauma types and negative affect. An exploratory analysis was conducted as part of the study to approximate the associations between emotion regulation difficulties and PTS symptom severity according to the DSM-5 negative alterations in cognitions and mood symptom cluster. The second step of the exploratory analysis accounted for 11.9% of the variance and confirmed that greater difficulties with emotional acceptance ( = .20, t = 2.62, p = .009) and greater difficulties accessing emotion regulation strategies when upset ( = .22, t = 2.27, p = .024) were the only two significant predictors of greater negative cognitions and mood symptom severity (O’Bryan et al. 2015). In addition to the results of the previous studies described above, further confirmation that PTSD symptoms have a clear relationship with emotion regulation difficulties was also demonstrated in a recent study conducted by Pickett, Barbaro, and Mello (2016). The study examined the relationships between PTSD specific sleep disturbance, poor sleep quality, and emotion regulation difficulties in a sample of college students (n = 947) from a large midwestern university (Pickett et al. 2016). The results of the study generated bivariate correlations indicating emotion regulation difficulties were positively correlated with PTS symptoms among study participants. Emotion regulation difficulties were assessed using the DERS in which the total score and subscale scores were used as measurements for detecting difficulties associated with emotion regulation. The scores were then correlated with the results of the Distressing Events Questionnaire 31 (DEQ), which is a 17-item self-report measure used to assess the severity of PTS symptoms experienced over the last 30 days and according to the DSM-IV-TR diagnostic criteria (Pickett et al. 2016). The multiple regression analyses confirmed that the DERS total score significantly correlated with PTSS (r = .41, p = .001), while all DERS subscale scores, with exception to subscale scores associated with a lack of awareness of emotion, significantly correlated with PTS symptoms within a range of measured correlations (r = .28-.42, p = .001). In addition to the bivariate correlations previously described, a hierarchical multiple regression analysis indicated that PTS symptoms positively predicted the emotion regulation difficulties as measured by the total DERS score (B = .72,  = .39, p < .001) and significantly accounted for a portion of the total variance (R2 = .17, p < .001; Pickett et al. 2016). Furthermore, and more specifically, empirical evidence demonstrated in a recent study that PTSD symptom severity is positively correlated with expressive suppression and inversely correlated with cognitive reappraisal as two different types of commonly identified emotional regulation strategies among military veteran patients (Boden et al., 2013). The overall results of the study suggest that those diagnosed with PTSD tend to overutilize expression suppression as a relatively ineffective emotion regulation strategy and alternatively under-utilize cognitive reappraisal as a relatively effective emotion regulation strategy. The sample of military veteran patients (n = 93) who participated in this study were admitted to a VA residential treatment program for PTSD between 2008 and 2010 with a mean age of 44.5 years (SD = 14.4). The correlation coefficients between expression suppression, cognitive reappraisal, and PTSD were examined at two temporal points, the first point was at treatment intake and the second point was at discharge. 32 Additionally, expression suppression and cognitive reappraisal were also investigated to determine whether significant differences occurred in pre to posttreatment use of the emotion regulation strategies from the initial point of intake for treatment as compared to eventual discharge from treatment. First, the results of the study indicated that expressive suppression was positively associated with higher PTSD total and cluster symptom severity at intake (PTSD total score: r = .32, p < 0.01 and reexperiencing, avoidance, hyperarousal subscale scores: r = .23 to .24, p < 0.05; numbing subscale score: r = .36, p = 0.01) and discharge (PTSD total score and all subscale scores: r = .39 to .49, p < 0.001) while cognitive reappraisal was negatively associated with lower PTSD total and cluster symptom severity at the same two temporal time points (PTSD total score and all subscale scores at initial intake: r = -0.38 to -.44, p < 0.01 and again at discharge: r = -0.35 to -.44, p < 0.01). In addition, the repeated measures ANOVAs were used to investigate whether use of expression suppression, cognitive reappraisal, and total PTSD symptom severity changed between pre and posttreatment. The results of univariate posthoc tests indicated that expression suppression significantly decreased and cognitive reappraisal significantly increased during treatment while total PTSD symptom severity significantly decreased during the same time period (expression suppression: F = 9.4, p < .0.01; cognitive reappraisal: F = 30.5, p < 0.01; and total PTSD symptom severity: F = 89.6, p < 0.01). Boden et al. (2013) also reported predictive statistics in which PTSD symptom severity at discharge was significantly predicted by change in emotion regulation. Examination of beta scores indicated that lower total PTSD symptom severity was significantly predicted by the reductions in the use of expression suppression ( = .26, p < 0.01) and by the increases in the use of cognitive reappraisal ( = -.21, p < 0.05). 33 Furthermore, the beta scores for individual PTSD symptom clusters indicated that lower PTSD avoidance and numbing symptom cluster severity was significantly predicted by a reduction in the use of expression suppression (avoidance symptom cluster:  = .31, p < 0.01 and numbing symptom cluster:  = .30, p < 0.01; Boden et al., 2013). While there is an extensive amount of research on the utilitarian value of emotion regulation strategies associated with decreasing negative affect, there is currently a limited amount of research regarding emotion regulation as a means of enhancing academic learning, maintaining academic engagement, and supporting positive academic outcomes (Strain & D’mello, 2015). However, Ivevic and Brackett (2014) recently examined positive and negative emotion dysregulation within learning environments and reported evidence indicating that mood dysregulation potentially contributes to poor academic outcomes. In their study comparing conscientiousness and emotion regulation ability, they concluded that emotion regulation ability was a significant and independent predictor of academic school related outcomes among private high school students (n = 213). They defined emotion regulation ability as consisting of knowledge associated with strategies that influence emotions for the purpose of reaching greater wellbeing, building successful relationships, and achieving important goals. The results of their study indicated that emotion regulation ability had significant correlations with school outcomes, which included rule violations associated with academic behavior related to infringement upon school rules such as tardiness, missing or incomplete work, and missing study hall (r = -.16, p < 0.05); formal recognitions for both academic work ethic and citizenship such as completing assignments, academic engagement, punctuality, advocating responsibility, modeling appropriate behavior, respect (r = .35, p < 0.001); 34 and academic honors such as student grades, level of coursework, academic citizenship (r = .27, p < 0.001), average GPA across three trimesters (r = .28, p < 0.001), and satisfaction with school (r = .30, p < 0.001). In addition, the results of the study reported betas with 95% confidence intervals and suggested that emotion regulation ability was a significant predictor of the same academic outcomes described above accounting for an additional 2% variance in academic rule violations ( = -.919 to -.029, R2 = 0.02, p < 0.05), 3% variance in satisfaction with school ( = .004 to .022, R2 = 0.03, p < 0.01), 4% variance in academic honors ( = .016 to .065, R2 = 0.04, p < 0.01), 5% variance in the average GPA across three trimesters ( = .002 to .009, R2 = 0.05, p < 0.001), and 6% variance in school recognitions ( = .030 to .093, R2 = 0.06, p < 0.001; Ivevic & Brackett, 2014). The results of the previously described study suggest the plausible notion that when college student servicemembers and veterans with limited emotion regulation ability experience and appraise emotional events as being emotionally overwhelming then their goals potentially shift from comprehending the immediate implications of their academic environment to a focus on managing their emotional response (Deffenbacher, Bornstein, Penrod, & McGorty, 2004; Levine & Edelstein, 2009). Additional research identified the use of cognitive reappraisal as an emotion regulation strategy being potentially salient to promoting learning and achieving academic success when college students are experiencing negative emotions associated within an academic setting and/or related to academic engagement. College students potentially experience an emotion after an appraisal is made regarding environmental events and the relevance an event has in relationship to their personal goals, values, and 35 wellbeing (Ellsworth & Scherer, 2003). The appraisals made between events and goals will potentially trigger strong emotions, which could then be decreased by reappraising the significance associated with an emotion eliciting event or the emotion could also be decreased by reappraising the outcome associated with the event. Either of these approaches to utilizing cognitive reappraisals has the potential to circumvent the appraisal process and plausibly decrease the intensity of the emotion being experienced (Gross, 1998). Strain and D’Mello (2015) conducted a study in support of utilizing cognitive reappraisals that examined the impact of affective experiences during learning among ethnically diverse adult learners. The study examined cognitive reappraisal as a mediator to increasing academic engagement and higher learning outcomes in comparison to controls in an online learning environment. Participants of the study who utilized cognitive reappraisal as an emotion regulation strategy while learning as much as they could about the U.S. Constitution demonstrated significant learning gains based upon change scores on an identical pretest and posttest as compared to an open ended appraisal condition, an expression suppression condition, and a control condition. An ANOVA yielded a significant main effect of condition on proportional learning gains, F (3, 102) = 2.73, MSE = 0.180, partial η2 = .074. The cognitive appraisal condition (M =. 0.30, SD = 0.17) achieved significantly higher scores in comparison to the openended appraisal (M =. 0.14, SD = 0.22), expression suppression (M = 0.15, SD = 0.29), and control conditions (M = 0.12, SD = 0.24; Strain & D’Mello, 2015). Furthermore, Levine and Edelstein (2009) provide qualitative support for the results of the study described above in an invited review article focusing on emotion and memory research. The authors stated that emotions have the capacity to impair memory when attention is 36 directed toward emotional stimuli. In other words, attentional resources are potentially depleted when attention is directed at emotional stimuli, which consequently interferes with the potential capacity for attention to be directed at neutral information and/or peripheral details thus preventing the same information and/or details from being encoded into short term and/or long-term memory (Levine & Edelstein, 2009). The practice of expression suppression as an emotion regulation strategy potentially relates to the current trend of college student servicemembers and veterans experiencing greater severity of PTSD symptoms and reporting lower academic gains due to its inefficiency associated with moderating physiological arousal and the subjective feeling of anxiety (Hofmann, Heering, Sawyer, & Asnaani, 2009). The insufficient academic outcomes of many Operation Iraqi Freedom (OIF), Operation Enduring Freedom (OEF), and Operation New Dawn (OND) college student servicemembers and veterans suffering from PTSD is possibly attributed to an overutilization of expressive suppression as an ineffective contextual emotion regulation coping strategy and an underutilization of cognitive reappraisal as an alternate coping strategy for regulating emotions related to past trauma (Sippel, Roy, Southwick, & Fichtenholtz, 2016). The potential for OIF, OEF, and OND college student servicemembers and veterans to overutilize expression suppression plausibly interferes with their posttraumatic growth associated with self-perceptions related to college self-efficacy, the ability to manage various academic demands, and the ability to direct their own academic development as a college student within an academic environment (Shepard & Wild, 2014; Wild & Paivio, 2004). 37 Self-Efficacy Social cognitive theory potentially provides a means to understanding the unique challenges that this population of college students encounter within college environments, particularly the interplay among military background, stress exposure, and various academic college campus experiences. The beliefs adopted by college student veterans and servicemembers regarding their capabilities to regulate thoughts and emotions through self-reactive influence is a core property of human agency within social cognitive theory. The perceived regulatory emotional self-efficacy of college student veterans and servicemembers has the potential to affect their perceived academic selfefficacy and academic functioning on a college campus when confronted with various stressful academic encounters based upon the quality of past military and/or combat related experiences (Bandura, 1997; Caprara et al., 2008). Perceived self-efficacy occupies a central role within a self-regulatory process based upon social cognitive theory and is a self-regulatory mechanism governed by the self-regulation of motivation and performance attainments (Bandura, 1988). Perceived personal self-efficacy is composed of a set of beliefs associated with an individual’s own capabilities to mobilize their motivation, cognitive resources, and courses of action required to exercise their control over various life events (Wood & Bandura, 1989). According to Bandura (1988), gaining and maintaining control over various challenging or difficult life events requires individuals to not only possess the necessary skills to respond to such events, but individuals must also exercise a resilient self-belief in their capabilities to gain control over these types of events for the purpose of accomplishing desired goals (Bandura, 1988). Perceived self-efficacy potentially enhances or impairs 38 motivation and performance attainment among individuals who possess and utilize the necessary skills to respond to challenging and/or difficult circumstances (Bandura, 1989a). According to Wood and Bandura (1989a) there are four sources of perceived selfefficacy which entail mastery experiences, modeling, social persuasion, and physiological states. For this study, physiological states will be considered and focused upon as a source of perceived self-efficacy in which individuals assess their capabilities based upon emotional arousal and their construal of somatic physiological information (Wood & Bandura, 1989). For example, individuals who believe that they cannot exercise control over stressors tend to experience higher levels of subjective distress and autonomic arousal (Bandura, Reese, & Adams, 1982), increased plasma catecholamine secretion (Bandura, Taylor, Williams, Mefford, & Barchas, 1985), and activation of endogenous opioid systems (Bandura, Cioffi, Taylor, & Brouillard, 1988). Within this previously described context, perceived self-efficacy operates in part and parcel through cognitive and affective processes as intervening influencers of action. Therefore, perceived self-efficacy is thought to be influenced by cognitive and affective processes which are perhaps self-hindering or self-aiding wherein predictions are often made regarding the occurrence of environmental events and creating the means to exercise control over the various events impacting daily living. This broad perspective of selfefficacy is fundamental to this study in which perceived affective self-efficacy plausibly affects cognitive functioning among college student servicemembers and veterans. In other words, individuals who believe in their capabilities to regulate affective processes remain highly efficient in their analytic thinking within the context of complex academic 39 environments, whereas those who tend to remain emotionally dysregulated become erratic in their analytic thinking within the same environments (Bandura, 1989b). Further explication also entails affective processes which are influenced by the match between aversive emotion stimuli associated with aspects of the environment and perceived coping capabilities among individuals. Thus, college student servicemembers and veterans who do not believe that they have the capability to exercise control over potential threats possibly experience autonomic and intrusive affective arousal, anxiety, and distress, which is essentially maintained due to a mismatch between perceived coping self-efficacy to control and manage intrusive and distressing emotions triggered by external and/or internal stimuli (Bandura, 1989b). College student servicemembers and veterans who have experienced the psychosocial aftermath of traumatic experiences associated with military combat are particularly susceptible to the effects of battlefield trauma on perceived coping selfefficacy (Benight & Bandura, 2004). Solomon, Weisenberg, Schwarzwald, and Mikulincer (1988) examined the effects of military related trauma on perceived selfefficacy among Israeli soldiers and established that the trauma seriously damaged the soldiers’ perceived self-efficacy to cope with combat situations. Additionally, Solomon, Benbenishty, and Mikulincer (1991) longitudinally examined the same sample of Israeli soldiers who continued to demonstrate lower levels of perceived self-efficacy; however, as levels of self-efficacy decreased there was also a correlation with an increase in the frequency of adaptation problems in their subsequent everyday lives. These previously described studies illustrate the pattern associated with the experience of military combat related trauma and its impact upon perceived self-efficacy which is also exemplified in 40 recent research. A similar pattern in the reduction of perceived PTSD symptoms as a function of increased levels of situation specific self-efficacy and general self-efficacy among returning OIF/OEF veterans (n = 216) is demonstrated in a recent study examining the relationships between peer support, self-efficacy, and PTSD symptoms in the context of veterans transitioning into civilian life (MacEachron & Gustavsson, 2012). The study considered self-efficacy to be a central component to exercising resilience towards the effects of combat traumatization and reported intercorrelations depicting significant inverse relationships between PTSD before and after peer support intervention with measurements of general self-efficacy (r = -.53, r = -.38) and situation specific selfefficacy (r = -.68, r = -.56; Bandura, 1994). Further, the results of the study demonstrated that both general self-efficacy and situation specific self-efficacy are significant predictors for reduced PTSD symptoms and function as significant mediators between peer support and reduced PTSD symptoms. The differences in variance explained between models for situation specific self-efficacy was small and significant (R2 = .01, p < .01) as was the difference in variance explained between models for general selfefficacy (R2 = .01, p < .01). The predictive values for situation specific self-efficacy and general self-efficacy were also significant ( = -.14, p < .01;  = -.12, p < .01). The results suggest that providing a supportive environment cultivates a change oriented process promoting self-efficacy beliefs regarding adaptive coping among OIF/OEF veterans and potentially reduces the impact of perceived PTSD symptoms (MacEachron & Gustavsson, 2012). Blackburn and Owens (2015) also reported similar results regarding the 41 relationships among combat exposure, general self-efficacy, and PTSD symptom severity. The sample of military veterans (n = 93) represented various service eras including Iraq (63%), Afghanistan (39%), Vietnam and post Vietnam (42%), Persian Gulf (31%), and other service eras (11%). A majority of study participants reported completing some college (44%) while some reported earning a college degree (29%) and the remainder of participants did not attend college (27%). The study considered general self-efficacy as a possible safeguard against PTSD symptom and defined the construct as confidence in the ability to produce desired outcomes with available resources (Bandura, 1994). The results of the study confirmed a significant negative correlation between general self-efficacy and PTSD severity (r = -.54) and a hierarchical linear regression confirmed the significance of the overall model for PTSD severity (F(6,81) = 15.37, p < .001; Adj. R2 = .01). The significant predictors of the model that are relevant to the current literature review included combat exposure ( = .20, p < .05) and general selfefficacy ( = -.42, p < .001). Furthermore, the interaction between combat exposure and general self-efficacy was also a significant predicator of PTSD severity ( = -.17, p < .05). The results of the study suggest that general self-efficacy is a protective factor for postcombat stress given that veterans reporting higher levels of combat exposure and higher levels of general self-efficacy also reported lower levels of PTSD symptom severity. Furthermore, the authors offered clinical suggestions which primarily focused on teaching strategies aimed at cultivating general self-efficacy among combat exposed veterans (Blackburn & Owens, 2015). Finally, beyond the previous two studies which establish self-efficacy as a significant predictor for PTSD symptom severity, Cieslak, Benight, and Lehman (2008) 42 suggest that the belief in one’s ability to manage posttraumatic recovery, coping selfefficacy, is a significant mediator between negative cognitions and posttraumatic distress. Coping self-efficacy was negatively related to posttraumatic distress, negative cognitions about self, negative cognitions about the world, and with the total score of the Post Traumatic Cognition Inventory (PTCI) among a total sample of women with a history of child sexual abuse (n = 66) consisting of three subsamples of university students (n = 22), inmates (n = 22), and women (n = 22). Approximately 20% of the sample reported high school completion, 57% some college education, and 23% earning a college degree. Correlational analyses indicated coping self-efficacy was negatively correlated with posttraumatic distress (r = -.57, p < .001), negative cognitions about self (r = -.35, p < .01), negative cognitions about the world, and the total score of the PTCI (r = -.26, p < .05). In addition, posttraumatic distress was regressed on coping selfefficacy ( = -.53, p < .001), negative cognitions about self ( = .30, p < .05) and negative cognitions about the world ( = .35, p < .01) with the regression analyses indicating that low coping self-efficacy predicted high levels of posttraumatic distress. However, when coping self-efficacy was removed from the model, negative cognitions of self and world no longer correlated significantly with posttraumatic distress ( = .10, .13) suggesting that coping self-efficacy serves as a full mediator between negative cognitions and posttraumatic distress (Cieslak et al. 2008). The results of the study begin to establish the notion that an example of self-efficacy that insinuates a means of coping with past trauma mediates the effects of negative cognitive appraisals associated with posttraumatic distress. In other words, posttraumatic distress is perhaps an indirect result of negative cognitive distortions that appear to operate through coping beliefs and self-efficacy 43 associated with posttraumatic distress. Regulatory Emotional Self-Efficacy Gross (2014) conceptualized emotion regulation flexibility as an adaptive ability to execute and apply emotion regulation strategies on a deliberate and automatic continuum in response to meeting congruent contextual demands in the environment. Additional constructs that are also similar to emotion regulation flexibility include effortful regulation, emotion related self-regulation, ego resiliency, and emotional intelligence. However, these constructs share a striking similarity to regulatory emotional self-efficacy (Caprara et al., 2008). The essential difference between regulatory emotional self-efficacy and previously identified constructs is that it is a perceived capability rather than a developed skill, ability, or quality of initiating, avoiding, inhibiting, maintaining, or modulating internal aspects related to negative or positive emotion. The distinction is between being one who is effectively able to manage emotional experiences versus one who feels and has a belief being competent to do so. For example, regulatory emotional self-efficacy would include the perceived capability to regulate physiological and/or cognitive processes so as to accomplish individual behavioral and cognitive adjustments, which eventually lead to accomplishing internal or external goals rather than how objectively well this process is ultimately deployed and managed by an individual. For the purpose of this study, regulatory emotional self-efficacy is in essence a social cognitive construct that is defined as one’s beliefs associated with a capability of managing negative emotion and expressing positive emotion (Alessandri, Vecchione, & Caprara, 2015). Regulatory emotional self-efficacy was recently examined as a plausible 44 predictor for future life satisfaction. More specifically, self-efficacy for regulating negative emotions (SERN) and in particular the self-efficacy for managing despondency and distress were considered as possible predictors of life satisfaction among college students (n = 127) attending a medium sized university in the Southern U.S. and a small college in the Northeastern U.S. (Lightsey et al., 2013). The study utilized the Regulatory Emotional Self-Efficacy Scale (RESES), which is a 12-item self-report instrument that was used as a means of measuring perceived self-efficacy for managing negative and positive emotions. The scale measures affective self-efficacy in the form of three dimensions or subscales reflecting important psychological factors related to emotion regulation; self-efficacy for managing despondency and distress (SEDes), self-efficacy for managing anger (SEAng), and the self-efficacy for experiencing and expressing positive emotions (SEPos). Additionally, Satisfaction with Life Scale (SWL), which is a 5-item self-report instrument was used to measure study participants’ overall satisfaction with life. The RESES was used to measure SEDes across four time periods ( = 0.76, 0.82, 0.78, 0.83) and correlated with measurements of life satisfaction using the SWL across the same four time periods ( = 0.85, 0.89, 0.91, 0.90). The life satisfaction of study participants measured by SWL at time period one was reported as being significantly correlated with SEDes at time periods one (r = 0.41, p < 0.01) and two (r = 0.37, p < 0.01). The study included a primary hierarchical regression analysis in which step 2 added SEDes and SEAng to step 1 consisting of SWL and negative affect as measured by the Positive and Negative Affect Schedule (PANAS) with a significant change in variance (R2 = 0.04, F (2, 22) = 7.71, p = 0.01). Furthermore, an examination of regression coefficients indicted that SEDes measured at time 1 was a significant 45 predictor of SWL ( = 0.21, t = 2.96, p = .004), but SEAng was not significant ( = 0.02, t = .31, p = .775; Lightsey et al. 2013). The predictive utility of self-efficacy for managing despondency and distress to predict future life satisfaction is consistent with the social cognitive theory and previous empirical evidence. For example, Bandura (1997) proposed the notion that self-efficacy of negative affect reduces one’s aversion to negative emotions. Additionally, Caprara et al. (2008) produced further evidence that the perceived self-efficacy in managing negative affect, particularly despondency and distress, tends to equate to individuals experiencing reduced frequency of problems associated with anxiety as well as depression and who also demonstrate characteristics of introversion and high self-esteem. The accumulation of the findings described above plausibly suggest that certain aspects and capabilities associated with experiencing negative emotions potentially enhances life satisfaction, which could plausibly be associated with and include academic performance and positive academic outcomes. College Self-Efficacy Solberg, O’brien, Villareal, Kennel, and Davis. (1993) developed the College Self-Efficacy Inventory (CSEI), which is a 20-item inventory consisting of three subscales (Course Efficacy, Social Efficacy, and Roommate Efficacy) measuring an individual’s perceived self-efficacy related to three important aspects of students’ college experience: academic coursework, college social environment, and living with roommates. The CSEI assesses college students’ confidence in their ability to accomplish academic related tasks and includes two subscales that operationally define a basic 46 conceptualization of academic self-efficacy (Gore, Luewerke, & Turley, 2006). For example, the course efficacy subscale measures college students’ perceived confidence in their ability to accomplish course related tasks such as researching and/or writing a course paper, while the social efficacy subscale measures the perceived confidence college students tend to experience in relationship to pro academic behaviors such as approaching and asking a professor to respond to a question outside the context of a classroom or independently engaging in a conversation with academic advising staff (Gore et al. 2006). For the purpose of this study, affective processes as described by Chemers, Hu, and Garcia (2001) are considered to be fundamental to the academic outcome of college student servicemembers and veterans. According to Chemers et al. (2001), self-efficacy has mediating effects upon cognitive, motivational and affective processes, which were included in their conceptualization of academic self-efficacy as a factor related to academic performance. According to their conceptualization, affective processes are thought to be impacted by self-efficacy through the influence that emotions have upon attention and construal of environmental demands, the choice of actions directing behavior, and the capability to regulate negative emotions and the potential to experience negative emotions. Chemers et al. (2001) presented factor loadings within their hypothesized model in which a significant and substantial direct effect of academic selfefficacy on academic performance was reported (= .34, p < .001) among first year college students attending residential colleges at a medium-sized campus university (n = 255). Furthermore, Hsieh, Sullivan, and Guerra (2007) reported a significant correlation between perceived academic efficacy and GPA (r = .36, p < .01) while the research study 47 also indicated that perceived academic efficacy significantly predicted GPA ( = .36, p < .001) among undergraduate students attending a large, metropolitan, Hispanic serving institution in the Southwest United States (n = 112). The results presented by Chemers et al. (2001) and Hsieh et al. (2007) are also consistent with more recent research in which psychosocial factors for college success were examined for the purpose of determining their predictive utility (Krumrei-Mancuso, Newton, Kim, & Wilcox, 2013). The study included academic self-efficacy as a psychosocial factor and was defined as the confidence in academic ability with an expectation for attaining success in college. The results of the study were reported based upon hierarchical regressions in which academic self-efficacy was significantly correlated with first semester and end of year semester GPA (r = .36, .34), while also being identified as a significant predictor of both first semester and end-of-year GPA ( = .30, .17 p < .01). The research literature reflects a predictable and consistent pattern in which the relationship between academic self-efficacy and academic outcomes are strongly correlated in which academic outcomes are consistently predicted by academic self-efficacy. The literature described above is clear in establishing consistent evidence for college students’ academic self-efficacy as a strong predictor for academic success as determined by GPA; however, the literature does not appear to directly consider the contribution and influence affective processes of self-efficacy have upon academic outcomes. The literature instead appears to concentrate its efforts on explicating the cognitive and motivation processes associated with self-efficacy and academic outcomes in lieu of emotional affective processes. 48 Purpose of the Present Study The purpose of the present study is to contribute to the limited research on factors impacting the academic success of college student servicemembers and veterans. Previous research has made connections between emotional regulation strategies such as cognitive reappraisal or expression suppression and the etiology and maintenance of PTSD symptoms (O’Bryan et al. 2015; Pickett et al. 2016; Tull et al. 2007). Research has also described the important role of self-efficacy in the reduction of PTSD symptoms (Blackburn & Owens, 2015; Cieslak et al. 2008; MacEachron & Gustavson, 2012) and its positive relationship with college student outcomes (Chemers et al. 2001; Hsieh et al. 2007; Krumrei-Mancuso et al. 2013). Thus, given that emotion regulation is considered to be a single overlapping response component within an overarching concept understood to be affect regulation (Gross, 1998), then a plausible linear path model can be conceptualized based on the variables described above to predict academic outcomes among college student servicemembers and veterans. The conceptualized linear path model can be described with an emphasis on regulatory emotional self-efficacy and its direct effect on college self-efficacy as variables that could potentially predict PTSD symptom severity and academic outcomes determined by study participant’s self-report. However, a structural model demonstrating the significant correlations between emotion regulation, specific forms of self-efficacy, and PTSD symptom severity as predictors of academic outcome appears to be absent from the literature. This study builds upon previous research and hypothesizes a linear path model that explicates the relationships between common emotion regulation strategies: cognitive appraisal and expression suppression, regulatory emotional self-efficacy, college self-efficacy, PTSD symptom 49 severity, and college GPA in a sample of college student servicemembers and veterans. The current study will include three linear path models. The first hypothesized path model will include observed variables previously described, which will include emotion regulation strategies: cognitive reappraisal and expression suppression, PTSD symptom severity, regulatory emotional self-efficacy, college self-efficacy, and academic outcome: cumulative GPA (see Figure 2.2). Additionally, the current study will present two alternative linear path models. The alternative path models are being presented for comparative purposes and to further explicate the combined influence from regulatory emotional self-efficacy and college self-efficacy in the context of the hypothesized path model described above. The two alternative path models included in the study will emphasize two forms of self-efficacy as observed predictor variables. Each additional and alternative linear path model will include a single form of self-efficacy so as to emphasize their influence on model fit indices and also highlight their combined effect when college self-efficacy and regulatory emotional self-efficacy are included in a linear path model to predict academic performance of college student servicemembers and veterans. First, regulatory emotional self-efficacy will be removed as an observed predictor variable to form an alternate linear path model. In the absence of regulatory emotional self-efficacy, the alternative linear path model will underscore college selfefficacy as an observed predictor variable (see Figure 2.3). Secondly, regulatory emotional self-efficacy will be reintroduced as means to replacing college self-efficacy as an observed predictor variable. This second alternative linear path model will designate regulatory emotional self-efficacy as the only form of self-efficacy to operate in the final path model (see Figure 2.4). The comparisons made between fit indices related to these 50 three linear path models will demonstrate the value of self-efficacy in predicting PTSD symptom severity and academic outcomes among a nationwide sample of college student servicemembers and veterans. Research Questions The following questions guide this research: • Question 1: To what degree do the data support the proposed models of correlates in predicting college student servicemember and veteran academic success as measured by self-reported college GPA. • Question 2: What is the strength of the relationship between PTSD symptom severity and college student GPA for a sample of college student servicemembers and veterans? • Question 3: What is the strength of the relationship between regulatory emotional self-efficacy and PTSD symptom severity and regulatory emotional self-efficacy and college self-efficacy for a sample of college student servicemembers and veterans? • Question 4: What is the strength of the relationship between college self-efficacy and college student GPA for a sample of college student servicemembers and veterans? • Question 5: What is the strength of the relationship between emotion regulation strategies and regulatory emotional self-efficacy and college self-efficacy for a sample of college student servicemembers and veterans? 51 Hypotheses It is hypothesized that self-efficacy and PTSD symptom severity will influence the relationship between emotion regulation and college GPA in a sample of college student servicemembers and veterans in the following ways. First, self-efficacy and PTSD symptom severity will influence the relationship between emotion regulation and college GPA in a sample of college student servicemembers and veterans. Second, regulatory emotional self-efficacy will have a direct effect on self-perceived college selfefficacy and PTSD symptom severity. Third, PTSD symptom severity and college selfefficacy will have a direct effect on college student GPA. Thus, regulatory emotional self-efficacy, college self-efficacy, and PTSD symptom severity serve as mediating variables between emotion regulation and college GPA for college student servicemembers and veterans. 52 Table 2.1 DSM-5 Research Criteria for PTSD A. Exposure to actual or threatened death, serious injury, or sexual violence in one (or more) of the following ways: 1. Directly experiencing the traumatic event. 2. Witnessing, in person, the event(s) as it occurred to others. 3. Learning that the traumatic event(s) occurred to a close family member or close friend. In cases of actual or threatened death of a family member or friend, the event(s) must have been violent or accidental. 4. Experiencing repeated or extreme exposure to aversive details of the traumatic event(s) (e.g., first responders collecting human remains; police officers repeatedly exposed to details of child abuse). Note: Criterion A4 does not apply to exposure through electronic media, television, movies, or pictures, unless this exposure is work related. B. Presence of one (or more) of the following intrusion symptoms associated with the traumatic events(s), beginning after the traumatic event(s) occurred: 1. Recurrent, involuntary, and intrusive distressing memories of the traumatic event(s). 2. Recurrent distressing dreams in which the content and/or affect of the dream are related to the traumatic event(s). 3. Dissociative reactions (e.g., flashbacks) in which the individual fells or acts as if the traumatic event(s) were recurring. (Such reactions may occur on a continuum, with the most extreme expression being a complete loss of awareness of present surroundings.) 4. Intense or prolonged psychological distress at exposure to internal or external cues that symbolize or resemble an aspect of the traumatic event(s). 5. Marked physiological reactions to internal or external cues that symbolize of resemble an aspect of the traumatic event(s). C. Persistent avoidance of stimuli associated with the traumatic event(s), beginning after the traumatic event(s) occurred, as evidenced by one or both of the following: 1. Avoidance of or efforts to avoid distressing memories, thoughts, or feelings about or closely associated with the traumatic event(s). 2. Avoidance of or efforts to avoid external reminders (people, places, conversations, activities, objects, situations) that arouse distressing memories, thoughts, or feelings about or closely associated with the traumatic event(s). D. Negative alterations in cognitions and mood associated with the traumatic event(s), beginning or worsening after the traumatic event(s) occurred, as evidenced by two (or more) of the following: 53 Table 2.1 continued 1. Inability to remember an important aspect of the traumatic event(s) (typically due to dissociative amnesia and not to other factors such as head injury, alcohol, or drugs). 2. Persistent and exaggerated negative beliefs or expectations about oneself, others, or the world (e.g., “I am bad,” “No one can be trusted,” “The world is completely dangerous,” “My whole nervous system is permanently ruined”). 3. Persistent, distorted cognitions about the cause or consequence of the traumatic event(s) that lead the individual to blame himself/herself or others. 4. Persistent negative emotional state (e.g., fear, horror, anger, guilt, or shame). 5. Markedly diminished interest or participation in significant activities. 6. Feelings of detachment or estrangement from others. 7. Persistent inability to experience positive emotions (e.g., inability to experience happiness, satisfaction, or loving feelings). E. Marked alteration in arousal and reactivity associated with the traumatic event(s), beginning or worsening after the traumatic event(s) occurred, as evidenced by two (or more) of the following: 1. Irritable behavior and angry outbursts (with little or no provocation) typically expressed as verbal or physical aggression toward people of objects. 2. Reckless or self-destructive behavior. 3. Hypervigilance. 4. Exaggerated startle response. 5. Problems with concentration. 6. Sleep disturbance (e.g., difficulty falling or staying asleep or restless sleep). 54 Examples of Hedonic and Counterhedonic Emotion Regulation. Negative Emotion Positive Emotion Decrease Increase Trying to reduce feelings of sadness after the death of a close friend. Trying to decrease feelings of joy at the funeral of someone who harmed you in the past. Trying to increase feelings of anger before entering a battle with those who are enemies. Trying to increase feelings of happiness when hearing about friend’s good fortune. Figure 2.1 Examples of Hedonic and Counterhedonic Emotion Regulation. Hedonic emotion regulation is represented as a decrease in negative emotion or an increase in positive emotion (white boxes). Counter hedonic emotion regulation is represented as an increase in negative emotion and a decrease in positive emotion (grey boxes; Gross, 2014). Figure 2.2 Proposed Original Path Model. Correlates for this model include emotion regulation strategies, regulatory emotional and college self-efficacy, PTSD symptom severity, and academic outcome: cumulative GPA. 55 Figure 2.3 Proposed College Self-Efficacy Inventory (CSEI) Path Model. Correlates for this model include emotion regulation strategies, PTSD symptom severity, regulatory emotional self-efficacy, and academic outcome: cumulative GPA. Figure 2.4 Proposed Regulatory Emotional Self-Efficacy Scale (RESES) Path Model. Correlates for this model include emotion regulation strategies, PTSD symptom severity, college self-efficacy, and academic outcome: cumulative GP CHAPTER 3 METHODS Nonexperimental survey research methods were used to gather data and path analyses were used to confirm the relationship between emotion regulation strategies, regulatory emotional self-efficacy, college self-efficacy, PTSD symptom severity, and college GPA in a sample of college student servicemembers and veterans across structural and latent models. Participants Participants for this study included female and male undergraduate and graduate college student servicemembers and veterans currently enrolled in college. For the purpose of this study, servicemember status is defined as a person with active membership with the “uniformed services” consisting of the armed forces otherwise known as Army, Navy, Air Force, Marine Corps, Coast Guard, and the commissioned corps of the National Oceanic and Atmospheric Administration and the Public Health Service. Additionally, according to Title 38 of the Code of Federal Regulations, Veteran status is defined as “a person who served in the active military, naval, or air service and who was discharged or released under conditions other than dishonorable.” This definition describes individuals who engaged in and completed service for any branch of 57 armed forces and classifies them as a veteran as long as they were not dishonorably discharged. However, inclusion for the purpose of the current study is based upon any single type of discharge from wartime and/or peacetime military service in the abovementioned armed forces and includes honorable discharge, discharge under honorable conditions or general discharge, discharge under other than honorable conditions or undesirable discharge, bad conduct discharge, and dishonorable discharge. Furthermore, inclusion criteria consist of college student servicemembers and veterans who discharged from the military during the following wartime eras and geographic wartime areas: Vietnam Era, Conflicts in Lebanon, Grenada, and Panama, Persian Gulf War: Operation Desert Storm, Afghanistan: Operation Enduring Freedom (OEF), Afghanistan: Operation Freedom’s Sentinel (OFS), Iraq: Operation Iraqi Freedom (OIF), Iraq: Operation New Dawn, and the Islamic State: Operation Inherent Resolve (OIR). Demographic variables collected for this study include gender, age, race/ethnicity, household income, military branch, years of active duty, the total number of deployments greater than four weeks, the reception of VA administered financial aid assistance, the reception of a service connected disability, number of credits completed, self-reported cumulative GPA, number of classes receiving a nonpassing grade, number of classes receiving an incomplete, number withdrawn classes, number of semesters on academic suspension, and endorsement of academic probation or dismissal. 58 Sampling Procedures Following approval from the University of Utah Institutional Review Board, an online survey was created to gather data. The online survey was created in the Qualtrics platform and had a hyperlink in which potential research participants were able to click to be taken to the online informed consent form. The online consent form explained the purpose of the study as well as describing the study procedures, confidentiality issues, and possible risks and benefits of participating in the study. After reading the online consent from, the potential research participant could choose whether or not to participate in the study. If they chose to participate, they would then click “yes” on the informed consent form and be taken to the survey: if they decline participation, then they would click “no” and be thanked for their interest. Participants were recruited nationwide through veteran support center administrators and academic administrators at colleges and universities. The principal investigator worked closely with a coinvestigator to contact these organizations and request permission to send an introduction email to enrolled college student servicemembers and veterans. The introduction email provided a brief introduction to the study and an invitation to participate. A hyperlink to the informed consent was included in this email. Potential research participants were invited to use the hyperlink to read more about the study and consent to participate. After reading the consent form, potential research participants could elect to participate in the study or decline participation. Additional methods of recruitment included hardcopy flyers, social media sites, and additional email invitations as necessary. The hardcopy flyers were made available to potential research participants via postings at military servicemember and veteran related 59 sites and events. Additionally, social media sites such as Facebook, Twitter, and Google+ were utilized to advertise a pdf version of the flyer through online postings. Finally, research participants were also be asked to forward the hyperlink to other student servicemembers and veterans upon completion of the online web-based survey. Sampling Size and Power A power analysis was conducted to determine the minimum sample size needed to detect an anticipated effect size (d = .15) with five predictors using conventional level of statistical power (1 -  = .80) and a 95% confidence interval (p < .05). The power analysis was conducted using an a priori sample size for multiple regression calculator, which yielded an estimated minimum sample size equal to 91 participants in order to detect an anticipated medium effect size (d = .15). Missing data were handled using full information maximum likelihood (FIML) as a method of estimation. As an extension of maximum likelihood, FIML takes advantage of all possible data points in analysis. Enders and Bandalos (2001) indicate that full information maximum likelihood is superior to listwise, pairwise, and similar response pattern imputations in handling missing data that may be considered ignorable. Measures The current study was comprised of four self-report measures to assess the use of expression suppression and cognitive reappraisal as methods for emotion regulation, regulatory emotional self-efficacy, college self-efficacy, and PTSD severity. These measures are described in further detail below and will include descriptions of the 60 psychometric properties of each. College student GPA was gathered by asking participants to self-report their cumulative GPA. All self-report measures were presented in the same order in which they are presented in the appendices. The order of which is presented here for the reader; Emotion Regulation Questionnaire, Regulatory Emotional Self-Efficacy Scale, College Self-Efficacy Inventory, and Posttraumatic Checklist-5. Emotion Regulation Questionnaire (ERQ) For the purpose of measuring the habitual use of emotion regulation processes, the Emotion Regulation Questionnaire (ERQ; Gross & John, 2003) was administered. The ERQ is a10-item self-report questionnaire, which captures two emotion regulation strategies, cognitive reappraisal as measured by six items and expression suppression as measured by four items. None of the items are reversed. Participants will be asked to rate the extent to which they typically attempt to think or behave in an effort to change their experienced emotions. They will respond to items that target emotion regulatory processes and strategies to determine how emotions are regulated and managed using a 7point Likert scale in which 1 means “strongly disagree”, 4 means “neutral”, and 7 means “strongly agree” across all items on the questionnaire. Scoring the ERQ utilizes an average of all scores on each subscale with each item being scored between 1 and 7 to capture the habitual use of each emotion regulatory process. Higher mean scores on either subscale indicates habitual use of the emotion regulatory process as a strategy for managing negative and/or positive emotion. The ERQ was developed with two samples of undergraduate college students and has been validated for use with undergraduate college students. 61 The ERQ is a well-established self-report measure based upon a process model of emotion regulation and is designed to assess individual differences in the use of two automatically executed emotion regulation processes: cognitive reappraisal and expression suppression. There is sufficient evidence for the ERQ’s 2-factor model accounting for both cognitive reappraisal and emotion suppression items. Evidence for a 2-factor model begins with the 6-item Reappraisal and the 4-item Suppression scales, which were found to be independent of one another based upon a scale intercorrelations mean (r = .01). Therefore, the ERQ Cognitive Reappraisal and ERQ Expressive Suppression subscales measure two separate emotion regulatory processes intended to down regulate negative emotion. Cognitive reappraisal is defined as an antecedent focused strategy to cognitively change one’s construal of a potentially emotional eliciting situation in such a way that it alters or lessens the potential for a subjective experience of emotional impact (Lazerus & Alfert, 1964). Alternatively, Expression Suppression is defined as a response focused strategy to down regulate negative emotion and involves the intentional inhibition of ongoing emotion expressive behavior (Gross, 1998). Each item in the ERQ was rationally derived and clearly measures the two independent emotion regulatory processes, such as “I control my emotions by not expressing them.” an example item from the Expression Suppression subscale and “I control my emotions by changing the way I think about the situation I’m in.” an example item from the Cognitive Reappraisal subscale. Furthermore, the general emotion items include attention to the valence of emotion associated with regulating both negative and positive emotions. For example, the Reappraisal scale and the Suppression scale include at least one item asking about regulating positive emotions such as joy and amusement as well as at least 62 one item asking about regulating negative emotions such as sadness and anger (Gross & John, 2003) The psychometric properties of the ERQ are based on normative samples and provide evidence to support the reliability and validity of the scale. The ERQ demonstrated internal consistency with alphas ranging from .75 - .82 for the reappraisal subscale and .68 - .76 for the suppression subscale (Gross & John, 2003). The test retest reliability across three months has also been reported (r =.69) for the ERQ subscales using four samples of undergraduate students (n = 1483). Gross and John (2003) also used four additional measures to determine the convergent and discriminant validity of the ERQ. First, a 13-item Inauthenticity scale measuring attempts to disguise the expression of one’s true inner experience of self, due to concerns of self-expression. Second, a 4-item Reinterpretation and Venting scales from the COPE inventory was used to measure coping styles. Third, to measure mood management, three scales form the Trait Meta Mood questionnaire were used and consisted of the 6-item Mood Repair scale, the 13-item Attention scale, and the 11-item Clarity scale. Fourth, the 30-item Negative Mood Regulation scale was used to measure the ability to influence mood change and alter negative mood. The most relevant results in support of convergent validity for the ERQ indicated strategies for emotion regulation are used to reach emotion regulation goals, inauthenticity generally predicts emotion suppression, and coping strategies as well as attempts to regulate negative mood generally predict emotion suppression versus cognitive reappraisal (Gross & John, 2003). First, modest betas for reappraisal (β = .20) and suppression (β = .18) significantly predicted (p < .05) the perception that emotion 63 regulation strategies (appraisal and suppression) help to achieve emotion regulation goals. In other words, both reappraisal and suppression are emotion regulation strategies positively correlated with the perception that subjective efforts to regulate emotion are successful in reaching emotional regulation goals rather than the perception that one emotion regulation strategy is perceived as being more successful in reaching emotion regulation goals above and beyond the other. Second, the measurement of inauthenticity significantly predicted emotion suppression as a strategy for emotion regulation (β = .47, p < .05), while cognitive reappraisal alternatively did not predict inauthenticity (β = -.05). This outcome is indicative of the convergent validity for the ERQ wherein the greater probability for inherent awareness regarding the lack of authentic emotion expression is similar to the attempt to disguise emotion among those who are more likely to utilize suppression as an emotion regulation strategy. Third, aspects of coping and mood also predicted both reappraisal and suppression as emotion regulation strategies. The practice of reappraisal was predicted by coping through reinterpretation (β = .43) while suppression was predicted by coping through venting (β = -.43). These results insinuate that individuals who typically rely upon reappraisal as an emotion regulation strategy are more likely to target something good during the course of a stressful event, whereas those who typically rely upon suppression are less likely to be aware of their negative emotion and also much less likely to express their affect. Moreover, reappraisal is associated with mood repair and that suppression involves the “shutting down” emotions in such way that the practice interferes with the amount of attention given to emotion, which leads to a decreased awareness and clarity of mood along with efforts to repair mood. Furthermore, the perceived efficacy associated with an individual’s ability to regulate mood was 64 positively correlated with reappraisal, but negatively correlated to suppression (Gross & John, 2003). The psychometric properties of the ERQ also demonstrate ethnic and gender differences in the measurement of expression suppression. Ethnicity effects were tested using one factorial analyses of variance (ANOVAs) with ethnicity as a between participants variable in two large samples of undergraduate students (n = 1127). The results of the ANOVAs were significant for both samples and indicated that European Americans are unlikely to use emotion suppression as an emotion regulation strategy in comparison to ethnic minority groups, F(3, 652) = 3.0, p = .03, and F(3, 303) = 5.2, p =.01 (Gross & John, 2003). Further confirmation of ethnicity effects was demonstrated by the planned contrast of mean ratings between two identity groups, European American and ethnic minority (Latino, Asian, and Black), using the same two samples described above in which significant differences were reported; t(654) = 3.0, p = .003 and t(305) = 3.3, p = .001 (Gross & John, 2003). Significant gender differences were also reported across all four samples with overall mean ratings of expression suppression for men (3.64, SD = 1.11) being greater in comparison to women (3.14, SD = 1.18); across samples all four ts = 3.0, all ps = .01. Regulatory Emotional Self-Efficacy Scale (RESES) For the purpose of measuring the perceived self-efficacy beliefs associated with managing negative (NEG) and expressing positive (POS) affect the Regulatory Emotional Self-Efficacy Scale (RESES) was administered (Caprara & Gerbino, 2001). The RESES is a 12-item self-report questionnaire, which assesses self-efficacy in regard 65 to emotional regulation and more specifically, the perceived self-efficacy in managing negative affect in response to adversities or challenging events as well as the capacity to express or manage positive emotions such as joy, enthusiasm, and pride (Bandura et al., 2003; Caprara & Gerbino, 2001). Self-efficacy in managing negative affect is described as an individual’s personal beliefs associated with the capacity to mitigate negative emotional arousal in response to distressing external events and to avoid experiences of becoming overwhelmed with negative emotions such as anger, irritation, despondency, and discouragement. Self-efficacy in expressing positive emotions is described as an individual’s personal beliefs associated with the capacity to experience or permit oneself to express positive emotions such as joy, enthusiasm, and pride in response to personal success or positive external events (Caprara et al., 2008). The RESES essentially measures self-efficacy beliefs associated with emotion regulation, which are presumed to modulate the intensity of emotions and sustain mechanisms of emotion self-regulation, often contributing to internalizing efforts directed at managing impulsive tendencies and thought to be indicative of low levels of externalizing/internalizing and psychopathic problems (Caprara et al., 2008; Eisenberg et al., 2001). Previous psychometric studies using exploratory factor analyses established two separate scales; self-efficacy beliefs in regard to regulating negative emotions (NEG) and self-efficacy beliefs in regard to expressing positive emotions (POS; Caprara & Gerbino, 2001). However, Caprara et al. (2008) also conducted additional exploratory factor analyses using three distinct samples of college students with mean ages and standard deviations to further assess the psychometric properties of the RESES; Italian sample (n = 768, M = 18.72, SD = 0.90), U.S. sample (n = 1,401, M = 18.86, SD = 1.00), and 66 Bolivian sample (n = 301, M = 19.49, SD = 1.46). On the basis of eigenvalues greater than one, three factors emerged from all three samples: the capacity to manage despondency/distress (DES), to express positive affect (POS), and to manage anger/irritation affects (ANG). Eigenvalues for the U.S. sample consisted of 3.48 (POS), 2.08 (DES), and 1.05 (ANG) along with respective cumulative percentages of variance explained and alphas (in parentheses), 29.03 (.69), 46.38 (.72), and 55.14 (.70). Furthermore, a confirmatory factory analysis using the covariance matrices from the same three samples were examined with consideration to testing the factor structure of three separate models. The model demonstrating adequate fit emerged from the U.S. sample (n = 1,401) with the following fit indices (χ2 = 263.3, p = .143, CFI = .95,) and is described as consisting of a single second order factor with two first order factors (despondency/distress and anger/irritation, each with four items) and a first positive order factor with four items (Caprara et al., 2008). Finally, the construct validity for the three factors contributing to the structure of the RESES was established with correlations between its three factors (DES, ANG, and POS) and five separate indices of maladjustment consisting of self esteem, hedonic balance (positive and negative affect), shyness, irritability, aggression and anxiety/depression as well as three indices of adjustment consisting of self esteem, positive affect, and prosocial behavior. All three factors were negatively and significantly (p < .01) correlated with indices of maladjustment, which respectively consist of negative affect (r = -.35, -.37, -.12); shyness (r = -.40, -.26, -.30); irritability (r = -.29, -.53, -.14); aggression (r = -.14, -.31, -.12) and anxiety/depression (r = -.44, -.34, -.25). Alternatively, all three factors were positively and significantly (p < .01) correlated with 67 the indices of adjustment: self esteem (r = .54, .30, .40); positive affects (r = .32, .19, .33); and prosocial behavior (r = .12, .19, .38; Caprara et al., 2008). College Self-Efficacy Inventory (CSEI) College self-efficacy was measured using the College Self-Efficacy Inventory (CSEI; Solberg et al. 1993), which is a 20-item inventory consisting of three subscales (Course Efficacy, Social Efficacy, and Roommate Efficacy) measuring an individual’s perceived self-efficacy related to three important aspects of students’ college experience: academic coursework, college social environment, and living with roommates. Measurements of college self-efficacy using the CSEI are derived from individuals responding to each item using a 10-point Likert type scale in which respondents mark anywhere in a range from 1 if they are not at all confident to a 10 if they are extremely confident (Solberg et al., 1993). Item responses indicate the degree of confidence a responder has in their ability to successfully perform various college related tasks. According to Bandura (1986), self-efficacy consists of the set of beliefs an individual has about their personal ability to organize behavioral, cognitive, and social skills for producing a desired behavioral outcome under a variety of circumstances. Further and more specifically, college self-efficacy is an individual’s perceived selfefficacy across three domains of the college experience: academic coursework, college social environment, and living with roommates (Wright, Jenkins-Guarneri, & Murdock, 2012). However, for this study, academic coursework and college social environment will be measured at the exclusion of living with roommates. The model being considered for this study consists of multiple predictors for academic outcomes; yet, the model does not 68 include college students’ roommate self-efficacy. Thus, for this study, efforts to include survey items related to this domain of an individual’s college experience is unnecessary. The CSEI was initially validated with a sample of 164 Mexican American and Latino American college students responding to a survey questionnaire consisting of 20 self-efficacy items considered to be representative of the college experience (response rate = 51.7%). The college-efficacy subscales were then established using a principal components analysis with the original 20 self-efficacy items. Factor loadings accounted for 70% of the total variance for items meeting eigenvalue criteria. This analysis resulted in a three factor solution based upon items with factor loadings greater than .50 and is represented by three efficacies: course (λ = 9.42), social (λ = 1.37), and roommate (λ = 2.10). Course self-efficacy consists of seven items related to aspects of course performance, which accounted for 44.8% of the variance with factor loadings ranging from .59 (understand your textbooks) to .91 (research a term paper). Social self-efficacy consists of eight items related to aspects of interpersonal and social adjustment, which accounted for 7.2% of the variance with factor loadings ranging from .56 (join a student organization) to .88 (participate in class discussions). Finally, roommate self-efficacy consists of four items related to aspects of interacting with roommates, which accounted for 11.5% of the variance with factor loadings ranging from .80 (divide chores with roommates) to .95 (get along with roommates). The college self-efficacy subscales, which are based upon this principle component analysis, provide a 3-factor model for measuring college self-efficacy (Solberg et al., 1993). The reliability for the CSEI was established with internal consistency using coefficient alpha estimates for the instrument score in its entirety (α = .93) and the three 69 subscales (α = .88). Convergent and discriminant validity were established based upon correlations between the three self-efficacy subscales and the Brief Symptom Inventory (BSI; Derogatis & Melisaratos, 1983), the College Stress Scale (CSS; Solberg et al., 1993), the Social Provision Scale (SPS; Russell & Cutrona, 1984) and the Acculturation Rating Scale for Mexican Americans (Cuellar, Harris, & Jasso, 1980). The findings based upon a principle components analysis with a varimax rotation indicate a pattern indicating that the college self-efficacy subscales were related to other aspects of adjustment and also unrelated to aspects of nonadjustment, which is suggests good convergent and discriminant validity for the CSEI (Solberg et al., 1993). Concurrent and predictive validity was confirmed using a sample of 257 freshman college students enrolled in a first-year experience course at a medium sized public midwestern university in which a confirmatory factor analysis was used to determine a three-factor model based upon significant factor loadings (p < .05) and significant correlations observed between all three subscales (p < .01; Gore et al., 2006). Gore et al. (2006) also confirmed the internal consistency with previously reported coefficient alpha estimates for the CSEI total score (α = .92) and the three selfefficacy subscales; course efficacy (α = .88), social efficacy (α = .86), and roommate efficacy (α = .83). Concurrent validity was further established with correlations between the CSEI measurement scores and two additional measures of college self-efficacy. CSEI scores were significantly correlated with scores on the short form of the Career Decision Making Self-Efficacy Beliefs Instrument (r = .62, p < .01), an alternate measure of selfefficacy (CDMSE-SF; Betz, Klein, & Taylor, 1996). Additionally, scores on the CSEI were compared with scores on the College Student Expectations Questionnaire (CSXQ; 70 Kuh & Pace, 1998). The CSEI total score was significantly correlated with all CSXQ subscales (p < .05), while various significant correlations (p < .05) were also established with CSEI subscale scores. First, the CSEI course self-efficacy subscale scores were significantly correlated with higher expectations for participation in various academic activities related to learning, writing, and reading. Secondly, scores on the CSEI social self-efficacy scale were significantly correlated with the majority of the CSXQ subscales. However, scores on this scale differentially predicted higher expectations for using campus facilities and joining university clubs, organizations, and service projects. Finally, the CSEI roommate self-efficacy scale scores were significantly correlated with course learning and writing scores and differentially correlated with higher expectations for interacting with faculty and establishing student acquaintances (Gore et al. 2006). Furthermore, the predictive validity of scores on the CSEI was determined based upon measures of college performance and persistence. Correlated CSEI scores with College GPA served as a predictive measure of performance, while comparisons of CSEI scores of retained students and unretained students were made to serve as a predictive measure of persistence. The predictive measure of performance based upon CSEI academic scores accounted for approximately 7% of the variance in cumulative GPAs (r = .25 to .29, p < .01), while the CSEI total score accounted for approximately 3% of the variance in cumulative GPAs (r = .17 to .19, p < .05). Additionally, statistically significant differences between retained versus unretained students’ CSEI obtained scores were observed at the end of their first semester to determine a predictive measure of persistence. Further, college students retained over a two-year period (M = 6.91, SD = 1.22) had significantly higher CSEI academic subscale scores compared to unretained 71 students (M = 6.22, SD = 1.56, F (1, 173) = 8.08, p < .01). Similarly, the observed CSEI total scores indicated that retained college students reported significantly higher levels of college self-efficacy (M = 7.15, SD = 1.08) as compared to unretained college students (M = 6.73, SD = 1.42; F (1, 173) = 4.52, p < .05; Gore et al. 2006). Posttraumatic Stress Disorder Check List for DSM-5 (PCL-5) Posttraumatic stress disorder (PTSD) symptom severity was measured using the posttraumatic stress disorder checklist for Diagnostic and Statistical Manual of Mental Disorders Fifth Edition (PCL-5; Weathers, Litz, Herman, Huska, & Keane, 1993 & Weathers et al., 2013), which is a commonly used self-report measure for assessing PTSD symptomatology and severity (Bovin et al., 2015). The PLC-5 consists of 20 items that correspond with the PTSD symptomatology as outlined in the Diagnostic and Statistical Manual of Mental Disorders Fifth Edition (DSM-5). Respondents indicate the extent to which PTSD symptoms have bothered them over the last month using a rating scale ranging from 0 to 4, which represents a total score ranging from 0-80 with higher scores indicating greater PTSD symptom severity (Bovin et al., 2015). There are studies that have evaluated the psychometric characteristics of the PCL5 utilizing treatment seeking veteran (Wortmann et al., 2016), active duty (Hoge, Riviere, Wilk, Herrell, & Weathers, 2014), and community samples (Armour et al., 2015). However, Blevins, Weathers, Davis, Witte, and Domino (2015) evaluated the psychometric properties of PCL-5 scores in two independent samples of trauma exposed undergraduate college students at a large public university in southeastern United States. The first study (n = 278), PCL-5 scores demonstrated high internal consistency (α = .94) 72 and was comparable to other measures of PTSD symptomatology. Moreover, a subset of the original participants (n = 53) was used to analyze the test retest reliability with a readministration of the PCL-5 using an approximately weeklong interval in between administrations (M = 6.14 days). The PCL-5 total-score demonstrated good test retest reliability (r = .82), while at the item level, the PCL-5 demonstrated overall good consistency with correlations ranging from r = .39 to .83 and a median r = .68. The convergent validity was demonstrated between the PCL-5 and other PTSD measures with a range of correlations (r = .74 to .85, p < 1), while correlations for the discriminant validity of the PCL-5 ranged from r = .31 to .60; the instrument was found to be moderately correlated with related constructs, such as depression (r = .60), and somewhat correlated with measures of unrelated constructs, such as antisocial personality features (r = .39) and mania (r = .31). Furthermore, confirmatory factor analyses indicated adequate fit with the 4-factor model of PTSD described in the DSM-5 (χ2 = 455.83, p < .001), standardized root mean square residual (SRMR = .07), root mean squared error of approximation (RMSEA = .08), and comparative fit index (CFI = .86). In the second study (n = 558), PCL-5 scores also demonstrated strong reliability and validity. Overall, results indicate that the PCL-5 is a psychometrically sound measure of PTSD symptoms (Blevins, Weathers, Davis, Witte, & Domino, 2015). Data Analysis This research study reports descriptive statistics and correlations for all major study variables. Additionally, a series of one-way analyses of variance (ANOVAs) was conducted to determine if there are significant mean differences in PTSD and college 73 GPA by age, gender, ethnicity/race, household income, military branch, and years of active duty. Significant mean differences were controlled for in subsequent analyses. Path Analysis A manifest path analysis was used to test hypothesized relationships of the independent variables, emotion regulation, regulatory emotional self-efficacy, college self-efficacy, and PTSD severity and the dependent variable, college GPA (see Figure 2.2). Path analysis is a form of multiple regression as it examines the influence of multiple independent variables on one dependent variable and a special case of structural equation modeling (SEM) as it tests a structural model, but not a measurement model (Kline, 1998). Path analysis is used to provide estimates of the magnitude and significance of hypothesized causal connections between sets of variables (Shumaker & Lomax, 1996). In path modeling, exogenous variables (or independent variables) are modeled as being correlated with and having both direct and indirect effects on the endogenous (or dependent variables). Endogenous may also be affected by measurement error, or factors outside of the model, and thus, they have additional error terms associated with them (Shumaker & Lomax, 1996). The advantage of using path analysis as an SEM technique is that it provides an opportunity to analyze a set of data representing the relationship among independent and dependent variables on a single occasion. Further, the path analysis allows for a decomposition of correlations in which total effect, direct effect, and indirect effect via mediation are revealed among variables. However, there is a significant limitation to using a path analysis based upon the difficult assumptions that must be satisfied, all of 74 which include the following: 1) variables must be related in a linear, additive, and causal manner; 2) residual variance is not correlated with variables preceding the model; 3) the flow of the model is linear so as to eliminate reciprocal causation; 4) only interval variables are measured; and 5) variables are measured without error (Shumaker & Lomax, 1996). Path Model Figure 2.2 illustrates the proposed path diagram that depicts the correlates between the observed variables being examined in this study. The hypothetical path diagram includes correlates represented by lines and observed variables represented by rectangles (Shumaker & Lomax,1996). The straight lines with a single arrow head represent the relationships with exogenous and endogenous variables being examined in this research study. The exogenous variables are those that are caused by unanalyzed forces outside the proposed model, while the endogenous variable is being influenced by variables within the model (Kline, 1998). The direction of the arrow represents the direction of causality supported by the relevant literature. The curved dotted line with bidirectional arrow heads represents the assumed correlations between cognitive appraisal and expression suppression that are assumed to covariate with one another (Gross, 1998; Kline, 1998). The observed exogenous variables represented by rectangles are being measured for this study and include emotion regulation strategies (cognitive reappraisal and expressive suppression), regulatory emotional self-efficacy, college selfefficacy, and PTSD symptom severity. The single endogenous variable is also represented by a rectangle located on the far right of the proposed model and is identified 75 as the academic outcome of college student servicemembers and veterans as determined by self-reported college GPA. Additional detail related to the path diagram model will be addressed in the data analysis section. Model Fit Several statistical tests and their indices were used to determine the adequacy of the proposed model’s fit to the data. The Pearson chi-square statistic (2) was the first statistical test used to determine the goodness of model fit. The Pearson chi-square statistic indicates the amount of difference between expected and observed covariance matrices, which is interpreted based upon a value in relationship to 0 and a probability value greater than 0.05 (p > 0.05). A chi-square close to 0 with a p > 0.05 indicates a goodness of fit with little difference between expected and observed covariance matrices. The second statistical test to be used was a Comparative Fit Index (CFI), which compares the performance of a proposed model to the performance on a baseline or null model that assumes zero correlation between all observed variables (Hu & Bentler, 1999). The CFI ranges from 0 to 1 with larger values indicating a better model fit; however, an acceptable model fit is indicated by a CFI of 0.90 or greater (Kline, 1998). The final statistical test to determine the proposed model fit was the Root Mean Square Error of Approximation (RMSEA), which is related to the amount of residual in a model. RMSEA values range from 0 to 1 with RMSEA values less than or equal to 0.05 being considered a close approximate fit, while RMSEA values greater than 0.05 to 0.08 are an acceptable fit. Any RMSEA value beyond 0.10 is a poor fit and unacceptable model (Kenny, 1979) CHAPTER 4 RESULTS The purpose of this study was to investigate and compare three different models, which include variables predicting academic performance among college student veterans and servicemembers. Academic performance is the outcome variable which remains constant across all three models. The comparative models in this study proposed three different variations in the direct and indirect relationships between exogenous and endogenous variables (see Figures 2.2, 2.3, and 2.4). The exogenous variables included in each model are defined as two separate emotion regulation strategies, expression suppression and cognitive reappraisal. The endogenous variables included across all three models predicting academic performance constitute PTSD symptom severity, college self-efficacy, and regulatory emotional self-efficacy. Based on previous research, this study hypothesized that self-efficacy and PTSD symptom severity will influence the relationship between emotion regulation and college GPA in a sample of college student servicemembers and veterans in the following ways. First, self-efficacy and PTSD symptom severity will influence the relationship between emotion regulation and college GPA in a sample of college student servicemembers and veterans. Second, regulatory emotional self-efficacy will have a direct effect on self-perceived college self-efficacy and PTSD symptom severity. Third, PTSD symptom severity and college self-efficacy 77 will have a direct effect on college student GPA. Thus, regulatory emotional selfefficacy, college self-efficacy, and PTSD severity serve as mediating variables between emotion regulation and college GPA for college student servicemembers and veterans. Statistical Data Analysis The results of this study are based upon a statistical analysis using a representative nationwide sample of college student servicemembers and veterans. The results are presented with a broad overview of the method for data collection, which will be briefly described and followed by a demographic description of the study sample. The descriptive results of the instruments used in this study will then be described based on this study sample and the results of a correlational analysis and analyses of variance for specific variables of interest will be reported. The results of the study will be outlined and presented in a sequential manner directly reflecting the questions and hypotheses guiding this research. First and foremost, the degree to which the data supports the proposed models of correlates in predicting academic success among college student servicemembers and veterans will be presented as comparisons of the goodness of fit indices across three hierarchical models. Additional comparisons across all three models will be made to address the remaining four questions guiding this research. The response to the second research question will be presented as a comparison of the path coefficients representing the predictive strength of the relationship between PTSD symptom severity and college student cumulative GPA across all three models. The next research question will be presented as a comparison of the three models with an emphasis on the strength of the relationship between regulatory emotional self-efficacy and PTSD symptom severity 78 as well as the strength between regulatory emotional self-efficacy and college selfefficacy. Fourth, the relationship between college self-efficacy and college student GPA will be examined across all three models and presented as a comparison of the predictive strength among the path coefficients from each model. Finally, the strength of the relationships between both emotion regulation strategies, cognitive reappraisal and expression suppression, in relationship to each form of self-efficacy, regulatory emotional self-efficacy and college self-efficacy, will be compared and presented across all three models. The University of Utah Institutional Review Board approved this study. The principle investigator and coinvestigator developed an online survey to gather data. A sample of nationwide participants (n = 412) were recruited through veteran support center administrators and academic administrators at colleges and universities. The principal investigator of this study worked closely with a coinvestigator who contacted these organization administrators and requested permission to send an introduction email to enrolled college student veterans and servicemembers. The introduction email provided a brief introduction to the study and in invitation to participate. A hyperlink to the informed consent was also included. The research participants were invited to use the hyperlink to read more about the study and consent to participate. The nationwide sample for this study included undergraduate and graduate college student servicemembers and veterans currently enrolled in college (n = 412). The frequencies and percentages of demographic variables for this sample of college student servicemembers and veterans appear in Table 4.1. The entire sample is largely made up of males (n = 315) and fewer females (n = 97). The racial identities within the sample 79 consists of Caucasian (n = 363), African American (n = 32), American Indian (n = 10), Asian (n = 2), and Other (n = 6) college student veterans and servicemembers with nearly half the sample reporting their identification with Hispanic/Latino ethnic origin (n = 186). A large proportion of the sample falls within an age range between 20 and 39 (n = 369) and with an average household income between $75,000 and $150,000 (n = 251). Each military branch is represented by study participants, Army being the military branch with largest representation of the sample (n = 212) and then followed by Air Force being the second largest military branch to be represented (n = 89). The number of years of active service ranges from no active duty whatsoever to 7 years or more with a large proportion reporting 2 to 4 years active duty (49.6%) and a substantial percentage reporting 1 to 10 deployments (63.9%). Data was available for 412 college student veterans and servicemembers. However, the results are based upon participant item survey responses made available from a convenient nationwide sample of college student veterans and servicemembers. Some individual items were missing since not all survey items were given a response by each study participant. Thus, the sample sizes across all the variables used in this study will vary based on the willingness of each study participant to respond to survey items. An analysis was conducted to determine descriptive statistics of the variables used in this study. Means and standard deviations from the instrument measurements appear in Table 4.2 First, the means and standard deviations for the Emotion Regulation Questionnaire’s (ERQ) two subscales, Cognitive Reappraisal (CR) and Expression Suppression (ES) are listed. The total scores for each subscale were available for 373 of the 412 individuals who comprised the entire sample. The average score for this sample on the CR subscale 80 was 26.27 (SD = 6.33; range = 9-42) and the average score for the ES subscale was 16.61 (SD = 4.36; range = 6-28). Second, the total scores for the Regulatory Emotional SelfEfficacy Scale (RESES) were available for 376 individuals from the sample. The average score on the RESES was 35.75 (SD = 8.24; range = 13-60). Third, total scores for the Colleges Self-Efficacy Inventory (CSEI) were available for 378 individuals in the sample. The mean CSEI score for this sample was 50.25 (SD = 10.72; range = 21-80). Next, total scores for the PTSD Checklist for DSM-5 (PCL-5) were available for 387 individuals with a mean score of 34.96 (SD = 15.81; range = 0-76). Finally, self-reported cumulative Grade Point Average (GPA) were provided by 295 individuals from the entire sample. The average GPA for this sample was 3.48 (SD = 0.429; range = 1.00-4.00). A correlational analysis of instrument composite scores was conducted to further understand the relationships between self-reported use of emotion regulation strategies (cognitive reappraisal and expression suppression), self-reported levels of regulatory emotional self-efficacy and college self-efficacy, self-reported severity of PTSD symptomatology, and self-reported academic college performance. The instrument composite scores were generated for each observable variable. All composite scores were calculated in the following manner. The items from each instrument were summed to determine the composite score for cognitive reappraisal, expression suppression, regulatory emotional self-efficacy, and college self-efficacy, and PTSD severity. There were no items from any of the instruments which required to be reversed scored. Thus, for example, all items pertaining to expression suppression from the Emotion Regulation Questionnaire (Gross & John, 2003) were summed to form a composite score for each study participant. Correlations between instrument measurements appear in Table 4.3. 81 The results of the correlational analysis indicated that self-reported use of cognitive reappraisal was significantly and positively correlated with reported levels of college self-efficacy (r = .17, p < .01) and regulatory emotional self-efficacy (r = .51, p < .01). Further, cognitive reappraisal was also significantly and positively correlated with selfreported cumulative Grade Point Average (r = .15, p < .01) and significantly and negatively correlated with PTSD severity (r = -.35, p < .01). The results of this correlational analysis indicate that as self-reported use of cognitive reappraisal increases then the regulatory emotion self-efficacy, college self-efficacy and cumulative GPA of study participants increases concurrently while PTSD severity decreases. In contrast, expression suppression was significantly and negatively correlated only with college selfefficacy (r = -.15, p < .01). This evidence suggests that higher levels of expression suppression are correlated with lower levels of college self-efficacy. Furthermore, the results of the correlational analysis demonstrate that self-reported levels of PTSD severity were minimally correlated with expression suppression (r = .007) as an endorsed strategy for regulating emotions and self-reported cumulative GPA (r = -.042). However, the results also indicate PTSD severity is significantly and negatively correlated with college self-efficacy (r = -.21, p < .01) as well as regulatory emotional self-efficacy (r = -.39, p < .01). Together these findings suggest that self-reported levels of expression suppression and cumulative GPA do not correspond with PTSD severity. However, the results also suggest that sample participants’ college self-efficacy and regulatory emotional selfefficacy both tend to decrease with a self-reported increase in PTSD severity. A one-way between subjects ANOVA was conducted to compare PTSD severity and cumulative GPA among college student veterans and servicemembers based upon 82 their ethnic origin as being Hispanic/Latino or non-Hispanic/Latino. Significant differences between ethnic origins were not found for PSTD severity while significant differences were found for cumulative GPA (see Table 4.4). However, the assumption of homogeneity was not met for both PTSD severity and cumulative GPA as indicated by Levene’s Test for Homogeneity of Variances, respectively reported as [F(1, 386) = 68.27, p = .000] and [F(1, 294) = 27.10, p = .000], therefore the Welch F-ratio is reported. There were not significant differences between groups denoting ethnic origin based on PTSD severity based on the Welch F-ratio [F(1, 356.68) = 1.316, p = .252] and cumulative GPA [F(1, 171.60) = 3.375, p = .055]. Mean score comparisons indicated that PTSD severity for the Hispanic/Latino condition (n = 179, M = 34.04, SD = 11.76) did not differ from the non-Hispanic/Latino condition (n = 209, M = 35.83, SD = 18.60; [F(1, 386) = 1.231, p = .268]; see Table 4.5). Similarly, comparisons between the mean cumulative GPA for the Hispanic/Latino condition (M = 3.53, SD = .54) and the non-Hispanic/Latino condition (M = 3.42, SD = .55) did not differ from one another (see Table 4.6). Taken together, these results suggest that sociocultural factors related to ethnicity do not necessarily influence academic performance based upon self-reported cumulative GPA, nor those do same factors related to ethnicity appear to significantly influence the severity of self-reported PTSD symptomatology. In similar fashion, one-way between subjects ANOVAs were conducted to compare PTSD severity and cumulative GPA among study participants based on gender (see Table 4.7). The assumption of homogeneity was met for both PTSD severity and cumulative GPA as indicated by Levene’s Test for Homogeneity of Variances, 83 respectively reported as [F(1, 386) = .000, p = .984] and [F(1, 294) = 1.01, p = .315]. Therefore, supporting both one-way between subjects ANOVAs being conducted to compare PTSD severity and cumulative GPA among college student veterans and servicemembers based upon their gender. There was a significant effect of PTSD severity on gender at the p < .05 level for this condition [F(1, 386) = 9.192, p = .003]. However, the second ANOVA conducted to compare cumulative GPA between the same two groups did not suggest a difference between means [F(1, 294) = .493, p =. 483]. Posthoc analyses were not performed for either of these ANOVAs due to their simple comparison of two different conditions associated with gender, female and male. Comparisons indicated the mean score for PTSD severity associated with the female condition (n = 94, M = 30.24, SD = 15.79) significantly differed from the male condition (n = 293, M = 36.31, SD = 15.61; see Table 4.8). In contrast, comparisons did not reveal significant mean score differences for cumulative GPA based on gender (see Table 4.9). The mean cumulative GPA among female college student veterans and servicemembers (n = 91, M = 3.51, SD = .44) did not significantly differ from the mean cumulative GPA among their male counterparts (n = 205, M = 3.47, SD = .42). Taken together, these results suggest that sociocultural factors related to gender perhaps do not necessarily influence academic performance based upon self-reported cumulative GPA, however, the sociocultural factors related to gender appear to significantly influence the severity of self-reported PTSD symptomatology with higher rates of severity among males. An additional one-way between subjects ANOVA was conducted to compare participants’ self-reported PTSD severity and cumulative GPA across all armed service branches of the U.S. military as well as the National Guard and Reserves (see Table 84 4.10). The assumption of homogeneity was not met for PTSD severity nor was the same assumption met for cumulative GPA. Levene’s Test for Homogeneity of Variances indicated the assumption for homogeneity had been violated for PTSD severity [F(6, 381) = 2.13, p = .049] and for cumulative GPA [F(6, 289) = 2.22, p = .041]. Therefore, a Welch’s F-ratio was used to compare potential differences between groups based on military branches. Based on the Welch F-ratio, significant differences in PTSD severity do exist across military branches [F(6, 41.51) = 2.838, p = .021], however, the Welch Fratio indicated significant differences between military branches do not appear to exist for cumulative GPA [F(6, 18.11) = 2.006, p = .073]. Posthoc analyses were conducted to compare the military branches based on PTSD severity given the significant Welch Fratio test. Specifically, Hochberg posthoc tests were conducted on all possible pairwise contrasts. The results of the Hochberg posthoc analysis (see Table 4.11) indicated that the mean PTSD severity score for the Reserves (n = 9, M = 15.44, SD = 19.55) significantly differs (p < .05) in comparison to the average mean PTSD severity score for all other service branches (n = 379, M = 36.11, SD = 16.29). In other words, based on this sample, colleges students serving or having served in the military reserves had significantly selfreported lower rates of PTSD severity on average as compared to all other branches of military service. A final pair of one-way between subjects ANOVAs were conducted to compare both self-reported PTSD severity and cumulative GPA across the five racial identities among study participants, which included White Caucasian, African American or Black, American Indian or Native American, Asian, and other (see Table 4.12). The test of homogeneity of variances was met for PTSD severity [F(4, 383) = 2.13, p = .076] 85 supporting the ANOVA indicating that a significant difference between racial groups does indeed exist [F(4, 383) = 3.56, p = .008]. Hochberg posthoc analyses indicated a difference between racial groups based on PTSD severity with the following racial groups being significantly different from the group of Asian college students (n = 2, M = 59, SD = 5.7; p < .05); Caucasian (n = 342, M = 35.34, SD = 15.9), African American (n = 31, M = 35, SD = 10.97), Native American (n = 8, M = 26.13, SD =17.05) and other (n = 5, M = 17, SD = 19.77). Alternately, the assumption of homogeneity of variances was violated based on measurements of cumulative GPA [F(4, 291) = 6.14, p = .000], which does not support the ANOVA indicating a significant difference between racial groups [F(4, 291) = 3.60, p = .007] when comparing academic outcome. Therefore, the Welch F-ratio test was alternatively used to compare racial groups so as to determine whether or not a significant difference between racial groups does indeed exist based on cumulative GPA. The Welch F-ratio comparing cumulative GPA across all racial identities indicated no significant difference between these groups [F(4, 6.065) = 1.869, p = .234]. In sum, the results appear to suggest that there are possible sociocultural factors influencing group differences when comparing PTSD severity across different racial groups; however, these results should be taken with caution due to the small number of participants identifying as Asian. Further, the same sociocultural factors do not appear to influence the academic performance of college students across the same racial identities included in this study. Structural Equation Model Path Analysis The results of this study are presented in a sequential manner to directly reflect the questions and hypotheses guiding this research. The hypothesized relations among 86 two emotion regulation strategies (cognitive reappraisal and expression suppression), college self-efficacy, regulatory emotional self-efficacy, posttraumatic stress symptom severity, and cumulative GPA are considered with a series of structural equation model (SEM) path analyses. This analysis confirmed the structural models (Figures 4.1, 4.2, and 4.3). The purpose of this analysis was to examine the hypothesized direct and indirect relationships among the observed variables and to describe the amount of explained and unexplained variance. A comparison of models was made to test hypotheses about path models and to examine the adequacy of fit for the just identified models predicting academic outcome among a sample of college student servicemembers and veterans. Comparisons were made between all three models in a hierarchical manner with two out of the three models being trimmed versions of the original model. Each of the two trimmed versions of the original model included a form of self-efficacy in the model to determine the predictive value of endogenous variables and the overall influence on model fit. A series of path analyses were conducted for this study using an MPLUS-7 (Version 1.4) statistical package with maximum likelihood estimates using covariance matrices to estimate and infer the presumed causal versus no causal correlations among observed variables. All participant responses were included in this analysis and any missing data was accounted for in MPLUS-7. The statistical package (MPLUS-7) is automatically set to a default by using full information maximum likelihood (FIML) in which all available data is used to directly estimate each parameter and analyze the entire, but incomplete data set. SEM path analysis is a valuable methodology for testing theories and the assumption of probabilistic causality between variables within a recursive path model. 87 SEM path analysis requires multiple fit indices to be used so as to evaluate the goodness of fit relative to a concurrent measurement of observed variables included in a given path model. This current study selected to use three fit indices and standardized residuals to determine the accuracy of model specification. The three fit indices selected for this current study are (1) Pearson chi-square (2), (2) Comparative Fit Index (CFI), and (3) Root Mean Square Error of Approximation (RMSEA). The 2 is an overall test of model fit used in this path analysis to assess the magnitude of the discrepancy between the observed covariance matrix and the covariance matrix specified by the model. A significant p value for a 2 suggests that the hypothesized model does not adequately describe the sample data. Bentler (1990) and Bentler and Bonett (1980) have cited problems associated with the use of the 2 statistic. Both authors have argued that this particular test statistic is sensitive to sample size. Bentler (1990) indicated that in small samples, T may not be 2 distributed and therefore distorts probability values associated with whether the null hypothesis can be rejected or not. Furthermore, Bentler and Bonnet (1980) state that an additional problem with the 2 statistic is a direct function of increasing sample sizes in which the probability of rejecting any model increases as N increases. The consequence is that virtually any model will be rejected as being statistically untenable if the sample size is large enough. Bentler (1990) has endorsed the use of additional fit indices such as the comparative fit index (CFI) to compare the target model to the fit of an independent, or null model. This index provides information about the relative reduction in the lack of fit as estimated by the noncentral 2 of a target model versus a baseline model. In larger samples, the CFI should be similar to other relative fit indices, however, the CFI should still be taken as a less biased indicator of fit. The CFI 88 will range between .00 and 1.00 with higher values representing better model specification. In addition to the CFI fit index is the RMSEA, which is a common absolute measure of goodness of model fit within structural equation modeling and is based on the noncentrality parameter (Kenny & McCoach, 2003). The RMSEA favors parsimony based on optimally chosen parameter estimates and indicates the degree to which the model fits that population covariance matrix. Recommendations for cut off points indicating “fair” fit have been historically suggested to be in the range of 0.05 to 0.10 and cut off points of 0.06 are more recently seen as being an acceptable indicator of good fit (Hu & Bentler, 1999). However, current literature suggests to use caution when assessing goodness of fit based upon the performance of the RMSEA in relation to models with small degrees of freedom (Kenny, Kaniskan, & McCoach, 2015). Furthermore, there is a growing concern related to the use of strict cut off values when evaluating the goodness of fit associated with a structural model based on small sample sizes. Alternatively, evaluation of model fit is improved when the characteristics of the data are considered and that multiple fit indices are used to assess and evaluate the fit of a given model (Nye & Drasgow, 2011). A set of comparisons of the goodness of fit indices across three hierarchical models is presented in Table 4.13 These indices describe the degree to which the current data supports the proposed models of correlates in predicting academic success among college student servicemembers and veterans. Further, the goodness of fit indices used in this study address the hypothesis that self-efficacy and PTSD symptom severity will influence the relationship between emotion regulation strategies and college student cumulative GPA. First, all three models produced significant Pearson chi-squared (2) 89 coefficients; Original Model (2 = 24.34, p = .05); RESES Model (2 = 11.20, p = .05); and CSEI Model (2 = 7.33, p = .05). However, trimmed models including only one form of self-efficacy produced Pearson chi-square difference scores comparative to the Original Model. The CSEI Model produced the greatest difference score (2diff = 17.01) and a chi-squared coefficient approaching nonsignificance (p = .02), while the RESES Model produced a smaller chi-square difference score (2diff = 13.14). Secondly, examination of all three target models based on the Comparative Fit Index (CFI) indicates good fit across all models (CFI = .90 - .92). More specifically, a comparison across all three models indicates that the CFI scores for both trimmed models, RESES (CFI = .92) and CSEI (CFI = .92), are comparatively improved relative to the Original Model (CFI = .90). Finally, the Root Mean Square Error of Approximation (RMSEA) indicates that all three models are approaching good fit with index scores slightly larger than the expected value (RMSEA < .08). All three models produced RMSEA fit index scores (RMSEA = .12 - .16) beyond an expected value (.05) that would suggest good overall model fit. The CSEI model yielded the lowest fit index score (RMSEA = .12) which is approaching the recommended cut off score of .10 for poor fitting models. Secondly, the three target models were tested to investigate the predictive strength of the relationship between PTSD symptom severity and college student cumulative GPA in a nationwide sample of college student servicemembers and veterans. PTSD symptom severity is positioned as a predictor variable in all three models with college student cumulative GPA as the criterion variable. Each model regresses cumulative GPA on PTSD symptom severity to produce a regression coefficient for the purpose of considering PTSD symptom severity’s predictive strength. The predictive strength of the 90 regression coefficient for PTSD symptom severity and college student cumulative GPA is based on a standardized beta weight. The results from a regression analysis across all three models suggests that PTSD symptom severity does not significantly predict college student cumulative GPA and accounts for very little of the variance among this nationwide sample of college student servicemembers and veterans. The standardized regression coefficients and the accounted variance for PTSD severity predicting college student cumulative GPA in the original model as well as the in CSEI model were equivalent with one another and nonsignificant (β = .13, SE = .095, p = .16) with a large percentage of the variance being left unexplained (1-R2 = .96). However, the regression coefficient for the RESES model was much smaller in comparison to both the original and CSEI models (β = .07 SE = .093, p = .45) with even less of the variance being accounted for in this model (R2 = .01). The next comparison across the three models in this study will focus on the strength of the relationship between regulatory emotional self-efficacy and PTSD symptom severity as well as the strength between regulatory emotional self-efficacy and college self-efficacy. It was hypothesized that regulatory emotional self-efficacy will have a direct effect on PTSD symptom severity and self-perceived college self-efficacy. Indeed, in consideration to the regression coefficients for the original model as outlined in Table 4.14, regulatory emotional self-efficacy does have a direct effect on PTSD severity as a significant predictor (β = -.398, SE = .062, p < .05) and accounts for a small portion of the variance with (R2 = .16). Furthermore, in addition to being a significant predictor of PTSD severity, regulatory emotional self-efficacy has a direct effect on college self-efficacy and was also found to be a significant predictor (β = .571, SE = .062, 91 p < .05) in the original model, with a moderate amount of variance being explained (R2 = .35). The statistical results associated with the original model can be compared to those associated with the RESES model (Table 4.15). It is important to note that the comparison of the original model is being made to the trimmed RESES model in which college self-efficacy was dropped as an endogenous variable. There is a notable difference in the value of the regression coefficient between the two models when comparing the direct effect that regulatory emotional self-efficacy has on PTSD severity. The value of the regression coefficient between PTSD severity and regulatory emotional self-efficacy decreased as a result of dropping college self-efficacy from the original model. Regulatory emotional self-efficacy has a direct effect on PTSD severity and remains a significant predictor for PTSD severity in the RESES model (β = -.240, SE = .076, p < .05). However, regulatory emotion self-efficacy significantly predicts PTSD severity with a smaller regression coefficient and explains only a small portion of the variance with (R2 = .22). Finally, a comparison of the strength of the relationship between regulatory emotional self-efficacy and PTSD symptom severity as well as the strength between regulatory emotional self-efficacy and college self-efficacy in the CSEI model cannot be made. This particular model does not include regulatory emotional selfefficacy as an endogenous variable and therefore a comparison is impossible due to the fact that regression coefficients do not exist for comparative purposes. The CSEI model is a trimmed model in which regulatory emotional self-efficacy was dropped from the original model in a hierarchical fashion for comparative purposes. More specifically, the trimmed models allow the influence of including a singular form of self-efficacy to be considered in the overall path analysis. As such, it is worth pointing out that within the 92 context of the CSEI model, college self-efficacy does not significantly predict PTSD severity (β = -.124, SE = .070, p = .08). The quantitative relationship between college self-efficacy and college student GPA will now be examined across all three models and will be presented as a comparison of the predictive strength among the regression coefficients from each manifest model included in this study. The hypothesized relationship college self-efficacy has to college student cumulative GPA was described as being a direct and significant predictor among a nationwide sample of college student servicemembers and veterans. A comparison between the original model and the CSEI model reveal regression coefficients for college student cumulative GPA on college self-efficacy (Table 4.16). The two regression coefficients from the original model and CSEI model are logically equivalent to one another with identical values (β = .18, SE = .085, p < .05) and an amount of variance explained (R2 = .04). These two regression coefficients depict college self-efficacy as having a direct effect and significantly predicting college student cumulative GPA. In addition, and for comparative purposes, it is worth noting that the RESES model does not include college self-efficacy. Instead, regulatory emotional self-efficacy is positioned in the model as being in direct relationship to college student cumulative GPA. The direct effect that regulatory emotional self-efficacy has on college student cumulative GPA is insubstantial and the regression coefficient is nonsignificant. Finally, the strength of the relationships between both emotion regulation strategies, cognitive reappraisal and expression suppression, will be addressed in relationship to each form of self-efficacy. In other words, regulatory emotional selfefficacy and college self-efficacy, will be compared and presented across all three 93 models. First, a path analysis of the statistical results from all three models show that emotion regulation strategies influence academic outcomes through college and regulatory emotional self-efficacies. Based on the standardized path coefficients, regulatory emotional self-efficacy was significantly predicted by both emotion regulation strategies in the original model and RESES model. The beta weights for each emotional regulation strategy are reported for these two models as follows, expression suppression (β = -.175, SE = .066, p < .05) and cognitive reappraisal (β = .578, SE = .057, p < .05). Comparatively, the beta weights for each emotional regulation strategy predicting regulatory emotional self-efficacy in the CSEI model are reported for expression suppression (β = -.218, SE = .073, p < .05) and cognitive reappraisal (β = .358, SE = .07, p < .05). In other words, both emotion regulation strategies, cognitive reappraisal and expression suppression significantly predict regulatory emotional selfefficacy and equivalently account for a small portion of the variance in both the original model and RESES model (R2 = .29) and with a lesser amount of variance explained in the CSEI model (R2 = .12). Furthermore, regulatory emotional self-efficacy was found to be a significant predictor for college self-efficacy (β = .571, SE = .062, p < .05) in the original model and accounted for a moderate amount of variance (R2 = .35) as previously stated above. Additionally, PTSD severity was significantly predicted by regulatory emotional selfefficacy (β = -.398, SE = .062, p < .05) and explained a small portion of the variance in the original model (R2 = .16). However, in regard to the RESES model, regulatory emotional self-efficacy did not significantly predict academic outcomes based on cumulative GPA (β = .075, SE = .086, p = .38), yet did significantly predict PTSD 94 severity (β = -.240, SE = .076, p < .05) with only a small portion of variance being explained (R2 = .22). This is in direct contrast to the CSEI model in which college selfefficacy significantly predicts cumulative GPA (β = .180, SE = .085, p < .05) with very little variance being explained (R2 = .04). Alternatively, college self-efficacy does not significantly predict PTSD severity (β = -.124, SE = .07, p = .08) yet accounts for a larger portion of variance (R2 = .19). Therefore, both emotion regulation strategies are exogenous variables which could plausibly be mediated by regulatory emotional selfefficacy through college self-efficacy in the original model so as to significantly predict college student cumulative GPA among college student veterans and servicemembers (β = .18, SE = .084, p < .05). However, for the purpose of further clarification, it is relevant to point out that although cognitive appraisal and expression suppression significantly predict regulatory emotional self-efficacy, neither of the emotion regulation strategies significantly predict self-efficacy in the original model. The first emotion regulation strategy being cognitive reappraisal (β = .028, SE = .076, p = .71) and the second strategy being expression suppression (β = -.118, SE = .065, p = .07). Finally, the plausible mediation in the original model described above does not appear to influence academic outcome through PTSD severity due to its path coefficient being a nonsignificant predictor (β = .134, SE = .095, p = .16) of cumulative GPA. Thus, it is plausible that regulatory emotional self-efficacy and college self-efficacy both serve as mediating variables between emotion regulation and college GPA among this sample of college student servicemembers and veterans based on the statistical results of the original model. These findings do not support the hypothesized model in its entirety and instead supports alternative mediational models predicting PTSD severity and academic 95 outcomes based on cumulative GPA. Latent Model Analysis This analysis considers a fourth and final latent model depicting emotion regulation strategies and college self-efficacy as predictors of academic outcome and takes into consideration the measurement model as well as the structural model. The measurement model consists of observed variables or instrument items, which are considered to be measured by valid and reliable instruments. These instrument items define and form the factors or latent variables in this final model (Figure 4.4). Observed variables from each measurement instrument are loaded on a factor and are correlated with the construct of interest, and share common variance with the other variables loaded on the same factor. In this case, four factors or latent variables were identified (cognitive appraisal, expression suppression, and college self-efficacy) with the corresponding observed variables loading onto these factors. College cumulative grade point average was treated as an observed variable solely based on participants’ self-reported GPA and hence a subjective estimation of their academic performance. Similar to the previous path analysis, this final analysis considers overall test of fit in structural equation modeling which assesses the magnitude of the discrepancy between the sample (observed) covariance matrix and the covariance matrix specified by the model (Bentler, 1990). All indexes of fit utilize a test statistic (T) that is the product of the sample size and a discrepancy function (Hoyle, 1995). For example, the 2 goodness of fit value was used for preliminary assessment of model fit in this study. A significant 2 value suggests that the hypothesized model does not adequately describe sample data. 96 Bentler (1990) cited problems associated with the use of the 2 statistic. He argued that this test statistic is sensitive to both sample size and violations of the assumption of multivariate normality and has endorsed the use of additional fit indices such as the comparative fit index (CFI; Bentler, 1990). A 2 change statistic (Bentler, 1980) can be used when one is able to specify alternative models that are nested within a hypothetical model. Two models are said to be nested if they both contain the same parameters, but the set of free parameters in one model is a subset of the free parameters in the other (Bentler, 1990). The 2 change statistic is used to determine which model better accounts for the observed data. Essentially, the observed 2 value for the nested model is subtracted from the less restrictive alternative model. The value obtained is also 2 distributed with degrees of freedom equal to the difference between degrees of freedom of the two models being compared (Hoyle, 1995). The hypothesized latent model predicting academic performance of college students is presented in Figure 4.4. Adequacy of model fit was determined using multiple goodness of fit indices. If a model provides adequate fit, a small 2 value and nonsignificant p value are expected. The 2 statistic for the model predicting persistence intentions was significant, suggesting poor fit. Quintana & Maxwell (1999) described the 2 statistic as being overly stringent in its evaluation of exact fit, therefore other indices were examined. In addition to 2, the Comparative Fit Index (CFI) and the Root Mean Square Error of Approximation (RMSEA) were interpreted, both of which also implied that the data fit the model poorly (see Table 4.17). The path coefficient or beta weight is the degree of change in the outcome variable for every single unit of change in the predictor variable. If the beta coefficient is 97 positive, the interpretation is that for every singular unit increase in the predictor variable, the outcome variable will increase by the beta coefficient value. If the beta coefficient is negative, the interpretation is that for every single unit increase in the predictor variable, the outcome variable will decrease by the beta coefficient value. For this sample of college student servicemembers and veterans, expression suppression was an emotional regulation strategy which negatively predicted college self-efficacy (β = -.443), while cognitive appraisal positively predicted college self-efficacy (β = .581). In other words, college student servicemembers and veterans endorsing cognitive appraisal were likely to have reported higher scores of self-efficacy as measured by the College Self-Efficacy Inventory, while those who endorsed expression suppression are more likely to have reported lower scores of self-efficacy. Notwithstanding these significant path coefficients, and contrary to expectations, emotion regulation strategies possibly mediated by college self-efficacy did not predict academic outcome. Even though the direct effect from both emotion regulation strategies to college self-efficacy were both significant as predicted, results indicated college self-efficacy may not mediate emotion regulation and significantly predict academic outcomes (β =.108; see Table 4.18). Further, given the statistical adequacy of the measurement model as previously described, a subsequent analysis of the structural fit of the latent model was conducted. One statistical model was tested with consideration given to regulation strategies and college self-efficacy as latent variables. Results suggest that this model provides inferior fit to the sample data when compared to CSEI structural model as described in the path analysis. While the 2 statistic was significant at an alpha level of .01, the RMSEA and CFI values were also somewhat unsupportive of the latent model (RMSEA = .10; CFI = 98 .70), when compared to the values of structural fit indices that would be considered more indicative of a good fitting model (RMSEA < .08; CFI  .90). A comparison of the fit indices between the latent model as described above to the previously described path analysis of the CSEI model revealed that the latent model clearly described the sample data with more accuracy. Further, the latent model also demonstrates less statistical support for the latent variables being used as predictors of academic outcome. Mediational Model Analysis Mediation models were generated to further explore college and emotional regulation self-efficacies as two mediating variables with cognitive reappraisal and expression suppression predicting cumulative GPA and PTSD severity. First, the relationship between expression suppression and cumulative GPA was found to be nonsignificant, which suggested that a mediational analysis between these two variables was not necessary. However, the relationship between cognitive reappraisal and cumulative GPA was significant and justified a mediational analysis with each mediating variable being included in two separate mediation models. The first model consisted of cognitive reappraisal and cumulative GPA being meditated by college self-efficacy and the second model consisted of the same predictor and criterion variables with regulatory emotional self-efficacy as the mediating variable. As to infer mediation, Baron and Kenny (1986) suggest that significant relationships are required between the predictor and the criterion variable as well as the predictor and the hypothesized mediator. Indeed, the direct path from cognitive reappraisal significantly predicted the criterion variable, cumulative GPA (β = 2.10, p < .05; see Figure 4.5) as well as both hypothesized mediator 99 variables, college (β = 28, p < .05; see Figure 4.6). and regulatory emotional selfefficacies (β = .51, p < .05; see Figure 4.7). However, neither mediating variable significantly predicated cumulative GPA as a criterion (see Figures 4.6 and 4.7). Therefore, the mediational model analysis presents evidence that neither college selfefficacy nor regulatory self-efficacy were significant mediating variables between cognitive reappraisal and cumulative GPA among this current sample of college student servicemembers and veterans. Secondly, the same two mediating variables, college self-efficacy and regulatory emotional self-efficacy, were used to mediate expression suppression and cognitive reappraisal with PTSD severity. These two mediating variables demonstrated significant beta weights between cognitive reappraisal and PTSD symptom severity, however, there were nonsignificant correlations between expression suppression and PTSD symptom severity. These correlational results suggested a mediational analysis could be conducted with both self-efficacies being introduced as mediating variables between cognitive reappraisal and PTSD symptom severity. Thus, consideration was first given to the direct path between cognitive reappraisal and PTSD symptom severity for the purpose of establishing the plausibility for mediation. The direct path between cognitive reappraisal and PTSD symptom severity was indeed found to be significant (r = -1.03; see Figure 4.8) therefore demonstrating a correlation between cognitive reappraisal and PTSD symptom severity (see Table 4.19). Furthermore, per the required criteria for mediation (Baron and Kenney, 1986) there are significant correlations between cognitive reappraisal and both college and regulatory emotional self-efficacies. The mediation models (Figures 4.9 and 4.10) both illustrate that the standardized regression coefficients 100 between PTSD symptom severity and both college (β = -.21, p < .05) and regulatory emotional (β = -.51, p < .05) self-efficacies are statistically significant, as were the standardized regression coefficients between cognitive reappraisal and PTSD severity described below. The mediational model analysis indicates two partial mediations between cognitive reappraisal and PTSD severity. First, cognitive reappraisal is partially mediated by college self-efficacy when predicting PTSD symptom severity. There is a small reduction in the direct path coefficient between cognitive reappraisal and PTSD symptom severity (β = -.93, p < .05) when college self-efficacy is introduced as a mediating variable (see Figure 4.9). This reduction in the direct path between cognitive reappraisal and PTSD symptom severity indicated a partial mediation with a small indirect effect size (d = -.10; see Table 4.20). Secondly, cognitive reappraisal is also partially mediated by regulatory emotional self-efficacy when predicting PTSD symptom severity. However, there is a larger reduction in the direct path coefficient between cognitive reappraisal and PTSD symptom severity (β = -.69, p < .05) when regulatory emotional self-efficacy is introduced as a mediating variable (see Figure 4.10). This reduction in the direct path between cognitive reappraisal and PTSD symptom severity indicated a partial mediation with an indirect effect size (d = -.34; see Table 4.20). The partial mediation between cognitive appraisal and PTSD symptom severity further defines the strength of the relationship between cognitive appraisal and regulatory emotional self-efficacy and college self-efficacy for this sample of college student servicemembers and veterans. Moreover, the partial mediation provides insight related to cognitive appraisal functioning as a significant predictor for PTSD symptom severity. 101 Table 4.1 Frequencies and Percentages for Sample Demographic Variables Demographic Variable n Gender 1. Male 315 2. Female 97 Race/Ethnicity 1. Caucasian/White 364 2. Hispanic/Latino/Spanish 186 3. African American/Black 32 4. American Indian or Native American 10 5. Asian 2 6. Native Hawaiian/Pacific Islander 1 7. Other 6 Age 1. 20 - 29 211 2. 30 - 39 158 3. 40 - 49 40 4. 50 - 59 1 5. 60 - 69 2 Income per household 1. Under $15,000 10 2. $15,000 - $24,999 24 3. $25,000 - $34,999 16 4. $35,000 - $49,999 30 5. $50,000 - $74,999 63 6. $75,000 - $99,999 112 7. $100,000 - $149,999 139 8. $150,000 - $199,999 14 9. $200,000 and Over 4 Military Branch 1. Army 212 2. Air Force 89 3. Navy 46 4. Marine Corps 31 5. National Guard 15 6. Coast Guard 10 7. Reserve 9 % 76.5 23.5 88.3 45.1 7.8 2.4 0.5 0.2 1.5 51.2 38.3 9.7 0.2 0.5 2.4 5.8 3.9 7.3 15.3 27.2 33.7 3.4 1.0 51.5 21.6 11.2 7.5 3.6 2.4 2.2 102 Table 4.1 continued Demographic Variable Active Duty 1. No active duty 2. One year 3. Two years 4. Three years 5. Four years 6. Five years 7. Six years 8. Seven years or more Total Deployments 1. None 2. 1-5 3. 6-10 4. 11-20 5. 21-30 6. 31-40 VA Administered Financial Aid 1. Yes 2. No Service Connected Disability 1. Yes 2. No Number of Completed Credits 1. 1-36 2. 37-72 3. 73-109 4. 110-146 5. 147-155 6. 156-164 7. 165-173 8. 174-182 9. 183 or more Self-reported cumulative GPA 1. 3.70 - 4.00 2. 2.70 - 3.69 3. 1.70 - 2.69 4. 1.69 and below n % 15 16 48 70 86 35 74 67 3.6 3.9 11.7 17.0 20.9 8.5 18.0 16.0 30 158 105 48 47 19 7.3 38.3 25.6 11.6 11.2 4.6 128 284 31.1 68.9 169 243 41.0 59.0 64 127 68 21 2 4 1 1 5 15.5 30.8 16.5 5.1 0.4 0.9 0.2 0.2 1.2 112 176 7 3 27.2 42.7 1.7 0.7 103 Table 4.1 continued Demographic Variable Nonpassing grades 1. 1 - 2 2. 3 - 4 3. 5 - 6 4. 7 or more Incomplete classes 1. 1-2 2. 3-4 3. 5-6 4. 7-8 Withdrawn classes 1. 1-2 2. 3-4 3. 5-6 4. 7-8 5. 8 or more Academic probation 1. Yes 2. No Semesters on academic suspension 1. 1-2 2. 3-4 3. 5-6 n % 26 12 13 11 6.3 2.9 3.2 3.1 21 15 11 6 6.0 3.7 2.6 1.4 31 21 6 8 2 7.5 5.1 1.5 1.9 .5 52 346 12.6 84.0 11 14 10 8.0 3.4 2.5 Table 4.2 Means and Standard Deviations for Measurement Sum Scores Measurement Instruments N 1. Emotion Regulation Questionnaire (ERQ) 1a. Cognitive Reappraisal (CR) 373 1b. Expression Suppression (ES) 373 2. Regulatory Emotional Self-Efficacy Scale (RESES) 376 3. College Self-Efficacy Inventory (CSEI) 378 4. PTSD Checklist-5 (PCL) 387 5. Cumulative GPA 295 M SD 26.27 16.61 35.75 50.25 34.96 3.48 6.33 4.36 8.24 10.72 15.81 0.429 104 Table 4.3 Correlations Between Measurement Sum Scores Variables 1 2 1. Cognitive Reappraisal 1.00 2. Expression Suppression .327* 1.00 ** 3. College Self-Efficacy .165 -.148** 4. Emotional Self-Efficacy .510** .069 5. PTSD Severity -.348** .007 6. Cumulative GPA Note. * p < .05 .149** .015 3 96928.977 SS .841 54.418 55.259 5 1.00 .550** -.207** 1.00 -.389** 1.00 .102* .043 -.042 Table 4.4 One-Way Analysis of Variance by Hispanic/Latino Ethnicity PTSD: Source df SS MS Between Groups 1 308.196 308.196 Within Groups 386 96620.623 250.313 Total 387 GPA: Source df Between Groups 1 Within Groups 294 Total 295 Note. * p < .05; ** p < .01 4 MS .841 .185 6 1.00 F 1.231 p .268 F 4.542 p .034* Table 4.5 PTSD Severity Means and Standard Deviations by Hispanic/Latino Ethnicity Ethnicity n M SD Non-Hispanic/Latino Origin 209 35.83 18.60 Hispanic/Latino Origin 179 34.04 11.76 Total 388 3.48 15.82 105 Table 4.6 Cumulative GPA Means and Standard Deviations by Hispanic/Latino Ethnicity Ethnicity n M SD Non-Hispanic/Latino Origin 116 3.42 .55 Hispanic/Latino Origin 180 3.53 .34 Total 296 3.48 .43 Table 4.7 One-Way Analysis of Variance by Gender PTSD: Source df SS Between Groups 1 2202.026 Within Groups Total GPA: Source Between Groups Within Groups 385 386 df 1 294 94326.227 96528.253 SS .093 55.167 Total Note. * p < .05 295 55.259 MS 2201.026 F 8.988 p .003* F .493 p .483 245.003 MS .093 .188 106 Table 4.8 PTSD Severity Means and Standard Deviations n Military Branch 1. Army 204 2. Navy 42 3. Air Force 86 4. Marine Corps 25 5. Coast Guard 9 6. National Guard 13 7. Reserve 9 8. Total 388 Race 1. White/Caucasian 342 2. Black African American 31 3. American Indian 8 4. Asian 2 5. Other 5 6. Total 388 Gender 1. Female 94 2. Male 293 3. Total 387 M SD 34.23 34.67 39.24 32.28 42.11 34.15 15.44 35.01 14.79 13.99 15.76 18.23 14.90 20.04 19.55 15.81 35.34 35.00 26.13 59.00 17.00 35.00 15.90 10.98 17.05 5.66 19.77 15.83 30.24 36.31 34.96 15.79 15.61 15.81 107 Table 4.9 GPA Means and Standard Deviations n Military Branch 1. Army 2. Navy 3. Air Force 4. Marine Corps 5. Coast Guard 6. National Guard 7. Reserve 8. Total Race 1. White/Caucasian 2. Black African American 3. American Indian/Alaskan Native 4. Asian 5. Other 6. Total Gender 1. Female 2. Male 3. Total M SD 193 36 25 24 3 7 8 296 3.52 3.25 3.59 3.37 3.60 3.47 3.64 3.48 .41 .49 .34 .50 .53 .59 .35 .43 250 28 10 2 6 296 3.49 3.58 3.31 2.62 3.18 3.48 .42 .36 .52 1.80 .29 .43 91 205 296 3.51 3.47 3.48 .44 .42 .43 Table 4.10 One-Way Analysis of Variance by Military Branch PTSD: Source df SS MS Between Groups 6 5765.756 960.959 Within Groups 381 91163.220 239.274 Total 387 96928.997 GPA: Source df SS MS Between Groups 6 2.937 .489 Within Groups Total Note. * p < .05 289 295 52.322 55.259 .181 F 4.016 p .001* F 2.703 p .014* 108 Table 4.11 Posthoc Analysis of PTSD Severity by Military Branch and Race Military Branch 1 2 3 4 5 1. Army 1.00 .22 1.00 .95 2. Navy .92 1.00 .99 3. Air Force .64 1.00 4. Marine Corps - 5. Coast Guard 6 1.00 7 .01* 1.00 .02* 1.00 .00* .89 1.00 .11 - 1.00 .01* - .11 6. National Guard 7. Reserve Race 1. White/Caucasian 1 2 3 4 5 - 1.00 .65 .29 .09 - .81 .31 .16 - .08 .97 - .01* 2. Black African American 3. American Indian 4. Asian 5. Other Note. * p < .05 (Significant difference in PTSD severity.) Table 4.12 One-Way Analysis of Variance by Race PTSD: Source df SS Between Groups 4 3441.447 Within Groups 383 93487.530 Total GPA: Source Between Groups Within Groups 387 df 4 291 96928.977 SS 2.603 52.656 Total Note. * p < .05 295 55.259 - MS 860.362 244.093 F 3.525 p .008* MS .651 .181 F 3.596 p .007* 109 Table 4.13 Goodness of Fit Comparison: Path Models Predicting Academic Outcomes Model df CFI RMSEA 2 * Original Model 24.34 6 .90 .13 RESES Model 11.20* 2 .92 .16 * CSEI Model 07.33 2 .92 .12 Note. *p = .05 Table 4.14 Analysis of Original Recursive Path Model With Multiple Regression Correlations, means, and standard deviations Variable 1 2 3 4 5 1. GPA 2. PCL -.04 3. CSEI .10 -.21** 4. RESES .04 -.39** .55** * ** ** 5. CR .15 -.35 .17 .51** 6. ES M SD .02 3.48 0.43 .01 34.96 15.81 -.15** 50.25 10.72 .07 35.75 8.24 .33** 26.27 6.33 Regressions to generate path coefficients and disturbance variances Regression Coefficients Criterion Predictors B R2  1. GPA PCL .004 .134 .04 * * CSEI .007 .180 2. PCL RESES -.789** -.398** .16 3. CSEI 4. RESES RESES CR ES .779** .048 -.290 .571** 16.61 4.36 (1-R2) .96 .84 .35 .65 .29 .71 .028 -.118 CR .736** .578** ES -.316** -.175** Note. * p < .05; ** p < .01 6 110 Table 4.15 Analysis of RESES Recursive Path Model With Multiple Regression Correlations, means, and standard deviations Variable 1 2 3 4 1. GPA 2. PCL -.04 3. RESES .04 -.39** * ** 4. CR .15 -.35 .51** 5. ES M SD .02 3.48 0.43 .01 34.96 15.81 .07 35.75 8.24 .33** 26.27 6.33 Regressions to generate path coefficients and disturbance variances Regression Coefficients Criterion Predictors B R2  1. GPA PCL .002 .071 .01 RESES .004 .075 2. PCL 3. RESES CR ES RESES CR Note. * p < .05; ** p < .01 -.794** .304 -.475** .736** -.316** -.314** .085 -.240** .578** -.175** 5 16.60 4.36 (1-R2) .99 .22 .78 .29 .71 111 Table 4.16 Analysis of CSEI Recursive Path Model With Multiple Regression Correlations, means, and standard deviations Variable 1 2 3 4 1. GPA 2. PCL -.04 3. CSEI .10 -.21** * ** 4. CR .15 -.35 .17** 5. ES M SD .02 3.48 0.43 .01 34.96 15.81 -.15** 50.25 10.72 .33** 26.27 6.33 Regressions to generate path coefficients and disturbance variances Regression Coefficients Criterion Predictors B R2  1. GPA PCL .004 .134 .04 CSEI .007* .180* 2. PCL CR -1.031** -.408** .19 3. CSEI ES CSEI CR ES .358 -.180 .622** -.536** .100 -.124 .358** -.218** .12 5 16.61 4.36 (1-R2) .96 .81 .88 Note. * p < .05; ** p < .01 Table 4.17 Goodness of Fit: Path Models Predicting Academic Outcomes Model df CFI 2 * Original Model 24.34 6 .90 * Latent Model 903.26 321 .70 Note. * p < .05 RMSEA .13 .10 112 Table 4.18 Regressions to Generate Path Coefficients and Disturbance Variances Regression Coefficients Criterion Predictors B R2  ** ** 1. CSEI CR .457 .581 .245 ES -.433** -.443** 2. GPA CSEI .057 .108 .012 (1-R2) .755 .988 Note. * p < .05; ** p < .01 Table 4.19 Regression to Establish Causal and Outcome Variable Correlation Causal Variable Outcome Variable Direct Effect (r) Cognitive Reappraisal PTSD Symptom Severity -1.03 Table 4.20 Mediational Model Analysis Results Mediational Model College Self-Efficacy Inventory (CSEI) Regulatory Emotional Self-Efficacy Scale (RESES) Figure 4.1. Original Path Analysis Model. Note. * p < .05; ** p < .01 Direct Effect (r) -.93 -.69 Indirect Effect (d) -.10 -.34 113 Figure 4.2 Regulation Emotional Self-Efficacy Scale (RESES) Path Analysis Model. Note. * p < .05; ** p < .01 Figure 4.3 College Self-Efficacy Inventory (CSEI) Path Analysis Model. Note. * p < .05; ** p < .01 Figure 4.4 Latent Model Predicting Academic Outcome. Predictor variables include emotion regulation strategies and college self-efficacy. Note. * p < .05; ** p < .01 114 Figure 4.5 Direct Effect Between Cumulative GPA and Cognitive Reappraisal. Note. * p < .05 Figure 4.6 College Self-Efficacy Mediation Model Predicting GPA. Note. * p < .05 Figure 4.7 Regulatory Emotional Self-Efficacy Mediation Model Predicting GPA. Note. * p < .05 115 Figure 4.8 Direct Effect Between PTSD Severity and Cognitive Reappraisal. Note. * p < .05 Figure 4.9 College Self-Efficacy Mediation Model Predicting PTSD Severity. Note. * p < .05 Figure 4.10 Regulatory Emotional Self-Efficacy Mediation Model Predicting PTSD Severity. Note. * p < .05 CHAPTER 5 DISCUSSION The purpose of the present study was to contribute to the limited research on factors impacting the academic success of college student servicemembers and veterans. This study builds upon previous research and hypothesized a linear path models that explicate the relationships between observed variables, which include two common emotion regulation strategies (cognitive appraisal and expression suppression), regulatory emotional self-efficacy, college self-efficacy, PTSD symptom severity, and college GPA in a sample of college student servicemembers and veterans. Research Questions • Question 1: To what degree do the data support the proposed models of correlates in predicting college student servicemember and veteran academic success as measured by self-reported college GPA. • Question 2: What is the strength of the relationship between PTSD symptom severity and college student GPA for a sample of college student servicemembers and veterans? 117 • Question 3: What is the strength of the relationship between regulatory emotional self-efficacy and PTSD symptom severity and regulatory emotional self-efficacy and college self-efficacy for a sample of college student servicemembers and veterans? • Question 4: What is the strength of the relationship between college self-efficacy and college student GPA for a sample of college student servicemembers and veterans? • Question 5: What is the strength of the relationship between emotion regulation strategies and regulatory emotional self-efficacy and college self-efficacy for a sample of college student servicemembers and veterans? Hypotheses It was hypothesized that regulatory emotional self-efficacy along with college self-efficacy and PTSD symptom severity would influence the relationship between emotion regulation and college GPA in a sample of college student servicemembers and veterans in the following ways. First, both self-efficacies and PTSD symptom severity will mediate the relationship between emotion regulation and college GPA in a sample of college student servicemembers and veterans. Second, regulatory emotional self-efficacy will have a direct effect on self-perceived college self-efficacy and PTSD symptom severity. Third, PTSD symptom severity and college self-efficacy will have a direct effect on college student GPA. Thus, regulatory emotional self-efficacy, college self-efficacy, and PTSD symptom severity serve as mediating variables between emotion regulation and college GPA for college student servicemembers and veterans. The degree to which the data supports the proposed structural equation models 118 and answers the hypotheses associated with the current study will now be discussed below. The current study includes comparisons between three linear path models and a single latent model. The first structural path model is considered to be the original model in which all manifest variables are included to predict academic outcomes. The two alternate structural models include only a singular form of self-efficacy at the exclusion of the other; the first model (CSEI model) includes college self-efficacy only and the second model (RESES model) only includes regulatory self-efficacy. Finally, the single latent model consists of important observed variables predicting academic outcomes among college student servicemembers and veterans. Model Comparisons The comparisons made between the four models were made based upon their structural equation model fit indices and path coefficients. The study demonstrated the overall value of emotion regulation and self-efficacy in predicting PTSD symptom severity and academic outcomes among a nationwide sample of college student servicemembers and veterans. An overview of the interpretation of the results from each statistical analysis for each model will be discussed below. This discussion will focus on the results obtained in the present study, their relationship to previous research, possible limitations of the present study, implications for future research, and potential practices derived from the present findings. 119 Original Structural Model The original model included and considered five observed variables based on composite scores for each participant with self-reported cumulative grade point average as the outcome variable being predicted by all others. Altogether, the specific fit indices for the original model partially supported an overall good fitting model and did not demonstrate adequate fit. However, results support a number of significant and meaningful paths. Based on the results of the path analysis conducted on the original model, both emotion regulation strategies were found to significantly predict regulatory emotional self-efficacy. The path analysis reveals that cognitive reappraisal positively predicted regulatory emotion self-efficacy, while expression suppression negatively predicted the same observed variable. This suggests that the confidence in the ability to regulate emotions for participants in this study is plausibly moderated and influenced by the ability to change the trajectory of an emotional response by cognitively reinterpreting the meaning of an internal emotional stimulus. Alternatively, expression suppression is a contrasting emotion regulation strategy in the original model in which the confidence in one’s ability to regulate emotions is being possibly moderated and influenced by concealing behavioral manifestations of internally experienced emotions. The path coefficients from the original model (Figure 4.1) suggest that college student servicemembers and veterans who conceal internal emotional experiences by suppressing behavioral manifestations linked to an emotion such as disguising facial expressions, are more likely to perceive a diminished confidence in their ability to regulate their emotions ( = -0.18, p < 0.05). Furthermore, regulatory emotional self-efficacy significantly and positively 120 predicted college self-efficacy. This statistical result is plausibly interpreted as a strong relationship between regulatory emotional self-efficacy and college self-efficacy in which both taken together are indicative of a more generalized self-efficacy. This global view of self-efficacy would suggest that personal agency and a capacity to generate an optimistic appraisal of self is influenced in part by an ability to recognize emotion and appraise the meaning of internal affective experiences. This internal cognitive affective process can be considered a reinterpretation of an internal emotional stimulus which appears to be a significant factor with consideration to generalized self-efficacy, but could also be comparatively and specifically applied to both regulatory emotional self-efficacy and college self-efficacy. The statistical results of the original model further suggest that confidence in regulating emotions plausibly moderates PTSD severity, while academic confidence related to college self-efficacy possibly moderates academic outcomes. This indicates that a pattern of practicing and utilizing emotion regulation strategies will predict the strength in the confidence underlying general self-efficacy and more specifically, that unique and specific forms of self-efficacy influence the symptomatic outcome associated with PTSD severity and academic performance associated with college grade point average. The fact that there is a significant relationship between both unique types of self-efficacy (college self-efficacy and regulatory self-efficacy) and their corresponding outcomes plausibly suggests that emotional regulation strategies are a component to determining the psychological, behavioral, and academic outcomes among college student servicemembers and veterans. Furthermore, the statistical results of the original model also suggest that regulatory self-efficacy and college self-efficacy are affective and 121 noncognitive processes which indeed coordinate together in developing and strengthening general self-efficacy. In the original model path analysis, both emotion regulation strategies significantly predict the self-efficacy to regulate emotions and alternatively do not predict academic confidence (Figure 4.1). The results from the original model further suggest there is a predictive relationship between regulatory selfefficacy and college self-efficacy in which academic confidence was positively predicted by the ability to effectively regulate emotions. This outcome is paired with the fact that college self-efficacy significantly predicted academic outcome, which leads to a plausible conclusion that the confidence or self-efficacy to regulate emotions as well as the confidence to academically perform and successfully navigate academic settings are both mediating academic outcomes in this sample of college student servicemembers and veterans. The current results partially support Cieslak et al. (2008) in which their study suggests that coping self-efficacy is a significant mediator between negative cognitions and posttraumatic distress. Taken altogether, the results demonstrate that affective regulation strategies are possibly conducive to constructing, developing, and strengthening a general self-efficacy as well as specific forms of self-efficacy and to more importantly serve to significantly predict academic outcomes. Additionally, the overall statistical results further demonstrate that emotion regulation and regulatory emotional self-efficacy will possibly moderate college selfefficacy, which ultimately predicts academic performance or academic outcomes based upon the current sample of college student servicemembers and veterans. The statistical outcomes of the original model path analysis indicate that some aspects of the structural model are indeed supported by the data. More specifically that academic outcomes 122 among this sample of college student servicemembers and veterans are influenced by emotion regulation strategies as well as regulatory emotional self-efficacy and colleges self-efficacy. The significant path coefficients in the original model suggest the relationship between emotion regulation strategies (cognitive reappraisal and expression suppression) and academic outcomes is mediated by both regulatory emotional selfefficacy and college self-efficacy. When taken into consideration altogether, a conclusion can be drawn in which both regulatory self-efficacy and college self-efficacy are significantly influenced by emotion regulation strategies and plausibly influence academic outcomes among college student servicemembers and veterans. Similar to the original model, the statistical data did not produce sufficient fit indices for the two alternate structural models (CSEI model and RESES model) in which only a singular form of self-efficacy was included in each model to predict college grade point average among college student servicemembers and veterans. The structural fit indices for each model are approaching good fit, however, neither model was established as a sufficiently specified and good fitting model. However, the results do appear to support a number of significant and meaningful path coefficients in both models. College Self-Efficacy Model The CSEI model produced an overall improved set of statistical fit indices in comparison to the RESES model. Both the RESES model and the CSEI model, suggest that cognitive reappraisal is a significant factor in predicting PTSD severity. In other words, cognitive reappraisal significantly predicted PTSD severity in a negative direction in both models in which higher scores of cognitive appraisal from the ERQ predicted 123 lower scores on the PCL-5. These results appear to support Tull et al. (2007) who explored the relationship between posttraumatic stress (PTS) symptom severity and difficulties with effective emotion regulation strategies among college students. Elevated PTS severity was found to be associated with lower levels of emotional acceptance, difficulty engaging in goal directed behavior when distressed, challenges associated with impulse control, restricted degree of emotional clarity, and limited access to effective emotion regulation strategies. Further, these results confirm the findings of O’Bryan et al. (2015) in a study examining the relationship between PTS symptoms and aspects of emotion regulation in a sample of undergraduate students. The results of the study indicated that greater difficulties with aspects of emotion regulation, such as emotional acceptance and accessing emotion regulation strategies when distressed, indeed predicted the cluster of PTS symptoms associated with negative alterations and mood while greater difficulties with emotional acceptance alone predicted avoidance and hyperarousal symptom severity above and beyond the number of trauma types and negative affect. This suggests that both models support cognitive reappraisal as an emotion regulation strategy which plausibly insulates college student servicemembers and veterans from elevated levels of PTSD symptom severity as measured by the PCL-5. Further, expression suppression significantly predicted both forms of self-efficacy in a negative direction suggesting that voluntary suppression of outward emotional expressions reduces college student servicemembers’ and veterans’ confidence in their abilities associated with both regulating emotions and successfully performing in academic settings. This evidence suggests that the ability to change the trajectory of an emotional response by reinterpreting the meaning of the emotional stimulus will perhaps 124 mitigate or diminish the constellation of clinical symptoms associated with PTSD symptomatology. This conclusion is further supported and reinforced by the regulatory emotional self-efficacy model in which regulatory emotion self-efficacy significantly predicted PTSD in a negative direction. This suggests the possibility that cognitive appraisal, along with perceived regulatory emotional self-efficacy among college servicemembers or veterans, will plausibly moderate levels of PTSD severity. In other words, this statistical outcome suggests that the two factors together reduce the severity of clustered posttraumatic symptoms associated with hyperarousal, negative alteration, avoidance, and intrusion. Regulatory Emotional Self-Efficacy Model The results associated with the college self-efficacy model diverge from those results related to the regulatory emotional self-efficacy model in which college grade point average is significantly predicted only by college self-efficacy. The path coefficient associated with the regulatory emotional self-efficacy model predicting academic outcomes does not support regulatory emotional self-efficacy as an observed variable being intentionally placed within the structural model to predict academic outcomes among college servicemembers and veterans. This particular difference between the two models suggests that while both forms of self-efficacy predict PTSD severity, it is only college self-efficacy rather than regulatory emotional self-efficacy that will perhaps mediate emotion regulation strategies and academic outcome based on its significant relationship to both emotion regulations strategies and academic outcomes. 125 Alternate Latent Model However, additional evidence based on the results of an alternate latent model analysis depict college self-efficacy as being a nonsignificant predictor of academic outcome among college student servicemembers and veterans. The latent model tested the relationships between the manifest or observed variables predicting academic outcomes. The fit indices associated with the latent model diagram suggest the data did not fit the model and do not fully support the proposed model of correlates in predicting academic performance among this sample of student servicemembers and veterans. Nevertheless, despite the apparent overall model misspecification, inspection of individual paths revealed interesting findings and support the significant relationships between both emotion regulations strategies and college self-efficacy. Overall, the results suggest that emotion regulation strategies significantly predict college self-efficacy without mediating the relationship between emotion regulation strategies and the selfreported college grade point average among college student servicemembers and veterans. In other words, the path coefficients between emotion regulation strategies and college self-efficacy indicate that cognitive reappraisal and expression suppression are significant predictors of college self-efficacy. However, even though the emotion regulation strategies as exogenous variables significantly predict college self-efficacy, it does not appear that the college self-efficacy mediates emotion regulations strategies and academic outcomes. Consequentially, these observed path coefficients partially support Strain and D’Mello’s (2015) findings in which an ANOVA yielded a significant main effect of condition on proportional learning gains, F (3, 102) = 2.73, MSE = 0.180, partial η2 = .074 with cognitive appraisal condition (M =. 0.30, SD = 0.17) achieving 126 significantly higher scores in comparison to the open-ended appraisal (M =. 0.14, SD = 0.22), expression suppression (M = 0.15, SD = 0.29), and control conditions (M = 0.12, SD = 0.24). These particular findings when combined together seem to imply that cognitive appraisal and expression suppression both play a significant role in the overall development of college self-efficacy among this sample of college student servicemembers and veterans. Significant Individual Relationships The results of the research study also highlight significant relationships between individual and specific observed variables contained in the three structural models, the latent model, and the mediational models included in this study. Regulatory Emotional Self-Efficacy The first series of relationships worth identifying and discussing are those between regulatory emotional self-efficacy and other observed variables across the three structural models. There are two significant relationships that regulatory emotional selfefficacy has with college self-efficacy and PTSD severity. The first significant relationship between regulatory emotional self-efficacy and its direct effect on selfperceived college self-efficacy is relevant to understanding the academic challenges faced by college student servicemembers and veterans. More specifically, results indicate that the relationship between the two observed variables positively predicts college selfefficacy. This relationship becomes important given that a small percentage of college student servicemembers and veterans across various service eras reported earning a 127 college degree (Blackburn & Owens, 2015). The relationship between the two specific forms of self-efficacy allude to a general self-efficacy in which regulating emotions is a factor associated with academic confidence and performing in academic settings. This is a crucial component for college student servicemembers and veterans in being able to successfully navigate academic and educational environments. This is the first study to consider college student servicemembers and veterans ability to regulate emotions as a factor contributing to their self-appraisal associated with succeeding and overcoming academic challenges on a college campus. More importantly is the second relationship regulatory emotional self-efficacy has with observed variables in the three models, and that is with PTSD severity. Regulatory emotional self-efficacy has a direct and negative relationship with the severity of PTSD symptoms. The original model as well as the RESES model both include regulatory emotional self-efficacy as a significant predictor of PTSD severity in which higher levels of regulatory self-efficacy equate to lower levels of PTSD severity (Figures 4.5 and 4.6). This significant relationship is further defined in the mediation models in which cognitive reappraisal is partially mediated by regulatory emotional self-efficacy. These particular findings in the current study seem to strengthen previous research in which O’Bryan et al. (2015) described college student servicemembers and veterans experiencing greater difficulties with aspects of emotion regulation, such as emotional acceptance and accessing emotion regulation strategies when distressed, which were positively correlated with PTSD symptoms and indeed predicted the cluster of PTSD symptoms associated with negative alterations and mood. Further, current results are congruent with findings from a recent study in which PTSD symptoms were demonstrated to have a clear 128 relationship with emotion regulation difficulties. The study confirmed significant relationships between PTSD specific sleep disturbance, poor sleep quality, and emotion regulation difficulties in a sample of college student (Pickett et al. 2016). However, PTSD symptom severity did not have a direct effect on academic outcomes in the current study. These results are consistent with previous research including populations of college student servicemembers and veterans. Bryan et al. (2014) conducted previous research which examined the relationship between selfreported PTSD symptom severity and grade point average in a sample of college student servicemembers and veterans. The outcome of the research found that PSTD severity was not a significant predictor of college grade point average ( = -0.040, p = 0.62). Consequently, the statistical results from the current study in which the severity of PTSD symptoms did not significantly predict academic outcome is congruent with previously established statistical evidence indicating that college grade point average is not significantly determined nor influenced by the severity of experienced PTSD symptoms among college student servicemembers and veterans. The statistical results described above contribute to an extremely small literature base associated with the relationship between PTSD severity and/or symptomatology and postsecondary academic performance. Accordingly, it is possible that college student servicemembers and veterans generally possess a unique set of characteristics or are implicitly exposed to other protective factors associated with military service that plausibly promote or interfere with academic performance and outcomes. Chemers et al. (2001) reported general self-efficacy as having mediating effects upon cognitive, motivational and affective processes related to academic performance. 129 According to their conceptualization, affective processes influence attention and construal of environmental demands, the choice of actions directing behavior, and the capability to regulate negative emotions. Research conducted by Chemers et al. (2001) yielded factor loadings in which affective processes had a significant and direct effect of academic self-efficacy on academic performance among first year college students. College Self-Efficacy The current results are somewhat consistent with the previously described findings in which college self-efficacy and the confidence to perform in academic settings significantly predicted academic outcomes under discrete methodological circumstances. The method to which variables were defined in the structural models versus the latent structural model is plausibly contributing to mixed results when predicting academic outcomes with college self-efficacy. College self-efficacy significantly predicted college student grade point average in the structural models in which composite scores were used to determine observed variables. The current study included two structural models in which college self-efficacy significantly predicted academic outcome; the original model ( = .18, p < .05) as well as the CSEI model ( = .18, p < .05). However, the same direct path was analyzed under alternate circumstances in which college self-efficacy was recognized as a latent variable and allowed to be determined by the measurement model in which CSEI items were used individually to define college self-efficacy. Under this methodological circumstance college self-efficacy did not significantly predict ( = .11, p > .05) the cumulative grade point average among a nationwide sample of student servicemembers and veterans. The 130 current results are mixed and also plausibly misleading due to the fact that the latent structural model is empirically under identified. It is worth pointing out that self-reported cumulative grade point average reported by college student servicemembers and veterans in this study is a singular indicator defining academic outcome. Consequently, academic outcome is a variable defined only by a self-reported grade point average rather than the outcome variable being defined by multiple indicators. The methodological practice for defining measurement models typically requires a minimum of two indicators per factor (Kline, 1998). The fact that academic outcome was defined by self-reported grade point average among college student servicemembers and veterans in this sample is perhaps distorting the statistical results in the study. There is potential improvement that can be made in future research in which academic outcome or academic performance as an outcome variable could be more effectively defined with additional indicators such as expected grade performance on exams or quizzes, class attendance, submitting assignments on time, adequate class preparation, effort toward academic endeavors, and ability to concentrate in class settings. Emotion Regulation However, this is the first study to consider the impact that emotion regulation would have on the confidence to perform in academic and educational settings using a structural equation model methodology for the purpose of predicting academic outcomes among college student servicemembers and veterans. The direct relationship between emotion regulation strategies and college self-efficacy are for the most part consistent across all models included in this study. Additionally, there are subtle differences across 131 all models and several significant findings worth discussing in consideration to the direct relationship that emotion regulation strategies have with two contrasting and comparable forms of self-efficacy. The current results of this study integrate with recent research in which regulatory emotional self-efficacy was examined as a plausible predictor for future life satisfaction. The research discovered several factors associated with affective selfefficacy as possible predictors of life satisfaction among college students (Lightsey et al., 2013). For example, and most relevant to current research, the study yielded significant results when utilizing the Regulatory Emotional Self-Efficacy Scale (RESES) as an instrument to measure affective self-efficacy across three different dimensions reflecting important psychological factors related to emotion regulation predicting overall satisfaction with life that is, managing despondency and distress, managing anger, and experiencing and expressing positive emotions (Lightsey et al., 2013). The study described above exemplifies the fact that the Emotion Regulation Questionnaire (ERQ) and the Regulatory Emotional Self-Efficacy Scale (RESES) are indeed measuring similar constructs related to the processes associated with regulating internal emotions. Thus, it can be expected for the path coefficients in the structural models (Figures 4.5 and 4.6) depicting the relationship between both emotion regulation strategies and regulatory emotional self-efficacy to be significant. Indeed, the current results indicate significant relationships between emotion regulation strategies and regulatory emotional selfefficacy in the expected directions: expression suppression predicts regulatory emotional self-efficacy in a negative direction while ( = -.18, p < .01) cognitive reappraisal is expressed in a significantly positive direction ( = .58, p < .01). This is further demonstrated in the current study when comparing cognitive appraisal across two 132 different forms of self-efficacy based on data driven beta weights in which cognitive appraisal proved to be an improved predictor of regulatory emotional self-efficacy ( = .58, p < .01) in comparison to college self-efficacy ( = .36, p < .01). In view of the fact that the beta weight is greater for regulatory emotional self-efficacy in comparison to college self-efficacy further suggests that there is a significant correlation between an effective emotion regulation strategy and the self-efficacy associated with regulating emotions. In fact, the correlation between regulatory emotional self-efficacy and cognitive appraisal is indeed significant and considered to be a moderate sized correlation (r = .51, p < .01). Additional evidence from the current study also suggests that emotion regulation strategies are related affective self-efficacy in which expression suppression was found to be a comparable predictor for both forms of self-efficacies: college selfefficacy ( = -.21, p < .01) and regulatory emotional self-efficacy ( = -.18, p < .01). College self-efficacy is a variable of particular interest in relationship to emotion regulation strategies due to the fact that college self-efficacy was found to be a significant predictor within the structure of the CSEI model for which college self-efficacy was isolated as a form of self-efficacy and measured as an observed variable (Figure 4.3). The fact both expression suppression ( = -.21, p < .01) and cognitive reappraisal ( = .36, p < .01) significantly predicted college self-efficacy suggests that emotion regulation strategies indeed influence college self-efficacy and significantly predict academic outcomes ( = .18, p < .05) among college student servicemembers and veterans. Similar results from the latent structural model supported the influence emotion regulation has on college self-efficacy. The results from previous structural models were replicated in the latent structural model in which expression suppression and cognitive reappraisal were 133 both emotion regulation strategies significantly predicting college self-efficacy (Figure 4.4). Expression suppression was an observed variable in the latent model that negatively predicted college self-efficacy ( = -.44, p < .05) while cognitive appraisal significantly predicted college self-efficacy in a positive direction ( = .59, p < .05). These results are consistent across all structural models in this study and suggest that emotion regulation strategies such as expression suppression and cognitive appraisal will partially determine the confidence college student servicemembers or veterans will have in themselves to successfully navigate academic and educational settings. However, there is an additional statistical outcome in the latent structural model worth noting in which emotion regulation strategies were not mediated by college self-efficacy and failed to significantly predict academic outcomes among college student servicemembers and veterans. This is plausibly attributed to the fact that academic outcome is a singular indicator defining academic outcome when methodological practice for defining measurement models typically requires a minimum of two indicators per factor (Kline, 1998). Despite this methodological weakness, the significant results from the statistical analysis of the measurement model as well as the latent structural model indicate that college selfefficacy is indeed determined in part by an ability to initiate, inhibit, or modulate an internal emotional state in a given academic situation. Limitations of the Present Study There are a several methodological limitations to the current research that are worth noting and describing for the purpose of improving future research related to the variables being examined in this study. The first limitation is an issue related the structure 134 of the conceptual models under investigation and the observed variables being arranged in a manner in which the causal mechanisms were potentially depicted with a greater degree of inaccuracy than what would be acceptable. The models investigated in this study were informed by empirical literature, however, it is plausible that the observed variables could be interpreted alternatively such that variables could have been structurally placed into the models so as to generate alternate causal mechanisms. For example, PTSD severity would have perhaps been a significant causal mechanism for either college self-efficacy and/or regulatory self-efficacy which then would have perhaps predicted emotion regulation strategies and ultimately been a stronger predictor for academic outcomes. Furthermore, there is plausibility in the fact that academic performance outcomes were grossly considered as an endogenous variable. It is possible that academic performance outcomes would have improved fit indices for models included in this study if it had been predicted by emotion regulation strategies and served as a mediating variable predicting different forms of self-efficacy. Consequently, this structural arrangement of observed variables depicts self-efficacy as an outcome variable of greater interest versus academic outcomes. Although the current models were informed by empirical literature, there is the potential for an alternate arrangement of observed variables which could plausibly improve overall structural fit of a justified model. Secondly, the current study is limited by the means to which the observed variables were measured and interpreted. There are two statistical weaknesses in which the data could perhaps be misinterpreted based on the methods used to create observed variables in the models. First, composite scores were used in the three structural models 135 (Original, CSEI, RESES) to define emotion regulations strategies, PTSD severity, and self-efficacy. A potential problem when using composite scores to define manifest variables is the accumulation of measurement error, which could lend to gross inaccuracy of estimates in the model. The use of composite scores is not necessarily an invalid approach, however, the inferences based on composite scores could possibly misrepresent the relationships in the model and plausibly lead to biased and unreliable estimates. Second, the measurement model is somewhat weak due to the endogenous outcome variable, college cumulative grade point average, being informed by only a single indicator. In other words, the structural models nor the latent model include a sufficient number of indicators for the endogenous outcome variable to be a sufficiently measured and subsequently cannot be considered an accurate representation of academic outcome in the models included in this study. Additional limitations to the current research study include issues related to demographics, size of the study sample, and potential alternate variables. First, the demographic characteristics are a representation of a population of college student servicemembers and veterans who are academically skilled and who are successfully retained college students. In other words, this sample population consists of college student servicemembers and veterans who academically persisted despite inherent challenges associated with earning a college degree. Consequently, the current sample population is perhaps not representative of the larger population of servicemembers and veterans who at one time attended college and were inadvertently not included in this sample due to a pattern of stopping or dropping out. This inherently eliminates any additional variance that could be accounted for in the measurement of observed variables 136 and does not give an accurate depiction of the composite scores or the outcome of the measurement model used in the latent model analysis. Furthermore, there was an insufficient sample size associated with demographic variables in order to accurately determine differences across gender, race and ethnicity as well as across military service branches. Finally, there is the possibility of alternative variables impacting the academic success of college student veterans that were not considered as observed variables in the current study. For example, measuring military identity to include the possibility that academic outcomes are determined by the academic cultural discordance between the military culture of college servicemembers and veterans and the dominant academic culture of college students. This is a potential variable which considers the cultural acceptance of military identities among civilian college students within academic settings. An observed variable such as military identity may be included in a model analysis to support alternate theories of academic performance among college student servicemembers and veterans whose demographics are more likely to be represented by nontraditional college students. This theory would perhaps suggest that older and more independent college student servicemembers and veterans who have received military education and training are more likely to endorse cognitive regulation as a preferred emotion regulation strategy and higher college self-efficacy as compared to younger veterans or younger nontraditional civilian college students. Implications From the Present Study The current findings may have important implications for individuals or groups who work closely with or on the behalf of college student servicemembers and veterans. 137 Given the significant relationships between emotion regulation strategies, self-efficacy beliefs, and posttraumatic symptom severity, it may be important for postsecondary academic institutions to consider methods and interventions aimed at strengthening emotion regulations strategies and cultivating self-efficacy beliefs among college student servicemembers and veterans. Individuals affiliated with academic institutions are in a position to promote emotion regulation strategies and self-efficacy beliefs within college academic settings. These individuals include university and college campus mental health professionals such as psychiatrists, psychologists, social workers, counselors, clinicians, and therapists. However, college academic advisors as well as university classroom instructors, faculty members, and administrators are also in positions to influence environments and provide interventions, activities, and instruction aimed at encouraging and strengthening cognitive appraisal and other evidenced based emotion regulation strategies. For example, colleges and universities could incorporate psychoeducational learning opportunities associated with emotion regulation into campus programs designed to retain and/or integrate college student servicemembers and veterans. In addition to various learning opportunities, college campus student veterans organizations could promote college self-efficacy through onsite practices and organizational programming that incorporate emotion regulation strategies. This may include providing university or college mental health professionals as well as postsecondary academic personnel with expertise in assisting college student servicemembers and veterans. These professional groups affiliated with universities would be trained in seminars or workshops to practice sensitivity toward the experiences and unique needs of college students who are serving or have served in the military armed forces. Additionally, the efforts described above 138 would perhaps include strategies to increase engagement with on campus health and wellness services among college student servicemembers and veterans. These particular approaches described above could perhaps increase various learning opportunities, supportive services, and valuable college student resources intended to advocate and foster emotion regulation skills and specific self-efficacy beliefs among college student servicemembers and veterans. Therefore, multiple approaches to encouraging, strengthening, and fostering emotion regulation, cognitive appraisal, and college self-efficacy on college campuses and universities can be considered within the context of two domains referred to here as 1) mental health and 2) academic health. Researchers have found higher prevalence rates of psychological disorders, such as PTSD, among college student servicemembers and veterans as compared to college students without military service experience. The imbalance associated with the mental health of college student servicemembers and veterans, particularly in relationship to PTSD severity, could be addressed on college campuses with strategies intended to cultivate social emotional learning and the mental wellbeing of college student servicemembers and veterans. Mental health professionals associated with colleges and universities are in a prime position to address the imbalance described above. Mental health professionals working closely with college student servicemembers and veterans are well positioned to explain the bidirectional relationship between emotions and behaviors and the impact that emotion regulation has on building academic relationships with others. The context in which college student servicemembers and veterans would be working with mental health professionals in the form of clinical treatment is an ideal opportunity to strengthen cognitive appraisal as a coping skill and 139 constructively promote emotion regulation skills and college self-efficacy beliefs. Furthermore, the approach described above could plausibly be supplemented through academic courses aimed at first-year college students in which social emotional and emotion regulation curricula could be delivered to college student servicemembers and veterans for the purpose of cultivating college self-efficacy. Promoting emotion regulation strategies and cognitive flexibility in this manner is a positive and proactive mental health practice targeting college student servicemembers and veterans. Altogether, this is an approach in which college student servicemembers and veterans are supported and encouraged to engage in healthy relationships and prevent or reduce negative alterations associated with internal conflict within themselves and external conflict with others. The academic health of college student servicemembers and veterans is another domain in which emotion regulation, cognitive appraisal, and college self-efficacy could be encouraged, strengthened, and fostered by academic personnel and program planning. Academic and educational professionals are placed in ideal roles to construct and influence academic environments that are conducive to promoting and utilizing emotion regulation and constructing self-efficacy beliefs associated with academic performance. This could perhaps include the provision of professional development and training for academic and educational personnel who regularly encounter and engage interpersonally with college student servicemembers and veterans. Providing professional development and training could promote emotion regulation strategies such as cognitive reappraisal as well as instructional strategies that foster regulatory emotion self-efficacy. Investing time and financial funding into professional development and training has the potential for 140 positive outcomes. Providing training and education related to social emotional development to academic and educational personnel who are working closely with college student servicemembers and veterans could potentially support overall efforts at integrating students with military service experiences into campus life and successfully interacting with diverse members of the college campus community. The professional development and training could be further incorporated and extended into conversational topics among professional learning communities sharing effective learning and teaching strategies in postsecondary learning environments and/or workshops that are organized and offered through on campus student veteran organizations and support centers. Overall, the promotion of academic health among college student servicemembers and veterans through proactive professional development and training related to emotion regulation and social emotional development could prove beneficial to the integration and retainment of undergraduate and graduate college students who have served in the armed forces. Future Research Directions Additional empirical research may lead to a more thorough understanding of the relationships among emotional regulation strategies, self-efficacy beliefs, posttraumatic symptom severity, and academic outcome expectations. Future research directions are based on five lessons learned from the current research study. First, additional indicators measuring academic outcome will be necessary in an effort to produce a well defined outcome variable included in a latent measurement model depicting emotion regulation strategies and self-efficacy beliefs associated with 141 academic performance. Approaching a research study with structural equation modeling as a methodology for explaining and predicating academic performance would need to consider additional indicators for academic performance such as factors related to attendance rates, discipline, and status in relationship to graduation. Secondly, future research related to predicting academic outcomes of college student servicemembers and veterans based on emotion regulation, self-efficacy beliefs, and PTSD severity will benefit from alternate models in which perhaps emotion regulation mediates PTSD severity and self-efficacy beliefs. In other words, research considering alternate structural models may perhaps yield better fitting models. A third suggestion for future research is to consider methodology in which the predictive strength of cognitive appraisal is compared with alternate emotion regulation strategies such as distress tolerance skills, mindfulness, and radical acceptance. This direction for future research could prove to be quite meaningful as it may include or exclude alternative emotion regulation strategies in relationship to college self-efficacy and academic performance outcomes among colleges student servicemembers and veterans. The fourth direction for future research entails the inclusion of additional variables associated with personality characteristics based on military service. The additional variables would be included into a measurement model of observed variables along with emotion regulation strategies and college self-efficacy beliefs among colleges student servicemembers and veterans. For example, it is possible that college student servicemembers and veterans generally possess a unique set of personality characteristics or are implicitly exposed to additional protective factors associated with military service 142 that plausibly promote or interfere with academic performance and outcomes. Further, there are potential masculinity traits and behaviors across the spectrum of gender identities that are perhaps related to military training and service which could be included in future measurement models considering emotion regulation and college self-efficacy. In other words, college student servicemembers and veterans who have served in the armed forces have perhaps developed social emotional skills related to survival and persistence in very specific military related environments in comparison to academic environments. These factors are inherent qualities of military service which are perhaps conducive to strengthening or otherwise weakening the cognitive flexibility and emotion regulation skills so as to reinforce or perhaps deteriorate self-efficacy beliefs related to academic performance. The personality and gender traits described above are also aspects of military service that could effectively be researched in a qualitative manner to specifically define the concepts, characteristics, and phenomenon related to military service. The themes discovered from qualitative research could then be included in future quantitative research as observed variables influencing academic and education related experiences of college student servicemembers and veterans. Lastly, a fifth and final direction for future research is to seek out and include a larger representation of the smaller demographic populations included in this study. For example, the statistical differences between racial and ethnic identities, gender identities, and military branches may be much more meaningful when larger representations of these identities are included in future research. 143 Conclusion In summary, additional research is required to draw significant and more meaningful conclusions about the predictive relationships that emotion regulation strategies and self-efficacy beliefs have with PTSD severity in relationship to academic outcomes among college student servicemembers and veterans. Future research will assist in making recommendations for specific interventions designed to enhance the mental health and academic health related outcomes for college student who have served in the United States Armed Forces. The results of this study confirm the value of including emotion regulation in relationship to college self-efficacy as well as regulatory emotional self-efficacy as it relates to PTSD severity. These findings add to the existing literature supporting the use of emotion regulation strategies in the context of reducing the severity of PTSD symptoms and strengthening college self-efficacy APPENDIX A EMOTION REGULATION QUESTIONNAIRE Instructions and Items: We would like to ask you some questions about your emotional life, in particular, how you control (that is, regulate and manage) your emotions. The questions below involve two distinct aspects of your emotional life. One is your emotional experience, or what you feel like inside. The other is your emotional expression, or how you show your emotions in the way you talk, gesture, or behave. Although some of the following questions may seem similar to one another, they differ in important ways. For each item, please answer using the following scale: 1-----------------2------------------3------------------4------------------5------------------6------------------7 strongly neutral strongly disagree agree 1. ____ When I want to feel more positive emotion (such as joy or amusement), I change what I’m thinking about. 2. ____ I keep my emotions to myself. 3. ____ When I want to feel less negative emotion (such as sadness or anger), I change what I’m thinking about. 4. ____ When I am feeling positive emotions, I am careful not to express them. 5. ____ When I’m faced with a stressful situation, I make myself think about it in a way that helps me stay calm. 6. ____ I control my emotions by not expressing them. 7. ____ When I want to feel more positive emotion, I change the way I’m thinking about the situation. 8. ____ I control my emotions by changing the way I think about the situation I’m in. 9. ____ When I am feeling negative emotions, I make sure not to express them. 10. ____ When I want to feel less negative motion, I change the way I’m thinking about the situation. APPENDIX B REGULATORY EMOTIONAL SELF-EFFICACY SCALE Instructions and Items: Respond to each item using the following 5 point scale: 1-not at all well to 5-very well. How well can you… 1. ____ Express joy when good things happen to you? 2. ____ Feel gratified over achieving what you set out to do? 3. ____ Rejoice over your successes? 4. ____ Express enjoyment freely at parties? 5. ____ Reduce your upset when you don’t get the appreciation you feel you deserve? 6. ____ Manage negative feelings when reprimanded by your parents or significant others? 7. ____ Avoid getting upset when others keep giving you a hard time? 8. ____ Get over irritation quickly for wrongs you have experienced? 9. ____ Avoid flying off the handle when you get angry? 10. ____ Keep from getting discouraged by strong criticism? 11. ____ Keep from getting discouraged in the face of difficulties? 12. ____ Keep from getting dejected when you are lonely? APPENDIX C COLLEGE SELF-EFFICACY INVENTORY Instructions and Items: The following 20 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 a ‘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. 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 01. ____ Make new friends at college. 02. ____ Divide chores with others you live with (omitted roommate self-efficacy). 03. ____ Talk with university academic and support (e.g., advising) staff 04. ____ Manage your time effectively. 05. ____ Ask a question in class. 06. ____ Participate in class discussions. 07. ____ Get a date when you want one. 08. ____ Research a term paper. 09. ____ Do well on your exams. 10. ____ Join a student organization. 11. ____ Talk to your professors/instructors. 12. ____ Join an intramural sports team. 13. ____ Ask a professor/instructor a question outside of class. 14. ____ Take good class notes. 15. ____ Get along with others you live with (omitted roommate self-efficacy). 16. ____ Divide space in your residence (omitted roommate self-efficacy). 17. ____ Understand your textbooks. 18. ____ Keep up to date with your schoolwork. 19. ____ Write course papers. 20. ____ Socialize with others you live with (omitted roommate self-efficacy) APPENDIX D PTSD CHECK LIST-5 (PCL-5) Instructions and Items: Below is a list of problems that people sometimes have in response to a very stressful experience. Please read each problem carefully and then use one of the numbers from the Likert scale described below to indicate how much you have been bothered by that problem in the past month. 0-Not at all, 1-A little bit, 2-Moderately, 3-Quite a bit, 4-Extremely 1. ___ Repeated, disturbing, and unwanted memories of the stressful experience? 2. ___ Repeated, disturbing dreams of the stressful experience? 3. ___ Suddenly feeling or acting as if the stressful experience were actually happening again (as if you were actually back there reliving it)? 4. ___ Feeling very upset when something reminded you of the stressful experience? 5. ___ Having strong physical reactions when something reminded you of the stressful experience (for example, heart pounding, trouble breathing, sweating)? 6. ___ Avoiding memories, thoughts, or feelings related to the stressful experience? 7. ___ Avoiding external reminders of the stressful experience (for example, people, places, conversations, activities, objects, or situations)? 8. ___ Trouble remembering important parts of the stressful experience? 9. ___ Having strong negative beliefs about yourself, other people, or the world (for example, having thoughts such as: I am bad, there is something seriously wrong with me, no one can be trusted, the world is completely dangerous)? 10. ___ Blaming yourself or someone else for the stressful experience or what happened after it? 11. ___ Having strong negative feelings such as fear, horror, anger, guilt, or shame? 12. ___ Loss of interest in activities that you used to enjoy? 13. ___ Feeling distant or cut off from other people? 14. ___ Trouble experiencing positive feelings (for example, being unable to feel happiness or have loving feelings for people close to you)? 15. ___Irritable behavior, angry outbursts, or acting aggressively? 16. ___Taking too many risks or doing things that could cause you harm? 17. ___Being “super alert” or watchful or on guard? 148 18. ___Feeling jumpy or easily startled? 19. ___Having difficulty concentrating? 20. ___Trouble falling or staying asleep REFERENCES Ackerman, R., DiRamio, D., & Mitchell, R. L. (2009). Transitions: Combat veterans as college students. New Directions for Student Services, 126, 5-14. Alessandri, G., Vecchione, G., & Caprera, G. V. (2014). Assessment of regulatory emotional self-efficacy beliefs: A review of the status of the art and some suggestions to move the filed forward. Journal of Psychoeducational Assessment, 33, 24-32. Armour, C., Tsai, J., Durham, T. A., Charak, R., Biehn, T. L., Elhai, J. D., & Pietrzak, R. H. (2015). Dimensional structure of DSM–5 posttraumatic stress symptoms: Support for a hybrid anhedonia and externalizing behaviors model. Journal of Psychiatric Research, 61, 106-113. Bandura, A. (1988). Self-efficacy conception of anxiety. Anxiety Research, 1, 1-77. Bandura, A. (1989a). Human agency in social cognitive theory. American Psychologist, 44(9), 1175-1184. Bandura, A. (1989b). Regulation of cognitive processes through perceived selfefficacy. Developmental Psychology, 25(5), 729-735. Bandura, A. (1994). Self-efficacy. In R. J. Corsini (Ed.), Encyclopedia of psychology (pp. 368-369). New York, NY: J. Wiley & Sons. Bandura, A. (1997). Self-efficacy: The exercise of control. New York, NY: Freeman. Bandura, A., Caprara, G. V., Barbaranelli, C., Gerbino, M., & Pastorelli, C. (2003). Role of affective self-regulatory efficacy on diverse spheres of psychosocial functioning. Child Development, 74, 769-782. Bandura, A., Cioffi, D., Taylor, C. B., & Brouillard, M. E. (1988). Perceived selfefficacy in coping with cognitive stressors and opioid activation. Journal of Personality and Social Psychology, 55, 479-488. Bandura, A., Reese, L., & Adams, N. E. (1982). Microanalysis of action and fear arousal as a function of differential levels of perceived self-efficacy. Journal of Personality and Social Psychology, 43, 5-21. 150 Bandura, A., Taylor, C. B., Williams, S. L., Mefford, I. N., & Barchas, J. D. (1985). Catecholamine secretion as a function of perceived coping self-efficacy. Journal of Personality and Social Psychology, 53, 406-414. Baron, R. M., & Kenny, D. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182. Barrett, L. F., Ochsner, K. N., & Gross, J. J. (2007). On the automaticity of emotion. In J. A. Bargh (Ed.), Social psychology and the unconscious: The automaticity of higher mental processes (pp. 173-217). New York, NY: Psychology Press. Barry, A. E., Whiteman, S. D., & MacDermid-Wadsworth, S. M. (2012). Implications of posttraumatic stress among military-affiliated and civilian students. Journal of American College Health, 60(8), 562-573. Benight, C. C., & Bandura, A. (2004). Social cognitive theory of posttraumatic recovery: the role of perceived self-efficacy. Behavior Research and Therapy, 42, 1129-1148. Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychology Bulletin, 107(2), 238-246. Bentler, P. M., & Bonnet, D. G. (1980). Significance tests and goodness of fit in the analysis of covariance structures. Psychology Bulletin, 88(3), 588-606. Betz, N. E., Klein, K. L., & Taylor, K. M. (1996). Evaluation of a short from of the career decision-making self-efficacy scale. Journal of Career Assessment, 4, 47-58. Blackburn, L., & Owens, G. P. (2015). The effect of self-efficacy and meaning in life on posttraumatic stress disorder and depression severity among veterans. Journal of Clinical Psychology, 71(3), 219-228. Blasco-Ros, C., Sánchez-Lorente, S., & Martinez, M. (2010). Recovery from depressive symptoms, state anxiety and posttraumatic stress disorder in women exposed to physical and psychological, but not to psychological intimate partner violence alone: A longitudinal study. BMC Psychiatry, 10, 98-109. Blevins, C. A., Weathers, F. W., Davis, M. T., Witte, T. K., & Domino, J. L. (2015). The posttraumatic stress disorder checklist for DSM-5 (PCL-5): Development and initial psychometric evaluation. Journal of Traumatic Stress, 28, 489-498. Boden, M. T., Westermann, S., McRae, K., Kuo, J., Alvarez, J., Kulkarni, M. R., & Bonn-Miller, M. O. (2013). Emotional regulation and posttraumatic stress disorder: A prospective investigation. Journal of Social and Clinical Psychology, 32(3), 296-314. 151 Bovin, M. J., Marx, B. P., Weathers, F. W., Gallagher, M. W., Rodriguez, P., Schnurr, P. P., & Keane, T.M. (2016). Psychometric properties of the PTSD checklist for Diagnostic and Statistical Manual of Mental Disorders-Fifth Edition in veterans. Psychology Assessment, 28(11), 1379-1391. Boyraz, G., Granda, R., Baker, C. N., Tidwell, L. L., & Waits, J. B. (2016). Posttraumatic stress, effort regulation, and academic outcomes among college students: A longitudinal study. Journal of Counseling Psychology, 63(4), 475486. Browne, M. W., & Cudeck, R. (1992). Alternative ways of assessing model fit. Sociological Methods & Research, 21, 230-258. Bryan, C. J., Bryan, A. O., Hinkson Jr., K., Bichrest, M., & Ahern, A. (2014). Depression, posttraumatic stress disorder, and grade point average among student servicemembers and veterans. Journal of Rehabilitation Research and Development, 51(7), 1035-1046. Caprara, G. V., & Gerbino, M. (2001). Affective perceived self-efficacy: The capacity to regulate negative affect and to express positive affect. In G. V. Caprara (Ed.), Self-efficacy assessment (pp. 35-50). Trento, Italy: Edizioni Erickson. Caprara, G. V., Di Giunta, L., Eisenberg, N., Gerbino, M., Pastorelli, C., & Tramontano, C. (2008). Assessing regulatory emotional self-efficacy in three countries. Psychological Assessment, 20(3), 227-237. Caprara, G. V., Fida, R. Vecchione, M., Del Bove, G., Vacchio, G. M., Barbaranelli, C., & Bandura, A. (2008). Longitudinal analysis of the role of perceived selfefficacy for self-regulated learning in academic continuance and achievement. Journal of Educational Psychology, 100(3), 525-534. Cate, C. A. (2014). Million Records Project: Research from Student Veterans of America. Student Veterans of America, Washington, DC. Cate, C. A., & Davis, T. (2016). Commonalities and intersections within the student veteran population. Student Veterans of America, 2, 1-7. Chemers, M. M., Hu, L., & Garcia, B. F. (2001). Academic self-efficacy and firstyear college student performance and adjustment. Journal of Educational Psychology, 93, 55-64. Cieslak, R., Benight, C., & Lehman, V. (2008). Coping self-efficacy mediates the effects of negative cognitions on posttraumatic distress. Behavior Research Therapy, 46(7), 788-798. Cuellar, I., Harris, L. C., & Jasso, R. (1980). An acculturation scale for Mexican American normal and clinical populations. Hispanic Journal of Behavioral Sciences, 2, 199-217. 152 Deffenbacher, K. A., Bornstein, B. H., Penrod, S. D., & McGorty, E. K. (2004). A meta-analytic review of the effects of high stress on eyewitness memory. Law and Human Behavior, 28, 687-706. Derogatis, L. R., & Melisaratos, N. (1983). The brief symptom inventory: An introductory report. Psychological Medicine, 13(3), 595-605. DiRamio, D., Ackerman, R., & Mitchell, R. L. (2008). From combat to campus: Voices of student-veterans. NASPA Journal, 45(1), 73-102. Eisenberg, N., Cumberland, A., Spinrad, T. L., Fabes, R. A., Shepard, S. A., & Reiser, M. (2001). The relations of regulation and emotionality to children’s externalizing and internalizing problem behavior. Child Development, 72, 1112-134. Elliot, M. (2015). Predicting problems on campus: An analysis of college student veterans. Analyses of Social Issues and Public Policy, 15(1), 105-126. Ellsworth, P. C., & Scherer, K. R. (2003). Appraisal processes in emotion. In R. J. Davidson, K. R. Scherer, & H. H. Goldsmith (Eds.), Handbook of affective sciences (pp. 572-595). New York, NY: Oxford University Press. Enders, C. K., & Bandalos, D. L. (2001). The relative performance of full information maximum likelihood estimation for missing data in structural equation models. Structural Equation Modeling: A Multidisciplinary Journal, 8, 430-457. Friedman, M. J. (2013). Finalizing PTSD in the DSM-5: Getting here from there and where to go next. Journal of Traumatic Stress, 26, 548-556. Gable, S. L., Reis, H. T., Impett, E. A., & Asher, E. R. (2004). What do you do when things go right? The intrapersonal and interpersonal benefits of sharing positive events. Journal of Personality and Social Psychology, 87, 228-245. Gore, P., Leuwerke, W. C., & Turley, S. E. (2006). A psychometric study of the college self-efficacy inventory. Journal of College Student Retention: Research, Theory, and Practice, 7, 227-244. Gross, J. J. (1998). Antecedent and response focused emotional regulation: Divergent consequences for experience, expression, and physiology. Journal of Personality and Social Psychology, 74, 224-237. Gross, J. J., & John, O. P. (2003). Individual differences in two emotion regulation processes: Implications for affect, relationships, and well-being. Journal of Personality and Social Psychology, 85(2), 348-362. Gross, J. J. (2014). Handbook of emotion regulation (2nd ed.). New York, NY: Guilford Press. 153 Helmer, D. A., Rossignol, M., Blatt, M., Agarwal, R., Teichman, R., & Lange, G. (2007). Health and exposure concerns of veterans deployed to Iraq and Afghanistan. Journal of Occupational and Environmental Medicine, 49, 475480. Hofmann, S. G., Heering, S., Sawyer, A. T., & Asnaani, A. (2009). How to handle anxiety: The effects of reappraisal, acceptance, and suppression strategies on anxious arousal. Behavior Research and Therapy, 47, 389-394. Hoge, C., Auchterlonie, J., & Milliken, C. (2006). Mental health problems, use of mental health services, and attrition from military service after returning from deployment to Iraq or Afghanistan. The Journal of the American Medical Association, 295(9), 1023-1032. Hoge, C., Riviere, L., Wilk, J., Herrell, R., & Weathers, F. (2014). The prevalence of post-traumatic stress disorder (PTSD) in US combat soldiers: A head-to-head comparison of DSM-5 versus DSM-IV-TR symptom criteria with the PTSD checklist. The Lancet Psychiatry 1, 269-77. Hoyle, R. H. (1995). The structural equation modeling approach: Basic concepts and fundamental issues. In R. H. Hoyle (Ed.), Structural equation modeling: concepts, issues, and applications (pp. 1-15). Thousand Oaks, CA: Sage Publications, Inc. Hsieh, P., Sullivan, J., & Guerra, N. (2007). A closer look at college students: Selfefficacy and goal orientation. Journal of Advanced Academics, 18, 454-476. Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1-55. Ivevic, Z., & Brackett, M. (2014). Predicting school success: Comparing conscientiousness, grit, and emotion regulation ability. Journal of Research in Personality, 52, 29-36. Kenny, D. A. (1979). Correlation and causality. New York, NY: Wiley. Kenny, D. A., Kaniskan, B., & McCoach, D. B. (2015). The performance of RMSEA in models with small degrees of freedom. Sociological Methods & Research, 44(3), 486-507. Kenny, D. A., & McCoach, D. B (2003). Effect of the number of variables on measures of fit in structural equation modeling. Structural Equation Modeling, 10, 333-351. Kim, Y., & Cole, S. (2013). Student veterans/servicemembers’ engagement in college and university life and education. Washington, DC: American Council on Education. 154 Kline, R. B. (1998). Principles and practice of structural equation modeling. New York, NY: Guilford Press. Krumrei-Mancuso, E. J., Newton, F. B., Kim, E., & Wilcox, D. (2013). Psychosocial factors predicting first-year college student success. Journal of College Student Development, 54(3), 247-266. Kuh, G. D., & Pace, C. R. (1998). College student expectations questionnaire (2nd Ed.). Center for Postsecondary Research and Planning, Bloomington: IN. Larsen, R. (2009). Toward a science of mood regulation. Psychological Inquiry, 11, 129-141. Lazerus, R. S., & Alfert, E. (1964). Short-circuiting of threat by experimentally altering cognitive appraisal. Journal of Abnormal and Social Psychology, 69, 195-205. LeDoux, J.E. (1994). Emotion, memory, and the brain. Scientific American, 270, 5057. Levine, L. J., & Edelstein, R. S. (2009). Emotion and memory narrowing: A review and goal-relevance approach. Cognition and Emotion, 23(5), 833-875. Lighthall, A. (2012). Ten things you should know about today’s student veteran. National Education Association Higher Education Advocate, 30(3), 12. Lightsey, O. R., McGhee, R., Ervin, A., Gharghani, G. G., Rarey, E. B., Daigle, R. P., Wright, K. F., Constantin, D., & Powell, K. (2013). Self-efficacy for affect regulation as a predictor of future life satisfaction and moderator of the negative affect-life satisfaction relationship. Journal of Happiness Studies, 14, 1-18. Livingston, W. G., Havice, P. A., Cawthon, T. W., & Fleming, D. S. (2011). Coming home: Student veterans’ articulation of college reenrollment. Journal of Student Affairs Research and Practice, 48, 315-331. MacEachron, A., & Gustavsson, N. (2012). Peer support, self-efficacy, and combatrelated trauma symptoms among returning OIF/OEF veterans. Advances in Social Work, 13(3), 586-602. May, C. L., & Wisco, B. E. (2016). Defining trauma: How level of exposure and proximity affect risk for posttraumatic disorder. Psychological Trauma: Theory, Research, and Policy, 8(2), 233-240. McBain, L., Kim, Y. M., Cook, B. J., & Snead, K. M. (2012). From soldier to student II: Assessing campus programs for veterans and servicemembers. Washington, DC: American Council on Education. 155 National Association of Student Personnel Administrators (2013). Measuring the success of student veterans and active duty military students. Washington, D.C.: The NASPA Research and Policy Institute. National Center for Veterans Analysis and Statistics (2014). Retrieved from https://www.va.gov/vetdata Nyaronga, D., & Toma, R. (2015). Student servicemembers/veterans’ mental health: Who is at greater risk for posttraumatic stress symptoms on college campuses? American Psychological Association, 27(4), 223-231. Nye, C. D., & Drasgow, F. (2011). Assessing goodness of fit: Simple rules of thumb simply do not work. Organizational Research Methods, 14(3), 548-570. O’Bryan, E. M., McLeish, A. C., Kraemer, K. M., & Fleming, J. B. (2015). Emotional regulation difficulties and posttraumatic stress disorder symptom cluster severity among trauma-exposed college students. Psychological Trauma: Theory, Research, Practice, and Policy, 7(2), 131-137. Panksepp, J. (1998). Affective neuroscience: The foundations of human and animal emotions. New York, NY: Oxford University Press. Parker, J. D. A., & Endler, N. S. (1996). Coping and defense: A historical overview. In M. Zeidner & N. S. Endler (Eds.), Handbook of coping: Theory, research, applications (pp. 3-23). Oxford, England: John Wiley & Sons. Parrott, W. G., & Smith, R. H. (1993). Distinguishing the experiences of envy and jealousy. Journal of Personality and Social Psychology, 64(6), 906-920. Peterson, A. L., Wong, V., Haynes, M. F., Bush, A. C., & Schillerstrom, J. E. (2010). Documented combat-related mental health problems in military noncombatants. Journal of Traumatic Stress, 23, 674-681. Pickett, S. M., Barbaro, N., & Mello, D. (2016). The relationship between subjective sleep disturbance, sleep quality, and emotion regulation difficulties in a sample of college students reporting trauma exposure. Psychological Trauma: Theory, Research, Practice, and Policy, 8(1), 25-33. Queen, B., & Lewis, L. (2014). Services and support programs for military servicemembers and veterans at postsecondary institutions, 2012–13 (NCES 2014-017). U.S. Department of Education. Washington, DC: National Center for Education Statistics. Quoidbach, J., Berry, E. V., Hansenne, M., & Mikolajczak, M. (2010). Positive emotion regulation and well-being: Comparing the impact of eight savoring and dampening strategies. Personality and Individual Differences, 49(5), 368373. 156 Quintana, S. M., & Maxwell, S. E. (1999). Implications of recent developments in structural equation modeling for counseling psychology. The Counseling Psychologist, 27(4), 485-527. Read, J. P., Bachrach, R. L., Wright, A., & Colder, C. R. (2016). PTSD symptom course during first-year of college. Psychological Trauma: Theory, Research, Practice, and Policy, 8(3), 393-403. Romero, D. H., Riggs, S. A., & Ruggero, C. (2015). Coping, family social support, and psychological symptoms among student veterans. Journal of Counseling Psychology, 62(2), 242-252. Rudd, M., Goulding, J., & Bryan, C. (2011). Student veterans: A national survey exploring psychological symptoms and suicide risk. Professional Psychology: Research and Practice, 42(5), 354-360. Rumann, C. B., & Hamrick, F. A. (2010). Student veterans in transition: Re-enrolling after war zone deployments. The Journal of Higher Education, 81(4), 431-458. Russell, D., Cutrona, C. E., Rose, J., & Yurko, K. (1984). Social and emotional loneliness: An examination of Weiss's typology of loneliness. Journal of Personality and Social Psychology, 46(6), 1313-1321. Samson, A. C., & Gross, J.J. (2012). Humour as emotion regulation: The differential consequences of negative versus positive humour. Cognition and Emotion, 26, 375-384. Sander, L. (2012). Out of uniform: At a half a million and counting, veterans cash in on post-9/11 GI bill. The Chronicle of Higher Education 3, 341-367. Scheppes, G., Suri, G., & Gross, J. (2015). Emotional regulation and psychopathology. Annual Review of Clinical Psychology, 11, 379-401. Schneider, W., & Shiffrin, R. M. (1977). Controlled and automatic human information processing: Detection, search, and attention. Psychological Review, 84(1), 166. Shepard, L., & Wild, J. (2014). Emotional regulation, physiological arousal and PTSD symptoms in trauma-exposed individuals. Journal of Behavior Therapy and Experimental Psychiatry, 45, 360-367. Shumaker, R. E., & Lomax, R. G. (1996). A beginner’s guide to structural equation modeling. Mahwah, NJ: Lawrence Erlbaum Associates. 157 Sippel, L. M., Roy, A. M., Southwick, S. M., & Fichtenholtz, H. M. (2016). An examination of the roles of trauma exposure and posttraumatic stress disorder on emotion regulation strategies of Operation Iraqi Freedom, Operation Enduring Freedom, and Operation New Dawn veterans. Cognitive Behaviour Therapy, 45(5), 339-350. Smith, A. C., & Kleinman, S. (1989). Managing emotions in medical school: Students’ contacts with the living and the dead. Social Psychology Quarterly, 52, 56-69. Solberg, V. S., O’Brien, K., Villareal, P., Kennel, R., & Davis, B. (1993). Selfefficacy and Hispanic college students: Validation of the College Self-Efficacy Instrument. Hispanic Journal of Behavioral Sciences, 15, 80-95. Solomon, Z., Benbenishty, R., & Mikulincer, M. (1991). The contribution of wartime, pre-war and postwar factors to self-efficacy: A longitudinal study of combat stress reaction. Journal of Traumatic Stress, 4, 345-361. Solomon, Z., Weisenberg, M., Schwarzwald, J., & Mikulincer, M. (1988). Combat stress reaction and posttraumatic stress disorder as determinants of perceived self-efficacy in battle. Journal of Social and Clinical Psychology, 6, 356-370. Strain, A. C., & D’Mello, S. K. (2015). Affect regulation during learning: The enhancing effect of cognitive reappraisal. Applied Cognitive Psychology, 29,119. Student Affairs Administrators in Higher Education (2013). Retrieved from https://.naspa.org/publications/data. Szczurek, L., Monin, B., & Gross, J. (2012). The rejections of affective deviants. Psychological Science, 23, 1105-1111. Thomas, J. L., Wilk, J. E., Riviere, L. A., McGurk, D., Castro, C. A., & Hoge, C. W. (2010). Prevalence of mental health problems and functional impairment among active component and National Guard soldiers following combat in Iraq. Archives of General Psychiatry, 67, 614-623. Tull, M. T., Barrett, H. M. McMillan, E. S., & Roemer, L. (2007). A preliminary investigation of the relationship between emotion regulation difficulties and posttraumatic stress symptoms. Behavior Therapy, 38, 303-313. Weathers, F. W., Litz, B. T., Herman, D. S., Huska, J. A., & Keane, T. M. (1993). The PTSD Checklist: Reliability, validity, and diagnostic utility. Paper presented at the annual meeting of ISTSS, San Antonio, CA. Weathers, F. W., Litz, B. T., Keane, T. M., Palmieri, P. A., Marx, B. P., & Schnurr, P. P. (2013). The PTSD Checklist for DSM–5 (PCL-5). Boston, MA: National Center for PTSD. 158 Wild, N., & Paivio, S. (2004). Psychological adjustment, coping, emotion regulation as predictors of posttraumatic growth. Journal of Aggression, Maltreatment, and Trauma, 8(4), 97-122. Wood, R., & Bandura, A. (1989). Social cognitive theory of organizational management. Academy of Management Review, 14(3), 361-384. Wortmann, J. H., Jordan, A. H., Resick, P. A., Foa, E. B., Yarvis, J. S., Mintz, J., . . . Peterson, A. L., (2016). Psychometric analysis of the PTSD Checklist-5 (PCL5) among treatment seeking military servicemembers. Psychological Assessment, 28(11), 1392-1403. Wright S. L., Jenkins-Guarnieri, M. A., & Murdock, J. L. (2012). Career development among first-year colleges students: College self-efficacy, student persistence, and academic success. Journal of Career Development, 40(4), 292-310.