! UNIVERSITY FITNESS CENTER PARTICIPATION AND COLLEGE STUDENT ACADEMIC SUCCESS By Samantha J. Deere A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirement s for the degree of Kinesiology ÑDoctor of Philosophy 2015 !ABSTRACT UNIVERSITY FITNESS CENTER PARTICIPATION AND COLLEGE STUDENT ACADEMIC SUCCESS By Samantha J. Deere University recreational sports departments are charged with promoting physical activity (PA) to college students. However, funding is necessary for the upkeep of equipment, quality programming, and continued p romotion of PA through recreational sports . Although PA is important to university adminis trators, it is likely secondary to student academic success. Therefore, the purpose of this investigation was to identify relationships between recreational sports participation and college student academic success. Study participants included all first t ime degree seeking freshmen students who graduated from high school in Spring, 2010, were in their first semester at the university in the Fall 2010, not student athletes, and had no prior college class experience. Academic variables were compared between recreational sports members (students who purchased a t least one fitness center membership) and nonmembers (students who did not purchase a membership). Diffe rences among levels (determined via identification card swipes) of recreational sports use were al so compared (never used, low, medium, high). Means ±SD and percentages were calculated for all variables of interest. Differences between members and nonmembers in cumulative GPA (cGPA) and cumulative credits completed (CCC) after four consecutive semester s were assessed via analysis of covariance (ANCOVA). Odds ratios (OR) and adjusted odds ratios (aOR) with 95% confidence intervals (CI) were calculated via logistic regression for reaching one -year retention, two -year retention, !and sophomore status accord ing to recreational sports member status (ref =nonmember). Differences among recreational sports use groups in four -year cGPA were calculated via ANCOVA. OR and aOR ratios with 95% CI were calculated via logistic regression for reaching one-year retention, two-year retention, and five -year g raduation according to recreational sports use level (ref=never used). A repeated measures analysis was utilized to assess a possible interaction between yea r in school and yearly recreational sports use on yearly GPA. After adjusting for covariates, members earned higher cGPA (3.17±0.48) and completed more credits (57.6±7.1) than nonmembers (3.01±0.55 and 55.7±9.0 respectively). Members were also more likely to enroll in a second (aOR=1.42, 95%CI: 1.10 -1.85) and third (a OR=1.39, 95%CI: 1.10 -1.75) year and achieve sophomore status (aOR=1.59, 95%CI: 1.14 -2.22) within two consecutive semesters than nonmembers. Medium (3.26±0.41) and high users (3.27±0.45) of recreational sports earned higher four -year cGPAs than low user s (3.20±0.42) and never used (3.08±0.47). Medium/high users were more likely to re ach one -year retention than never used (aOR=1.94, 95%CI: 1.11 -3.38). No differences were found among recreational sports use levels in two -year retention or five -year BachelorÕs degree attainment. Yearly recreational sports use positively related to yearly GPA , but the relationship did not differ by year in school. Results of this study indicate a positive relationship between recreational sports and college student academic success and should be shared with campus administrators so that they appreciate the benefits of this department and fund it sufficiently. Further, students should participate in recreational sports early and often in their aca demic careers. Future researchers should investigate these relationships, and their mechani sms, within other components of recreational sports (e.g., club/intr amural sports) and should collaborate with campus departments to identify additional areas of stu dent success that relate to recreational sports participation. !! Copyright by SAMANTHA J. DEERE 2015 !"!ACKNOWLEDGEMENT S First I must thank my committee members, the people who made this dissertation happen. To Dr. Lanay Mudd, thank you for your down to earth attitude and for answering endless statistical questions. To Dr. Richard McNeil, thank you for your unwavering trust in my research abilities, for putting me in contact with MSU data gurus, and for your guidance in all things life the last four years. To Dr. Kristen Renn , thank you for your enthusiasm; I heard you present at my PhD orientation and you have been an inspir ation ever since. This dissertation took much more than my fingers typing; you were all critical to its completion and I cannot thank you enough. I would also like to acknowledge Dr. Ira Washington, who spent a great deal of time helping me navigate all things MSU data. Ira, you were instrumental in the completion of this dissertation, thank you. I would also like to acknowledge and sincere ly thank the many undergraduate and graduate students who provided unwavering support over the last four years. Particu larly, I must thank Catherine Gammon, for being a phenomenal friend, letting me bounce ideas off her, listening to me vent, and providing me with endless tea and chocolate (both of which are critical to graduate school). I must also thank my very dear frie nds, Christopher Connolly and Rebecca Schlaff, the two people who encouraged me to keep going and who provided comedic relief at a ll the right moments. Finally, I must thank my biggest supporters. To Dr. Jim Pivarnik, my advisor, thank you for your patie nce, friendship, knowledge, and support. I am beyond grateful for your guidance, willingness to meet at a momentÕs notice, and for considering me a member of your family. To mom and dad, thank you for believing in me since the day I was born Éand for sending cookies . I would not be where I am today without you. To Matt, thank you for your confidence in me, for cooking dinner while I yelled at my computer, for the high fives along the way, and for your love and support through all the ups and downs. You are my rock, you keep me sane, and you bring a happiness to my !"#!life that is untouchable. Last, but certainly not least, a big thank you to Carson and Charlie, my beloved fur babies. The tail wags, butt wiggles, and nightly snuggles kept a smile on my face every single day. !"## !TABLE OF CONTENTS LIST OF TABLES ......................................................................................................................... ix LIST OF FIGURES ....................................................................................................................... xi CHAPTER 1 ....................................................................................................................................1 INTRODUCTION ...........................................................................................................................1 RECREATION AL SPORTS AND ACADEMIC SUCCESS .............................................5 Recreational Sports and Grade Point Average .........................................................5 Rec reational Sports and Retention ...........................................................................6 Recreational Sports an d BachelorÕs Degree Attainment .........................................8 SUMMARY .........................................................................................................................9 RESEARCH AIMS ............................................................................................................10 Definition ...............................................................................................................10 Specific A ims .........................................................................................................11 ORG ANIZATION OF THE DISSERTATION .................................................................13 REFERENCES ..................................................................................................................14 CHAPTER 2 ..................................................................................................................................19 REVIEW OF THE LITERATURE ...............................................................................................19 INTRODUCTION .............................................................................................................19 Foundational Theory ..............................................................................................23 AstinÕs Theory of Involvement ..............................................................................23 TintoÕs Theory of Departure ..................................................................................24 RECREATIONA L SPORTS AND ACADEMIC SUCCESS ...........................................25 Grade Point Average ..............................................................................................25 Retention ................................................................................................................30 BachelorÕs Degree Attainment ...............................................................................36 SUMMARY .......................................................................................................................38 REFERENCES ..................................................................................................................40 CHAPTER 3 ..................................................................................................................................46 MANUSCRIPT ONE ....................................................................................................................46 ABSTRACT .......................................................................................................................46 INTRODUCTION .............................................................................................................47 BACKGROUND ...............................................................................................................48 Academic Success and Retention ..........................................................................48 Recreationa l Sports and Student Services ..............................................................48 METHODS ........................................................................................................................51 Academic Success ..................................................................................................51 Recreational Sports Fitness Center Membership ...................................................52 Data Analysis .........................................................................................................53 RESULTS ..........................................................................................................................53 !"### ! DISCUSSION ....................................................................................................................55 ADDENDUM ....................................................................................................................58 Covariates ..............................................................................................................58 Statistical Analysis .................................................................................................59 Results ....................................................................................................................60 Conclusion .............................................................................................................68 REFERENCES ..................................................................................................................69 CHAPTER 4 ..................................................................................................................................72 MANUSCRIPT TWO ....................................................................................................................72 INTRODUCTION .............................................................................................................72 METHODS ........................................................................................................................76 Data Collection ......................................................................................................76 Academic Success ..................................................................................................77 Recreational Sports ................................................................................................77 Covariates ..............................................................................................................78 Statistical Analysis .................................................................................................79 RESULTS ..........................................................................................................................80 DISCUSSION ....................................................................................................................84 REFERENCES ..................................................................................................................88 CHAPTER 5 ..................................................................................................................................92 MANUSCRIPT THREE ................................................................................................................92 INTRODUCTION .............................................................................................................92 METHODS ........................................................................................................................97 Data Collection ......................................................................................................97 Academic Success ..................................................................................................97 Recreational Sports ...............................................................................................98 Covariates ..............................................................................................................99 Statistical Analysis ...............................................................................................100 RESULTS ........................................................................................................................102 DISCUSSION ..................................................................................................................108 Limitations and Str engths ....................................................................................111 Summary ..............................................................................................................112 REFERENCES ................................................................................................................114 CHAPTER 6 ................................................................................................................................117 SUMMARY AND CONCLUSIONS ..........................................................................................117 LIMITATIONS, STRE NGTHS, AND FUTURE DIRECTIONS ...................................123 REFERENCES ................................................................................................................126 !#$!LIST OF TABLES Table 3.1. Academic success variables for recreational sports fitness center Ñmembers and n onmembers. ...........................................................................................................................54 Table 3.2 Means, standard deviations (SD) , frequencies, and percentages of dependent and covariate variables separated by recreational sports membership. ........................61 Table 3.3. Means, standard deviat ions (SD) , frequencies, and percentages of independent and covariate variables separated by one -year retention status. ................................62 Table 3.4. Means, standard deviations (SD) , frequencies, and percentages of independent and covariate variables separated by two -year retention status. ...............................63 Table 3.5. Means, standard deviations (SD) , frequencies, and percentages of independent and covariate variab les separated by class standing. .................................................64 Table 3.6. Analysis of covariance for cumulative grade point average after four consecutive semesters. ...........................................................................................................65 Table 3.7. Analysis of covariance for cumulative credits completed after four consecutive semesters ............................................................................................................65 Table 3.8. Odds ratios (OR), adjus ted odds ratios (aOR), and 95% c onfidence intervals (CI) for the effect of independent and covariate variables on one-year retention status. ..........................................................................................................66 Table 3.9. Odds ratios (OR), adjus ted odds ratios (aOR), and 95% confidence intervals (CI) for the effect of independent and covariate variables on two-year retention status. ...............................................................................................................66 Table 3.10. Odd s ratios (OR), adjus ted odds ratios (aOR), and 95% c onfidence intervals (CI) for the effect of independent and covaria te variables on class standing. ................67 Table 4.1. Means, standard deviations (SD) , frequencies, and percentages of independent and covariat e variables separated by one -year retention. ..........................................82 Table 4.2. Means, standard deviation s (SD) , frequencies, and percentages of independent and covariate variables separated by two -year retention. ..........................................83 Table 4.3. Odds ratios (OR), adjus ted odds ratios (aOR), and 95% confidence intervals (CI) for the effect of independent and covariate variables on one-year retention. ........................................................................................................................84 !$!Table 5.1. Means, standard deviations (SD) , and percentages of dependent and covariate variables separated by recreational sports use lev el. ............................104 Table 5.2 . Analysis of covariance for four -year cumulative grade point average . ......................105 Table 5.3 . LSD post hoc comparisons of four -year grade point average among recreational sports use levels. ......................................................................................................105 Table 5.4 . Means, standard deviations (SD) , frequencies, and percentages of independent and covariate variables separated by five -year graduation. ....................................106 Table 5.5 . Means ± standard errors (SE) of cumulative grade point average by year in school. 106 Table 5.6 . Means ± standard errors (SE) of cumu lative grade point average by recreational sports use level. ............................................................................................................................107 !$#!LIST OF FIGURES Figure 1.1. Vi ncent TintoÕs theory of departure. .............................................................................4 Figure 5.1. Graphical depiction of the interaction between year in school and recreational sports use l evel on yearly grade point average. .......................................................107!!%!CHAPTER 1 INTRODUCTION Previous literature indicates a strong relationship among regular participation in physical activity (PA), current health, and chronic disease prevention .1 In 2008, the United States Department of Health and Human Services (DHHS) issued PA guidel ines indicating adults should participate in at least 150 minutes a week of moderate - or 75 minutes a week of vigorous -intensity aerobic PA, or an equivalent combination .2 Adults should also participate in ÒÉmuscle strengthening activities that are moderate or high intensity and involve all major muscle grou ps on two or more days a weekÉ Ó 2 Approximately 50% of adults do not meet aerobic guidelines and 76% do not meet strength training guidelines .3 Public health programming is necessary to increase the percentage of adults meeting PA guidelines. Previously, PA programming has targeted various populations including children, adolescents, college students, and adults .4,5 However, the most appropriate population to address has not been established. Researchers consistently find an inverse relationship between age and PA participation. 6 The decline in PA participation from childhood to adulthood largely occurs between 13 and 18 years of age. 7 The decline in PA participation results in fewer adults meeting PA guidelines, including those who are college aged (18 -24 years). In 2011, 56.8 percent of college aged adults met aerobic PA guidelines, 44.1 met strength, and 30.0 met both. 3 College is a critical time for development of healthy lifestyle behaviors , including regular participation in PA, with the hope of carryover into later adult life. 8,9 Therefore PA programmi ng directed towards college students has the potential to be of great public health benefit. !!&!University recreational s ports departments are designed to promote PA and provide ample opportunities for college students to develop healthy lifestyle behaviors .10 Recreational sports departments often include amenities and facilities such as, personal training, relaxation rooms, pools, gyms, courts, and fitness centers including weight lifting and cardio equipment. Previous literature indicates that over 85% of students nationwide participate in recreational sports on college campuses. 11 The frequency of participation in recrea tional sports during college is directly related to PA levels post -college. 12 Funding is necessary to ensure the upkeep of equipment, quality programming , and the continued promotion of PA through recreational sport s. To obtain adequate funding, u niversity departments must provide academic administrators with data that illustrate the departmental worth in student success. Par ticipation in recreational sports may have a direct relationship with PA participation, student health , and academic success. Further, participation in recreational sports may indirectly impact academic success through the mediator of student health. University administrators may view student health as beneficial yet ancillary to college studen t academic su ccess. Therefore, recreational sports departments should prove their contribution towards student success by identifying the relationships between participation in recreational sports and college student academic success. Previous studies investigating th e relationships between participation in recreational sports and college student academic success are grounded in two well -established higher education theories, Alexander AstinÕs Theory of Involvement 13 and Vincent TintoÕs Theory of Departure 14,15 . AstinÕs 13 theory suggests that the more a student is involved at an institution, the more likely the student is to learn and succeed. In his theory, involvement is defined as, ÒÉ the amount of physical and psychological energy that the student devotes to the academic experience.Ó 13 Astin13 notes that uninvolved students are more likely to drop out of college than !!'!involved students. Similarly, T intoÕs Theory of Departure 14,15 focuses on student integration. His theory suggests that student pre -entry attributes (family background, skills and abilities, prior schooling) influence student goals and his/her commitment to the institution prior to attending college, which in turn impact institutional experiences and academic/social integration while attending college. The level of academic/social integration in college influences student goals and institutional commitment during college and ultimately affects the decision to continue attending the institution (Figure 1 .1). Recreational sports researchers hypothesize that college students engage with recreational sports facilities and programming, thus directly impacting student involvement . Therefore, based on the theories of Astin 13 and Ti nto 14,15 , increased involvement in recreational sports may positively impact college student academi c success. It is important to note that while student involvement is critical to college student success, too much involvement can be detrimental. Astin 13 and TintoÕs 14,15 theories are purposefully vague to fit the structure of a variety of academic instit utions. However, researchers have also focused on specific factors , which occur during college, that influence college student success (e.g. hours spent working for pay 13, living on/off campus 13,16 , feeling a sense of belonging to the inst itution 13-17). Specifically , authors have linked effective time management to college student success .17 Students who are under/ over -involved and who do not effectively manage their time are less successful in college than students who identify time to effectively study and ways to become involved as part of a balanced lifestyle. 17 Therefore, the amount of student involvement in recreational sports may be critical to the potential impact of recreational sports on college student academic success , that is, a moderate amoun t of recreational sports participation may have a more positive impact on academic success than a low or high level of involvement in recreational sports . !!(!Figure 1. 1. Vinc ent TintoÕs t heory of departure. Source: Tinto, Leaving College: Rethinking the Causes and Cures of Student Attrition (2 nd ed.), p. 114.15 Historically college student academic success has been defined by student gr ade point average 18 (GPA), retention 19, and degree attainment 19. The majority of literature investigating the relationships among recreational sports participation and academic success focuses on student retention .20,21 Few investigators have studied the associations among partic ipation in recreational sports, GPA, and degree attainment; however, empirical evidence of these relationships trends towards a positive association. 20-23 Despite this trend, t he contribution of recreational sports programs to college student academic succ ess remains largely understudied and the literature available is somewhat contradictory . !!)!RECREATIONAL SPORTS AND ACADEMIC SUCCESS Recreational Sports and Grade Point Average College student GPA is an excellent predictor of college student retention 24-26 and de gree attainment 25. Students with higher GP As, particularly in the first year of college are more likely to be retained and graduate than students with lower GPAs .28-30 Researchers have investigated factors that may influence college student GPA, including race 28, gender 28,29, and high school GPA28-30. Identified variables that predict college student GPA are primarily unmodifiable (e.g. race and gender) and tend to explain little variance so there is a need to identify additional , particularly modifiable variables. Research investigating the relationship between participation in recreational sports and GPA is scarce. Four studies have assessed this relationship .20,22,31 ,32 Three found positive associations between recreational sports and GPA 20,22,31, and one found no association 32. Belch, Gebel, and Maas 20 utilized u niversity records to compare the number of times a student entered recreational sports fitness facilities and GPA. Kampf and Teske 22 also utilized u niversity records to assess the relationship between recreational sports fitness center use and college student GPA. Further Kampf and Teske 22 assessed the difference in GPA between 1) recreational sports student employees and non -employees, and 2) club sport participants and non -participants. Recently, we took a slightly different approach and assessed differences in GPA between recreational sports members and non -members. 31 Although all three studies 20,22,31 utilized large sample sizes, n=11,076, n=3,308, and n=4,843 respectively, and performed similar analyses, no investigative group accounted for potential confounding variables. In contrast, Frauman 32 found no difference in GPA between recreational sports users and non -users using a smaller sample size (n=385). !!*!Recreational Sports and Retention Retention is a very important measure to academic administrators. Thousands of dollars are lost for each stu dent who leaves an institution prior to degree attainment. 33 The risk of student attrition declines after the first year of college, thus retention research primarily focuses on the students' first to second year . One -year retention is usually defined by two consecutive fall semester enrollments. 34 The national retention rate for first time degree seeking students at four year public institutions is 79.9%. 35 Previous investigators have identified many variables that predict retention (age 25,36, sex24,36, race 24,25,36, residency 25, high school GPA 24,25,36,37, SAT/ACT scores 36, first semester college GPA 24,25,36, socioeconomic status 36, sense of belonging 16,17,38). Race, sex, socioeconomic status, and high school GPA are the most relia ble predictors of retention , however they are all unmodifiable, including high school GPA .19 Specifically, students who are white , female , of high socioeconomic status , and who earn high high school GPAs are the most likely to be retained. 19 Several investigators have posit ed that the most important factor may be student integration into university culture. 15-17,37,38 Recreational s ports offers a means for all students to become part of the u niversity and is a modifiable factor that may positively influence student retention. Students who participate in recreational sports facility activities can increase their self -esteem, develop social relationships, and develop communication and leadership skills. In addition to conventional programs (fitness classes, in tramural sports), Recreational s ports departments provide wellness programs, outdoor recreation, and a building space that appeals to a variety of clubs, and organizations, and facilitates a variety of physical activities. Recreational s ports departments are versatile and offer a sense of community for a diverse group of students. !!+!Researchers who have assessed the relationships between recreational sports participation and retention have primarily ut ilized self -report 11,21 ,39-42; two groups of researchers implemented the National Intramural Recreational Sports AssociationÕs (NIRSA) Quality and Importance of Recreational Sports Survey (QI RS). 21,39 The QIRS is a battery of questions that assess es the basic benefits of participation in recreational sports. Specifically, one question relates directly to retention and is listed below: ÒIn deciding to continue at [insert institution], how important to you was the availability of recreational facilities and programs?Ó Answer choices include d not important, somewhat important, important, and very important. Investigators report ed that 31 -40% of students report ed that the availability of recreational facilities and programs was important or very important in their decision to continue at the institution. 21,39 However, all studies assessing the relationship be tween recreational sports participation and retention, including those utilizing the QIRS survey, are limited by their cross -sectional designs, self -report, simplistic statistics , and many by non -random sampling. Four studies to date have utilized univers ity databases to address the self -report limitation. 20, 22,23,31 Results of this research positively enhance the findings of literature utilizing self-report . Studies utilizing self -report indica te that students believe the availability of recreational sports is important in their decision to continue at an ins titution 21,39, and research utilizing university databases indicate that recreational sports participation is positively related to retention 10,22 -23,31. However, studies utilizing university databases differ in study design, which reduces comparability o f results 20,22,23, 29, and are limited by simplistic statistics 20,31. Therefore, !!,!more research, that utilizes university databases, is needed to increase comparability, improve statistical analyses , and further understand how student beliefs relate to actual retention . Although previous investigations are widely considered as preliminary evidence, researchers should be cautious when interpreting self -reported, convenience sampled data that fails to control for possible confounding variables. There is a need to investigate the relationship between recreational sports participation and student retention by utilizing university data, considering many potential confounders in a systematic fashion, and by implementing more sophisticated analytic techniques. Recreational Sports and Bachelor 's Degree Attainment International graduation rates are calculated by dividing the Ônumber of graduates at each year of ageÕ by the Ôpopulation at that age Õ; the values are then averaged . Currently, the United StatesÕ international graduation rate is 37.7%, which is ranked 1 3th, among 30 countries participating in the Organization for Economi c Co -operation and Development .43 FinlandÕs graduation rate is first in the world at 62.6% followed by the Slovak Republic at 57.1%. 43 President Obama has pushed for an increase in college graduates. He has set a new goal for the country, indicating, ÒÉthat by 2020, America would once again have the highest proportion of college graduates in the world.Ó 44 Previous research indicates a number of variables that partially explain why college stud ents do or do not graduate ( e.g., AP credits 45, college major 27, first year GPA27, race 45-47, gender 45-47, socioeconomic status 45,49). However, the accuracy of each variable in predicting Bachelor 's degree attainment varies across studies .47,48 Scholars must continue searching for variables to better explain and predict Bachelor Õs degree attainment. !!-! Despite the importance placed on Bachelor 's degree attainment in the United States, only one published study 23 has assessed the relationship between participat ion in re creational sports and B achelor Õs degree attainment. Huesman et al. 23 utilized logistic regression to assess the relationship between semester participation in recreational sports (Fall semester of Freshman year) and five -year degree attainment (n=5,211). The authors included th e ratio of credits attempted to credits completed, CÕs received, DÕs received, course withdraw count, ACT/SAT score, remedial classes taken, remedial classes failed, athlete status, gender, race, socioeconomic status, and off -campus housing as covariates. Recreational sports use was entered in the model as a continuous variable. In a secondary analysis, recreational sports use was entered as a categorical variable (10 times vs. 25 times ). These groups were chosen based on the mean and standard deviation of recreational sports use. After controlling for all covariates , results suggested that utilizing the recreational sports fitness center 25 times in the first semester of freshmen year increased the likelihood of 5 -year graduation by 2% as compared to studen ts who used the fitness center 10 times in the first semester. However, it should be noted that recreational sports participation ranged form zero to 154 uses of the fitness centers in the first semester. The authorsÕ statistical analysis compared 10 to 25 uses and therefore did not address the extreme ends of recreational sports participation. There is a clear need for more research in this area. SUMMARY Previous literature assessing the relationships between recreational sports participation and 1) GPA, 2) retention, and 3) Bachelor 's degree attainment is scarce and the existent literature is limited by self -report and lacks specificity of recreational sports use (time and type) and !!%.!sophisticated study design and analysis. More specifically , previous literature includes very diverse definition s of recreational sports participation . Some authors focused on a single semester, while others included multiple semesters of participation. Few authors focused on a single component of recreational spor ts (i.e. intramural sports, group fitness, fitness center use, club sports, or recreational sports student employees), while most included all aspects of recreational sports in their Ôrecreational sports participationÕ definition. To ensure continued fundi ng for recreational sports, data are necessary to further investigate the relationships between participation in recreational sports and college student academic success . This dissertation will focus on modifiable factor s related to college student success . Specifically, we will address student membership and utilization of recreational sports fitness centers as factor s that may impact college student academic success defined by, GPA, retention, and B achelor 's degree attainment. The unique analysis of these relationships will strengthen previous literature and may help solidify the argument for a positive relationship between recreational sports and academic success. Findings may encourage college students to participate in regu lar PA to enhance stu dent success and may encourage u niversity administrators to allocate funding resources to recreational sports departments for the promotion of PA. RESEARCH AIMS Definition : Recreational Sports Participation : Recreational sports participation is defined in two ways throughout this dissertation. 1. In Specific Aim 1 recreational sports participation is defined by purchasing a recreational sports fitness center membership . This membership includes access to !!%%!two on campus weigh t rooms and cardio centers, which are not accessible to students who do not purchase a fitness center membership. Use of other recreational sports components (i.e. club sports, intramural sports, group fitness, and building use other than the f itness centers) was not considered. 2. In specific aims 2 -4 recreational sports participation is defined as use of recreational sports fitness centers. Use of on campus fitness centers is recorded via identification swipes. Use of other recreational sports co mponents (i.e. club sports, intramural sports, group fitness, and building use other than the fitness centers) was not considered. Specific Aims: Specific Aim 1 : To investigate the differences in college student semester grade point average (GPA), credits completed, class standing, and retention between students who purchased a recreational sports fitness center membership their freshmen year and students who did not. Hypothesis 1: Students who purchased a recreational sports fitness center membership in their first semester will earn higher cumulative GPAs, complete more credits, and are more likely to be retained a second and third year than students who di d not purchase a recreational sports fitness membership in their first semester. Specific Aim 2 : To investigate the relationship between recreational sports fitness center use and college student one - and two -year retention. !!%&!Hypothesis 2: There will be a direct relationship between use of recreational sports fitness centers and one - and two -year retention. Specific Aim 3 : To investigate the relationships between recreational sports fitness center use over four years and college student academic success defined by cumulative GPA and BachelorÕs degree attainment . Hypothesis 3.1: There will be an inverted U shape relationship between recreational sports fitness center use and college student cumulative GPA. Students using the recreational sports fitness center the least and most often will earn lower cumulative GPAs than those who use the facilities a moderate amount. Hypothesis 3.2: There will be an inverted U relationship between recreational sports fitness cent er use and college student five -year Bac helorÕs degree attainment . Students using the recreational sports fitness center the least and most often will be less likely to complete their degree s in five years compared to students who use recreational sports fitness centers a moderate amount. Specific Aim 4 : To investigate the impact of year in school on the relationship between recreational sports use and GPA. !!%'!Hypothesis 4.1: There will be an interaction between year in school and recreational sports use on yearly GPA. First year students will experience the greatest effect of recreational sports use on GPA. ORGANIZATION OF THE DISSERTATION This dissertation is organized into six chapters. Chapter one includes the introduction, specific aims, and hypotheses. Chapter two is a review of the literature related to th e specific aims. Chapter three is organized as a manuscript ( abstract, introduction, background, me thods, results, discussion , and references ) published previously in the Recreat ional Sports Journal, and includes an addendum to the published manuscript. Chapter s four and five are organized as manuscripts (introduction, methods , results, discussion , and references). Chapter three address es spec ific aim one, chapter four address es specific aim two, and chapter five address es specific aims three and four. All findings are summarized in chapter six. ! !!!!%(!REFERENCES !!%)!REFERENCES 1. Li J, Siegrist J. Physical activity and risk of cardiovascular disease Ña meta -analysis of prospective cohort studies. International journal of environmental research and public health. 2012;9(2):391 -407. 2. U.S. Department of Health and Human Services. 2008 physical activit y guidelines for Americans. Be active, healthy, and happy. Washington. DC: U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion, Office of the Assistant Secretary for Health, Office of the Secretary; 2008. 3. Centers for Disease Control and Prevention (CDC). Adult participation in aerobic and muscle -strengthening physical activities --United States, 2011. MMWR. Morbidity and mortality weekly report. 20 13; 62(17):326 -330. 4. Dishman R, Buckworth J. Increasing physical activity: a quantitative synthesis. Medicine and science in sports and exercise. 1996;28(6):706 -719. 5. Kahn EB, Ramsey LT, Brownson RC, et al. The effectiveness of interventions to increase physical activity: A systematic review . American journal of preventive medicine. 2002;22(4):73 -107. 6. Caspersen CJ, Pereira MA, Curran KM. Changes in physical activity patterns in the United States, by sex and cross -sectional age. Medicine & Sci ence in Sports & Exercise. 2000(32):1601 -1609. 7. Sallis JF. Age -related decline in physical activity: a synthesis of human and animal studies. Medicine and science in sports and exercise. 2000;32(9):1598 -1600. 8. Pascarella ET. How college affects stude nts: Ten directions for future research. Journal of College Student Development. 2006;47(5):508 -520. 9. Arterberry CM. An epidemic hits campus: The challenge of obesity for future students and campus recreation. Recreational Sports Journal. 2004;28(2):19 -29. 10. Haderlie BM. Influences of Campus Recreation Programs and Facilities on Student Recruitment and Retention. Recreational Sports Journal. 1987:4. 11. NIRSA/NASPA. (2010). NIRSA/NASPA consortium campus recreation impact study. descriptive summary re port. Retrieved from : http://www.nirsa.org/docs/Discover/ Research/Campus_Rec_Impact_Study.PDF !!%*!12. Forrester S, Ross CM, Hall S, Geary C. Using past campus recreational sports participation to explain current physical activity levels of alumni. Recreational Sports Journal . 2007;31:83 -94. 13. Astin AW. Student involvement: A developmental theory for higher education. Journal of college student personnel. 1984;25(4):297 -308. 14. Tinto V. Leaving college: Rethinking the causes and cures of student attrition. Chicago: University of Chicago Press ; 1987. 15. Tinto V. Leaving College: Rethinking the Causes and Cures of Student Attrition. (2nd Ed.). Chicago: University of Chicago Press; 1993. 16. Peltier GL, Laden R, Matranga M. Student persistence in college: A review of research. Journal of college student retention. 2014; 1(4):357 -375. 17. Britton BK, Tesser A. Effects of time -management practices on college grades. Journal of educational psychology. 1991; 83(3):405. 18. May MA. Predicting Acad emic Success. Journal of educational psychology. 1923;14(7):429. 19. Renn KA, Reason RD. College Students in the United States: Characteristics, Experiences, and Outcomes: Characteristics, Experiences, and Outcomes. San Francisco: Jossey -Bass; 2012. 20. Belch HA, Gebel M, Maas GM. Relationship between student recreation complex use, academic performance, and persistence of first -time freshmen. Journal of Student Affairs Research and Practice. 2002;38(2):220 -234. 21. Lindsey R, Sessoms E. Assessment of a campus recreation program on student recruitment, retention, and frequency of participation across certain demographic variables. Recreational Sports Journal. 2006;30(1):30 -39. 22. Kampf S, Teske EJ. Collegiate Recreation Participation and Retention. Recreational Sports Journal. 2013;37(2):85 -96. 23. Huesman R, Brown AK, Lee G, Kellogg JP, Radcliffe PM. Gym Bags and Mortarboards: Is Use of Campus Recreation Facilities Related to Student Success? Journal of Student Affairs Research and Practice. 2009;46(1) :50 -71. 24. Reason RD. Student variables that predict retention: Recent research and new developments. Journal of Student Affairs Research and Practice. 2009;46(3):850 -869. 25. Murtaugh PA, Burns LD, Schuster J. Predicting the retention of university students. Research in Higher Education. 1999;40(3):355 -371. !!%+! 26. DesJardins SL, McCall BP, Ahlburg DA, Moye MJ. Adding a timing light to the Òtool boxÓ. Research in Higher Education. 2002;43(1):83 -114. 27. DesJardins SL, Kim D -O, Rzonca CS. A nested analysis of factors affecting bachelor's degree completion. Journal of College Student Retention: Research, Theory and Practice. 2003;4(4):407 -435. 28. Ting S -MR, Robinson TL. First -year academic success: A prediction combining cognitive and psychosocial variables for Caucasian and African American students. Journal of College Student Development. 1998. 29. DeBerard MS, Spielmans G, Julka D. Predictors of academic achievement and retention among college freshmen: A longitudinal study. College student journal. 2004;38(1):66 -80. 30. Cohn E, Cohn S, Balch DC, Bradley J. Determinants of undergraduate GPAs: SAT scores, high -school GPA and high -school rank. Economics of Education Review. 2004;23(6):577 -586. 31. Danbert SJ, Pivarnik JM, McNeil RN, Washington IJ. Academic success and retention: the role of recreational sports fitness facilities. Recreational Sports Journal. 2014;38(1):14 -22. 32. Frauman E. Differences between participants and non -participants of campus recreation offerings across demographic variables and perceptions of the college experience. Recreational Sports Journal. 2005;29(2):156 -165. 33. Raisman NA. The Cost of College Attrition at Four -Year Colleges & Universities. Educational Policy Insitute;2013. 34. Integrated Post Secondary Education Data System Glossary [Internet]. Washington (DC): National Center for Education Statistics . [cited 2015 april 20]. Available from : http://nces.ed.gov/ipeds/glossary/ 35. Institutional retention and graudation rates for undergraduate students [Internet]. Washington (DC): National Center for Education Statistics. [cited 2015 april 20]. Available from: https: //nces.ed.gov/programs/coe/indicator_cva.asp 36. Ishler J, Upcraft ML. The keys to first -year student persistence. Challenging and supporting the first -year student: A handbook for improving the first year of college. 2005:27 -46. 37. Astin AW. What matters in college?: Four critical years revisited. San Francisco (CA):Jossey -Bass ; 1993. !!%,!38. Astin AW. Preventing students from dropping out. San Francisco (CA):Jossey -Bass ; 1975. 39. Bradley J, Phillipi, RH., Bryant, JA. Minorities benefit from their association with campus recreation programs. NIRSA Journal. 1992;16(3):46 -50. 40. Henchy A. The influence of campus recreation beyond the gym. Recreational Sports Journal. 2011;35(2):174 -181. 41. Hall D. Participation in a campus recreation program and its effect on retention. Recreational Sports Journal. 2006;30(1):40 -45. 42. Mallinckrodt B, Sedlacek WE. Student Retention and the Use of Campus Facilities by Race. NASPA journal. 1987;24( 3):28 -32. 43. Youth Indicators 2011 [Internet]. Washing (DC): National Center for Education Statistics [cited 2015 april 20} Available from: https://nces.ed.gov/pubs2012/2012026/tables/table_23.asp 44. Obama BH. Remarks by the president on the American g raduation initiative. Speech presented at Macomb Community College, Warren, MI. 2009. 45. Adelman C. The Toolbox Revisited: Paths to Degree Completion From High School Through College. US Department of Education. 2006. 46. DeAngelo L, Franke R, Hurtado S, Pryor JH, Tran S. Completing college: Assessing graduation rates at four -year institutions. Higher Education Research Institute, Graduation School of Education & Information Studies, University of California, Los Angeles; 2 011. 47. Astin AW, Oseguera L. Degree attainment rates at American colleges and universities: Revised edition. Higher Education Research Institution, University of California Los Angeles. 2005. 48. Astin AW. Degree Attainment Rates at American Colleges a nd Universities: Effects of Race, Gender, and Institutional Type. ERIC; 1996. 49. Walpole M. Socioeconomic status and college: How SES affects college experiences and outcomes. The review of higher education. 2003;27(1):45 -73. !!!!%-!CHAPTER 2 REVIEW OF THE LITERATURE INTRODUCTION The United States (US) Department of Health and Human Services (DHHS) recommends that all adults participate in at least 150 minutes of moderate - or 75 minutes of vigorous aerobic physical activity (PA) per week, or an equivalent combination of moderate and vigorous intensity aerobic activity .1 In addition, the DHHS recommends that all adults participate in strengt h training activities that are at least moderate intensity and include all major muscle groups on two or more days per week.1 Research indicates that few adults meet aerobic and strength training recommendations .2 The 2012 National Health Interview Survey (NHIS) revealed that 50.0% of adults met DHHS aerobic PA guidelines, 23.9% met DHHS strength training guidelines, and 20.6% met both, which were marginal increases from t he 2008 NHIS. 2 To bring attention to the importance of regular PA, Healthy People 2020 2 personnel defined objectives related to decreasing sedentary activity and increasing PA participation . The second Healthy People 2020 2 PA objective relates directly to the DHHS PA recommendations. PA-2 Increase the proportion of adults who meet current Federal PA guidelines for aerobic PA and for muscle -strengthening activities. 2 Effective public health programming is necessary to increase the percentage of adults meeting PA guidelines, which will help meet the Healthy People 2020 objective. However, the most appropriate population to focus on has not been established. PA participation declines !!&.!throughout the lifespan. 3 Researcher s have identified the sharpest decline between the ages of 13 and 18 years. 4 In an effort to minimiz e the sharp decline of PA in the teenage year s, researchers have developed PA interventions specific for school aged children. Dobbins, Husson, DeCorb, & LaRocca 5 reviewed 44 studies assessing the effect of PA interventions in children and adolescents aged six to 18 years. They found that school based PA interventions led to increased participation in PA during scho ol hours (Odds Ratio=2. 74, 95% Confidence Interval=2.01 -3.75).5 While PA interventions are often effective in children and adolescents, the relationship between child PA and adult PA participation is weak to moderate at best .6 Telama et al., 6 utilized a questionnaire to measure the relationship between child/adolescent PA and adult PA. Subjects (n=2309) were 3, 6, 9, 12, 15, or 18 years old at ini tial assessment, secondary assessment took place 21 years later . Spearman correlation s between initial PA assessment and secondary PA assessment varied from 0.33 to 0.44 in males and 0.14 to 0.26 in females, indicating a weak to moderate relationship between PA in childhood and PA in adult life. Telema et al., 6 assessed the relationship between child/adolescent PA and adult PA, but did not include college student PA (ages 18 -24 years). Researchers have categori zed young adults as Òemerging adults.Ó 7 Positive PA behaviors may emerge during college and carry over into adult life. The relationship between PA in college and adult life is largely unde rstudied. It is well known that college impacts student behavior development .8,9 For decades researchers have investigated behavior change in college students. 8,9 Many theories have been developed that describe the impact of college on various aspects of student development such as psychosocial development 10, identi ty development 11-13, and intellectual development 14,15 . Despite the effects of college on various components of student development, few studies have !!&%!investigated PA intervention during college as a means of developing healthy lifestyle behaviors such as regul ar PA participation. Plotnikoff et al., 16 recently published the first review of interventions that target PA, nutrition, and healthy weight in college students. The authors identified studies that 1) were written in English, 2) included students attending post secondary institutions, 3) included interventions aimed at improving PA and/or dietary intake and/or we ight, and 4) were quantitative. The authors selected 41 studies. Twenty -nine of the 41 studies examined PA, 11 assessed PA exclusively, and 18 in combination with other health outcomes. Eighteen of the 29 studies examining PA showed significant improvemen ts in PA or fitness behaviors from pre - to post -intervention. Improvements in PA and fitness behaviors included increases in total minutes of PA, number of day s participating in PA, MET levels, and decreases in exercise barriers. Plotnikoff et al., 16 fol lowed their systematic review with three meta -analyses. The meta -analyses addressed the impact of intervention on 1) total PA participation 17-21, 2) vigorous PA participation 17,21,22 -24, and 3) moderate PA participation 17,21,22 -24. The results indicated no differences in total PA or vigorous PA participation between intervention and control groups. 16 However, the results of the meta -analysis did indicate that intervention groups participated in significantly more moderate PA post intervention than control groups. 16 Together the results of the three meta-analyses indicate that previous PA interventions in college students increased moderate PA, but had no effect on vigorous or total PA participation during c ollege . More research is needed in this area to identify effective means of increasing all types of PA participation in college students. University recreational s ports departments are vehicle s for PA promotion in college students. Recreational s ports departments offer students a diverse array of opportunities to !!&&!participate in PA, such as pools, courts, tracks, intramural sports, group fitness classes, and fitness centers including strength and cardio equipment. NIRSA: Leaders in Recreation state that approximately 85% of college students who live on campus utilize recreational sports in some capacity .25 Participation in recreational s ports in college may c arry over into post -college life. Forrester et al., 26 investigated the relationship between participation in campus recrea tional sports and post -college PA. The authors utilized multiple regression and a convenience sample of health, physical education, and recreation alumni who graduated from a single university between May 2001 and May 2005 (n=310). The independent variabl e was recreational sports use in college and the dependent variable was current PA participation, both assessed with a survey. Forrester et al., 26 noted that the frequency of recreational sports use in college was positively related to post -college self -reported PA participation ( t=2.51 , p=0.013). PA interventions developed through recreational sports have the opportu nity to impact the majority of college students in the nation. However, the successes of such interventions depend on the funding necessary to ensure the upkeep of equipment, quality programming , and the continued promotion of PA through recreational sport s. University administrators may consider the development of healthy lifestyles as important, but ancillary to college student academic success. To obtain adequate funding, recreational sports departments must provide academic administrators with data that illustrate the departmental worth in student academic success. !!&'!Foundational Theory Historically, researchers have fixated on student recruitment and student retention as definitions of academic success when analyzing the relationship between recreational sports participation and academic success. More recently investigations have focused on the relationship between recreational sports participation and 1) college student grade point average (GPA) and 2) Bachelor Õs degree attainment. Investigations assessing the relationship between recreational sports and college student academic s uccess are grounded in two well -establish ed theories of college student success, AstinÕs 27 Theory of Involvement and TintoÕs 28,29 Theory of Departure. AstinÕs Theory of Involveme nt. AstinÕs 27 Theory of I nvolvement was established in 1984 and is based on student involvement as a behavior. Astin27 stated, ÒIt is not so m uch what the individual thinks or feels, but what the individual does, how he or she behaves, that define and identifies involvement.Ó Astin27 further defined student involvement by the following phras es: attach oneself to, commit oneself to, devote oneself to, engage in, go in for, incline toward, join in, partake of, participate in, plunge into, show enthusiasm for, tackle, take a fancy to, take an interest in, take on, take part in, take to, take up, undertake. He believed student success was not defined entirely by academics, but also by the extent to which the student engaged with the school. By AstinÕs 27 philosophy, student s are more likely to be retained if they feel a sense of belonging at the school, if they identify with other students, and/ or identify with a particular place at the institution. !!&(!TintoÕs Theory of Departure AstinÕs 27 Theory of Involvement identified multiple variables related to student success and retention, but did not uncover the magnitude of the relationship. Tinto 28,29 addressed the ÒÉreasons for, magnitude of, and mediating aspects of retention .Ó30 Tinto 28,29 agreed with AstinÕs 27 findings, that student involvement was a large indicator for student retention; however, he further broke AstinÕs 27 theory down into academic and social involvement. Tinto 28,29 stated that student retention depended on the studentsÕ integration into the academic and social communiti es of the institution. Some departures from the institution are considered involuntary, for example the student did not meet the required GPA. Tinto 28,29 argues departures that are voluntary often are due to student perception of a problem related to academic or social belonging. Institutions have the ability to assist students in feeling a sense of belo nging to both the academic and social constructs. Tinto 29 states, ÒThe point of retention efforts is not merely that individuals be kept in college. Education, the social and intellectual development of individuals, rather than just their continued presence on campus should b e the goal of retention efforts Ó (p. 145). Astin27 and TintoÕ s28,29 models are widely accepted in student success research. They are considered the foundation of research assessing the relationship between recreational sports and college student academic success. Future research assessing this relationship should continue to consider variables identified by Astin 27 and Tinto 28,29 , such as, precollege attributes (sex, race), precollege experience (high school GPA), university commitment, and academic and social integration while in college. !!&)!RECREATIONAL SPORTS AND ACADEMIC SUCCESS Of the few studies investigating the relationship between recreational sports and college student academic success, most utilize self -report for both recreational sports use and academic success variables. Additionally, few utilize proper statistical analy sis to control for confounding variables. Despite these limitations, data trend towards a positive relationship between recreational sports participation and 1) GPA, 2) retention, and 3) Bachelor's degree attainment. However, the contribution of recreation al sports programs to college student academic success remains largely understudied and the literature available is somewhat contradictory. Grade Point Average Previous research indicates a variety of factors that influence college student GPA including, but not limited to: high school GPA 31-33, gender 31,32 , race 31, ACT and SAT scores 32, first generation status 34, socioeconomic status (SES) 35, and smoking 32. High school GPA is considered the best predict or of college academic success and has been shown to account for 19% of the variance in college student GPA .27,36 However, investigators have examined a variety of models to predict college student GPA. At best, these models account for 50% of the variance in GPA . Future research should focus on additional variables to better predict college student GPA. The authors of four studies to date have investigated the relationship between participation in recreational sports and college student GPA .37-40 Three studies agree on a positive relationship 37-39 and one study found no association 40. Researchers utilized different definitions of recreational sports use and measured college student GPA at varying time points. !!&*!No researchers controlled for the effects of confounding variables and all utilized univariate statistical methods ( StudentÕs t -tests or Chi -square ). Belch et al, 38 were the first to inve stigate the relationship between participation in recreational sports and college student GPA. The sample included all first time freshmen in the Fall semesters of 1993, 1994, and 1995 (n=11,076). Students who swiped their student ID to enter the recreatio nal sports facility in their first semester at the institution (Fall semester 1993, 1994, or 1995) were considered recreational sports users (73% of the sample) . Academic information was obtained via university Registrar. The authors compared recreational sports user GPA to non -user GPA. StudentÕs t -tests revealed that recreational sports users had lower mean high school GPAs, SAT scores, and ACT scores, but earned higher first semester GPAs (Mean=2.53) compared to non -users (mean=2.44, p<.001). However, no differences were found in first year GPA between users and non -users. Strengths of this study included the objective measurement of recreational sports use (ID swipe) and obtaining academic information from the university R egistrar. However, the study did not control for confounding variables and did not address the homogenous sample by account ing for level of recreational sports use. Kampf and Teske 39 addressed one of the limitations of th e Belch et al. 38 study by accounting for the number of times a student entered the recreational sports building. The study sample included first -time full -time students (n=3,809). Re searchers obtained data from a university database that included club sport enrollment (yes/no), campus recreation student employment sta tus (employed/not employed), student recreation center entry counts measured via ID swipes into the recreational sports building, and latest term cumulative GPA. Latest term cumulative GPA was the last recorded GPA after the Fall, Spring, or Summer semeste r during the first academic year. Researchers utilized last recorded GPA because they did not know if or !!&+!when a student dropped out of the institution. StudentÕs t -tests revealed no differences in latest term cumulative GPA between club sport participants and non -participants. Correlation analysis revealed a weak to moderate correlation and a direct relationship between student recreation center entry counts and latest term cumulative college GPA (r=.204, p<0.001). In addition, campus recreation student employees earned higher latest term cumulative GPAs (3.05±0.72) than non -employees (2.56±0.99, p=0.021). These findings agree with and expand those of Belch et al., 38 and suggest that the type of recreational sports participation (i.e. , club sports participation, student recreational center entry counts, recreational sports employment) may mediate the relationship between recreational sports participation and college student GPA. Strengths of this study include the large sample size, various measures of recreational sports participation, and the obtainment of data fro m a u niversity database. However, latest term cumulative GPA is not a valid measure of student success as time of attendance at the University is not consistent among subjects . Additional ly, the authors did not report a range or mean for student recreation center entry counts, which inhibits interpretation. Finally , similar to Belch et al., 38 this study did not control for confounding variables. Danbert et al. 37 was the most recent group to assess the relationship between recreational sports and college student GPA (see Chapter 3 for complete manuscript). The study sample included all first time degree seeking f reshmen students who graduated fr om high school in the Spring 2010, who in the Fall 2010 were in their first semester at the university, and who had no prior college class experience (n=4,843). All data were gathered from a u niversity database. Most institutions utilize a fee -based model to fund their recreational s ports department. In a fee -based model, each student is assessed a recreational sports fee as pa rt of enrollment expenses. The u niversity utilized in the Danbert et al. ,37 study employs a membership -model. Access to !!&,!some recreational sports facilities is included in tuition costs; however, a membership fee is charged to students who wish to use the recreational sports fitness centers , which house all cardio and weight training equipment. Recreational sports use was de fined by fitness center member status (member/non -member) , which included approximately twenty -five percent of the sample (n=1,138). Cumulative GPA was obtained from the Registrar and assessed after four completed semesters. StudentÕs t -tests indicated th at recreational sports users earned higher cumulative GPAs (3.13±0.52) after four semesters than non -users (3.00±0.52, p<0.0001). Danbert et al.Õs 37 finding of a positive relationship between recreational sports membership and college student GPA support the findings of Belch et al., 38 and Kampf & Teske 39. Similar to Belch et al., (2001) and Kampf & Teske, 39 strengths of the Danbert et al., 37 study include a large sample, and the collection of data from a u niversity database. Further, the impact of fitness center membership was assessed as opposed to global recreational sports use. A major limita tion is the lack of potential confounding variable inclusion. In addition, level of recreational sports use was not accounted for. It is possible that some users had a fitness center membership, but never actually utilized the facilities. In contrast to the previously discussed studies, Frauman 40 did not find a relationship between recreational sports and college student GPA. The sample included 389 students enrolled in twenty -five different classes at the institution. Study participants completed a one -time survey and reported demographic information, recreational sports user status, and GPA. Students who reported participating or using programs and service s associated with the recreational s ports department were considered recreational sports users (80.7% of the sample). GPA was reported by choosing one of the following categories: first semester student, 1.99 or below, 2.0 -2.49, 2.5-2.99, 3.0-3.49, or 3.5 or above. Chi -square analyses indicated no relationship between !!&-!recreational sports user status and GPA. This finding contradicts that of all other studies. FraumanÕs 40 study is limited by the relatively small, homogenous sample and self -report for all variables. Further, Frauman 40 did not account for confounding variables or level of recreational sports use . FraumanÕs 40 negative find ings might be attributable to the small, homogeneous sample. In summary, most authors have shown a positive relationship between a dichotomized measure of recreational sports use and college student GPA. While each of the four studies contributed valuable information to assessment of the relationship between recreational sports and college student GPA, all had similar limitations that must be addressed. Limitations include a lack of confounding variables and inconsistencies in the defini tions of recreation al sports participation . Further most studies had a homogenous sample due to most subjects participating in at least one aspect of recreational sports. Future research should identify, measure, and assess the impact of confounding variables through more sophisticated statistical analyses. Investigators should measure college student GPA at similar time -points to increase inter -study comparability and should increase the variability in the independent variable by identifying levels and/or type of recreatio nal sports use. Finally, researchers should continue to investigate the relationship between recreational sports and GPA in specific entities of recreational sports such as fitness centers, intramurals, group fitness, and student employment. Future researc h will help recreational sports practitioners understand the most desired level and type of participation for maximal positive impact on college student GPA. !!'.!Retention The National Center for Education Statistics defines retention as consecutive Fall-to-Fall enrollment .41 A student who enrolls full time in a Fall semester and again enrolls full or part time in the following Fall seme ster is classified as meeting one -year retention. Spring enrollment is not considered in Fall -to-Fall retention statistics. The risk of student attrition declines after the first year of college, thus retention research focuses primarily on the students' f irst and second years. The national one -year retention average at all four -year institutions in the United States is 78.8%.42 Thousands of dollars are lost for each student who leaves an institution prior to graduation. 43 To minimize funding loss, academic administrators are constantly working to increase college student retention rates. Previous investigators have iden tified many variables that predict retention (age 30,44 , sex 30,45 , race 30,44,45 , residency 44, high school GPA 30,44 -46, SAT/ACT scores 30, first quarter college GPA 30,44,45 , SES 30, sense of belonging 28,29,47 ). These variables consist primarily of precollege attributes, or factors that exist prior to the student entering college (i.e., gender, race, high school GPA). TheoristÕs Astin 27 and Tinto 28,29 would argue that college student retention is more complex than prec ollege attributes. Instead, they believe college student retention is affected by a studentÕs level of involvement in social and academic settings while attending college. Research is needed to understand various types of involvement at the university leve l; of particular interest is the relationship between the level of involvement in recreational sports participation and college student retention. Researchers have typically used survey data to investigate the relationship between recreational sports and college student retention. 25,48 -51 Most investigators utilize self -report for both the independent variable of recreational sports participation and the outcome variable of retention. 25,48 -51 Four groups of researchers took the investigation a step further and have !!'%!assessed this relationship with data from Registrars. 37-39,52 Only two studies assessed the impact of confounding variables on the relationship between recreational sports and college student retention. 39,52 To date, data from all but one 50 study indicate a positive relationship between recreational sports and college student retention. Bradley, Phillipi, & Bryant 49 and Lindsey and Sessoms 48 utilized the National Intramural Quality and Importance of Recreational Sports Survey (QIRS). The QIRS is a battery of questions that assess es the basic benefits of participation in r ecreational sports. Specifically, one question , rated on a 5 -pt Likert scale, relates directly to retention and is listed below: ÒIn deciding to continue at [insert institution], how important to you was the availability of recreational facilities and p rograms?Ó Approximately 31 -40% of students reported that recreational sports was either important or very important in deciding to continue at the institution. 48,49 Lindsey and Se ssoms 48 (n=244) further stated that the level of importance of the availability of recreational facilities and programing in deciding to continue at the institution differed by class standing. However, the authors did not indicate which class (freshman, sophomore, junior, senior) indicated the most and/or least importance. Bradley, Phillipi & Bryant 49 (n=2,000) reported racial differences in level of importance. African Americans were more likely to report that recreational sports were important or very important in their decision to remain at the university than their Caucasian counterpar ts. Both studies found positive relationships between recreational sports and college student retention. However, conclusions from the Lindsey and Sessoms 48 study should be drawn !!'&!cautiously due to their convenience sampling of students from a single Department of Physical Education and Health. Additionally, both studies are limited by self -report. Henchy 51 (n=237) used a variation of the QIRS, the National Intramural Recreational Sports Association Recreational Impact Survey. Similar to the previously discussed literature, she found that 31% of students indicated that recreationa l sports had a strong or moderate influence on their decision to remain at the institution. Beginning in 2006, NIRSA Leaders in Recreation partnered with NASPA: Student Affairs Professional s in Higher Education to assess the impact of recreational sports on various aspects of college student development. The NIRSA/NASPA consortium is utilized by recreational sports departments all over the United States. In 2010, NIRSA: Leaders in R ecreation released a report on the NIRSA/NASPA consortium data. 25 The authors indicated that 45% of specifically junior and senior students reported recreational sports being moderately important to their decision to remain at the institution, marginally higher than the 31 -40% of all students reported in previously discussed literature. Like the studies utilizing the QIRS, results from Henchy 51, and the NIRSA/NASPA consortium 25 are limited by self -report. It is not possible to ascertain to what degree students utilized recreational sports or if the students were act ually retained. Instead, results show studentÕs beliefs in the importance of recreational sports on retention. Mallicrodt & Sedlacek 50 utilized their own 24 -item questionnaire to assess the relationship betw een recreational sports use and college student retention. A random sample , stratified by race, of 207 undergraduates completed the survey in their second semester at the institution. Eighty percent of the sam ple was retained the following F all semester. T he authors found use of recreational sports facilities did not predict retention for all students, but did positively predict retention in Black students. Mallicrodt and SedlacekÕs 50 findings add to !!''!previous literature by assessing the impact of recreational sports use on student retention by race. However the authorsÕ null finding in all students contradicts the findings of previous literature. The contradiction may be due to a difference in the measurement of retention. M ost researchers who used self -report assessed retention as a perception of the importance of recreational sports in the decision to remain at the institution. Mallicrodt & Sedlacek 50 utilized data from the university Registrar to measure retention. Further , the contradict ory findings may be due to Mallicrodt & SedlacekÕs 50 small sample . To address the issue of self -report, investigators of two studies 37,38 utilized data from a university Registrar to assess the relationship between re creational sports participation and college student retention. Belch et al., 38 identified recreational sports users by recording identification card swipes into the recreational sports centers . Students who entered the facility in their first semester at the institution were considered users; those who did not were co nsidered non-users. Danbert et al., 37 separated recreational sports users and n on-users via membership. Students who purchased a membership to t he recreational sports fitness center in their first semester at the university were considered users, and those who did not purchase a membe rship were considered non -users . Together, the stu dies indicate a positive relationship with a 2 -7% difference in one -year retention rates between recreational sports users and non -users. However, neither study accounted for level of use of the recreational sports facilities. It is possible that a stude nt entered the facility for reasons other than exercise and it is possible that a student purchased a fitness center membership, but never actually utilized the facility. Additionally, neither study accounted for possible confounding variables. Recently, researchers have utilized logistic regression to account for possible confounding variables in the relationship between recreational sports participation and college !!'(!student retention. Huesman et al., 52 utilized a sample of 5,211 first time degree seeking freshmen. Recreational sports use was assessed through identification swipes into the recreational sports facilities and analyzed on a continuous scale . The authors included 1 4 variables (ratio of credits attempted to credits completed, CÕs received, DÕs received, course withdraw count, ACT/SAT score, remedial classes taken, remedial classes failed, athlete status, gender, race, tuition -reciprocity, tuition non -reciprocity, socioeconomic st atus, and off -campus housing) in their model and found a positive relationship between recreational sports participation and college student retention ( aOR=1.009; 95% CI=1.003, 1.015, p<0.01). Although statistically significant, the adjusted odds ratio for recreational sports use was small and suggests low practical significance . Kampf et al., 39 assessed the impact of recreational sports use on student retention with a sample of 3,809 first ti me degree seeking freshmen. The authors used logistic regression to assess the impact of three entities of recreational sports use (club sports, recreational sports student employment, and recreation center use measured through identification swipes into t he recreational sports center) on student retention. Kampf et al., 39 assessed a range of confounding variables in their model (club sports participant, Fall semester rec enter entry count, A CT score, high s chool GPA, latest term cumulative college GPA, sex, first generation st atus, citizenship, White race, Native American race, Asian race, Black/African American race, Hawaiian/Pacific Island race, and non specified race ). Students who participated in club sports were 2. 4 times more likely to be retained than students who did not participate in club sports. Additionally, each one -unit increase in recreational center use increased the odds of being retained by 1. 4 times. Recreational s ports student employment was not included in the logistic regression because 100% of the recreational sports employees were retained. After controlling for confounding variables, the findings of both Huesman et al., 52 and !!')!Kampf and Teske 39 indicate a positive relationship between recreational sports participation and college student retention. Huesman et al., 52 and Kampf &Teske 39 extended previous literature by accounting fo r confounding variables. However, study limitations include the assessment of recreational sports use in the first semester only and possible multicollinearity within logistic regression models. Variables that may have contributed to multicollinearity include the academic variables (Cs received, Ds received, withdraw count, remedial class taken, and remedial class failed), and the socioeconomic variables (tuition reciprocity, tuition non -reciprocity, and Pell grant eligibility). Future investigators sh ould assess recreational sports use on a yearly basis and address multicollinearity when building models. Future researchers should aim to replicate studies by utilizing the same independent, dependent, and confounding variables to increase comparability between studies. Replication of research may help to narrow the gap in effect sizes that were reported by previous literature. Each study discussed concluded a positive relationship between recreational sports participation and college student retention, with the exception of Mallinckrodt and Sedlacek 50. The null findings of Mallinckrodt and Sedlacek 50 are likely attributable to the authorÕs relatively small, homogenous sample. Despite the null findings of Mallinckrodt and Sedlacek 50, most research shows that recreational sports participation is in some w ay related to student retention. However , drawing conclusions from previous literature should be done ca utiously , due to study design limitations and limited data analysis. For example, j ust two of the cited studies assessed recreational sports use as a continuou s variable 39,52 , most relied on self -report of both independent and dependent variables 25,48 -51, and few studies included confounders 39,52 . Therefore , future research that better defines and assesses both independent and dependent variables is !!'*!needed to further substantiate the findings and to uncover the level of recreational sports associated with student retention. !!BachelorÕs Degree Attainment Bachelor 's Degree attainment is the primary goal of college students. Students who complete a BachelorÕs degree have high job satisfaction and earn an income two times that of high school graduates and six times that of high school dropout s.53,54 The unemployment rate of college graduates is much lower than that of high school graduates. 55 Additionally , higher income is associated with a variety of health related outcomes, such as increased life expectancy, improved quality of life, and lower risk of disease .56 On an individual level, college graduation is important for student overall wellbeing in adult life; on a national level , college graduation is important for colleges and universities as it is considered in institution rankings; on a global perspective , college graduation is important when comparing countries. The United StatesÕ graduation rate is currently ranked 13 th among 30 countries participating in the Organization for Economic Co -operation and Development. 57 Pre sident Obama advocates for an increase in United States Õ college graduation rates. His goal is ÒÉthat by 2020, America would once again have the highest proportion of college graduates in the world.Ó 58 Despite the economic reasons to graduate college, the university driven need to achieve high college graduation rates, and President ObamaÕs goals, few students who begin college act ually complete a BachelorÕs degree in timely fashion. Of all first time degree -seeking students attending four -year degree granting institutions in 2006, 39% completed a degree in four years, 54.9% in five years, and 58.7% in six years. 59 Research is nece ssary to understand why college students do and do not graduate and how to increase graduation rates. !!'+! Decades of research ha ve identified multiple variables that contribute to college student degree attainment, including, but not limited to: gender 60-63, race 60-62, SES 35,60 , first generation status 35, high school GPA 63, first year college GPA 64, and college major 64. However, similar to college student GPA and retention, these variables together account for minimal variance in college student degree attainment. Females, Asian and Pacific Islanders, students with a high SES, families with experience in higher education, those majoring in the social sciences, and those with high - high school GPAs and first year college GPAs, are more likely to earn a college degree than al l other groups. In short, a myriad of factors interact and subsequently affect whether or not a student graduates college. More research is necessary to better understand college student degree attainment, to identify additional variables to may impact the success of students, and to meet President ObamaÕs goal of becoming the country with the highest proportion of college graduates in the world. Huesman et al., 52 appear to be the only investigators who have assessed the relationship between recreational sports use and col lege student degree attainment. The authors utilized a sample of 5,211 first time degree -seeking students who began college in the Fall of 20 01. Recreational sports use was defined as the number of times the student entered the fitness facilities in the fi rst semester of college. Students entered the facility an average of 9.9 ±15.4 times in the first semester at the University , entry counts ranged from 0 -154. Students entering the facility at leas t one standard deviation above the mean (25 times) were compa red to those who entered the facility 10 or less times. Students who earned a BachelorÕs degree within five years of their first semester of enrollment were considered to achieve degree attainment. Utilizing logistic regression and controlling for 1 4 poss ible covariates (ratio of credits attempted to credits completed, CÕs received, DÕs received, course withdraw count, ACT/SAT score, !!',!remedial classes taken, remedial classes failed, athlete status, gender, race, tuition -reciprocity, tuition non -reciprocity, socioeconomic status, and off -campus housing) the authors found a statistically significant positive relationship between recreational sports fitness center use and five -year graduation ( aOR=1.007; 95%CI=1.003,1.012) . Although statistically significant, t he adjusted odds ratio for recreational sports use was small and may not be practically significant. However, the authors concluded that utilizing the fitness center approximately 25 times over the course of the first semester in college increased the pred icted probability of five -year graduation by 2%. The authors did not account for recreational sports use after the first semester of college nor did they address the wide range of recreational center entry counts (0 -25). It is possible that multicollineari ty was present in the final logistic regression model. Variables that may have contributed to multicollinearity include the academic variables (Cs received, Ds received, withdraw count, remedial class taken , and remedial class failed), and the socioeconomi c variables (tuition reciprocity, tuition non -reciprocity, and Pell grant eligibility). Despite some limitations, this study stands alone in assessing the relationship between recreational sports and college student degree attainment. More research is need ed in this area. SUMMARY Few adults meet DHHS aerobic and strength PA guidelines. Students develop lifestyle behaviors while attending college. PA interventions may be effective in a college student population. Recreational sports is an excellent platfor m to build PA interventions as r ecreational sports offers an opportunity for college students to participate in regular PA. The successes of such interventions depend on the funding necessary to ensure the upkeep of equipment, quality programming an d the continued promotion of PA through recreational sports. University !!'-!administrators may consider regular PA participation as beneficial, yet ancillary to college student academic success. Therefore, research investigating the relationships between recre ational sports participation and college student academic success are necessary. Previous research trends towards a positive relationship between a dichotomized measure of recreational sports participation and college student GPA, retention, and Bachelor Õs degree attainment. However, previous literature is limited by self -reported data, lack of assessing confounding variables, and broad measures of recreational sports participation (type and amount) . Future research should utilize data from a university da tabase, should investigate the impact of confounding variables on the relationship between recreational sports participation and academic success, and should assess recreational sports use via student identification swipes into facilities to account for ac curate assessment of recreational sports participation and specificity of participation in various components of recreational sports. Research assessing the specific relationships between recreational sports participation and 1) GPA, 2) retention, and 3) Bachelor 's degree attainment may help academic administrators understand the importance of recreational sports. A positive association between recreational sports participation and academic success will strengthen the argument for increased funding for recreational sports departments, which in turn may help to increase the proportion of adults meeting DHHS PA guidelines. !!(.!REFERENCES !!(%!REFERENCES 1. U.S. Department of Health and Human Services. 2008 physical activity guidelines for Americans. Be active, healthy, and happy. Washington. DC: U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion, Office of the Assistant Secretary for Health, Office of the Secretar y; 2008. 2. U.S. Department of Health and Human Services. Office of Disease Prevention and Health Promotion. Healthy People 2020 . Washington, DC. Available at http://www.healthypeople.gov/2020/topicsobjectives2020/objectiveslist.aspx?topicId=33. Accessed August 18, 2014. 3. Caspersen CJ, Pereira MA, Curran KM. Changes in physical activity patterns in the United States, by sex and cross -sectional age. Medicine & Science in Sports & Exercise. 2000(32):1601 -1609. 4. Sallis JF. Age -related decline in physica l activity: a synthesis of human and animal studies. Medicine and science in sports and exercise. 2000;32(9):1598 -1600. 5. Dobbins M, DeCorby K, Robeson P, Husson H, Tirilis D. School !based physical activity programs for promoting physical activity and fi tness in children and adolescents aged 6 !18. The Cochrane Library. 2009. 6. Telama R, Yang X, Viikari J, V−lim−ki I, Wanne O, Raitakari O. Physical activity from childhood to adulthood: a 21 -year tracking study. American journal of preventive medicine. 2005;28(3):267 -273. 7. Arnett JJ. Emerging adulthood: A theory of development from the late teens through the twenties. American psychologist. 2000;55(5):469. 8. Pascarella ET, Terenzini PT. How college affects students: Findings and insights from twenty years of research. San Fransisco (CA): Jossey -Bass; 1991. 9. Pascarella ET, Terenzini PT. How College Affects Students: A Third Decade Of Research (3rd Eds). San Fransisco (CA): Jossey -Bass; 2005. 10. Chickering AW, Reisser L. Education and Identity (2nd eds.) . San Fransisco (CA): Jossey -Bass; 1993. 11. Helms JE. Black and White racial identity: Theory, research, and practice. Westport, CT:Greenwood Press; 1990. !!(&!12. Renn KA, Shang P. (Eds.). Biracail and multiracial college students: Theory, research, ad best practices in student affairs (New Directions for Student Services #123.) San Fransisco (CA): Jossey -Bass; 2008. 13. Abes ES, Jones SR, McEwen MK. Reconceptualizing the model of multiple dimensions of identity: The role of meaning -making capacity i n the construction of multiple identities. Journal of college student development. 2007;48(1):1 -22. 14. Belenky MF, Clinchy BM, Goldberger NR, Tarule JM. WomenÕs way of knowing: The development of self, voice, and mind. New York (NY): Basic Books; 1986. 15. Perry WG Jr. Forms of Intellectual and Ethical Development in the College Years: A Scheme. New York (NY): Holt, Rinehart and Winston; 1970. 16. Plotnikoff RC, Costigan SA, Williams RL, et al. Effectiveness of interventions targeting physical activity, nutrition and healthy weight for university and college students: a systematic review and meta -analysis. International Journal of Behavioral Nutriti on and Physical Activity. 2015;12(1):45. 17. Gow RW, Trace SE, Mazzeo SE. Preventing weight gain in first year college students: an online intervention to prevent the Òfreshman fifteenÓ. Eating behaviors. 2010;11(1):33 -39. 18. Huang S -J, Hung W -C, Chang M, Chang J. The effect of an internet -based, stage -matched message intervention on young Taiwanese women's physical activity. Journal of health communication. 2009;14(3):210 -227. 19. Sk„r S, Sniehotta FF, Molloy GJ, Prestwich A, Araujo -Soares V. Do brief online planning interventions increase physical activity amongst university students? A randomised controlled trial. Psychology and Health. 2011;26(4):399 -417. 20. Wadsworth DD, Hallam JS. Effect of a web site intervention on physical activity of college females. American journal of health behavior. 2010;34(1):60 -69. 21. Werch CE, Moore MJ, Bian H, et al. Efficacy of a brief image -based multiple -behavior interventio n for college students. Annals of Behavioral Medicine. 2008;36(2):149 -157. 22. Buscemi J, Yurasek A, Dennhardt A, Martens M, Murphy J. A randomized trial of a brief intervention for obesity in college students. Clinical obesity. 2011;1(4 !6):131 -140. 23. Claxton D, Wells GM. The effect of physical activity homework on physical activity among college students. Journal of physical activity & health. 2009;6(2):203. 24. Grim M, Hortz B, Petosa R. Impact evaluation of a pilot web -based intervention to increase physical activity. American Journal of Health Promotion. 2011;25(4):227 -230. !!('!25. NIRSA/NASPA. (2010). NIRSA/NASPA consortium campus recreation impact study. descriptive summary report. Retrieved from http://www.nirsa.org/docs/Discover/ Research/Campus_Rec_Impact_Study.PDF 26. Forrester S, Ross CM, Hall S, Geary C. Using past campus recreational sports participation to explain current physical activity levels of alumni. Recreational Sports Journal . 2007;31:83 -94. 27. Astin AW. Student i nvolvement: A developmental theory for higher education. Journal of college student personnel. 1984;25(4):297 -308. 28. Tinto V. Leaving college: Rethinking the causes and cures of student attrition. Chicago: University of Chicago Press; 1987. 29. Tinto V. Leaving College: Rethinking the Causes and Cures of Student Attrition. (2 nd Ed.). Chicago: University of Chicago Press; 1993. 30. Ishler J, Upcraft ML. The keys to first -year student persistence. Challenging and supporting the first -year student: A handbook for improving the first year of college. 2005:27 -46. 31. Ting S -MR, Robinson TL. First -year academic success: A prediction combining cognitive and psychosocial variables for Caucasian and Afric an American students. Journal of College Student Development. 1998. 32. DeBerard MS, Spielmans G, Julka D. Predictors of academic achievement and retention among college freshmen: A longitudinal study. College student journal. 2004;38(1):66 -80. 33. Cohn E, Cohn S, Balch DC, Bradley J. Determinants of undergraduate GPAs: SAT scores, high -school GPA and high -school rank. Economics of Education Review. 2004;23(6):577 -586. 34. Strayhorn TL. Factors influencing the academic achievement of first -generation col lege students. Journal of Student Affairs Research and Practice. 2006;43(4):1278 -1307. 35. Walpole M. Socioeconomic status and college: How SES affects college experiences and outcomes. The review of higher education. 2003;27(1):45 -73. 36. Wolfe RN, Johnson SD. Personality as a predictor of college performance. Educational and psychological measurement. 1995;55(2):177 -185. 37. Danbert SJ, Pivarnik JM, McNeil RN, Washington IJ. Academic success and retention: the role of recreational sports fitness facilities. Recreational Sports Journal. 2014;38(1):14 -22. !!((!38. Belch HA, Gebel M, Maas GM. Relationship between student recreation complex use, academic performance, and persistence of first -time freshmen. Journal of Student Affairs Research and P ractice. 2002;38(2):220 -234. 39. Kampf S, Teske EJ. Collegiate recreation participation and r etention. Recreational Sports Journal. 2013;37(2):85 -96. 40. Frauman E. Differences between participants and non -participants of campus recreation offerings acro ss demographic variables and perceptions of the college experience. Recreational Sports Journal. 2005;29(2):156 -165. 41. Integrated Post Secondary Education Data System Glossary [Internet]. Washington (DC): National Center for Education Statistics. [cited 2015 april 20]. Available from : http://nces.ed. gov/ipeds/glossary/ 42. Institutional retention and graudation rates for undergraduate students [Internet]. Washington (DC): National Center for Education Statistics. [cited 2015 april 20]. Available from: h ttps://nces.ed.gov/programs/coe/indicator_cva.asp 43. Raisman NA. The Cost of College Attrition at Four -Year Colleges & Universities. Educational Policy Insitute;2013. 44. Murtaugh PA, Burns LD, Schuster J. Predicting the retention of university students. Research in Higher Education. 1999;40(3):355 -371. 45. Reason RD. Student variables that predict retention: Recent research and new developments. Journal of Student Affairs Research and Practice. 2009;46(3):850 -869. 46. Astin AW. What matters in college?: Four critical years revisited. San Francisco (CA):Jossey -Bass; 1993. 47. Preventing students from dropping out. San Francisco (CA):Jossey -Bass ; 1975. 48. Lindsey R, Sessoms E. Assessment of a campus recreation program on student recruitment, retention, and frequency of participation across certain demographic variables. Recreational Sports Journal. 2006;30(1):30 -39. 49. Bradley J, Phillipi, RH., Bryant, JA . Minorities benefit from their association with campus recreation programs. NIRSA Journal. 1992;16(3):46 -50. 50. Mallinckrodt B, Sedlacek WE. Student retention and the use of campus facilities by r ace. NASPA journal. 1987;24(3):28 -32. 51. Henchy A. The influence of campus recreation beyond the gym. Recreational Sports Journal. 2011;35(2):174 -181. !!()!52. Huesman R, Brown AK, Lee G, Kellogg JP, Radcliffe PM. Gym Bags and Mortarboards: Is Use of Campus Recreation Facilities Related to Student Success? Journal of Student Affairs Research and Practice. 2009;46(1):50 -71. 53. Murphy KM, Welch F. Wages of college graduates. The economics of American higher education : Springer; 1992:121 -140. 54. Murphy KM, Welch F. Inequality and relative wages. The America n Economic Review. 1993:104 -109. 55. Employment projections [Internet]. Washington (DC): U.S. Beureau of Labor Statistics [cited 2015 april 20] Available from: www.bls.gov/emp/ep_chart_0001.htm 56. Pascarella ET. How college affects students: Ten directi ons for future research. Journal of College Student Development. 2006;47(5):508 -520. 57. Youth Indicators 2011 [Internet]. Washing ton (DC): National Center for Education Statistics [cited 2015 april 20] Available from: https://nces.ed.gov/pubs2012/2012026 /tables/table_23.asp 58. Obama BH. Remarks by the president on the American graduation initiative. Speech presented at Macomb Community College, Warren, MI. 2009. 59. Digest of education statistic [Internet]. Washington (DC): National Center for Education Statistics [cited 2015 april 20] Available from: https://nces.ed.gov/programs/digest/d13/tables/dt13_326.10.asp 60. Adelman C. The Toolbox Revisited: Paths to Degree Completion From High School Through College. US Department of Education. 2006. 61. DeAngelo L, Franke R, Hurtado S, Pryor JH, Tran S. Completing college: Assessing graduation rates at four -year institutions. Higher Education Research Institute, Graduation School of Education & Information Studies, University of California, Los Angeles; 2011. 62. Astin A, Oseguera L. Pre -college and institutional in ! uences on degree attainment. College student retention: Formula for student success. 2005:245 -276. 63. Astin AW. Degree Attainmen t Rates at American Colleges and Universities: Effects of Race, Gender, and Institutional Type. Higher Education Research Insitute, University of California at Los Angeles. 1996. 64. DesJardins SL, Kim D -O, Rzonca CS. A nested analysis of factors affecti ng bachelor's degree completion. Journal of College Student Retention: Research, Theory and Practice. 2003;4(4):407 -435. !!(*!CHAPTER 3 MANUSCRIPT ONE Chapter three addresses specific aim one and was completed in 2014. The manuscript titled, Academic Success and Retention: The Role of Recreational Sports Fitness Facilities , was published in the Recreational Sports Journal , April 2014 .1 Specific Aim 1 : To investigate the differences in college student semester grade point average (GPA), credits completed, class standing, and retention between students who purchased a recreational sports fitness center membership their freshmen year and students who did not. ABSTRACT This study evaluated the role of a university recreational sports and fitness center, in studentsÕ academic success. Study participants included freshmen at a large midwestern university (n=4843; 56% women; 67% white). Recreational sports fitness facility members (students who purchased a recreational sports fitness facilities membership in their first semester; n=1138) were compared to non -members (students who did not purchase a recreational sports fitness facility membership in their first semester; n=37 05). Means ± SD and percentages were calculated for all variables of interest. Differences between groups were analyzed using t -tests and percentages. Members had significantly higher high school grade point averages (GPA) (p=0.002). After four consecutive semesters, members had significantly higher cumulative college GPA (p " 0.0001) and cumulative credits completed (p " 0.0001). Significantly more members than non -members were enrolled in school after two completed years, 89% and 85% !!(+!respectively. Results sho w recreational sports fitness facility membership is associated with, and may be beneficial to, college students' academic success. INTRODUCTION Based on a meta analysis of 21 prospective studies leisure time physical activity is important to current health and chronic disease prevention for all po pulations. 2 However, physical activity tends to decline with age, including the college years, which bridge adolescence and adulthood. While attending college, participation in recreational sports is one way for students to engage in daily leisure time physical activity. Stable funding is necessary for the continued survival of recreational sports departments. University administrators may view recreational sports departments as beneficial, yet ancillary to the academic success and retention of college students. This perception may result in underfunding of recreational sports departments. Therefore, recreational sports departments must illustrate their contribution to academic success and retention to help de monstrate their importance and ensure stability. 3 Through cross sectional studies using self -reported data, researchers have found recreational sports departments to be a deciding factor in whether students attend and continue at a given University. 3-4 How ever, few studies have used longitudinal data collected from university databases to analyze the relationships among college student academic success, retention, and participation in recreational sports. Therefore, current data are needed to explore the im pact of recreational sports on student academic success and retention. !!(,!BACKGROUND Academic Success and Retention Student academic success and retention research is founded primarily on two key theories, Asti nÕs Theory of Involvement 5, and Ti ntoÕs Theor y of Departure 6. Astin 7 indicated a positive relationship among academic success, retention, and student involvement. He found that student success is not defined entirely by academics, but also by the extent to which the student engages with the instituti on. Students enhance their engagement by participating in many aspects of college life, such as academics, social events, and activities related to Student Affairs. Specifically related to r ecreational sports, Astin 7 indicated that participation in sports has a profound positive effect on college student persistence. Tinto 8 stated that student retention depends upon the studentsÕ integration into the academic and social communities of the institution. Some student de partures from the institution are considered involuntary; for example, the student did not meet the required grade point average (GPA). Tinto 8 argued that voluntary departures are often due to the studentÕs perception of a problem related to academic or so cial belonging. Institutions have the ability to assist students in feeling a sense of belonging to both the academi c and social constructs. Tinto 8 stated, ÒThe point of retention efforts is not merely that individuals be kept in college. Education, the social and intellectual development of individuals, rather than just their continued presence on campus should be the goal of retention effortsÓ (p. 145). Recreational sports departments can capitalize on the theories of Astin 5 and Tinto 8 by creating an envi ronment that connects students to the institution. This may result in improved academic success, retention, and current health of the student. !!(-!Researchers have identified variables related to student academic success and retention, including, but not limi ted to, first semester college student GPA, gender, race, high school achievement, participation in student support services, socioeconomic status, financial aid, interpersonal intera ctions, and place of residence .7-9 Previous literature indicates varying impacts of each factor related to academic success and retention. 9 The number of potential confounders related to these outcomes, and the notion that students may be involved in multiple areas of campus concurrently, pose a difficult problem when researchi ng the specific contribution of recreational sports to college student academic success and retention. Recreational Sports and Student Services Currently, over 85% of students who live on campus nationwide participate in some form of recreational sports, 62% of students w ho live off campus participate. 10 Males participate more often than females. The majority of recreational sports participants are fresh men and sophomores .10 At the university in this study, 65% of all students participate in recreational sports and fitness services, 90% of the participants are undergraduate students, and freshmen and sophomores make up 60% of the students who participate. Previous literatur e relating academic success, retention, and recreational sports and fitness services participation lacks consistency in study design and exposure variables. However, most literature indicates a small, but positive, relationship between participation in rec reational sports and academic success and retention .3,11 -17 In contrast, Frauman 18 found no relationship , and Lindsey and Sessoms 19 found a positive relationship only in junior and senior students, but not freshmen and sophomores. !!).!Studies assessing this r elationship are primarily cross -sectional, posing an issue in establishing causality between recreational sports and fitness services participation and academic success and retention. In 2001, Belch and colleagues 11 used a prospective cohort study design (n=11,076) and reported positive relationships among recreational sports use, student retention, and academic success outcomes such as GPA and earned credit hours. Recreational sports users had lower high school GPAs and lower ACT scores, but consistently h ad higher first year college GPAs and completed more credit hours than their non -user counterparts. Huesman et al. 15 reported participation in recreational sports accounted for only 1% of the variance in student retention. They stated that students who used the fitness center just 25 times in a single semester, increased their predicted probability of 1 -year retention b y 1%, and predicted probability of 5 -year graduation by 2%. Although encouraging, this relationship is small, and the logistic model that Huesman and colleagues 15 developed was designed to predict retention in retained students, but it did not address thos e who left the university. With the exception of Belch et al. 11 and Huesman et al. 15, studies have used self -report to evaluate both the exposure variable of recreational sports and outcome variables related to academic success and retention. There are d iscrepancies of exposure and outcome definitions within the literature. Some universities require an identification card swipe to enter the fitness center; others consider the exposure of recreational sports to be all recreational sports facilities -- incl uding outdoor facilities that are not monitored. Potentially, the largest limitation to any observational study of this type is that of unmeasured confounders. Some studies have accounted for several confounding variables, such as high school GPA, gender, race, and socioeconomic status Ñbut left out others, such as place of residence. Despite these limitations, !!)%!studies cited are accepted as preliminary data displaying a positive relationship between student academic success, retention, and participation in recreational sports. Recreational sports participation may be related to student success and retention; however, we found no studies that have assessed the relationship between purchasing a fitness membership at a university and student academic success an d retention. The purpose of this study was to determine the relationship between recreational sports membership and college student academic success and retention. Based on previous literature it was hypothesized that students who purchase a fitness membe rship in their first semester at university will have higher GPAs, complete more credits, and are more likely to be retained than their non -member counterparts. METHODS Data were extracted from a university database at a large midwestern university. A sample of all first time degree seeking freshman -- students who had graduated from high school in the spring of 2010, who in the fall of 2010 were in their first semester at the university, and who had no prior college class experience were used in analyses (n = 4843, 56% women, 67% white). We chose these criteria to eliminate students who had earned college grade points, experienced a college classroom, or had been exposed to recreational sports previously. The cohort made up approximately 70% of the univers ityÕs 2010 freshmen class. Academic Success Seven variables were identified as measures of academic success and analyzed at two time points, the first after two completed semesters (spring semester 2011), and the second after !!)&!four completed semesters (sp ring semester 2012). Variables included, high school GPA (HsGPA), cumulative college GPA (cGPA), cumulative college credits (ccc), one -year retention, two-year retention, and class standing. HsGPA was defined as the studentÕs cumulative GPA in high school; this information was reported on a 4.0 scale to the university through high school transcripts. A studentÕs cGPA was calculated as an average of the studentÕs semester GPAs. Cumulative college credit was defined as the total number of credits the student completed with a passing grade; on average, a full -time takes between 12 and 17 credits each semester. If a student enrolled full or part time at the university for two consecutive fall semesters, one -year retention was achieved. If a student enrolled full or part time at the university for three consecutive fall semesters, two -year retention was achieved. Class standing was defined by the number of credits completed with a passing grade. Students with 0 -27 credits completed were considered freshmen status, those with 28 -55 credits completed were considered sophomore status, those with 56 -87 credits completed were considered juniors, and those with 88 or more credits completed, seniors. Recreational Sports Fitness Center Membership The university recreatio nal sports department was a multidimensional department including fitness centers, athletic courts and fields, group fitness classes, intramural sports, and other fitness activities. Most institutions assess recreational sports and fitness services fees within tuition costs, giving students ÒfreeÓ access to all or some of the facilities. At the university studied, use of most recreational sports facilities was included in the semesterÕs tuition. An additional fee was assessed to students who chose to use th e campus fitness centers, and to those who participated in intramural sports and/or group fitness. At the time of study, two recreational !!)'!sports fitness centers (~20,000 square feet) were available on campus. Each center consisted of standard cardio and st rength exercise equipment, including treadmills, cycles, steppers, free weights, and strength machines. Purchasing a membership was voluntary for students and could be done on a one -semester ($85) or two -semester ($160) basis. Students who purchased a recr eational sports fitness center membership in their first semester at the university were considered recreational sports fitness center members (n= 1138; 24%); those who did not purchase a membership in their first semester were considered recreational spor ts fitness center non-members (n=3705; 76%). Data Analysis Descriptive variables were reported using means ± standard deviations. Our exposure variable was recreational sports fitness center membership. StudentÕs t -tests were used to analyze differences between recreational sports fitness center members and non -members in HsGPA, as well as cGPA and CCC, across four consecutive semesters. Percentages and confidence intervals were used to assess group differences in one - and two - year retention rates, as well as class standing following two consecutive semesters. RESULTS The study sample was very similar to the demographics of the university; women made up 56% of the sample and 67% of the sample were white. International students made up 13% of the sample, which is consistent with institutions similar to the universi ty studied. Table 1 shows outcome variables of interest by recreational sports and fitness membership status. Recreational sports and fitness center members achieved higher GPAs and completed more credits than non -!!)(!members. As seen in Table 1, HsGPA was si gnificantly higher (p = 0.002) in members (3.55 ± .38) compared to non -members (3.52 ± .30). After four consecutive semesters, cGPA was significantly higher (p " 0.0001) in members (3.13 ± 52) compared to non -members (3.00 ± 59), and CCC was significantly higher (p " 0.0001) in members (56.6 ± 8.9) compared to non -members (54.1 ± 11.3). Although not statistically significant, more members than non -members were still enrolled at the university after one completed year, which may hold practical significance, 91% and 88%, respectively, and significantly more members than non -members were still enrolled after two completed years, 89% [86.6 - 90.3%] and 85% [83.9 - 86.2%], respectively. After two completed semesters, sophomore status was achieved by more members (74%) than non -members (60%). Table 3.1. Academic success variables for recreational sports fitness center Ñmembers and nonmembers. Success Variables RSFS fitness center members (n=1,138) RSFS fitness center nonmembers (n=3,705) HsGPA M (SD) 3.55 (0.38)* 3.52 (0.30) CGPA a M (SD) 3.13 (0.52)* 3.00 (0.59) CCCa M (SD) 56.6 (8.9)* 54.1 (11.3) 1-Year Retention % 95% CI 90.7 [89.0-92.4] 88.0 [86.9-89.0] 2-Year Retention % 95% CI 88.5 [86.6-90.3]** 85.0 [83.9-86.2] Class Standing b % 95% CI 73.6 [71.0-76.1]** 60.4 [58.8-62.0] Abbreviations: HsGPA = high school grade point average; CGPA = cumulative college grade point average; CCC = cumulative credits completed; Class Standing = % of students who reach sophomore status; CI = confidence interval; a = assessed after four completed semester; b = assessed after two completed semesters. Note. *p<0.05, **Significantly different by 95% Confidence Intervals. !!))!DISCUSSION Maintenance and growth of recreational sports is necessary to ensure continued participation of college students in daily physical act ivity. Positive relationships between recreational sports participation and academic success may help to increase funding for recreational sports and fitness center memberships because university administrators may see this as an essential service to provi de to students. Previous research shows that recreational sports may have a small, but positive impact on academic success and retention. However, limitations of previous studies include a lack of longitudinal data, dissimilar study designs, self -report fo r exposure and outcome variables, and insufficient information regarding how students accessed the campus fitness facilities. The present investigation extends previous findings by utilizing longitudinal data, assessing specifically fitness center members hips, and by including multiple measures of academic success to help provide comparability to other literature. The results support the proposed hypothesis and illustrate a positive association between recreational sports fitness center membership and aca demic success and retention. Although study findings with respect to GPA are encouraging, it is important to analyze their practical significance. Cumulative GPA was found to be 0.13 points higher for recreational sports and fitness center members compar ed to non -members. Depending on an individual student's future goals and actual GPA, this small difference may not be meaningful. However, an increase of GPA of 0.13 points due to recreational sports fitness center membership would be more meaningful for a student whose absolute GPA is near a cutoff point for acceptance into graduate school. Perhaps a more important finding was the relationship between recreational sports fitness center membership and student retention. Previous studies have noted recrea tional sports fitness center participation to increase predicted probability of one -year retention by 1%. 15 These results !!)*!show a significant increase of 3.5% in two -year retention among recreational sports and fitness services members, compared to non-members. Although not statistically significant, these data also suggest a positive relationship between recreational sports fitness center membership and 1 -year retention. The current study was performed at a large university with approximately 49,00 0 students. A difference in retention of 3.5% equates to approximately 1575 students. The financial impact on a university of a single student departure may equate to thousands of dollars. 20 This loss includes unrealized tuition, fees, and alumni contribut ions. 20 While these data presented are not able to confirm a causal relationship between recreational sports and fitness center membership and student retention, they do illustrate a compelling argument for the as sociation of the two variables. This study did not statistically account for potential confounding variables. High school GPA is a known predictor of college student academic success. It may be argued that the difference in cumulative college GPA between recreational sports members and n on-members is due to the confounding influence of high school GPA. However, cumulative college GPA decreased from high school GPA for both members and non -members. This decrease was greater in non -members, which indicates that membership in recreational sp orts may have a positive influence on college student academic success. Additionally, gender may influence the relationship between recreational sports membership and academic success. Research indicates that females tend to earn higher GPAs than males. 21 This is also true at the university of study where females consistently earn higher GPAs than men (Office of the Registrar, personal communication, September 25, 2013). Specifically, cumulative college GPA for this study was assessed in the spring semeste r of 2012 when the university average GPA for females (3.09) was higher than males (2.97) (Office of the Registrar, personal communication, September 25, 2013). !!)+!There were more male recreational sports members, and more female non -members. Therefore the re lationship of recreational sports membership and academic success may be stronger than these data illustrate due to the confounding nature of gender. Factors contributing to college student success and retention make up a complex framework. Many environme ntal variables not assessed in this study may influence the relationship between recreational sports and fitness membership and student academic success and retention. Previous literature indicates that if students can feel a sense of community with the university they are more lik ely to be retained and succeed. 5 It is possible that students participating in recreational sports are forming bonds with students from other areas of campus, or teachers and friends on campus, which may also contribute to higher academic success. Socioeconomic status (SES) may play a key role in this relationship as well. Recreational sports and fitness center members in this study made the voluntary decision to purchase a membership. Therefore, student SES may confound the relati onship between recreational sports membership, student academic success, and retention. This study did not account for SES; future research should investi gate this potential confounder. Despite a number of limitations, this investigation has many strengths . For example, a large sample and multiple variables were used to assess academic success. A concrete definition of recreational sports exposure was utilized. This study did not specifically assess use of recreational sports, however it may be argued bas ed on Self Determination Theory 22 that students who choose to purchase a recreational sports membership are more motivated to use the facility than students who have a fee for recreational sports included in tuition costs. All results, regardless of size, w ere in the hypothesized direction. Future research should continue to include multiple variables to assess academic success, which will allow for better comparison among !!),!studies. Environmental variables must be assessed in future research to help establish a causal relationship. Our study assessed membership as a way to access the recreational sports fitness facilities; the next step is to assess use of the facilities and the relationship of use to academic success and retention. As previously noted, recrea tional sports and fitness services is a means for students to participate in daily physical activity. Future research should also address overall physical activity behaviors of students in relation to recreational sports fitness center use, academic succes s, and student retention. ADDENDUM The previously published manuscript did not include an assessment of possible confounding variables. To address this limitation we chose to identify possible confounding variables and utilize more sophisticated statisti cs. Our methods and results are below. Covariates Based on previous literature 6-9,22 and information available from the u niversity Registrar, we identified s ix possible covariates that may impact the relationship between recreational sports use and college student academic success; these varia bles included: HsGPA , first year college GPA, ACT score, gender , race , and first semester Pell grant eligibility. HsGPA a nd first year college GPA were based on a 4.0 scale. ACT score could range from 1 -36. All other covariates were categorical: gender (male/female), race ( white/other, which included : American Indian/Alaskan Native , Asian, Black (non -Hispanic) , Hawaiian/Paci fic Islander, Hispanic , and Multiracial ), and Pell grant eligibility ( eligible/not eligible ). !!)-!Statistical Analysis The independent variable was recre ational sports membership (members vs. non -members ). Dependent variables included: cGPA, CCC, one- and two -year retention, and class standing. Description information for all variables is reported with means ± standard deviations and percentages. The relationship between recreational sports membership and 1) cGPA and 2) CCC was assessed via analy sis of covariance (ANCOVA). The relationships between recreational s ports membership (ref: non -member ) and 1) one -year retention, 2) two -year retention, and 3) class standing were assessed via logistic regression. Possible covariates for both sets of analy ses included: HsGPA, ACT score, gender (ref: male), race (ref: other ), and Pell grant eligibility (ref: eligible ). Recreational s ports membership was entered into the models first. Covariates were entered based on the strength of relationship to the dependent variable determined via chi -square and one-way ANOVA . Variables related to the dependent variable (p<0.1) were considered in model building. Unadjusted odds ratios (OR) and adjusted odds ratios (aOR) were calculated with associated 95% confidence intervals for reaching one -year retention versus not, reaching two-year retention versus not, and achieving sophomor e status versus not, accor ding to recreational sports member status. Variables were removed from the model if they did not significantly (p<0.05) impact the overall model or did not change the odds ratio associa ted with recreational sports member status by more than ten percent. As a final step, f irst year college GPA was added to two of our adjusted models to identify if recreational sports remained significantly associated with two -year retention and class standing, after controlling for the robust predictor of academic success. First yea r college GPA was not entered into the model !!*.!predicting one -year retention as it is well known that students who have low GPAs are most likely to drop out in the first year as compared to students with high GPAs. Results Means, standard deviations, and percentages for all variables of interest are listed in Tables 3.2 -3.5. Descriptive information for the relationship between recreational sports member status and 1) cGPA and 2) CCC is listed in Table 3.2. Descriptive inform ation for the relationship between recreational sports member status and 1) one-year retention, 2) two -year retention, and 3) class standing is listed in T ables 3.3, 3.4 , and 3. 5 respectively. Univariate analyse s revealed significant relationships betwee n membership status and 1) all covariates (except ACT) , 2) cGPA, and 3) CCC (p<0.05). After adjusting for gender, Pell grant eligibility, race, ACT score, and high school GPA, cGPA was significantly higher (p<0.001) in members (3.17 ± 0.48) compa red to nonmembers (3.01 ± 0.55) and CCC was significantly higher (p<0.001) in members (57.5 6± 7.1) compared to non -members (55.7 ± 9.0) (Tables 3.6 and 3.7 ). Univariate analyses showed one -year retention, two -year retention, and class standing were all si gnificantly related to membership status and all covariates with the exception of gender (Tables, 3.3 -3.5). Therefore, membership status, Pell grant eligibility, race, high school GPA, and ACT score were entered into logistic regression analyses. As seen i n Tables 3. 8-3.10, the adjusted odds of achieving one-year retention, two -year retention, and sophomore status differed only slightly from unadjusted models when comparing rec reational sports members to non-members . Specifically, the adjusted odds were sig nificantly higher for members compared to non -members to reach one -year retention (aOR= 1.42, 95%CI: 1.10 -1.85), two-year retention !!*%!(aOR= 1.39, 95%CI: 1.10 -1.76), and to achieve sophomore status (aOR= 1.59, 95%CI: 1.14 -2.22). When entered into log istic regression models for two -year retention and class standing, first year cumulative GPA was the on ly significant predictor of two -year retent ion (aOR=4.26, 95%CI: 3.66 -4.96), and class standing (aOR=14.96, 95%CI: 11.57 -19.35); recreational sports membership was no longer significant (data not shown in tables) . Table 3.2. Means, standard deviations (SD) , frequencies, and percentages of dependent and covariate variables separated by recreational sports membership. Variable Total Member Nonm ember n 3470 899 2571 frequency (percentage) Gender* Male 1467 (42.3) 509 (56.6) 958 (37.3) Female 2003 (57.7) 390 (43.4) 1613 (62.7) Pell* Eligible 1070 (30.8) 174 (19.3) 896 (34.9) Not Eligible 2400 (69.2) 725 (80.7) 1675 (65.1) Race* White 2721 (78.4) 786 (87.4) 1935 (75.3) Other a 749 (21.6) 113 (12.6) 636 (24.7) Mean+SD High School GPA* 3.55±0.31 3.57±0.29 3.55±0.32 ACT 24.25±3.12 24.71±2.60 24.09±3.25 cGPAb,c* 3.05±0.54 3.17±0.48 3.01±0.55 CCCb,c* 56.19±8.59 57.56±7.12 55.71±9.01 Abbreviations: GPA = grade point average; cGPA = cumulative college grade point average; CCC = cumulative credits completed; a= Black (non-Hispanic) , Hispanic, Asian, Hawaiian/Pacific Islander, American Indian /Alaskan Native , and Multiracial ; b = assessed after four completed semesters, c = dependent variable. Note. *Statistical difference between recreational sports membership status at p<0.05. !!*&!Table 3.3. Means, standard deviations (SD) , frequencies, and percentages of independent and covariate variabl es separated by one -year retention status. Variable Total Retained Not Retained n 3992 3551 441 frequency (percentage) Recreational sports membership a* Member 990 (24.8) 911 (25.7) 79 (17.9) Non member 3002 (75.2) 2640 (74.3) 362 (82.1) Gender Male 1687 (42.3) 1505 (42.4) 182 (41.3) Female 2305 (57.7) 2046 (57.6) 259 (58.7) Pell* Eligible 1291 (32.3) 1106 (31.1) 185 (42.0) Not Eligible 2701 (67.7) 2445 (68.9) 156 (58.0) Race* White 3073 (77.0) 2767 (77.9) 306 (67.1) Other b 919 (23.0) 784 (22.1) 135 (30.6) mean ±SD High School GPA* 3.54±0.32 3.55±0.31 3.44±0.35 ACT* 24.14±3.21 24.24±3.15 23.42±3.62 Abbreviations: GPA = grade point average; a = independent variable; b = Black (non -Hispanic), Hispanic, Asian, Hawaiian/Pacific Islander, Ameri can Indian/Alaskan Native, and M ultiracial. Note. *Statistical d ifference between retention groups at p<0.05 . !!*'!Table 3.4. Means, standard deviations (SD) , frequencies, and percentages of independent and covariate variables separated by two -year retention status. Variable Total Retained Not Retained n 3875 3293 582 frequency (percentage) Recreational sports membership a* Members 968 (25.0) 864 (26.2) 104 (17.9) Nonmember 2907 (75.0) 2429 (73.8) 478 (82.1) Gender Male 1639 (42.3) 1395 (42.4) 244 (41.9) Female 2226 (57.7) 1889 (57.6) 338 (58.1) Pell* Eligible 1244 (32.1) 983 (29.9) 261 (44.8) Not Eligible 2631 (67.9) 2310 (70.1) 321 (55.2) Race* White 2983 (77.0) 2607 (79.2) 376 (64.6) Other b 892 (23.0) 686 (20.8) 206 (35.4) Mean ±SD High School GPA* 3.54±0.32 3.56±0.30 3.41±0.37 ACT* 24.16±3.19 24.31±3.05 23.33±3.78 First year cGPA c* 2.95±0.69 3.07±0.57 2.29±0.92 Abbreviations: GPA = grade point average; cGPA = cumulative grade point average; a = independent variable; b = Black (non -Hispanic), Hispanic, Asian, Hawaiian/Pacific Islander, American Indian/A laskan Native, and M ultiracial; c = assessed after two complet ed semesters. Note.*Statistical d ifference between retention groups at p<0.05. !!*(!Table 3.5. Means, standard deviations (SD) , frequencies, and percentages of independent and covariate variables separated by class standing. Variable Total Sophomore Not Sophomore n 3875 3541 334 frequency (percentage) Recreational sports membership a* Members 968 (24.9) 921 (26.0) 47 (14.2) Nonmember 2907 (75.1) 2620 (74.0) 287 (85.8) Gender Male 1639 (42.3) 1489 (42.1) 150 (44.9) Female 2236 (57.7) 2052 (57.9) 184 (55.1) Pell* Eligible 1244 (32.1) 1069 (30.2) 175 (52.4) Not Eligible 2631 (67.9) 2472 (69.8) 159 (47.6) Race* White 2983 (77.0) 2803 (79.2) 180 (53.9) Other b 892 (23.0) 738 (20.8) 154 (46.1) Mean ±SD High School GPA* 3.54±0.32 3.56±0.31 3.31±0.38 ACT* 24.16±3.19 24.33±3.07 22.27±3.83 First year cGPA * 2.95±0.69 3.06±0.55 1.75±0.84 Abbreviations: GPA = grade point average; cGPA = cumulative grade point average; a = independent variable; b = Black (non -Hispanic), Hispanic, Asian, Hawaiian/Pacific Islander, American Indian/Alaskan Native, and Multiracial. Note. * Statistical di fference between class standing groups at p<0.05 !!!!*)!Table 3.6. Analysis of covariance for cumulative grade point average after four consecutive semesters . Effects Sum of squares df Mean square F Value P value Recreational sports membership a 7.61 1 7.61 36.19 <0.001* Genderb 4.27 1 4.27 20.32 <0.001* Pellc 5.52 1 5.52 26.25 <0.001* Race d 6.80 1 6.80 32.36 <0.001* ACT 22.25 1 22.25 105.83 <0.001* High School GPA 96.56 1 96.56 459.37 <0.001* *Significant difference at p<0.05 aMember, nonmember (ref) bFemale, Male (ref) cNot eligible, Eligible (ref) dWhite, Other(ref) Note. dOther includes: Black (non-Hispanic) , Hispanic, Asian, Hawaiian/Pacific Islander, American Indian /Alaskan Native , and Multiracial. Table 3.7. Analysis of covariance for cumulative credits completed afte r four consecutive semesters . Effects Sum of squares df Mean square F Value P value Recreational sports membership a 856.84 1 856.84 14.54 <0.001* Genderb 2721.14 1 2721.14 46.16 <0.001* Pellc 3280.54 1 3280.54 55.65 <0.001* Race d 1608.43 1 1608.43 27.29 <0.001* ACT 5623.90 1 5623.90 95.40 <0.001* High School GPA 10562.02 1 10562.02 179.17 <0.001* *Significant difference at p<0.05. aMember, nonmember (ref) bFemale, Male (ref) cNot Eligible, Eligible (ref) dWhite, Other (ref) Note. dOther includes: Black (non-hispanic) , Hispanic, Asian, Hawaiian/Pacific Islander, American Indian /Alaskan Native , Multiracial) !!!!**!Table 3.8. Odds ratios (OR), adjusted odds ratios (aOR), and 95% confidence intervals (CI) for the effect of independent and covariate variables on one -year retention status. Variable OR [95% CI] aOR [95% CI] b Recreactional sports membership c Member 1.58 [1.23-2.04]* 1.42 [1.10-1.85]* Nonmember a a High School GPA 2.73 [2.03-3.64]* 2.25 [1.63-3.09]* ACT 1.08 [1.05-1.11]* 1.02 [0.99-1.06] Pell Grant Not Eligible 1.59 [1.31-1.96]* 1.29 [1.03-1.61]* Eligible a a Race White 1.56 [1.25-1.93]* Other d a Abbreviations: GPA = grade point average; a = reference group; b=adjusted model, which includes recreational sports membership status, high school GPA, ACT, and Pell grant eligibility ; c = independent variable; d = Black (non -Hispanic), Hispanic, Asian, Hawaiian/Pacific Islander, American Indian /Alaskan Native , and Multiracial. Note. *Significantly difference by 95% CI. Table 3.9. Odds ratios (OR), adjus ted odds ratios (aOR), and 95% confidence i ntervals (CI) for the effect of independent and covariate variables on two -year retention status. Variable OR [95% CI] aOR [95% CI] b Recreational sports membership c Member 1.64 [1.30-2.05]* 1.39 [1.10-1.76]* Nonmember a a High School GPA 3.98 [3.07-5.18]* 3.10 [2.35-4.09]* ACT 1.10 [1.07-1.13]* Pell Grant Not Eligible 1.91 [1.60-2.29]* 1.43 [1.17-1.75]* Eligible a a Race White 2.08 [1.72-2.52]* 1.33 [1.07-1.66]* Other d a a Abbreviations: GPA = grade point average; a = reference group; b=adjusted model, which includes recreational sports membership status, high school GPA, Pell grant eligibility, and race; c = independent variable; d = Black (non -Hispanic), Hispanic, Asian, Hawaiian/Pacific Islander, American Indian /Alaskan Native , and Multiracial. Note. *Significant difference by 95% CI . !!*+!Table 3.10. Odds ratios (OR), adjus ted odds ratios (aOR), and 95% confidence i ntervals (CI) for the effect of independent and covariate variables on class standing. Variable OR [95% CI] aOR [95% CI] b Recreational sports membership c Member 2.15 [1.56-2.95]* 1.59 [1.14-2.22]* Nonmember a a High School GPA 8.52 [6.17-11.77]* 4.73 [3.30-6.79]* ACT 1.22 [1.18-1.26]* 1.07 [1.03-1.12]* Pell Grant Not Eligible 2.55 [2.03-3.19]* 1.35 [1.02-1.77]* Eligible a a Race White 3.24 [2.58-4.09]* 1.48 [1.11-1.98]* Other d a a Abbreviations: GPA = grade point average; a = reference group; b=adjusted model, which includes recreational sports membership status, high school GPA, ACT score, Pell grant eligibility, and race; c = independent variable; d = Black (non -Hispanic), Hispanic, Asian, Hawaiian/Pacific Islander, American Indian /Alaskan Native , and Multiracial. Note. *Significant difference by 95% CI. !!*,!Conclusion Our original manuscript showed that recreation al sports members earned higher cGPAs, completed more credits, and were more likely to reach two -year retention compared to non -members .1 No differences were found between members and non -members in one -year retention. The results of this addendum support those of our original manuscript. 1 After controlling for confounding variables, re creational sports members earned higher cGPAs, completed more credits, were more likely to reach one - and two - year retention, and were more likely to achieve sophomore status after two consecutive semesters as compared to non -members. These results provid e additional evidence that supports the positive relationship between recreational sports membership and college student academic success. However, it is important to note that first year GPA is a stronger predictor of two -year retention and class standing than all other predictors entered into the models, including recreational sports membership (data not shown) . Therefore, future research should investigate the relationship between recreational sports participation and first year GPA. Further , this addend um addresses only a few of the limitations of our original study ; it does not address all. In particular, future research should investigate the relationship between recreational sports use and college student academic success as well as the relationship between recreational sports use and physical activity behaviors. !!*-!REFERENCES !!+.!REFERENCES 1. Danbert SJ, Pivarnik JM, McNeil RN, Washington IJ. Academic success and retention: the role of recreational sports fitness facilities. Recreational Sports Journal. 2014;38(1):14 -22. 2. Li J. & Siegrist J. Physical activity and risk of cardiovascular disease - a meta -analysis of prospective cohort studies. International Journal of Environmen tal Research and Public Health. 2012;9(2):391-407. 3. Haines DJ. Undergraduate student benefits from university recreation. Recreational Sports Journal. 2001;25(1):25-33. 4. Moffitt J. Recreating retention. Recreational Sports Journal . 2010;34(1):24-33. 5. Astin AW. Preventing students from dropping out. San Francisco(CA):Jossey -Bass; 1975. 6. Tinto V. Dropout from higher education: a theoretical synthesis of recent research. Review of educati onal research. 1975;45(1):89-125. 7. Astin AW. Student involvement: A developmental theory for higher education. Journal of college student development . 1999;40(5):518-29. 8. Tinto V. Leaving College: Rethinking the Causes and Cures of Student Attrition . (2nd Ed.). Chicago: University of Chicago Press; 1993. 9. Reason RD. Student variables that predict retention: recent research and new developments. NASPA. 2009;46(3):482-501. 10. NIRSA/NASPA. (2010). NIRSA/NASPA consortium campus recreation impact study. descriptive summary report. Retrieved from http://www.nirsa.org/docs/Dis cover/Research/Campus_Rec_Impact_Study.PDF 11. Belch HA, Gebel M, & Maas GM. Relationship between student recreation complex use, academic performance, and persistence of first -time freshmen. Journal of Student Affairs Research and Practice . 2001;38(2):254-268. 12. Bryant JA, Banta TW, & Bradley JL. Assessment provides insight into the impact and effectiveness of campus recreation programs. NASPA . 1995;32(2):153-160. 13. Churchill WD, & Iwai SI. College attrition, student use of campus facilities, and a consid eration of self -reported personal problems. Research in higher education . 1981;14(4):353-365. !!+%!14. Hall DA. Participation in a campus recreation program and its effect on retention. Recreational Sports Journal . 2006;30(1):40-45. 15. Huesman RL, Brown AK, Lee G, Kellogg JP, & Radcliffe PM. Gym bags and mortarboards: Is use of campus recreation facilities related to student success? NASPA. 2009;46(1):50-71. 16. Miller JJ. Impact of a university recreation center on social belonging and student retention. Recreational Sports Journal . 2011;35(2):117-129. 17. Windschitl MR. The relationship of participation in recreational sports with retention rates and academic success of first -year college students [dissertation] . Minneapolis (MN): University of Minnesota; 2008. 158 p. 18. Frauman E. Differences between participants and non -participants of campus recreation offerings across demographic variables and perceptions of the college experience. Recreational Sports J ournal. 2008;29(2):156-165. 19. Lindsey R, & Sessoms E. Assessment of a campus re creation program on student recruitment, retention, and frequency of participation across certain demographic variables. Recreational Sports Journal. 2006;30(1):30-39. 20. DeBerard S, Julka DL , & Sp elmans GI. Predictors of academic achievement and retention among college freshmen: a longitudinal study. College Student Journal . 2004; 38(1):66-80. 21. Conger D, & Long MC. Why are men falling behind? Gender gaps in college performance and persistence. The ANNALS of th e American Academy of Political and Social Science. 2010;627:184-214. 22. Deci EL, & Ryan RM. Self-Determination Theory: A macrotheory of human motivation, development, and health. Canadian Psychology. 2008;49(3):182-185. 23. Murtaugh PA, Burns LD, Schuster J. Predicting the retention of university students. Research in Higher Education. 1999;40(3):355 -371. !!+&!CHAPTER 4 MANUSCRIPT TWO Chapter four addresses Specific Aim two. Specific Aim 2 : To investigate the relationship between recreational sports fitness center use and college student one - and two -year retention. INTRODUCTION Regular participation in physical activity (PA) is important for current health and chronic disease prevention. However, few adults meet PA guidelines .1 Decades of research indicate that students develop many lifestyle behaviors in college. 2,3 However, data assessing the relationship between PA behaviors during and post college are limited, yet the existing literature point s in a positive direction. Plotni koff et al., 4 recently published the first review of interventions that target PA, nutrition, and healthy weight in college students. The authors identified 29 studies that assessed the impact of intervention on PA participation in college students. Eighteen of the 29 studies showed significant improvements in PA or fitness behaviors from pre - to post -intervention. Improvements in PA and fitness behaviors included increases in total minutes of PA participation, number of days participating in PA, acti vity intensity, and decreases in exercise barriers . Further, a survey of 380 college alumni (2 -10 years post graduation) indicated a positive relationship between exercise behaviors during the senior year of college and exercise behaviors post graduation. 5 Specifically, 85% of participants who indicated they were active as college seniors were as or more active at the time of survey, and, 82% of those who were not active as college seniors indicated they were the same or less active at the time of survey. 5 More research is needed in this area; however, the literature discussed suggests that the development of positive !!+'!PA behaviors during college may help increase the proportion of college -aged adults meeting PA guidelines. Recreational sports departments are a potential avenue for PA promotion to college students. R esearchers have investigated the relationship between recreational sports use during college and PA participation post college in a sample of 310 health, physical education, and recreation alumni w ho graduated from a single university between May 2001 and May 2005. 6 The authors 6 reported that participation in recreational sports during college directly related to PA participation post college (p=0.013). These findings provide evidence that PA promotion through university recreational sports departments may result in increased PA participation not only during college, but also later in life. University recreational sports departments are charged with promot ing positive physical activity behavio rs to college students. NIRSA: Leaders in Recreation report that 85% of all students engage in some form of university recreational sports at least once during an academic career. 7 In addition to reaching a large number of students, recreational sports fa cilities house the equipment necessary to support PA, including, cardio and resistance training equipment. However, funding is necessary for continued PA promotion and equipment upkeep. Although development of student healthy lifestyle behaviors is import ant to university administrators, it is secondary to their interest in student retention. Therefore, recreational sports departments should seek to identify relationships between recreational sports and college student retention to help justify their exis tence to university administrators . Not only is r etention a very important measure to academic administrators , it is a bottom line issue . Thousands of dollars are lost for each student who leaves an institution prior to degree attainment .8 The risk of student attrition declines after the first year of colleg e, thus past retention !!+(!research has focused primarily on students' first to second years. One -year retention is usually defined by two consecutive Fall semester enrollments and two year retention by three consecutive Fall semester enrollments. 9 Previous investigators have identified many variables that predict retention including age 10,11, sex 11,12, race 10-12, residency 10, high school GPA 10-13, SAT/ACT scores 11, first semester college GPA 10-12, socioeconomic status (SES) 11, and sense of belonging 14-16. Race, sex, SES, and high school GPA appear to be the most reliable predictors; White /Asian female students with high GPAs in high school and high socioeco nomic status are most likely to be retained. 17 However, variables known to impact college student retention are primarily unmodif able, meaning the student cannot change their status. To increase retention rates, researchers much identify variables that are modifiable, i.e. , the student can alter their status within the variable. Several investigators have posited that the most important variable predicting student retention may be student integration into university culture. 14-16,18 Recreational sports offers a means for stud ents to become part of the u niversity. Students who participate regularly in recreational sports facility activities can develop self-esteem, social relationships, and communication and leadership skills. In addi tion to conventional programs (fitness class es, intramural sports), recreational sports departments provide wellness programs, outdoor recreation, and building space that appeals to various clubs, organizations, and other activities. Recreational sports departments can be versatile and offer a sense of community for a diverse group of students. Researchers who have assessed relationships between recreational sports participation and retention have primarily utilized self-report measures obtained with the Q uality of Intramural Recreational Sports survey (QIRS). 19,20 Results indicate that 31 -35% of students report that the availability of recreational facilities and programs is important or very important in their decision !!+)!to continue at the institution. 19,20 However, the se studies are limited by cross -sectional designs, self-report of both independent and dependent variables, and simplistic statistics and sampling strategies. Four studies to date have utilized university databases to overcome the self -report limitation. 21-24 Results of this research corroborate the previous findings, indicating that recreational sports participation is positively related to retent ion. However, corroboration does not indicate that self -report is sufficient to measure retention. Researchers that utilized self -report measured studentsÕ beliefs that they would continue at the institution the following year. Researchers utilizing university databases were able to determine whether the student was in fact retained. Further, to our knowledge, studies utilizing the QIRS have not compared student belief of continuing at the institution and actual retention. Therefore, the two methods provide dif ferent information and should not be considered interchangeable. The four studies utilizing university databases also have limitations. Two of them 21,24 dichotomized recreational sports use into users and non -users. This method results in a largely homogenous sample as the majority of college students participate in some form of recreational sports. Further, both studies evaluated their data using percentages and 95% confidence intervals, which limits precise assessment of the complex nature of stude nt retention. Two additional studies utilized logistic regression 22,23 and each controlled for a variety of confounding variables . However, these studies used varying measures of recreational sports use, and included differing confounding variables (only ACT score, gender and race were consistently used), making their results difficult to compare. Although previous investigations are widely considered as preliminary evidence of a positive relationship between recreational sports use and retention , !!+*!research ers should be cautious when interpreting results that include homogenous samples and self-reported, convenience sampled data that fails to control for possible confounding variables. There is a need to investigate the relationship between recreational sports participation and student retention by utilizing university data, considering many potential confounders, and implementing more sophisticated analytic techniques. Therefore, the pu rpose of this study was to investigate the relationship between university recreational sports fitness center use and college student one - and two -year retention. It was hypothesized that after controlling for potential covariates, there would be a direct relationship between use of recreational sports fitness centers and one - and two -year retention. METHODS Data Collection Data were extracted from a university database housed at a large Midwestern university. Study participants included all first time degree seeking freshmen students who graduated from high school in the Spring of 2010 . Participants were in their first semester at the university in Fall 2010 , were not student athletes, and had no prior college class experience. Students who purchased a membership in at least one of their first two semesters (Fall 2010 and Spring 2011) at the university were utilized to assess one -year retention, and students who purchased a membership in at least one of their first four semesters (Fall 2010/11 and Sprin g 2011/12) at the university were utilized to assess two -year retention. Students who did not purchase a membership to th e recreational sports fitness centers were excluded from this study. Few students enroll in the summer semester and even fewer elect to purchase a summer fitness !!++!membership. Therefore, summer semester academic variables and recreational sports use were not evaluated. Academic Success Acade mic success was defined by one - and two -year retention. As per the National Center for Education Statistics, Fall -to-Fall retention was utilized to assess one - and two -year retention. 8 Students who enrolled full time in the Fall 2 010 semester and full or pa rt time in the Fall 2011 semester were considered to have reached one-year retention . Students who enrolled full time in the Fall 2010 semester and full or part time in Fall 2 011 and Fall 2012 semesters were considered to have reached two-year retention. Enrollment in Spring semesters was not considered for Fall -to-Fall retention rates. Recreational Sports The u niversity recreational sports department offers a variety of opportunities for students to participate in PA including outdoor facilities, indoor pools, courts, cardio equipment, free weights, and strength training machines. Access to some university re creational sports facilities is included in tuition costs. However, an additional fee is charged to students who wish to purchase a fitness center membership, which provides access to all recreational sports fitness centers. There are two fitness centers on campus total ing ~20,000 square feet of space . These fitness centers house cardio machines and resistance training equipment. To verify membership, students swipe their student ID s each time they enter a recreational sports fitness c enter. This information is recorded electronically . For this investigation, total number of times s tudy !!+,!participants swiped int o the fitness center each semester was gathered and utilized to identify level of recreational sports use. To our knowledge, there are no well -accepted standards used to establish cut points to define levels of recreational sports use. Therefore we identi fied cut points for four groups (never used, low use, medium use, and high use) by utilizing data from all students enrolled at the institution between Fall 2010 and Spring 2014. The level Ônever usedÕ was defined by purchasing a membership, but never util izing the recreational sports fitness centers. To identify cut points for low, medium, and high use, recreational sports fitness center use per semesters of membership was calculated for all students who purchased a membership and used the recreational spo rts fitness facilities. We utilized a quartile split to first identify four groups of recreational sports fitness center use. The middle two groups were combined, resulting in three levels of recreational sports use (low = lowest quartile of use , medium = middle two quartiles of use , high = highest quartile of use ). This allowed us to identify the extreme ends of the recreational sports use curve, thus improving comparability and reducing potential type two errors. Study participants were categorized into on e of the four population derived levels based on the number of times they swiped their ID into a recreational sports fitness center per semesters of enrollment . Swipes in the first two semesters (Fall 2010, Spring 2011) were considered to evaluate one -yea r retention and swipes in the first four semesters (Fall 2010/11 and Spring 2011/12) to evaluate two -year retention. Covariates Based on previous literature 10-13 and information available from the u niversity R egistrar, we identified six possible covariates that may impact the relationship between recreational sports !!+-!use and college student retention; these variables included: high school GPA (HsGPA), first year college GPA, ACT score, gender , race , and first semester Pell grant eligibili ty. HsGPA and first year college GPA were based on a 4.0 scale. ACT score could range from 1 -36. All other covariates were categorical: gender (male/female), race (white/other, which inc luded: Asian, Black (non -Hispanic), Hispanic, Hawaiian/Pacific Islande r, and American Indian/Alaskan Native ), and Pell grant eligibility (eligible/not eligible). Statistical Analysis Our exposure variable was recreational sports use level (ref: never used) and our dependent variables were 1) one- and 2) two- year retention. Possible covariates included : HsGPA, ACT score, gender (ref: male) , race (ref: other) , and Pell grant eligibility (ref: eligible) . Descriptive information is reported with means ± standard deviations and percentages. Chi -square and one -way ana lysis of variance were utilized to assess the univariate relationship s between each variable and 1) one- and 2 ) two -year retention. Logistic r egression models were developed to describe the relationship between recreational sports use and 1) one - and 2) tw o-year retention. Recreational s ports use was entered into the models first, followed by HsGPA, race, and gender. This order of entry was b ased on previous research. 10-13 Additional covariates, if related to retention (p<0.10), were entered based on the st rength of relationship. Unadjusted odds ratios (OR) and adjusted odds ratios (aOR) were calculated with associated 95% confidence intervals for reaching one -year retention versus not, and reaching two -year retention versus not, according to recreational sp orts level. Variables were removed from the model if they did not significantly (p<0.05) impact the prediction of retention or did not change the odds ratio associated with recreational sports use by more than ten percent. After evaluating the final !!,.!adjusted model for two -year retention we planned to assess first year college GPA as a covariate to identify if recreational sports remained significantly associated with two -year retention after controlling for the robust predictor of academic success. Fi rst year college GPA was not entered into the model predicting one -year retention as it is well known that students who have low GPAs are most likely to drop out in the first year as compared to students with high GPAs. RESULTS The sample utilized to ass ess one -year retention (n=1344 , 94% retained ) was 50% female, 87% white, and 21% Pell grant eligible (Table 4.1) . Mean±standar d deviation Hs GPA, ACT score, and first year cumulative GPA were 3.57±0.29, 24.65±2.66, and 3.09±0.61, respectively (Table 4.1) . The sample utilized to assess two -year retention (n=1583, 90% retained) was 51% female, 87% white, and 22% Pell grant eligible (Table 4.2) . Mean±standard deviation high school GPA, ACT score, and first year cumulative GPA were 3.57±0.29, 24.63±2.74 and 3.09 ±0.61, respectively (Table 4.2) . Recreational sports use cut points established from population data were defined as never used (purchased a membership, but never used the facilities), low use (0 -5 times/membership), medium use (>5 -26 times/membership), and high use (>26 times/membership). The study sample was categorized into one of the four groups based on number of uses/semesters of enrollment. Medium use and high use categories were combined due to low frequencies of subjects who were not retained and c ategorized as medium/ high users, which resulted in three categories for analyses, never used, low use (0 -5 times/semester of enrollment), and medium/ high use (> 5 times/semester of enrollment). It should be noted that there was a low frequency of students categorized as Ônever usedÕ who did not reach two -year retention. However, we chose to continue the analysis without combining never !!,%!used and low use groups, as these two groups are inherently different. All descriptive information is listed in Tables 4.1 and 4.2. Chi-square analyses revealed significant global differences in one -year retention among level s of recreational sports use. One -way analysis of variance indicated a positive relationship between HsGPA and one -year retention (Table 4.1). Based on previous literature, level of recreational sports use, HsGPA, race, and gender were consecutively entered into logistic regression models to predict one -year retention. Pell grant eligibility, and ACT score were not considered for model building, as they did not meet criteria to enter the model (p<0.1). The most parsimonious model identified to predict one -year retention included recreational sports use level and high school GPA. As seen in Table 4.3 , the unadjusted odds of achieving one-year retention were significantly higher among medium/ high users of recreational sports compared to those categorized as nev er used (OR=1.98, 95% CI: 1.12 -3.44). Results were minimally affected and remained significant after adjusting for HsGPA (aOR for medium/high users vs. never used =1.94, 95%CI: 1.12 -3.38). No differences were identified in one -year retention between low users and never used. Univariate analyses showed a significant relationship between two -year retention and 1) race, 2) HsGPA, and 3) first year cumulative GPA. No relationship was found bet ween two-year retention and 1) gender , 2) Pell grant eligibility, or 3) recreational sports use levels (Table 4.2). Therefore, due to the null findings between independent and dependent variables, we did not build a logistic regression model to predict two -year retention. !!,&!Table 4.1. Means, standard deviations (SD) , frequencies, and percentages of independent and covariate variables separated by one -year retention. Variable Total Retained Not Retained n 1344 1259 85 frequency (percentage) Level of RS Use a*€ Never Used 322 (24.0) 296 (23.5) 26 (30.6) Low 363 (27.0) 332 (26.4) 31 (36.5) Medium/ High 659 (49.0) 631 (50.1) 28 (32.9) Gender Male 676 (50.3) 631 (50.1) 45 (52.9) Female 668 (49.7) 628 (49.9) 40 (47.1) Race White 1165 (86.7) 1097 (87.1) 68 (80.0) Other b 179 (13.3) 162 (12.9) 17 (20.0) Pell Grant Eligibility Eligible 278 (20.7) 257 (20.4) 21 (24.7) Not Eligible 1066 (79.3) 1002 (79.6) 64 (75.3) mean±SD High School GPA* € 3.57±0.29 3.58±0.29 3.46±0.28 ACT composite score 24.65±2.66 24.66±2.64 24.52±2.96 Abbreviations: RS = Recreational sports; GPA = grade point average; cGPA = cumulative grade point average; a = independent variable; b = Black (non -Hispanic), Hispanic, Asian, Hawaiian/Pacific Islander, American Indian/Alaskan Native, and Multiracial. Note. *Global statistical difference between retention groups at p<0.05 . €Meets criteria to enter logistic regression model, p<0.1 !!,'!Table 4.2. Means, standard deviations (SD) , and percentages of independent and covariate variables separated by two -year retention. Variable Total Retained Not Retained n 1583 1420 163 frequency (percentage) Level of RS Use a Never Used 67 (4.2) 59 (4.2) 8 (4.9) Low 637 (40.2) 570 (40.1) 67 (41.1) Medium/ High 879 (55.5) 791 (57.5) 88 (54.0) Gender Male 770 (48.96 685 (48.2) 85 (52.1) Female 813 (51.4) 735 (51.8) 78 (47.9) Race* € Other b 214 (13.5) 184 (13.0) 30 (18.4) White 1369 (86.5) 1236 (87.0) 133 (81.6) Pell Grant Eligibility Yes 346 (21.9) 307 (21.6) 39 (23.9) No 1237 (78.1) 1113 (78.4) 124 (76.1) mean±SD High School GPA* € 3.57±0.29 3.58±0.29 3.47±0.32 First Year c GPA*€ 3.09±0.61 3.15±0.54 2.53±0.86 ACT composite score 24.63±2.74 24.65±2.68 24.46±3.18 Abbreviations: RS = Recreational sports; GPA = grade point average; cGPA=cumulative grade point average; a = independent variable; b = Black (non -Hispanic), Hispanic, Asian, Hawaiian/Pacific Islander, American Indian/Alaskan Native, and Multiracial. Note. *Global statistical difference between retention groups at p<0.05 €Meets criteria to enter logistic regression model, p<0.1 !!,(!Table 4.3 . Odds ratios (OR), adjus ted odds ratios (aOR), and 95% confidence i ntervals (CI) for the effect of independent and covariate variables on one -year retention. Variable OR [95% CI] aOR [95% CI] b Level of RS use c Never Used a a Low 0.94 [0.55-1.62] 1.01 [0.58-1.75] Medium/ High 1.98 [1.12-3.44]* 1.94 [1.11-3.38]* High School GPA 4.16 [2.09-8.23]* 3.59 [1.79-7.24]* Gender Male a Female 1.12 [0.72-1.74] NA Race Other d a White 1.69 [0.97-2.95] NA Abbreviations: RS = Recreational sports; GPA = grade point average; a = referent group; b= adjusted model which includes recreational sports use level and high school GPA; c = independent variable; d = Black (non -Hispanic), Hispanic, Asian, Hawaiian/Pacific Islander, American Indian/Alaskan Native, and Multiracial. Note. *Significant difference 95% CI DISCUSSION The results of this study s upport most previous literature 19-25 and our hypothesis of a direct relationship between recreational sports participation and college student one -year retention. After controlling for HsGPA, students who participated in recreational sports more than five times per semester were nearly twice as likely to continue at the institution after one year of enrollment than students who purchased a recreational sports membership, but never used the facilities. Based on this finding, students should be encouraged to utilize the recreational sports fitn ess centers early and often in their academic careers. Results of this study do not support our hypothesis that recreational sports participation positively relates to two -year retention. The majority of study participants in our sample reached two-year r etention (n=1420, 90% of the sample) and few students purchased a membership but never used the facilities (n=67, 4% of the sample), which resulted in a low frequency of students !!,)!not retained and categorized as Ônever usedÕ (n=8, 0.5% of the sample). The l ow frequency may have resulted in a type two error. Further, previous research suggests that integration into the university is a primary predictor of student retention and is most important during the first year of college .2,3,13 -16 Therefore, it is possi ble that students in this study became integrated into the institution their first year of college, and therefore involvement through recreational sports in their second year of college wa s not critical to achieving two -year retention. Previously, we inve stigated the relationship between purchasing a membership to recreational sports in the first seme ster at the institution and two -year retention. 24 Although we did not control for confounding variables, results showed that first semester recreational sport s members were more likely to reach two -year retention than non -members (p< 0.05).24 The finding suggests that institutional involvement through recreational sports membership in the first semester of college is important to two -year retention. However, our previous study did not assess actual use of the recreational sports facilities. Therefore, to fully understand the relationship between recreational sports use and two -year retention, future research should utilize one-year recreational sports use as the independent variable. Additionally, we believe that this is the first study to investigate the relationship between recreational sports use and two -year retention; our results await replication in other college samples before final conclusions can be drawn . Theorists Alexander Astin 18 and Vincent Tinto 14,15 suggest that student integration into university culture, particularly in the first year, is extremely important for student success. More specifically, students who become involved both academically and socially in the ir first year of college are more likely to persist than students who are not involved .14,15 ,18 Further, the risk of student drop out declines drastically after th e first year of college .14,15 Taken together, the results !!,*!of our 2014 stud y and the pres ent investigation support Astin 18 and TintoÕs 14,15 theories. Involvement, through recreational sports, during the first semester and first year of college is positively related to college student one -year retention. However, involvement throu gh recreational sports across two years, does not relate to two -year retention. Therefore, in an effort of increasing retention rates, recreational sports practitioners should encourage college students to purchase and utilize a fitness membership early in their academic career s. Results of univariate analyses generally agree with recent research that reflects the ever -changing college student population s.12 However, t he findings of our study do not support previous literature relating soc ioeconomic status to retention. 13,26,27 This contradictory finding may be explained by a possible type II error resulting from the low frequency of students who were Pell grant eligible and not retained (n=39, 2.5% of the sample). Further, we chose to use Pell grant eligibi lity as a proxy for student SES. Previous investigators have used several proxy variables to estimate SES (e.g. parental educati on, income, and/or occupation) .26,27 Therefore the contradictory findings may be due to differences in what variable is used to estimate SES. More research is needed to understand how SES impacts retention in a sample of recreational sports members. Although results of this study are encouraging, there are some limitations. This investigation was conducted at a university where s tudents must purchase a membership to utilize the recreational sports fitness centers; more commonly, all students have access to fitness centers by paying a fee as part of tuition and fees. Further, an average of 92% of study participants were retained, w hich is much higher than the 2011 national average at four year public institutions for one -year retention (78.7%) .28 Therefore, results may not be generalizable to institutions that assess a recreational sports fee to all students or that have lower reten tion !!,+!rates. Additionally, use of recreational sports was calculated as use per semesters of enrollment. This method was chosen to create a variable that represents Ôregular useÕ. It is possible that a study participant utilized recreational sports a very h igh amount in one semester, and not in the next, but was categorized as a high user. Finally, covariates addressed in this study are primarily pre-entry attributes. We did not account for other variables that may impact a studentÕs decision to remain at th e institution such as hours per week working for pay or whether the students lived on or off campus. Therefore, the interpretation of recreational sports use levels and the covariates measured should be considered when comparing the results of this study to previous literature. Despite its limitations, this study has many strengths. We utilized a large sample, investigated six possible confounding variables, and investigated recreational sports use over more than one semester. This study also included defi nitions of recreational sports use levels via population data . Finally, we restricted the analysis to a single component of recreational sports, fitness centers, which allows for increased specificity compared to previous literature. In summary, results of this study show significant differences in one -year retention among recreational sports use levels; students who use the facilities more than five times per semester are most likely to be retained. No differences were identified among recreati onal sports use levels and two -year retention. These results should be shared with academic administrators to bring attention to the positive impact of recreational sports participation on student retention. Future research should investigate the relationshi p between college student retention and additional components of recreational sports, such as intramurals and club sport participation, in a variety of college student population s. Further, future investigation should identify and assess add itional confounding variables !!,,!REFERENCES !!,-!REFERENCES 1. U.S. Department of Health and Human Services. Office of Disease Prevention and Health Promotion. Healthy People 2020 . Washington, DC. Available at http://www.healthypeople.gov/2020/topicsobjectives 2020/objectiveslist.aspx?topicId=33. Accessed August 18, 2014. 2. Pascarella ET, Terenzini PT. How college affects students: Findings and insights from twenty years of research. San Fransisco (CA): Jossey -Bass; 1991. 3. Pascarella ET, Terenzini PT. How College Affects Students: A Third Decade Of Research (3rd Eds). San Fransisco (CA): Jossey -Bass; 2005. 4. Plotnikoff RC, Costigan SA, Williams RL, et al. Effectiveness of interventions targeting physical activity, nutrition and healthy weight for university and col lege students: a systematic review and meta -analysis. International Journal of Behavioral Nutrition and Physical Activity. 2015;12(1):45. 5. Sparling PB, Snow TK. Physical activity patterns in recent college alumni. Research Quarterly for Exercise and Sport . 2002;73(2):200 -205. 6. Forrester S, Ross CM, Hall S, Geary C. Using past campus recreational sports participation to explain current physical activity levels of alumni. Recreational Sports Journal . 2007;31:83 -94. 7. NIRSA/NASPA. (2010). NIRSA/NASPA consortium campus recreation impact study. descriptive summary report. Retrieved from http://www.nirsa.org/docs/Discover/ Research/Campus_Rec_Impact_Study.PDF 8. Raisman NA. The Cost of College Attrition at Four -Year Colleges & Universities. Educational Policy Insitute;2013. 9. Integrated Post Secondary Education Data System Glossary [Internet]. Washington (DC): National Center for Education Statistics. [cited 2015 april 20]. Available from : http://nces.ed.gov/ipeds/glossary/ 10. Murtaugh PA, Burns LD, Schuster J. Predicting the retention of university students. Research in Higher Education. 1999;40(3):355 -371. 11. Ishler J, Upcraft ML. The keys to first -year student persistence. Challenging and supporting the first -year student: A ha ndbook for improving the first year of college. 2005:27 -46. !!-.!12. Reason RD. Student variables that predict retention: Recent research and new developments. Journal of Student Affairs Research and Practice. 2009;46(3):850 -869. 13. Astin AW. What matters in college ?: Four critical years revisited. San Francisco (CA):Jossey -Bass; 1993. 14. Tinto V. Leaving college: Rethinking the causes and cures of student attrition. Chicago: University of Chicago Press; 1987. 15. Tinto V. Leaving College: Rethinking the Causes and Cures of Student Attrition. (2 nd Ed.). Chicago: University of Chicago Press; 1993. 16. Astin AW. Preventing students from dropping out. San Francisco(CA):Jossey -Bass; 1975. 17. Renn KA, Reason RD. College Students in the United States: Characteristics, Experiences, an d Outcomes: Characteristics, Experiences, and Outcomes. San Francisco: Jossey -Bass; 2012. 18. Astin AW. Student involvement: A developmental theory for higher education. Journal of college student personnel. 1984;25(4):297 -308. 19. Lindsey R, Sessoms E. Assessme nt of a campus recreation program on student recruitment, retention, and frequency of participation across certain demographic variables. Recreational Sports Journal. 2006;30(1):30 -39. 20. Bradley J, Phillipi, RH., Bryant, JA. Minorities benefit from their as sociation with campus recreation programs. NIRSA Journal. 1992;16(3):46 -50. 21. Belch HA, Gebel M, Maas GM. Relationship between student recreation complex use, academic performance, and persistence of first -time freshmen. Journal of Student Affairs Research and Practice. 2002;38(2):220 -234. 22. Kampf S, Teske EJ. Collegiate Recreation Participation and Retention. Recreational Sports Journal. 2013;37(2):85 -96. 23. Huesman R, Brown AK, Lee G, Kellogg JP, Radcliffe PM. Gym Bags and Mortarboards: Is Use of Campus Recreation Facilities Related to Student Success? Journal of Student Affairs Research and Practice. 2009;46(1):50 -71. 24. Danbert SJ, Pivarnik JM, McNeil RN, Was hington IJ. Academic success and retention: the role of recreational sports fitness facilities. Recreational Sports Journal. 2014;38(1):14 -22. 25. Henchy A. The influence of campus recreation beyond the gym. Recreational Sports Journal. 2011;35(2):174 -181. !!-%!26. Walpole M. Socioeconomic status and college: How SES affects college experiences and outcomes. The review of higher education. 2003;27(1):45 -73. 27. Cabrera F, Stampen JO, Hansen WL. Exploring the effects of ability to pay on persistence in college. Review of H igher Education. 1990:13(3):303 -336. 28. U.S. Department of Education, National Center for Education Statistics, Integrated Postsecondary Education Data System (IPEDS), Spring 2008 through Spring 2014. !!-&!CHAPTER 5 MANUSCRIPT THREE Chapter five addresses Specific Aims three and four. Specific Aim 3 : To investigate the relationships between recreational sports fitness center use over four years and college student academic success defined by cumulative grade point average (GPA) and degree completion . Specific Aim 4 : To investigate the impact of year in school on the relationship between recreational sports use and grade point average. INTRODUCTION Currently, the United StatesÕ has a college graduation rate of 37.7% and is ranked 13 th, among 30 cou ntries participating in the Organization for Economic Co -operation and Development .1 FinlandÕs graduation rate is first in the world at 62.6% follo wed by Slovak Republic at 57.1%. 1 President Obama has pushed for an increase in college graduates. He has set a new goal for the country, indicating, ÒÉthat by 2020, America would once again have the highest proportion of college graduates in the world.Ó 2 Previous research indicates a number of variables that partially explain why college students do or do not gr aduate (e.g., AP credits 3, college major 4, first year grade point average (GPA) 4, race 4-7, gender 4-8, socioeconomic status 5,9). However, the accuracy of each variable in predicting Bachelor 's degree attainment varies across studies 7,8 , although there is g eneral agreement that college GPA is the best predictor of college student degree attainment .4 Students with higher GPAs, particularly in the first year of college are more likely to be retained and graduate than students with lower GPAs .10-12 Many factors !!-'!shown to predict college degree attainment also relate to college student GPA, including race 10, gender 10,11 , and high school GPA10-12. Theorists Alexander Astin 13 and Vincent Tinto 14,15 suggest that college student success depends more on s tudent involvement than pre entry attributes (e.g., race, gender, and high school GPA ). AstinÕs 13 theory suggests that the more a student is involved at an institution, the more likely the student is to learn and succeed. In his theory, involvement is defined as, ÒÉ the amount of physical and psychological energy that the student devotes to the academic experience.Ó Astin13 notes that uninvolved students are more likely to drop out of college than involved students. Similarly, TintoÕs Theory of Departure 14,15 focuses on student integration. His theory suggests that student pre -entry attributes (family ba ckground, skills and abilities, prior schooling) influence student goals and his/her commitment to the institution prior to attending college, which in turn impact institutional experiences and academic/social integration while attending college. The level of academic/social integration in college influences student goals and institutional commitment during college and ultimately affects the decision to continue attending the institution. It is important to note that the level of involvement in the first ye ar of college is considered most important in college student success. Additionally, while student involvement is critical to college student success, too much involvement can be detrimental. Astin and Tinto both specify the importance of meaningful involv ement, especially that which relates specifically to academics. Students who are under/over -involved and who do not manage their time effectively are less successful in college than students who identify time to study effectively and ways to become involve d as part of a balanced lifestyle. 16 Similar to the goals of President Obama, academic administrators often define success by academic measures such as GPA and degree attainment. However , student affairs professionals !!-(!focus more broadly on overall student development. The stu dent affairs department at our university states, ÒThe programs, services, and facilities providedÉare designed to create a stimulating and supportive environment that enhances the personal development, learning, educational suc cess, and caree r preparation of all students.Ó 17 Their mission is to ÒÉprepare graduates for participation and leadership in an increasingly diverse and complex global society.Ó 17 One area of personal development is the development of healthy lifestyle beh aviors such as regular participation in physical activity (PA). In 2008, the United States Department of Health and Human Services (DHHS) issued PA guidelines indicating adults should participate in at least 150 minutes a week of moderate - or 75 minutes a week of vigorous -intensity aerobic PA, or an equivalent combination. 18 Adults should also participate in ÒÉmuscle strengthening activities that are moderate or high intensity and involve all major muscle group s on two or more days a weekÉ Ó 18 In 2011, 56.8 percent of college aged adults met aerobic PA guidelines, 44.1% met strength -training guidelines , and 30.7 percent met both. 19 Promotion of PA to college students is critical to increase the proportion of adults meeting PA guidelines. Recreational sports is one avenue for students to develop the healthy lifestyle behavior of regular participation in PA. Further , based on the theories of Astin13 and Tinto 14,15 , participation in recreational sports is an opportunity for involvement at the university, which may help increase student academic success . However, f unding is necessary to ensure the continued promotion of PA through recreational sports. Although academic administrators may believe PA is important, it is likely secondary to academic success. Therefor e, research is necessary to investigate t he relationship between recreational sports participation and academic success . !!-)!Research investigating the relationship between participation in recreational sports and GPA is scarce. Four studies 20-23 have assessed this relationship . Three 20-22 found positive associations between recreational sports and GPA, and one 23 found no association . Belch, Gebel, and Ma as20 and Kampf and Teske 21 utilized u niversity records to compare recreational sports fitness facility use and GPA. Further, Kampf and Teske 21 assessed the difference in GPA between 1) recreational sports student employees and non -employees, and 2) club sport participants and non -participants. Recently, we 22 took a slightly different approach and assessed differences in GPA between recreational sports members and non -members . Differences in college GPA between recreational sports participants and non -part icipants of all three studies ranged between 0.09 and 0.13, with participants earning a higher GPA than non -participants. Although all three studies 20-22 utilized large sample sizes, n=11,076, n=3,308, and n=4,843 respectively, and performed similar analys es, no investigative group accounted for potential confounding variables or assessed levels of recreational sports use . In contrast, Frauman 23 found no difference in GPA between recreational sports users and non -users using a smaller sample (n=385). Theref ore, due to limitations of previous investigations and conflicting findings, more research is needed. Despite the importance placed on Bachelor's d egree attainment in the United States, only one published study 24 has assessed the relationship between par ticipation in recreational sports and Bachelor's degree attainment. Huesman et al. ,24 utilized logistic regression to assess the relationship between first semester participation in recreational sports (Fall semester of freshman year) and five -year degree attainment (n=5,211). The authors included the ratio of credits attempted to credits completed, CÕs received, DÕs received, course withdraw count, ACT/SAT score, remedial classes taken, remedial classes failed, athlete st atus, gender, race, socioeconomic !!-*!status, and off -campus housing as covariates. Campus recreation use was entered in the model as a continuous variable. In a secondary analysis, the authors 24 compared ten entries into the recreational sports center to 25 entries . These values were chosen based on the mean and standard deviation of campus recreation use in the first semester of college . Results suggested that utilizing the recreational sports fitness center 25 times in the first semester of freshmen year inc reased the likelihood of 5 -year graduation by 2% compared to students who used the fitness center 10 times or less in the first semester. Although results of the study are in a positive direction , there is a clear need for more research in this area. Due t o a limited number of investigations and conflicting findings, research is needed to investigate the relationships between recreational sports and 1) college student GPA and 2) Bachelor's degree attainment. The purpose of this study was t hree -fold. The first purpose was to investigate the relationships between four -year recreational sports fitness center use and four -year cumulative GPA ; the second to investigate the relationships between four -year recreational sports fitness center use and fiv e year Bachelor's degree attainment. Due to the potential negative effects of over/under involvement on college student success, w e hypothesized that students using the recreational sports fitness center the least and m ost often would earn lower four -year cumulative GPAs and be less likely to graduate in five -years than those who used the facilities a moderate amount. Previous authors have largely focused on recreational sports use in the first semester of college, therefore, o ur third purpose was to invest igate the impact of year in school on the relationship between recreational sports use and yearly GPA. We hypothesized an interaction between year in school and recreational sports use on yearly GPA, with first year students experiencing the greatest effe ct of recreational sports use on GPA. !!-+!METHODS Data Collection Data were extracted from a university database housed at a large Midwestern university. Study participants included all first time degree seeking freshmen students who graduated from high school in Spring, 2010. Participants were in their first semester at the university in Fall, 2010, were not student athletes, had no prior coll ege class experience, and purchased a fitness center membership in a Fall or Spring semester at least once in their academic careers. These criteria ensured all subjects had made similar academic progress, had the goal of achieving a college degree, and ha d access to the university recreational sports facilities for all earned college credits. Since varsity athletes have different college schedules than typical college students, they were excluded from the sample. Academic Success Acade mic success was def ined by four -year cumulative GPA (4.0 point scale) and five -year Bachelor's degree attainment (yes/no). Four -year cumulative GPA was calculated as the average of all previous enrolled semester GPAs. University curricula are generally structured for four -year BachelorÕs degree attainment. However, universities report, to the National Center for Education Statistics, the number of students who graduate within 150% of normal time to completion. Further, the majority of previous literature in higher education a ssesses five - and/or six-year graduation rates. Therefore, to allow for comparability, we assessed five -year degree attainment; students who completed a BachelorÕs degree by May, 2015 were considered to have achieved five -year degree attainment. Academic s uccess was defined by comparing yearly cumulative GPAs among study participants, calculated as the average of all enrolled Fall and !!-,!Spring semester GPAs in a given year, to assess the impact of year in school on the relationship between recreational sports use and GPA. Recreational Sports The u niversity recreational sports department offers a variety of opportunities for students to participate in PA including outdoor facilities, indoor pools, courts, cardio equipment, free weights, and strength training machines. Access to some of the university recreational sports facilities is included in tuition costs. However, an additional fee is charged to students who wish to purchase a fitness membership, which provides access to the campus recreation fitness ce nters. There are two fitness centers on campus totaling ~20,000 square feet of space. These fitness centers house cardio machines and resistance training equipment. To verify membership, students swipe their student IDs each time they enter a recreational sports fitness center, and their information is recorded electronically. For this investigation, total number of times the study participants swiped into the fitness center during each semester (Fall or Spring) of enrollment was gathered and utilized to i dentify level of recreational sports use (never used, low use, medium use, high use). To our knowledge, there are no established cut points to define levels of recreational sports use. Therefore we identified cut points for four groups (never used, low us e, medium use, and high use) by utilizing data from all students enrolled at the institution between Fall 2010 and Spring 2015. The level Ônever usedÕ was defined by purchasing a membership, but never utilizing the recreational sports fitness centers. To i dentify cut points for low, medium, and high use, recreational sports fitness center use per semesters of membership was calculated for all students who purchased a membership and used the recreational sports fitness centers . We !!--!utilized a quartile split t o first identify four groups of recreational sports fitness center use. The middle two groups were combined resulting in three levels of recreational sports use (low use, medium use, high use). This allowed us to identify the extreme ends (low and high) o f the recreational sports use curve, thus improving comparability and reducing potential type two errors. Study participants were categorized into one of the four population derived levels based on the number of times they swiped their ID into a recreation al sports fitness center per semesters of enrollment . Semesters of enrollment over a four -year period were utilized to assess four -year GPA and five -year BachelorÕs degree attainment. Semesters of enrollment per year were utilized to address our secondary purpose, which assesses the impact of year in school on the relationship between recreational sports and GPA. Covariates Based on previous literature 4,6,8 -12 and information available from the u niversity Registrar, we identified six possible covariates that may impact the relationship between recreational sports use and 1) four -year cumulative GPA and 2) five -year Bachelor's degree attainment; these variables included: high school GPA (HsGPA), ACT score, gender , race , and first semester Pell grant eligibility. HsGPA and first year college GPA were based on a 4.0 scale. ACT score could range from 1 -36. All other covariates were categorical: gender (male/female), race (white/other, which included American Indian/Alaskan Native , Asian, Black [non -Hispanic], Hawaiian/Pacific Islander, Hispanic ethnicity , and Multiracial ), and Pell grant eligibility (eligible/not eligible). !!%..!Statistical Analysis Recreational sports use cut points established from popu lation data were defined similarly for all purposes, categories included: never used (purchased a membership, but never used the facilities), low use (0 -5 times/membership), medium use (>5 -26 times/membership), and high use (>26 times/membership). The stud y sample was categorized into one of the four groups based on number of uses/ total semesters of enrollment over four years for our first and second purposes, and number of uses/semester of enrollment per year for our third purpose . To address our first purpose, the relationship between recreational sports use and GPA, we utilized four -year recreational sports use level as the independent variable and four -year cumulative GPA as the outcome variable. Possible covariates included: HsGP A, ACT score, gender, race , and Pell grant eligibility. Descriptive information is reported with means ± standard deviations and percentages. We utilized an analysis of variance (ANOVA) and Pearson correlations to assess the unadjusted relationship betwee n four -year cumulative GPA and 1) recreational sports use level and 2) all covariates . Variables related to four year cumulative GPA (p<0.10) were considered for adjusted analyses. To adjust for possible covariates, we utilized an analys is of covariance (ANCOVA). LSD post h oc tests were completed to identify specific differences in four -year cumulative GPA between recreational sports use groups. Four -year recreational sports use (ref: medium use) and five -year bachelorÕs degree attainment were used as our independent and dependent variables to address our second purpose. The Ômedium useÕ category of recreational sports use levels was chosen as a reference category based on our inverted U shape hypothesis, which suggested that students using the recreational sports fitness center the least and most often would be less likely to graduate in five -years than !!%.%!those who used the facilities a moderate amount. Possible covariates included: HsGPA, first year college GPA, ACT score, gender (ref: male), race (ref: othe r), and Pell grant eligibility (ref: eligible). Descriptive information is reported with means ± standard deviations and percentages. Chi-square and one -way ANOVA were utilized to assess the independent relationship between each variable (independent and covariates) and five -year degree attainment. We developed binary logistic regression models to describe the relationship between recreational sports use a nd five -year BachelorÕs degree attainment. Variables that were related to five -year BachelorÕs degree attainment (p<0.10) were entered into the models. Recreational sports use was entered into the models first, followed by each covariate based on the stren gth of relationship to four -year degree attainment. Variables were removed from the model if they did not significantly (p<0.05) impact the prediction of five -year degree attainment or did not change the odds ratio associated with recreational sports use b y more than ten percent. We calculated unadjusted odds ratios (OR) and adjusted odds ratios (aOR) with associated 95% confidence intervals for achieving five -year degree attainment vs. not, according to recreational sports level. Our third purpose was to i nvestigate the impact of year in school on the relationship between recreational sports use and yearly GPA. We utilized a repeated measures analysis and included yearly recreational sports use level (never used, low, medium, high) and year in school (1-4) as independent variables. The dependent variable was yearly GPA. Year in school was considered the repeated time point. LSD post hoc tests were assessed to identify differences in GPA among year in school and recreational sports center use levels. !!%.&!RESU LTS The sample (n=1548) utilized to assess differences in four -year cumulative GPA among recreational sports use levels was 52% female, 23% Pell grant eligible, and 86% White. Mean±standard deviation high school GPA, ACT, first year cumulative GPA, and fo ur year cumulative GPA were, 3.58±0.29, 24.61±2.72, 3.14±0.52, and 3.22±0.42 respectively. Females earned hi gher four -year GPAs than Males (3.29±0.39 vs. 3.15± 0.45), students not eligible for Pell grants earned higher four -year GPAs than eligible students (3.23± 0.40 vs. 3.14±0.47), and Whites earned higher four -year GPAs than other races (3.26± 0.40 vs. 3 .00±0.46), Descriptive information by level of recreational sports use is listed in Table 5.1. Univariate analyses revealed that four year GPA wa s significantly related to 1) recreational sports use level, 2) gender, 3) Pell grant eligibility, and 4) race ; ACT score and HsGPA were also significantly related to four -year GPA, r=0.21, p<0.001 and r=0.41, p<0.001, respectively. As seen in Table 5. 2, four -year cumulative GPA differed by recreational sports use level after adjusting for gender, Pell grant eligibility, race, ACT, and HsGPA (p<0.05). Students who were high (3.27±0.45) and medium (3.26±0.41) recreational sports users earned significantly h igher four -year cumulative GPAs than the low users (3.20±0.42) and those who never used the facilities (3.08±0.47). There was no significant difference in four -year cumulative GPA between low users and never used or high and medium recr eational sports user s (Table 5.3). The sample (n=1788) utilized to assess the relationship between recreat ional sports use level and five -year BachelorÕs degree attainment was 52% female, 23% Pell grant eligible, and 86% White. Mean±standard deviation high school GPA, ACT, and first year cumulative GPA were 3.57±0.30, 24.6±2.76, and 3.08±0.60 respectively ( Table 5. 3). As seen in Table 5. 4, !!%.'!univariate analysis showed significant differences in five -year graduation between genders, Pell grant eligibility, and races . High schoo l GPA and first year cumulative GPA were positively related to five year graduation. AC T score was not related to five -year graduation . Most importantly for our study purpose, no differences were found among four -year recreational sports use level and five -year graduation. Therefore, due to the null finding between four -year recreational sports use levels and five -year graduation, logistic regression analysis was not performed. The sample utilized to assess an interaction between year in school and recreational sports use level on yearly GPA was 49% female, 77% white, and 22% Pell grant eligible. Mean±standard error GPA by year in school and by recreational sports u se level is listed in tables 5.6 and 5.7 respectively. Results of the repeated measures analysis indicated a significant relationship between year in school and yearly GPA (F 3, 2047 =7.143, p<0.001), and between yearly level of recreational sports use and yearly GPA (F 3,5530 =4.60, p=0.003). Specifically, the highest GPAs were earned in the first and fourth years, and the lowest in the second and third years of college (Table 5. 4). High users of recreational sports facilities earned the highest GPAs and students who purc hased a membership but never used it earned the lowest (Table 5. 5). The interaction of year in school * yearly level of recreational sports use on yearly GPA was not statistically significant (F9,2601 =1.73, p=0.08). As seen in F igure 5.1, the interaction t rended towards significance; the largest difference in GPA among re creational sports use levels was in the first year of college and the smallest in the fourth year of college. !!%.(!Table 5.1 . Means, standard deviations (SD) , and percentages of dependent and covariate variables separated by recreational sports use level. Variable Total Never Used Low Medium High n 1548 64 820 538 126 frequency (percentage) Gender* Male 735 (47.6) 25 (40.0) 330 (40.4) 278 (51.7) 102 (81.1) Female 813 (52.4) 39 (60.0) 490 (59.6) 260 (48.3) 24 (18.9) Pell Eligible 349 (22.5) 15 (23.1) 198 (24.1) 111 (20.7) 25 (19.7) Not Eligible 1199 (77.5) 49 (76.9) 622 (75.9) 427 (79.3) 101 (80.3) Race White 1333 (85.7) 53 (83.1) 692 (84.2) 471 (86.9) 116 (91.3) Other a 215 (14.3) 11 (16.9) 128 (15.8) 67 (13.1) 10 (8.7) Mean+SD High School GPA 3.58±0.29 3.53±0.30 3.57±0.30 3.60±0.26 3.58±0.31 ACT 24.61±2.72 24.50±3.03 24.57±2.81 24.67±2.59 24.61±2.52 First year cGPA* 3.14±0.52 2.87±0.59 3.10±0.53 3.22±0.49 3.25±0.53 Four year GPAb* 3.22±0.42 3.08±0.47 3.20±0.42 3.26±0.41 3.27±0.45 Abbreviations: GPA = grade point average, cGPA= cumulative GPA; a = Black (non -Hispanic), Hispanic, Asian, Hawaiian/Pacific Islander, American Indian/Alaskan Native, Multiracial; b = dependent variable. Note. *Globa l statistical difference among recreational sports use level groups at p<0.05. !!!!%.)!Table 5. 2. Analysis of covariance for four -year cumulative grade point average. Effects Sum of squares df Mean square F Value P value RS Use Level a 3.04 3 1.01 7.36 <0.001* Genderb 7.22 1 7.22 52.42 <0.001* Pellc 0.54 1 0.54 3.94 0.047* Race d 3.34 1 3.34 24.24 <0.001* ACT 2.28 1 2.28 16.56 <0.001* High School GPA 27.65 1 27.65 200.73 <0.001* Abbreviations: RS = Recreational sports; GPA = grade point average aNever used, Low, Medium, High bMale, female cEligible, not eligible dWhite, Other (Black [non Hispanic ], Hispanic, Asian, Hawaiian/Pacific Islander, American Indian/Alaskan Native , Multiracial) *Significant difference at p<0.05. Table 5. 3. LSD post h oc comparisons of four -year grade point average among recreational sports use levels. Four year Recreational Sports Use Level 1 Four Year Recreational Sports Use Level 2 Mean Difference 1-2 Standard Error p-value Never Used Low Use -0.09 0.07 0.057 Medium Use -0.15 0.08 0.002* High Use -0.21 0.09 <0.001* Low Use Medium Use -0.06 0.03 0.003* High Use -0.12 0.06 0.001* Medium Use High Use -0.06 0.07 0.12 Note. *Statistical difference between recreational s ports use levels at p<0.05. !!%.*!Table 5. 4. Means, standard deviations (SD) , frequencies, and percentages of independent and covariate variables separated by five -year graduation. Variable Total Graduated Did not Graduate n 1788 1352 436 frequency (percentage) Level of RS Use a Never Used 81 (4.5) 58 (4.3) 23 (5.3) Low 926 (52.8) 704 (52.1) 222 (50.9) Medium 622 (34.8) 479 (35.4) 143 (34.8) High 159 (8.9) 111 (8.2) 48 (11.0) Gender*€ Male 866 (48.4) 595 (44.0) 271 (62.2) Female 922 (51.6) 757 (56.6) 165 (37.8) Pell Grant Eligibility* € Eligible 414 (23.2) 273 (20.2) 141 (32.3) Not Eligible 1374 (76.8) 1079 (79.8) 295 (67.7) Race* € White 1529 (85.5) 1197 (88.5) 332 (76.1) Other b 259 (14.5) 155 (11.5) 1704 (23.9) mean±SD High School GPA* € 3.57±0.30 3.60±0.29 3.47±0.31 ACT composite score 24.60±2.76 24.64±2.59 24.46±3.24 First Year c GPA*€ 3.08±0.60 3.21±0.48 2.69±0.74 Abbreviations: RS = Recreational sports; GPA = grade point average; cGPA = cumulative grade point average; a = independent variable; b = Black (non -Hispanic), Hispanic, Asian, Hawaiian/Pacific Islander, American Indian/Alaskan Native, Multiracia l. Note. *Global statistical difference between retention groups at p<0.05 €Meets criteria to enter logistic regression model, p<0.1 Table 5. 5. Means ± standard e rrors (SE) of cumulative grade point average by year in school. Year in school Mean ±SE Year 1 3.02±0.02 Year 2* 2.98±0.02 Year 3* 2.94±0.02 Year 4 € 3.02±0.02 Note. *Significantly different than year one (p<0.05) €Significantly different than year three (p<0.05) !!%.+! Table 5. 6 Means ± standard e rrors (SE) of cumulative grade point average by recreational sports use l evel. Recreational Sports Use Level Mean ±SE Never Used € 2.97±0.02 Low use* € 2.84±0.02 Medium use 3.00±0.02 High use 3.04±0.03 Note. *Significantly different than m edium Use (p<0.05) €Significantly different than high use (p<0.05) Figure 5.1. Graphical depiction of the interaction between year in school and recreational sports use level on yearly grade point average. 2.60 2.70 2.80 2.90 3.00 3.10 3.20 3.30 1 2 3 4 Mean GPA Year in School High Medium Low Never Recreational Sports Use !!%.,!DISCUSSION In regards to our first purpose, p revious research indicates a positive relationship between a dichotomized measure (user/non -user) of recreational sports use and first/second year college student GPA. The results of the present study extend previous literature by investigating four -year G PA and indicate that study participants who participated in recreational sports regularly (>5 times per semester) over four years, earned higher four -year cumulative GPAs than study participants who purchase d a recreational sports membership, but d id not u se the facilities. Taken together, previous literature and the present study indicate strong evidence for a positive relationship between recreational sports use and cumulative GPA. Although positive, the results of the present investigation do not suppo rt our hypothesis. We hypothesized that students participating a low and high amount in recreational sports would earn lower GPAs than students participating a moderate amount. This hypothesis was based upon previous literature that indicates low/high leve ls of student involvement may be detrimental to student success. Our recreational sports use level development was based on population data and indicated that high users entered the fitness facilities 26 or more times per semester. There are 16 weeks in a semester. Therefore, if a stu dent was a high level and consistent user then they entered the fitness facilities less than two times per week. We anticipated the high level of recreational sports use category to include a larger number of fitness facility e ntries, which was an incorrect assumption. This level of involvement (26+ fitness center entries per semester) is not likely to induce detrimental effects on student success. Researchers have not investigated possible mechanisms for th e positive relations hip between recreational sports fitness center use and college student GPA . It is possible that students who participate in recreational sports develop a sense of belonging to the university, !!%.-!which in turn increases their commitment to the university and student success. Additionally, previous research indicates a positive relationship between participation in PA an d college student grade average. 25 Students who participate in recreational spor ts may engage in more PA than students not participating, and thus may experience increased cognitive functioning. Future research should investigate the relationship between PA and recreational sports as a possible mechanism for the relationship between r ecreational sports use and cumulative GPA. Based upon student success literature and one previous investigation related to recreational sports 24, we hypothesized that low and high users of recreational sports would be less likely to graduate in five years than moderate users. Results of the present investigation do not support our hypothesis. Previously, Huesman et al. 24, completed the only study that investigated the relationship between recreational sports use and five -year graduation. Their results indi cated that students who participated in recreational sports 25 times in their first semester at the University were more likely to graduate in five years than students who participated 10 times . No relationship was found between four -year recreational spor ts use and five -year graduation in the present study. However, our methodology differed from that of Huesman et al. 24 We assessed recreational sports use across four years. Never used, in the present study, was a single group defined as purchasing a membership, but never using the facilities. We also controlled for different confound ing variables than Huesman et al .24 Additionally, student success theorists argue that institutional involvement in the first year of college, as opposed to subsequent years, is critical for retention and BachelorÕs degree attainment .13-15 Therefore, it is possible that regular recreational sp orts participation across four years is not essential to complete a BachelorÕs degree within five years. Future research should investigate the role of first year recreational sports use in five -year BachelorÕs degree attainment. !!%%.!Previous literature sugge sts that student experiences in the first year of college are the best predictors of college success .13-15 Therefore , we hypothesized an interaction between year in school and yearly recreational sports use on yearly GPA, with first year students experienc ing the greatest effect of recreational sports use on GPA. Results do not support our hypothesis. Participation in recreational sports positively related to GPA similarly across all years of college. However, results trended towards a significant relations hip (p=0.08). Based on our hypothesis, we would expect a larger range of mean GPAs across recreational sports use levels in the first year of college compared to subsequent years . Figure 5.1 illustrates that recreational sports use may impact GPA in the first, second, and third years of college similarly, but have little impact on the fourth year of college. The similar effects on GPA in years one, two, and three, may explain t he lack of statistica l finding. This trend does not support our hypothesis and instead indicates that recreational sports use has a positive impact on GPA across the first three years of college. It is possible that recreational sports participation correlates directly to participation in PA. If so, students participating in recr eational sports may experience the benefits of PA, such as increased cognitive ability and decreased stress, which in turn may positively influence student academic success. Additionally, recreational sports participation may relate to a sense of belonging similarly across year one, two, three, and four. Therefore , the continued positive relationship between recreational sports and GPA over four years may be due to an increased sense of belonging for recreational sports users as compared to non -users. Furth er, the identified relationship may be due to the combined, positive cognitive -, social -, and emotional - benefits associated with physical activity and feeling a sense of belonging at the university. This is the first study, to our knowledge, that investig ated an interaction between year in school and recreational sports use on GPA. Future research should !!%%%!investigate this interaction in a variety of college populations, and aim to identify possible mechanism s for the relationship. Limitations and Strength s Although the results of this study are encouraging and extend previous literature, there are some limitations that should be discussed . All study participants chose to purchase a membership to utilize the recreational sports fitness facilities at some point in their academic careers. Most institutions do not implement a membership model, but instead provide access to all students by asse ssing a fee that is included in the studentsÕ cost of attendance. Therefore, caution should be taken when generalizing these results to all universities. Although we accounted for a number of possible covariates when assessing the relationship between recr eational sports use and four -year GPA, additional variables may impact academic success. Examples could include place of residence (on/off campus), college major, and hours per week working for pay. Additionally, recreational sports use was calculated to r epresent regular use across four years. Given our method of determining facility use, it was not possible to determine whether it was spread out evenly across the membership period, or whether it clustered in some meaningful fashion. It appears that we ar e the first investigators to analyze the relationship between recreational sports use level and GPA over time and the findings of this investigation may only be applicable to large, public universities. Therefore, future investigations should repeat this s tudy at various types of institutions. Despite limitations, the present investigation has many strengths. This study included a large sample size and was the first to investigate the relationship between four -year recreational sports use levels and 1) four -year GP A and 2) five year graduation. We also addressed multiple !!%%&!covariates in analyses to assess these relationships, which substantiates our findings. Finally, we focused on a specific component of recreational sports (fitness centers), which focuses th e interpretation and future comparison of our findings. With respect to our third purpose, this is the first study to investigate whether an interaction exists between year in school and yearly recreational sports use on GPA. Studies relating recreational sports use to college student academic success primarily utilize analysis of covariance or logistic regression to control for potential covariates. Although beneficial, it is not possible to measure and control for all potential covariates. Our repeated m easures analysis, utilized to address our secondary purpose, inherently controls for all covariates within our given sample, and therefore is a substantial strength to the present study. Summary In summary, our results show that four -year recreational sports use is positively related to four -year GPA. However, there are no apparent differences in five -year BachelorÕs degree attainment among f our-year recreational sports use levels . Finally, while yearly level of recreational sports use is positively related yearly GPA this relationship does not differ significantly by year in school. Based on our st udy results, we recommend that r ecreationa l sports departments share these findings with academic adm inistrators to increase awareness of this modifiable factor (recreational sports participation) related to college student academic success . In addition, students should aim to participate in recreational sports at least five times per semester, as this ma y play a role in increased yearly GPAs, and increased four -yea r cumulative GPA. Future studie s should be designed to investigate possible mechanisms for the relationship between recreational sports use and GPA, should assess the relationship between !!%%'!first year recreational sports use and five -year bachelorÕs degree attainment, and should implement our methods in samples derived from various types of higher education institutions. !!%%(!REFERENCES !!%%)!REFERENCES 1. Youth Indicators 2011 [Internet]. Washing (DC): National Center for Education Statistics [cited 2015 april 20} Available from: https://nces.ed.gov/pubs2012/2012026/tables/table_23.asp 2. Obama BH. Remarks by the president on the American graduation initiative. Speech presented at Macomb Community College, Warren, MI. 2009. 3. Adelman C. The Toolbox Revisited: Paths to Degree Completion From High School Through College. US Department of Education. 2006. 4. DesJardins SL, Kim D -O, Rzonca CS. A nested analysis of facto rs affecting bachelor's degree completion. Journal of College Student Retention: Research, Theory and Practice. 2003;4(4):407 -435. 5. Adelman C. The Toolbox Revisited: Paths to Degree Completion From High School Through College. US Department of Education. 2006. 6. DeAngelo L, Franke R, Hurtado S, Pryor JH, Tran S. Completing college: Assessing graduation rates at four -year institutions. Higher Education Research Institute, Graduation School of Education & Information Studies, University of California, Los Angeles; 2011. 7. Astin AW, Oseguera L. Degree attainment rates at American colleges and universities: Revised edition. Higher Education Research Institution, University of California Los Angeles. 2005. 8. Astin AW. Degree Attainment Rates at American Colleges and Universities: Effects of Race, Gender, and Institutional Type. ERIC; 1996. 9. Walpole M. Socioeconomic status and college: How SES affects college experiences and outcomes. The review of higher educati on. 2003;27(1):45 -73. 10. Ting S -MR, Robinson TL. First -year academic success: A prediction combining cognitive and psychosocial variables for Caucasian and African American students. Journal of College Student Development. 1998. 11. DeBerard MS, Spielmans G, Jul ka D. Predictors of academic achievement and retention among college freshmen: A longitudinal study. College student journal. 2004;38(1):66 -80. 12. Cohn E, Cohn S, Balch DC, Bradley J. Determinants of undergraduate GPAs: SAT scores, high -school GPA and high -school rank. Economics of Education Review. 2004;23(6):577 -586. !!%%*!13. Astin AW. Student involvement: A developmental theory for higher education. Journal of college student personnel. 1984;25(4):297 -308. 14. Tinto V. Leaving college: Rethinking the causes and cures of student attrition. Chicago: University of Chicago Press; 1987. 15. Tinto V. Leaving College: Rethinking the Causes and Cures of Student Attrition. (2 nd Ed.). Chicago: University of Chicago Press; 1993. 16. Britton BK, Tesser A. Effects of time -management prac tices on college grades. Journal of educational psychology. 1991; 83(3):405. 17. Maybank DB. About student affairs and services. n.d. Retrieved from http://www.vps.msu.edu/ 18. U.S. Department of Health and Human Services. 2008 physical activity guidelines for Americans. Be active, healthy, and happy. Washington. DC: U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion, Office of the Assistant Secretary for Health, Office of the Sec retary; 2008. 19. Centers for Disease Control and Prevention (CDC). Adult participation in aerobic and muscle -strengthening physical activities --United States, 2011. MMWR. Morbidity and mortality weekly report. 2013; 62(17):326 -330. 20. Belch HA, Gebel M, Maas G M. Relationship between student recreation complex use, academic performance, and persistence of first -time freshmen. Journal of Student Affairs Research and Practice. 2002;38(2):220 -234. 21. Kampf S, Teske EJ. Collegiate Recreation Participation and Retention. Recreational Sports Journal. 2013;37(2):85 -96. 22. Danbert SJ, Pivarnik JM, McNeil RN, Washington IJ. Academic success and retention: the role of recreational sports fitness facilities. Recreational Sports Journal. 2014;38(1):14 -22. 23. Frauman E. Differences between participants and non -participants of campus recreation offerings across demographic variables and perceptions of the college experience. Recreational Sports Journal. 2005;29(2):156 -165. 24. Huesman R, Brown AK, Lee G, Kellogg JP, Radcliffe PM. Gym Bags and Mortarboards: Is Use of Campus Recreation Facilities Related to Student Success? Journal of Student Affairs Research and Practice. 2009;46(1):50 -71. 25. Danbert SJ, Mudd LM, Pivarn ik JM. Meeting ACSM physical activity recommendations and college student academic s uccess. Journal of American College Health. In Review. !!%%+!CHAPTER 6 SUMMARY AND CONCLUSIONS University recreational sports departments are designed to promote physical activity (PA) and provide ample opportunities for college students to develop healthy lif estyle behaviors. Recreational s ports departments often include amenities and facilities such as, personal training, relaxation rooms, pools, gyms, courts, and fitness centers including weight lifting and cardio equipment. Previous research indicates that participation in recreational sports during college positively relates to PA participation post -college .1 Therefore, PA promotion to college students, through recreati onal sports, may help to increase the proportion of adults meeting PA guidelines .2 Adequate funding is necessary for the promotion of PA through recreational sports. However, in times of tight university budgets, obtaining adequate funding is not an easy task and university administrators may consider PA participation beneficial yet ancillary to student academic success. Administrators at the institution assessed in this dissertation have charged university departments with proving their worth in student success. Therefore, to increase the likelihood of funding, the university recreational sports department must explain their impact on student success. Previous literature indicates a positive relationship between recreational sports participation and 1) college student grade point average (GPA) 3-5, 2) retention 4,5 , and 3) five -year BachelorÕs degree attainment 6. However, previous literature is limited by simplistic statistics and self-report of the independent and dependent variables. Additionally, few studies address potential confounding variables, and most dichotomize samples into recreational sports users an d non-users. Finally, most investigators assess recreational sports use in the first year of college. !!%%,!The research presented in this dissertation extends previous literature by 1) identifying levels of recreational sports use within a single component of r ecreational sports (fitness centers), 2) measuring recreational sports use over time, 3) addressing potential confounding variables, and 4) implementing more complex analytics. Results of th e research presented in this dissertation indicate that students who purchase a recreational sports fitness center membership in their first semester of college are more likely to 1) achieve sophomore status after two consecutive semesters, 2) earn higher GPAs and complete more credits after four consecutive semesters, and 3) return for a second and third year of college, than students who do not purchase a membership. Further, among recreational sports fitness center members, GPA and likelihood of one -year retention vary by recreational sports use level. Members who ut ilize the fitness center more than five times per semester over four years earn higher four -year cumulative GPAs than members who use the facility zero to five times. Additionally, recreational sports use level was found to be positively related to yearly GPA. Stude nts using the fitness centers more than five times per semester earn higher GPAs than those using the fitness centers one to five times. However, this relationship does not differ by year in school. With respect to retention, members wh o utilize the fitness centers more than five times per semester throughout their first year of college are more likely to return for a second year of school than members who never used the facilities. Two -year retention does not differ by recreational sports use le vel among members. The relationship between level of recreational sports use among members and 1) achieving sophomore status and 2) credits completed were not assessed. Finally, among members, four -year recreational sports use level is not related to Bache lorÕs degree attainment within five years. !!%%-! The results of the studies presented in this dissertation generally agree with previous literature. 3-5 However, two -year recreational sports use was not related to two -year retention, nor was four -year recreatio nal sports use related to five -year BachelorÕs degree attainment. These findings do not agree with previous investigations. Few students did not reach two -year retention (n=163, 10% of sample), and few students purchased a membership and never used the fac ilities (n=67, 4% of sample), which resulted in a low frequency of students who were not retained and categorized as never used (n= 8, 0.5% of the sample). Additionally, only 25% of our sample failed to earn a BachelorÕs degree in five years and few purchas ed a membership, but did not use the facilities (n=81, 5% of sample), which resulted in a low frequency of students who did not graduate and were categorized as never used (n=23, 1.3% of the sample). Therefore, the difference of findings may be due to a ty pe two error due to low statistical power. Application of theoretical beliefs may also provide insight into our findings. Theorists Alexander Astin 7 and Vincent Tinto 8,9 suggest that student integration into u niversity culture, particularly in the first y ear, is extremely important for student retention and persistence. Previous literature, indicating a positive relationship between recreational sports use and 1) retention 3-5 and 2) BachelorÕs degree attainment 6, assess recreational sports use in the firs t semester or first year of college. Based on the theories of Astin 7 and Tinto 8,9 , use of the recreational sports fitness centers in the second, third, and fourth years of college may not be as critical for two -year retention and five -year BachelorÕs degre e attainment. Mechanisms for the relationships between recreational sports and college student academic success have not been explored. Previous research indicates a positive relationship between PA participation and GPA among college students. 10 In this dissertation, level of recreational sports use was calculated as total use of fitness centers (during a specified time !!%&.!frame) divided by semesters of enrollment. Therefore, high users in this study theoretically utilized the recreational sports fitne ss centers more regularly than never, low, and medium users. If we assume that recreational sports fitness center use directly relates to PA participation, then high users of recreational sports participate in more regular PA than never, low, and medium users. Based on this assumption , participation in recreational sports may positively relate to PA participation, which in turn may positively impact college student academic success. This relationship may be explained by previous research that indicates posi tive effects of PA on cognitive function ing such as increased reaction time, and impro ved short and long term memory. 11 A second potential mechanism is student integration into the university. Previous literature suggests that students who are socially and academically integrated into the university are more successful. 7-9 University recreational sports fitness centers are rich in social possibility. Students may develop meaningful relationships with recreational sports employees and/or other recreational sports users. Level of recreational sports in this study may represent social integration. If so, then high users of recreational sports fitness centers may be more socially integra ted than never, low, or medium users. Therefore, participation in recreational sports may positively relate to social integration, which in turn may positively influence college student success. Students participating in recreational sports may also diff er from students who do not participate. For example, students who have time to participate in recreational sports may be early risers, or have better time management skills (both factors are related to academic success 12,13). Further, we are not able to d etermine a causal relationship between recreational sports use and college student academic success due to the design of the studies presented in this !!%&%!dissertation. It is possible that students who are more academically successful are also more self -motiva ted than students who are less academically success ful and thus choose to participate in PA through recreational sports more readily than students who have less self -motivation. Anecdotally, we hypothesize that recreational sports fitness center use leve ls directly relate to participation in PA, which in turn positively influences college student academic success. In our opinion, recreational sports fitness centers are more suited for PA participation than social opportunity. Students are socially isolate d by the use of headphones when working out in the fitness centers. Students not socially isolated often work out with a single partner, who they likely met outside of the recreational sports fitness center. For these reasons, it is less likely that use of the recreational sports fitness centers directly relates to social integration, and instead relates to participation in PA. However, other aspects of recreational sports (e.g. , intramurals, club sports) lend themselves to social integration more readily t hat fitness centers. Future research should explore the role of PA and social integration in the relationships found in this dissertation. Further, future research should investigate these relationships within additional components of recreational sports. Although mechanisms for the relationships found in this study are not clear, the results are encouraging for those who have stake in college stud ent success . Moreover, recreational sports participation is a modifiable factor, unlike most variables related to college student success (e.g. race, gender, socioeconomic status, and high school GPA). Recreational sports departments work to develop programming that positively enhances the college student experience. The studies discussed in this dissertation focus specifically on college student academic success. However, success can be defined in a number of ways and should not be limited to academics only. Several departme nts within student affairs share a similar mission, !!%&&!ÒÉ (to prepare) graduates for participation and leadership in an increasingly diverse and complex global society Ó.14 Participation in PA has been shown to positively influence physical health 15 as well as emotional wellbeing 16, and cognitive functioning 11. Recreational sports departments promote regular participation in PA and therefore participation in recreational sports may increase PA participation and subsequently positively relate to other aspects of success (e.g. , emotional wellbeing, physical health) beyond academics . Recreational sports departments should be encouraged to share the results of this dissertation and collaborate with other groups on campus (e.g. , faculty, resources for persons with disabilities, women resource centers, leadership development groups, health centers, and graduate student programs) to improve students success and uncover other ways in which recreational sports participation might positively impact student success. College student success does not depend on a single department, program, or event. Instead student success demands a collaborat ive effort among many members of the campus community, including the student. The American College of Sports Medicine (ACSM) recently began Exercise is Medicine on Campus TM (EIM -OC), an initiative aimed at increasing PA participation and awareness on college campuses. Universities that join the EIM -OC initiative recognize the power in collaborat ion among departments for the betterment of student health, wellness, and success. Recreational sports in particular has the capacity to impact studentsÕ lives in a variety of ways, including academic success, social development, emotional development, and physical health. Recreational sports is not simply a place for students to play sports, it is a department that provides students with opportunities to develop their whole sel ves and engage with the University. Those interested in enhancing college studen t experiences and outcomes !!%&'!should consider becoming an active member of the EIM -OC initiative and should engage with recreational sports as it is a critical department in student health and success . The results of this dissertation are encouraging for recreational sports departments and student affairs professionals. Further, there are ample opportunities to explore additional areas of college student success and their relationship to recreational sports participation. University administrators from t he institution presented in this dissertation requested that all departments prove their worth in student success. Results of this study clearly illustrate how recreational sports participation relates to college student academic success and should be shar ed with campus administration so that adequate funding continues for their operation. In addition, results should be shared with recreational sports administrators and other campus wellness personnel to help promote memberships and regular participation . LIMITATIONS, STRENGTHS, AND FUTURE DIRECTIONS While the results of this study are very encouraging to recreational sports practitioners, there are some limitations that should be discussed. Most universities assess a fee to all students through tuition and fees, which provides open access for students to al l recreational sports facilities. Although the majority of recreational sports facilities at the university in this dissertation were available to all students, a membership fee was assessed to those who wished to utilize the recreational sports fitness ce nters. Therefore, due to differences in payment models the results of this study may not be generalizable to all universities. Further, academic success variables are incredibly complex. Although we controlled for multiple covariates, we could not address all. For example, we did not measure factors such as hours per week working for pay, or !!%&(!place of residence, which are factors that may influence college student academic success. Further, a large proportion of our samples chose to continue at the instituti on for a second (94%) and third (90%) year, which is higher than most colleges, and may have resulted in a type two error, particularly when assessing the relationship between recreational sports use level and two -year retention. Finally, we calculated lev el of recreational sports use as number of fitness center uses divided by semesters of enrollment. This method was utilized to create a variable that represented regular use. However, it is possible that a study participant utilized the fitness center many times in one semester, few in the next, and still was classified as a high user. These limitations should be considered when generalizing the findings of this study to other university populations. Despite limitations, there are multiple strengths to th is study. We utilized a large sample and assessed recreational sports use over multiple semesters. We also addressed multiple covariates in each assessment and utilized complex statistics to control for covariates when necessary. Further, this was the firs t investigation to assess recreational sports use over time in relation to GPA, retention, and BachelorÕs degree attainment. Finally, we investigated a single component of recreational sports, fitness centers, which allowed for more precise interpretation of findings. This study included a sample of students from a large public university. It is well known that student experiences diff er by type of institution 17,18 , therefore our results await confirmation before strong conclusions can be drawn . Further, in an effort to better understand the relationship between recreational sports use and BachelorÕs degree attainment, researchers should investigate the relationship between recreational sports use during the first year of college and five year BachelorÕs degree attainment. Additionally, this study focused on one aspect of recreational !!%&)!sports, fitness centers , and did not investigate mechanisms of relationship. Future research should investigate these relationships , and their mechanisms, within in tramural sports, club sports, and other components of recreational sports. Finally, researchers should collaborate with other departments on campus to identify additional areas of student success (e.g. emotional development, social development, and physica l health) that relate to recreational sports participation. !!%&*!REFERENCES !!%&+!REFERENCES 1. Forrester S, Ross CM, Hall S, Geary C. Using past campus recreational sports participation to explain current physical activity levels of alumni. Recreational Sports Journal . 2007;31:83 -94. 2. U.S. Department of Health and Human Services. 2008 physical activity guidelines for Americans. Be active, healthy, and happy. Washington. DC: U.S. Department of Health and Human Services, Office of Disease Preve ntion and Health Promotion, Office of the Assistant Secretary for Health, Office of the Secretary; 2008. 3. Belch HA, Gebel M, Maas GM. Relationship between student recreation complex use, academic performance, and persistence of first -time freshmen. Journal of Student Affairs Research and Practice. 2002;38(2):220 -234. 4. Danbert SJ, Pivarnik JM, McNeil RN, Washington IJ. Academic success and retention: the role of recreational sports fitness facilities. Recreational Sports Journal. 2014;38(1):14 -22. 5. Kampf S, Teske EJ. Collegiate Recreation Participation and Retention. Recreational Sports Journal. 2013;37(2):85 -96. 6. Huesman R, Brown AK, Lee G, Kellogg JP, Radcliffe PM. Gym Bags and Mortarboards: Is Use of Campus Recreation Facilities Related to Student Success? Journal of Student Affairs Research and Practice. 2009;46(1):50 -71. 7. Astin AW. Student involvement: A developmental theory for higher education. Journal of college student personnel. 1984;25(4):297 -308. 8. Tinto V. Leaving college: Rethinking the c auses and cures of student attrition. Chicago: University of Chicago Press; 1987. 9. Tinto V. Leaving College: Rethinking the Causes and Cures of Student Attrition. (2 nd Ed.). Chicago: University of Chicago Press; 1993. 10. Danbert SJ, Mudd LM, Pivarnik JM. Mee ting ACSM Physical Activity Recommendations and College Student Academic Success. Journal of American College Health. In Review. 11. Hillman CH, Erickson KI, Kramer AF. Be smart, exercise your heart: exercise effects on brain and cognition. Nature reviews neuroscience. 2008;9(1):58 -65. !!%&,!12. Trockel MT, Barnes MD, and Egget DL. Health -related variables and academic performance among first -year college students: implications for sleep and other behaviors. Journal of American College Health. 2000;49(3):125 -131. 13. Gortner Lahmers A, Zulauf CR. Factors associated with academic time use and academic performance of college students: A recursive approach. Journal of College Student Development. 2000;41(5):544 -556. 14. Maybank DB. About student affairs and services. n.d. Retrieved from http://www.vps.msu.edu/ 15. Li J, Siegrist J. Physical activity and risk of cardiovascular disease Ña meta -analysis of prospective cohort studies. International journal of environmental research and public health. 2012;9(2):391 -407. 16. Salovey P, Rothman AJ, Detweiler JB, Steward WT. Emotional states and physical health. American Psychologist. 2000;55(1):110. 17. Pascarella ET, Terenzini PT. How college affects students: Findings and insights from twenty years of research. San Fransisco (CA): Jossey -Bass; 1991. 18. Pascarella ET, Terenzini PT. How College Affects Students: A Third Decade Of Research (3rd Eds). San Fransisco (CA): Jossey -Bass; 2005.