{"id":6322,"date":"2019-02-21T06:30:34","date_gmt":"2019-02-21T12:30:34","guid":{"rendered":"http:\/\/thesportjournal.org\/?p=6322"},"modified":"2020-06-02T11:48:36","modified_gmt":"2020-06-02T16:48:36","slug":"student-athletes-vs-athlete-students-the-academic-success-campus-involvement-and-future-goals-of-division-i-student-athletes-who-were-university-bound-compared-to-those-who-would-not-have-attended","status":"publish","type":"post","link":"https:\/\/thesportjournal.org\/article\/student-athletes-vs-athlete-students-the-academic-success-campus-involvement-and-future-goals-of-division-i-student-athletes-who-were-university-bound-compared-to-those-who-would-not-have-attended\/","title":{"rendered":"Student-Athletes vs. Athlete-Students: The academic success, campus involvement, and future goals of Division I student athletes who were university bound compared to those who would not have attended a university had they not been an athlete."},"content":{"rendered":"\n

Authors:<\/strong> Brenda L. Vogel, Jeff Kress, and Daniel R. Jeske<\/p>\n\n\n\n

Corresponding Author:<\/strong><\/em>
Jeff Kress, Ph.D.
Department of Kinesiology
1250 Bellflower Blvd. \u2013 MS 4901, HHS2-103
Long Beach, CA 90840
jeff.kress@csulb.edu
949-375-3958<\/p>\n\n\n\n

Brenda L.Vogel is a Professor of\nCriminology and Criminal Justice and the Director of the School of Criminology,\nCriminal Justice, and Emergency Management at California State University, Long\nBeach. She served as the CSULB NCAA Faculty Athletics Representative from\n2007-2015.<\/p>\n\n\n\n

Jeff Kress is an Associate\nProfessor in the Department of Kinesiology at California State University, Long\nBeach and teaches in the area of Physical Education Teacher. His research\ninterests have been in the area of sport performance enhancement through\npsychological methods.<\/p>\n\n\n\n

Daniel Jeske is a Professor, in the Department of\nStatistics at the University of California, Riverside.  He has served as the UCR NCAA Faculty\nAthletics Representative. He\nis an elected Fellow of the American Statistical Associationand an\nElected Member of the International Statistical Institute. He has published\nover 100 peer-reviewed journal articles and is a co-inventor on 10 U.S. Patents\nand is currently the Editor-in-Chief of The American Statistician.<\/p>\n\n\n\n

Student-Athletes vs. Athlete-Students: The academic success, campus involvement, and future goals of Division I student athletes who were university bound compared to those who would not have attended a university had they not been an athlete.
<\/h3>\n\n\n\n

ABSTRACT<\/strong><\/p>\n\n\n\n

This study examined the\ndifferences between two groups of Division I student athletes: those who would\nhave attended a 4-year university regardless of their participation in\nathletics and those who would not have attended a 4-year university had it not\nbeen for the opportunity afforded them through their athletic ability.  The researchers examined a number of academic\nfactors including GPA, participation in intensive academic experiences, class\nparticipation and preparation, perception of academic experience, importance of\ngraduation, major selection, and participation in extracurricular activities, future\ngoals, and identification as an athlete or student. The data from the NCAA\u2019s\nGrowth, Opportunities, Aspirations and Learning of Students in College (GOALS)\nsurvey that was administered to a nationwide, random sample of NCAA student\nathletes in 2006 are discussed. Our results suggest that there were significant\ndifferences between the two groups in several of the domains measured.  For example, our findings suggest that student\nathletes who identify as athletes first and students second think less about\nacademics when choosing a college, are less likely to major in mathematics and\nscience, are less likely to select a major to prepare for graduate school or a\nspecific career, have lower GPAs, are less likely to participate in classes,\nare less likely to be involved in extracurricular activities, are less willing\nto sacrifice on athletics participation for academics, feel graduation is less\nimportant to them and to their families, and believe becoming a professional\nathlete is more likely. Implications for the NCAA and college athletics\nprograms are discussed.<\/p>\n\n\n\n\n\n\n\n

Keywords: <\/strong> <\/em>Student-Athletes, Athlete Students, G.O.A.L.s survey
<\/p>\n\n\n\n

INTRODUCTION<\/strong><\/p>\n\n\n\n

\u201c\u2019Running\nopened up a lot of opportunities for me, coming from a family that didn\u2019t know about\ncollege,\u2019 he says. \u2018If it wasn\u2019t for running, I probably wouldn\u2019t have gone to\ncollege and I wouldn\u2019t be where I am today\u2019\u201d (15).  <\/em><\/p>\n\n\n\n

This\nquote, from a cross-country student athlete from Willamette University,\nillustrates one of the more compelling arguments in support of intercollegiate\nathletics: athletic competition can provide educational opportunities to young\npeople who would not have otherwise attended college. It is common for coaches\nto recruit a prospective student athlete because of her\/his exceptional\nathletic ability even if s\/he has marginal academic qualifications.  Indeed, most universities have provisions for\n\u201cspecially\u201d admitting students who do not qualify for regular admission.  Those students include not only athletes, but\nalso artists, the children of donors or political figures, celebrities, or\nfriends of campus administrators and legacies. \nNonetheless, some of those prospective student athletes may have never\neven considered attending a college or university, let alone be prepared for\nattendance.<\/p>\n\n\n\n

The reasons<\/em> some high-school athletes do not\nintend to attend a university are many. Some come from disadvantaged\nbackgrounds where attending a university is not a consideration (6).  Some may have intended to enroll in a junior\ncollege. Some of them may have lacked appropriate academic counseling while in\nhigh school or concentrated their energies on competing with their high school\nteam or an elite club team rather than on their studies. Still others are\ninternational students who spent their high school years competing on junior\nnational teams and neglected their studies. <\/p>\n\n\n\n

Regardless\nof the reasons why they were not university-bound in the first place, many of\nthese students end up on university campuses competing and attending class\nalongside other student athletes and non-athletes who were university-bound all\nalong.  While\nseveral studies have examined student athletes\u2019 academic performance and the\npredictors of their academic success (13,24,25,27), there is no research that compares\nthe academic success of student-athletes who were university-bound regardless\nof their involvement in sports, to those who would not have attended university\nwere it not for their athletic talent. This study aims to address this gap in\nthe literature. <\/p>\n\n\n\n

LITERATURE REVIEW<\/strong><\/p>\n\n\n\n

The\nchallenge of balancing academic standards with athletic competitiveness is not\nnew.  Since its inception in 1906, National\nCollegiate Athletic Association (NCAA) member intuitions have\nwrestled with the same issues that we face today: the extreme pressure to win,\nwhich is compounded by the commercialization of sport, and the need for\nregulations and a regulatory body to ensure fairness and safety (26).  Today,\nthe NCAA\u2019s primary responsibility is to protect student athletes\u2019 wellbeing and\nto provide them with the skills to succeed on the playing field, in the\nclassroom, and throughout life.  According\nto the most current NCAA Division I Manual (20) under the heading \u201cPurposes,\u201d\nsection 1.2 reads: \u201cTo initiate, stimulate and improve\nintercollegiate athletics programs for student-athletes and to promote and\ndevelop educational leadership, physical fitness, athletics excellence and\nathletics participation as a recreational pursuit. (p. 1)\u201d And in the Principle\nof Student Athlete Well Being section 2.2 reads, \u201cIntercollegiate athletics\nprograms shall be conducted in a manner designed to protect and enhance the\nphysical and educational well-being of student-athletes. (p. 3)\u201d  <\/p>\n\n\n\n

Despite the stated purpose, with\nbillions of dollars on the line (through TV contracts, salaries, incentives,\netc.) athletic programs, athletic directors, and coaches face tremendous\npressure to build and maintain winning programs. A series of scandals rocked\ncollege sports in the 1980\u2019s in which 109 colleges and universities were\ncensured, sanctioned, or put on probation by the NCAA.  That number included more than half the\nuniversities playing at the NCAA\u2019s top competitive level (57 institutions out\nof 106).  At that time, nearly a third of\npresent and former professional football players responding to a survey near\nthe end of the decade said they accepted illicit payments while in college and\nmore than half said they saw nothing wrong with the practice (17). Add to that\nanother survey showing that among 100 big-time schools, 35 had graduation rates\nunder 20 percent for their basketball players and 14 had the same low rate for\ntheir football players (17).<\/p>\n\n\n\n

As a result, the Knight Commission\non Intercollegiate Athletics (known simply as the Knight Commission) was formed\nin 1989.  Its founding co-chairmen were Reverend Theodore\nM. Hesburgh<\/a>, president of the University of\nNotre Dame<\/a>, and William C. Friday<\/a>, former\npresident of the University\nof North Carolina<\/a>. Shortly\nbefore the establishment of the Commission, Time Magazine (10) ran an article\nquestioning whether student athletes were really getting an education at all.  \u201cThere is an obsession with winning and\nmoneymaking that is pervading the noblest ideals of both sports and education\nin America.\u201d Its victims, Time went on to say, were not just athletes who found\nthe promise of an education a sham but \u201cthe colleges and universities that\nparticipate in an educational travesty — a farce that devalues every degree\nand denigrates the mission of higher education.\u201d (p. 54)<\/p>\n\n\n\n

Among its initial recommendations,\nthe 1991 Knight Commission report advocated the \u201cone-plus-three\u201d model, a new\nstructure of reform in which the \u201cone\u201d — presidential control — is directed\ntoward the \u201cthree\u201d — academic integrity, financial integrity and independent\ncertification. With respect to academic integrity, the commission wrote that athletes\n\u201cshould not be considered for enrollment at a college or university unless they\ngive reasonable promise of being successful at that institution in a course of\nstudy leading to an academic degree Despite the fact that the Commission held\nno formal authority, nearly two-thirds of its specific recommendations were\nendorsed by the NCAA by 1993. Moving forward, \u201clikelihood of graduation,\u201d\nbecame the standard against which universities would measure the admission of\nstudent athletes (17).<\/p>\n\n\n\n

In 2001, The Knight Commission\nreconvened to assess what had transpired during the decade following their\ninitial report and to assess the state of college athletics at the beginning of\nthe new century. While they found some progress had been made, their findings\nwere equally if not more disturbing than those found in the first report. They\nfound that the graduation rates of athletes in Division 1A basketball and\nfootball at top institutions to be \u201cdismally low-and in some cases failing.\u201d In\nbasketball, the five-year graduation rate was 34 percent. The graduation rate\nfor white football players was 55 percent and for black football players 42\npercent.  Part of the problem according\nto the report was: \u201cAthletes are often admitted to institutions where they do\nnot have a reasonable chance to graduate. They are athlete-students, brought\ninto the collegiate mix more as performers than aspiring undergraduates. Their\nambiguous academic credentials lead to chronic classroom failures or chronic\ncover-ups of their academic deficiencies. As soon as they arrive on campus,\nthey are immersed in the demands of their sports. In light of these\ncircumstances, academic failure, far from being a surprise, is almost\ninevitable\u201d (16).<\/p>\n\n\n\n

Academic Progress Rate<\/em><\/p>\n\n\n\n

In 2003, the NCAA instituted the Academic Progress Rate\n(APR). The APR was\ndesigned to improve the academic performance of student athletes (24) and to\nincrease graduation rates (21).  It \u201cholds\ninstitutions accountable for the academic progress of their student-athletes through\na team-based metric that accounts for the eligibility and retention of each\nstudent-athlete for each academic term\u201d (20).  Teams must maintain a minimum APR or face an\narray of penalties including reduced practice time, competition reductions,\nscholarship reductions, and post-season bans (20).  In practice, the APR has changed the way\ncoaches recruit, making them less likely to \u201ctake a chance\u201d on a recruit who\nmay be less academically prepared and who has the potential to cost them an APR\npoint.  It has also changed the way\nathletic administrators make decisions. \n\u201cHow will this affect our APR?\u201d is a question administrators ask when\nconsidering resource allocation, travel schedules, recruiting budgets, etc. (C.\nMasner, personal communication, March 21, 2017).<\/p>\n\n\n\n

Since\nits inception, research suggests that the APR has increased eligibility,\nretention, and graduation of student athletes (21). Other reforms, including\nchanges to initial eligibility requirements and progress toward degree rules\nhave certainly had a part in the academic gains as well (21). Nonetheless,\noverall student athlete academic success has improved since the advent of the\nAPR system.<\/p>\n\n\n\n

Special Admissions<\/em><\/p>\n\n\n\n

Despite the institution of the APR\nand the emphasis on academic success, nearly all universities offer some form\nof \u201cspecial admission\u201d to select students whether they are the children of\ndonors or influential community members, exceptional musicians, talented\nactors, skilled athletes, or members of another targeted group.  In fact, the NCAA addresses this practice directly\nthrough its Bylaw\n14.1.1.1 that states, \u201cA student-athlete may be admitted under a special\nexception to the institution\u2019s normal entrance requirements if the\ndiscretionary authority of the president or chancellor (or designated\nadmissions officer or committee) to grant such exceptions is set forth in an\nofficial document published by the university (e.g., official catalog) that\ndescribes the institution\u2019s admissions requirements\u201d (20).<\/p>\n\n\n\n

Nonetheless,\nstudent-athletes who are specially admitted characteristically enter with\nborderline or below SAT, ACT and or academic records (18, 21). While it is not known\nhow many student athletes are admitted each year under special circumstances\nnationally, according to Knobler (18) \u201cmore than half of scholarship athletes\nat the University of Georgia, the University of Wisconsin, Clemson University,\nUCLA, Rutgers University, Texas A&M, University, and Louisiana State\nUniversity were special admits.\u201d<\/p>\n\n\n\n

University presidents, athletic\ndirectors, and coaches struggle with the recruitment of top athletes whose\nacademic qualifications are marginal (or worse). Many coaches make the argument\nthat \u201cif we don\u2019t admit him, then our conference rival will!\u201d A senior\nadministrator from a prominent Division IA university explained, \u201c\u2026we are going\nto participate in athletics, and we will recruit students who have a good or\nreasonable chance of succeeding here in order to be competitive in the NCAA\u201d (7).<\/p>\n\n\n\n

Admitting student-athletes with a\n\u201creasonable chance\u201d of graduation while remaining competitive in a national\narena is a prevalent pattern as typified by the University of Michigan\nPresident Mark Schlissel who said, \u201cWe admit students who aren\u2019t as qualified,\nand it\u2019s probably the kids that we admit that can\u2019t honestly, even with lots of\nhelp, do the amount of work and the quality of work it takes to make\nprogression from year to year. An individual\u2019s academic deficiencies are often\noverlooked to fill competitive rosters\u201d (29).\nTom Lifka, chairman of the committee that handles athlete admissions at the\nUniversity of California, Los Angeles, a program that has won more national\nchampionships in all of sports than any other school, was quoted as saying: \u201cIf\nyou\u2019re going to mount a competitive program in Division I-A, and our\ninstitution is committed to do that, some flexibility in admissions of athletes\nis going to take place.  Every\ninstitution I know in the country operates in the same way. It may or may not be\na good thing, but that\u2019s the way it is\u201d (18).<\/p>\n\n\n\n

The\nreality of \u201cflexibility in admissions of athletes\u201d is having noticeable\nconsequences. A CNN investigation in 2014 was able\nto obtain the public records of several schools that revealed that most have\nbetween 7% and 18% of revenue sport athletes reading at an elementary school\nlevel (8, 23). Many student athletes in that investigation were scoring well\nbelow the SAT reading threshold for being college literate of 400, several in\nthe 200-300\u2019s which is an elementary reading level and too low for college\nclasses.  The national average that year\nwas 497.  The same investigation noted\nthat on the ACT, which has 36 as its highest score and a national average of\n20, most teams had an average score in the teens.  <\/p>\n\n\n\n

Former and current academic\nadvisers, tutors, and professors report that it is nearly impossible to jump\nfrom an elementary to a college reading level while juggling a hectic schedule\nas an NCAA athlete (8). While\nconducting research on the academics of student athletes, Bimper (12) wrote,\n\u201cDumb jocks’ are not born; they are being systematically created and\ninstitutionally accommodated\nby the culture of sport that is creating this disparity we see between academic\nperformance and graduation rates\u201d (p. 1).<\/p>\n\n\n\n

Anecdotal evidence suggests that some\nathletes are being recruited and admitted to universities as athletes first and\nas students second.  These students could\nessentially be called \u201cathlete-students\u201d while the majority of college athletes\nare commonly referred to as \u201cstudent-athletes.\u201d Ironically, the term \u201cstudent athlete\u201d was created by former Executive Director of the\nNCAA Walter Byers<\/a>\nin the 1950\u2019s to counter attempts to require universities to pay workers’\ncompensation<\/a> after the widow of a college football\nplayer died during a game and sued for benefits (4).<\/p>\n\n\n\n

Aside\nfrom journalistic reports, there is no comprehensive, academic research\ncomparing the academic success of \u201cstudent-athletes\u201d who were university-bound\nregardless of their involvement in sports, to \u201cathlete-students\u201d who would not\nhave attended university were it not for their athletic talent.  This study aims to address this gap in the\nliterature.<\/p>\n\n\n\n

Research\nQuestions<\/em><\/p>\n\n\n\n

With\nthis research, the researchers sought to address ten specific research\nquestions. The researchers used several items from the GOALS survey to represent\neach research question.  The ten\nquestions listed with the specific GOALS items used for each, and listed in\nAppendix 1.<\/p>\n\n\n\n

  1. Do\nstudent-athletes differ from athlete-students with respect to whether or not athletics\nparticipation influenced university\/college choice, major choice, and class\nselection?<\/li>
  2. Do\nstudent-athletes differ from athlete-students with respect to how athletic\nparticipation has affected their GPA?<\/li>
  3. Do\nstudent-athletes differ from athlete-students with respect to their class\nparticipation and preparation?<\/li>
  4. Do\nstudent-athletes differ from athlete-students with respect to their involvement\nin intensive academic experiences?<\/li>
  5. Do\nstudent-athletes differ from athlete-students with respect to their level of\nparticipation in extracurricular activities and campus events?<\/li>
  6. Do\nstudent-athletes differ from athlete-students with respect to the degree to\nwhich college contributed to their personal growth and development?<\/li>
  7. Do\nstudent-athletes differ from athlete-students with respect to whether they\nconsider themselves more of an athlete than a student?<\/li>
  8. Do\nstudent-athletes differ from athlete-students with respect to their perception\nof their academic experience?<\/li>
  9. Do\nstudent-athletes differ from athlete-students with respect to how important graduation\nis to them?<\/li>
  10. Do\nStudent-athletes differ from Athlete-students with respect to their future\ngoals?<\/li><\/ol>\n\n\n\n

    METHODOLOGY<\/strong><\/p>\n\n\n\n

    Data\nand Sample<\/em><\/p>\n\n\n\n

    The\ndata for this study comes from the 2006 NCAA Growth, Opportunities, Aspirations\nand Learning of Students in College (GOALS) survey.  The survey yielded responses from over 21,000\nstudent athletes at 627 Division I, II, and III member institutions.  The current study, however, is based only on\ndata from Division I institutions. Respondents provided information about their\nlives as student athletes across a spectrum of domains, including academic\nengagement and success, athletics experiences, social experiences, career\naspirations, health and well-being, campus and team climate, and time\ncommitments. The NCAA Research Division selected one to three teams per\ninstitution to be surveyed in order to provide representative samples within\neach division. The Faculty Athletics Representative (FAR) on each campus was\nasked to administer the surveys to the selected teams. The response rate for\nDivision I institutions was 66%.<\/p>\n\n\n\n

    Variables<\/em><\/p>\n\n\n\n

    Our\nresearch focused on the views and behaviors of two groups: student-athletes and\nathlete-students. The GOALS survey asked respondents to respond on a six-point\nLikert scale to the following statement: \u201cI would have gone to a 4-year university\nsomewhere even if I hadn\u2019t been an athlete.\u201d \nWe defined student-athletes as those who indicate (responded either\nstrongly agree, agree, or somewhat agree) that they would have attended a\n4-year university regardless of their status as an athlete.  Athlete-students were defined as those who\nindicated (responded either strongly disagree, disagree, somewhat disagree)\nthey would not<\/em> have attended a 4-year\nuniversity had they not been an athlete. \nBased on these definitions, the sample included 7110 students, 6350\n(89.3%) of whom were Division I student-athletes and 760 (10.7%) of whom were Division\nI athlete-students.[i]<\/a>\n<\/p>\n\n\n\n

    The 10\ncontrol and 65 dependent variables and their frequencies and percentages for\neach group (athlete-students and student-athletes) are provided in Appendix\nOne.  The 65 dependent variables were grouped\nby research question; we used several individual GOALS survey items to measure\neach of our ten research questions.  A\nreview of Appendix One reveals that the distribution of student-athletes (89.3%\nof the overall sample) and athlete-students (10.7% of the overall sample) varies\nsignificantly across several of the dependent variables. This suggests that student-athletes\nand athlete students differ on a number of important measures.  Many of these apparent differences are\nconfirmed through the analysis discussed below.<\/p>\n\n\n\n

    Analysis<\/em><\/p>\n\n\n\n

    The analysis\nproceeded in two steps. First, we examined the differences between\nstudent-athletes and athlete-students in the 10 control variables. Second, controlling\nfor those 10 variables, we examined the differences between the two groups on the\n65 dependent variables. <\/p>\n\n\n\n

    We\nused a goodness of fit chi square analyses to determine if athlete-students and\nstudent-athletes differ on the ten control variables. The control variables\nincluded sport, academic class, level of participation, athletic aid, recruited\nathlete, would still attend this university if different coach, gender,\nrace\/ethnicity, father\u2019s educational level, and mother\u2019s educational\nlevel.  The chi square distribution was\nused to test whether observed data differed significantly from theoretical\nexpectations (9).  In this study, the\ntheoretical expectation was that each subgroup (e.g., women, students on full\naid, etc.) would include 89.3% (.893) student-athletes (SA) and 10.7% (.107) athlete-students,\n(AS).  This tested the null hypothesis\nthat the distribution on each control variable attribute was: AS=.107 and\nSA=.893.<\/p>\n\n\n\n

    The\nsecond step analyzed each dependent variable using a multinomial logistic\nregression model.  The coefficient on the\ngrouping variable was of primary interest for exploring the ten research\nquestions outlined above.  The grouping\nvariable was encoded as a binary variable, taking the value of 0 for\nstudent-athletes and 1 for athlete-students. \nThe control variables were similarly encoded using corresponding sets of\nbinary variables.  All binary variables\nused to encode the grouping variable and the control variables were included in\nthe multinomial logistic regression model as explanatory variables.  <\/p>\n\n\n\n

    The\nmultinomial logistic regression model framework can handle both nominal and\nordinal dependent variables.  In the case\nof nominal dependent variables, the model delivered estimated probabilities for\neach possible category of the variable, and showed how these probabilities vary\ndepending on the covariate values.  In\nthe case of an ordinal dependent variable, the model delivered the cumulative\nprobabilities of being at or below each category level of the variable.  <\/p>\n\n\n\n

    For\neach dependent variable, the difference between the student-athlete and athlete-student\npopulations were assessed by examining the sign and magnitude of the maximum\nlikelihood estimate of the coefficient on the grouping variable.  The null hypothesis of no difference between\nthe two populations was tested using a Wald test (12) to determine if the\nestimated coefficient on the grouping variable was statistically different from\nzero.  All analyses were carried out\nusing the PROC LOGISTIC procedure in the software package SAS\/STAT Version 9.2\nCopyright \u00a9 2008 by SAS Institute Inc.  <\/p>\n\n\n\n

    To\nillustrate the proposed analyses, consider the analysis of GPA.  GPA is an ordinal variable with 9\ncategories.  The researchers collapsed\nthe 9 categories into 4 categories:  the\nA category, with GPA between 3.5 and 4.0, the B category with GPA between 2.5\nand 3.49, the C category with GPA between 1.5 and 2.49 and finally, the D\ncategory with GPA less than 1.5.  After\ncollapsing, GPA was distributed as a multinomial distribution with 4 outcome\ncategories, which was analyzed with a multinomial logistic regression\nmodel.  The independent variables in the\nregression were:<\/p>\n\n\n\n

    \"\"<\/figure>\n\n\n\n

    Multinomial\nlogistic regression models the probability of each of the four GPA\ncategories.  This was different from\ntypical least squares regression which models the expected value of the\ndependent variable.  Here, GPA was a\ncategorical variable, taking on the values A, B, C and D.  So the expected value of GPA, coded this way,\ndoes not mean anything.  It is more\nappropriate to predict the probability of each GPA category than it is to\npredict the mean GPA.  The independent\nvariables were used in the same way as least squares regression, except they tried\nto explain influences they have on the probability of each GPA category.  <\/p>\n\n\n\n

    GPA was expressed in the model through so-called logit equations, which modeled the probability of each GPA category.\u00a0 Let \"\" denote the vector of all 40 indicator variables that collectively represent the grouping variable and the control variables.\u00a0 Let \"\" denote the vector of slopes on these variables \"\", \"\" and \"\" let, and be three intercept parameters.\u00a0 Since GPA is an ordinal variable, the logit equations used are so-called cumulative logit equations and are:<\/p>\n\n\n\n

    \"Figure<\/figure>\n\n\n\n

    which can be solved for GPA category probabilities as follows:<\/p>\n\n\n\n

    \"Figure<\/figure>\n\n\n\n

    The maximum likelihood algorithms in the SAS procedure PROC LOGSITIC produced estimates and significance tests for the \"\" parameters, and our interest was primarily in the estimate of coefficients on the indicator variables \"\" and \"\" corresponding to the grouping variable.\u00a0 The null hypothesis of no effect for the grouping variable was \"\".<\/p>\n\n\n\n

    There\nare cases where the dependent variable was nominal, such as the items asking\n\u201cwhich of the following experiences have you or will you be involved in during\ncollege,\u201d where there are four outcomes (yes, not yet, no I don\u2019t have time,\nand no I have no interest) that do not have a natural mathematical\nordering.  In cases such as this,\nso-called generalized logits were used when modeling the dependent\nvariable.  With generalized logits, each\nequation can have its own slope vector. \nFor question 8 in part 2, there are three generalized logit equations\nand they are:<\/p>\n\n\n\n

    \"\"<\/figure>\n\n\n\n

    This can be solved for outcome\nprobabilities as follows:<\/p>\n\n\n\n

    \"\"<\/figure>\n\n\n\n

    Again, maximum likelihood algorithms in the SAS procedure PROC LOGSITIC produce estimates and significance tests for the \"\", \"\" and \"\"\u00a0parameters.\u00a0 The null hypothesis for no effect of the grouping variable is \"\".<\/p>\n\n\n\n

    RESULTS<\/strong><\/p>\n\n\n\n

    Table\n1 provides the results of a series of one-way chi-squares testing if the\ndistribution of athlete-students and student-athletes across each control variable\nattribute (e.g. females, seniors, etc.) differed significantly from their\ndistribution in the overall sample. Athlete-students make up 10.7% of the\noverall sample and student-athletes make up 89.3% of the overall sample. A\nreview of Table 1 suggests that the proportion of athlete-students to\nstudent-athletes varied significantly across several attributes of the ten\ncontrol variables.  <\/p>\n\n\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n

    Table 1<\/p>\n\t

    Athlete-students to student-athletes across several attributes of the ten control variables<\/em><\/p><\/td>\n <\/tr>\n

    Control Variable<\/strong><\/p><\/td>\n

    Attribute<\/strong><\/p><\/td>\n

    Chi-Square1<\/strong><\/p><\/td>\n

    p value<\/strong><\/p><\/td>\n

    Interpretation<\/strong><\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Men’s Baseball<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Men’s Basketball<\/p><\/td>\n

    10.48<\/p><\/td>\n

    0.001<\/p><\/td>\n

    AS overrepresented<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Men’s Football<\/p><\/td>\n

    6.30<\/p><\/td>\n

    0.012<\/p><\/td>\n

    AS overrepresented<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Men’s Golf<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Men’s Ice Hockey<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Men’s Lacrosse<\/p><\/td>\n

    3.89<\/p><\/td>\n

    0.048<\/p><\/td>\n

    AS underrepresented <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Men’s Soccer<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Men’s Swimming<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Men’s Tennis<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Men’s Track (Indoor or Outdoor)<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Men’s Wrestling<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Women’s Basketball<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Women’s Field Hockey<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Women’s Golf<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Women’s Gymnastics<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Women’s Ice Hockey<\/p><\/td>\n

    5.86<\/p><\/td>\n

    0.015<\/p><\/td>\n

    AS underrepresented <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Women’s Lacrosse<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Women’s Softball<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Women’s Soccer<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Women’s Swimming<\/p><\/td>\n

    10.61<\/p><\/td>\n

    0.001<\/p><\/td>\n

    AS underrepresented <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Women’s Tennis<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Women’s Track (Indoor or Outdoor)<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Women’s Volleyball<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

    Academic Class<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Freshman<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Sophomore<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Junior<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Senior<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Graduate Student<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

    Level of participation<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    First Team<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Second Team<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Third Team<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Practicing, not competing<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

    Athletic Aid<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    No<\/p><\/td>\n

    18.06<\/p><\/td>\n

    0.000<\/p><\/td>\n

    AS underrepresented <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Yes, partial aid<\/p><\/td>\n

    3.85<\/p><\/td>\n

    0.049<\/p><\/td>\n

    AS underrepresented <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Yes, full aid<\/p><\/td>\n

    35.98<\/p><\/td>\n

    0.000<\/p><\/td>\n

    AS overrepresented<\/p><\/td>\n <\/tr>\n

    Recruited Athlete<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Yes<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    No<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

    Would still attended this college if different coach?<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Very likely<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Likely<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Somewhat likely<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Somewhat unlikely<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Unlikely<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Very unlikely<\/p><\/td>\n

    8.99<\/p><\/td>\n

    0.002<\/p><\/td>\n

    AS overrepresented<\/p><\/td>\n <\/tr>\n

    Gender<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Female<\/p><\/td>\n

    5.41<\/p><\/td>\n

    0.02<\/p><\/td>\n

    AS underrepresented <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Male<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

    Race\/Ethnicity<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    White, non-Hispanic<\/p><\/td>\n

    27.74<\/p><\/td>\n

    0.000<\/p><\/td>\n

    AS underrepresented <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    African American<\/p><\/td>\n

    36.60<\/p><\/td>\n

    0.000<\/p><\/td>\n

    AS overrepresented<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Other<\/p><\/td>\n

    12.62<\/p><\/td>\n

    0.000<\/p><\/td>\n

    AS overrepresented<\/p><\/td>\n <\/tr>\n

    Father’s Educational Level<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    No HS<\/p><\/td>\n

    39.15<\/p><\/td>\n

    0.000<\/p><\/td>\n

    AS overrepresented<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Completed HS<\/p><\/td>\n

    27.02<\/p><\/td>\n

    0.000<\/p><\/td>\n

    AS overrepresented<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Attended college no degree<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Associate’s degree<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Bachelor’s degree<\/p><\/td>\n

    21.98<\/p><\/td>\n

    0.000<\/p><\/td>\n

    AS underrepresented <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Master’s degree<\/p><\/td>\n

    24.11<\/p><\/td>\n

    0.000<\/p><\/td>\n

    AS underrepresented <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Doctoral degree<\/p><\/td>\n

    15.26<\/p><\/td>\n

    0.000<\/p><\/td>\n

    AS underrepresented <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Don’t know<\/p><\/td>\n

    6.73<\/p><\/td>\n

    0.009<\/p><\/td>\n

    AS overrepresented<\/p><\/td>\n <\/tr>\n

    Mother’s Educational Level<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    No HS<\/p><\/td>\n

    42.98<\/p><\/td>\n

    0.000<\/p><\/td>\n

    AS overrepresented<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Completed HS<\/p><\/td>\n

    28.51<\/p><\/td>\n

    0.000<\/p><\/td>\n

    AS overrepresented<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Attended college no degree<\/p><\/td>\n

    10.83<\/p><\/td>\n

    0.001<\/p><\/td>\n

    AS overrepresented<\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Associate’s degree<\/p><\/td>\n

     <\/p><\/td>\n

    NS<\/p><\/td>\n

     <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Bachelor’s degree<\/p><\/td>\n

    35.63<\/p><\/td>\n

    0.000<\/p><\/td>\n

    AS underrepresented <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Master’s degree<\/p><\/td>\n

    19.94<\/p><\/td>\n

    0.000<\/p><\/td>\n

    AS underrepresented <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Doctoral degree<\/p><\/td>\n

    4.21<\/p><\/td>\n

    0.040<\/p><\/td>\n

    AS underrepresented <\/p><\/td>\n <\/tr>\n

     <\/p><\/td>\n

    Don’t know<\/p><\/td>\n

    3.98<\/p><\/td>\n

    0.046<\/p><\/td>\n

    AS overrepresented<\/p><\/td>\n <\/tr>\n

    1 The calculated value of chi-square is corrected for continuity.<\/p><\/td>\n <\/tr>\n<\/table>\n\n\n\n

    The\nfindings indicated that athlete-students were overrepresented<\/em> in the following groups:<\/p>\n\n\n\n