Factors Related to Attrition and Retention of Under-Represented Minority Students: National and Regional Trends

Transcription

1 OEA Report Factors Related to Attritio ad Retetio of Uder-Represeted Miority Studets: Sebastia Lemire ad Cydy Syder November 2006 INTRODUCTION This report summarizes aalyses of selected data from files of the Natioal Postsecodary Studet Aid Study (NPSAS:04). The NPSAS:04, coducted by the Natioal Ceter of Educatioal Statistics (NCES) durig the school year, is a comprehesive atiowide study desiged to determie how studets ad their families pay for postsecodary educatio ad to describe characteristics of those erolled. It captures extesive iformatio o studets educatioal circumstaces, ad the resultig dataset affords the opportuity to describe the atioal cotext relative to a variety of educatioal issues. Our purpose is to preset a limited umber of educatioal variables, both atioally ad regioally, relatig to attritio ad retetio of uder represeted miority studets. Of particular iterest are patters ad relatioships that might be used to iform UW academic programs we iclude such variables as istitutioal characteristics, studets fiacial status, pre college factors, post secodary academic performace, ad extra curricular obligatios. Results from the aalysis of atioal ad regioal data cosistetly showed that uderrepreseted miority studets were more likely tha White ad Asia studets to be at risk for attritio. Notable fidigs related to studet retetio from the NPSAS:04 were as fols: The NPSAS:04 idex of risk is based o the sum of seve characteristics thought to affect persistece adversely: delayed erollmet, o high school diploma, part time erollmet, fiacial idepedece, havig depedets, havig sigle paret status, ad workig full time while erolled. Usig this idex, we foud that Asia studets at least atioally were the least at risk, with the majority (81%) havig oly oe or two of the seve risk idicators. Coversely, two out of five of the Black studets (41%) had three or more risk factors, appearig to be most at risk of attritio. While overall, the majority of studets atteded college full time, at both the atioal ad the regioal levels, a higher percetage of Black ad Latio studets, as well as multi racial studets, atteded school exclusively part time, compared with White ad Asia studets. At the atioal level the percetage of studets who had take remedial courses was higher amog Pacific Islader (39%), Black (35%), ad America (33%) studets tha amog White (24%), Asia (26%), ad multi racial (23%) studets. Uder represeted miority studets eared slightly er GPAs i college (ragig from ) tha did White ad Asia studets (about 3.0). Natioal ad regioal data distiguish betwee the marital status of studets parets ad the kids of households i which studets live. Natioally, Black studets had the est percetage of Copyright 2006 Uiversity of Washigto Office of Educatioal Assessmet uw.edu/assessmet/reports

2 parets who were married (50%) ad came from the highest percetage of sigle paret households (19%), compared with all other ethicities. The percetage of studets i divorced/separated households was much higher amog Black (27%), Latio (22%), America (25%), ad multiracial studets (22%), tha amog White (17%), Asia (11%), ad Pacific Islader studets (8%). At both atioal ad regioal levels, most studets parets had bachelor s degrees, with the exceptio of parets of Black ad Latio studets, whose parets were more likely to have a high school diploma or the equivalet. At the atioal level, Black studets had the highest fiacial aid applicatio rate (90%), the highest average amout of aid awarded ($10,015), ad the highest ratio of aid to budget (70%). The est applicatio rates were amog White (75%), Asia (75%), ad other studets (75%), ad these studets also had some of the est ratio of aid to studet budget (about 55%). Natioally the average amout of total loas take out durig the academic year was higher amog Black ($6,147) ad White ($5,826) studets tha amog Latio ($5,049), Asia ($5,177), ad America studets ($5,331). The amout still owed o all udergraduate educatio loas was highest for Black studets ($15,047), foled by multi racial studets ($13,637), ad White studets ($13,075). Natioally, a higher percetage of Hawaiia ad Asia studets received help from their parets to pay tuitio ad fees (58% ad 48%), while a smaller proportio of Black (29%) ad America (27%) studets received such help. Latio ad America studets teded to sped more time tha White ad Asia studets at work. Black ad Latio studets were more likely tha other groups to cite college tuitio, fees, or livig expeses as their primary reaso for workig ad to deem their work as havig a egative impact o their grades. THE NATIONAL POSTSECONDARY STUDENT AID STUDY (NPSAS:04) The NPSAS:04 was based o a atioally represetative sample of udergraduate, graduate, ad first professioal studets erolled i public baccalaureate istitutios i across the Uited States at ay time betwee July 1 st, 2003 ad Jue 30 th, It icluded approximately 80,000 studets from about 1,300 istitutios. ii Data were collected from multiple sources, icludig: Studet records: Data from istitutioal fiacial aid ad registrar records at the istitutios curretly atteded. Studet iterviews: Data collected directly from sampled studets via web based selfadmiistered or iterviewer admiistered questioaires. Cetral Processig System (CPS): U.S. Departmet of Educatio Database or federal fiacial aid applicatios for the academic year. Natioal Studet Loa Data System (NSLDS): U.S Departmet of Educatio database of federal Title IV loas ad Federal Pell Grats. Itegrated Postsecodary Educatio Data System (IPEDS): U.S. Departmet of Educatio, Natioal Ceter for Educatio Statistics, database of descriptive iformatio about idividual postsecodary istitutios. 2

3 SELECTED DATA From the comprehesive NPSAS:04 dataset, we extracted iformatio pertaiig to all udergraduate studets erolled i doctoral four year public uiversities atioally (a sample of approximately 15,000 studets ad 130 istitutios). We were able to extract aalogous data for schools i the far west regio (Alaska, Califoria, Hawaii, Nevada, Orego, ad Washigto), although we could ot obtai the umbers of studets ad istitutios sampled i this area. Because of the for may of the ethicity groups sampled withi the far west regio (e.g. studets idetifyig as America / Alaska or Hawaiia / Pacific Islader), we were uable to aalyze several of the uder represeted miority subgroups at the regioal level. Accordigly, our aalyses at the regioal level are limited. Sice the purpose of the NPSAS:04 dataset is to capture data o studet aid, a relatively large umber of the variables relate to studet fiacial status. However, a umber of demographic ad academic variables are also icluded; of particular iterest, is a Risk Idex computed by NPSAS:04 to demark threats to college persistece. This is described more fully be. Additioal iformatio icludes: Istitutioal Selectivity ad Locatio Fiacial Factors Studet Backgroud Studet Academics Other Factors I costructig the comparative tables throughout the remaider of this report, we defie uder represeted miorities i the same maer as the UW Office of the Registrar. This defiitio icludes studets who self idetify as Africa America/Black, Hispaic/Latio, America or Alaska, Hawaiia or other Pacific Islader, but excludes those who report their ethicity as Asia. For this reaso, iformatio o Caucasia ad Asia studets is provided together, foled by that for uder represeted miority studets ad those of combied ethicity. Risk Idex The NPSAS:04 dataset icludes a computed variable defied as the sum of seve characteristics iversely related to college persistece. These factors are: (a) havig o high school diploma, (b) delayig college erollmet, (c) erollig part time, (d) beig fiacially idepedet, (e) havig depedets, (f) beig a sigle paret, ad (g) workig full time while erolled. It is importat to ote that most, if ot all, of these are either directly or idirectly related to studets fiacial circumstaces a critical factor i studet post secodary retetio. No ethicity was reported as subject to all seve risk factors, either atioally or regioally (see Table 1). Natioally, Asia studets were least at risk, with the majority (81%) experiecig oly oe or two of the seve idicators. However, two out of five of the Africa America 3

4 studets (41%) ad studets i the Other category (46%) had three or more risk factors, appearig to be most at risk of attritio. At the regioal level, Africa America ad Hispaic studets teded to have a higher umber of risk factors tha Asia ad Caucasia studets. Table 1. Risk Idex (percetage) Natio (Nat) ad Regio (Reg) Number of Risk Factors Oe Two Three Four Five Six Nat Reg Nat Reg Nat Reg Nat Reg Nat Reg Nat Reg Caucasia Asia Africa America / Black Hispaic / Latio America or Alaska Hawaiia / Pacific Islader Other More tha oe ethicity Overall Istitutioal Selectivity ad Locatio Literature o retetio suggests that istitutioal characteristics may be relevat to retetio (Hurtado, 1998). For example, the level of selectivity of the istitutio a studet atteds is ofte coected to studets levels of academic preparatio. At the atioal level, as Table 2 shows, the majority of studets, across most ethic groups, atteded a istitutio that the NPSAS:04 determied was either moderately selective (61%) or very selective (23%), iii with the exceptio of America studets who typically atteded either miimally selective (49%) or moderately selective (41%) istitutios. Asia studets teded to eroll at very selective (37%) ad most selective schools (12%) at a much higher rate tha ay of the ethicity groups. The same is true i the far west regio where close to oe fourth of the Asia studets (22%) atted some of the most selective istitutios i the regio. 4

5 Table 2. Selectivity of istitutio curretly attedig (percetage) Natio (Nat) ad Regio (Reg) Most selective Very selective Moderately selective Miimally selective Ope admissio Nat Reg Nat Reg Nat Reg Nat Reg Nat Reg Caucasia Asia Africa America / Black Hispaic / Latio America or Alaska Hawaiia / Pacific Islader Other More tha oe ethicity Overall I additio, the locatio of istitutios might play a role i retetio. As show i Table 3, at the atioal level, two thirds of the studets erolled at four year public istitutios atteded istitutios i either mid sized (havig a populatio less tha 250,000) or large cities (havig a populatio greater tha or equal to 250,000). At the regioal level, these studets made up a eve higher 83%. Asia studets at both the atioal ad regioal levels were more likely tha ay other group to atted college i a large city. I marked cotrast, more tha two thirds of America studets atteded istitutios i a tow (populatio aroud 25,000). Thus, the locatio of the UW i a large city may play ito the decisios of some America studets about whether to atted college here. Table 3. Type of locatio of istitutio curretly attedig (percetage) Natio (Nat) ad Regio (Reg) Urba frige of large city Urba frige of mid-size city Large city Mid-size city Tow Rural Not assiged Nat Reg Nat Reg Nat Reg Nat Reg Nat Reg Nat Reg Nat Reg Caucasia Asia Africa America / Black Hispaic / Latio America or Alaska Hawaiia / Pacific Islader Other More tha oe ethicity Overall Regardig distaces from home, however, Asia studets appear to have the largest average distace gap betwee their homes ad the istitutios they are curretly attedig at both the atioal (522 miles) ad regioal levels (438 miles), as show i Table 4. 5

6 Table 4. Distace betwee home ad istitutio curretly attedig (average distace i miles) Natio (miles) Regio (miles) Caucasia Asia Africa America / Black Hispaic / Latio America or Alaska 224 Hawaiia / Pacific Islader 96 Other 365 More tha oe ethicity Overall Fiacial Factors Nearly all research o retetio ackowledges the importace of studets fiacial eeds i their decisios about whether to stay or leave college (Upcraft, Mulledore, & Fidler, 1994; Redo, 1994; Mohammadi 1994; Tito, 1994). Similarly, the NPSAS:04 reported sigificat differeces i studets fiacial situatios which we summarize here. I aalyzig aspects of studets fiacial status, the NPSAS:04 defied a studet s total budget as her complete educatioal expeses, icludig tuitio ad fees, room ad board, books ad supplies, trasportatio, ad other livig expeses. Furthermore, the NPSAS:04 used istitutioal reports to determie budget estimates. All average amouts of studet aid, credit card balaces, tuitio, or mothly alace reported by the NPSAS:04 were reported as weighted meas, based o the recipiets of the specified aid, tuitio, alace. I other words, orecipiets of the specified type of aid, tuitio or alace were excluded from these calculatios. As show i Table 5, Asia studets reported the highest amout spet o educatio, both atioally ad regioally. This is cosistet with this group s higher rate of attedace at the most select schools. Far west regioal average budgets ted to be slightly higher tha atioal averages for all groups except for Africa America studets. Table 5. Studet total budget (average adjusted for attedace) Natio ($) Regio ($) Caucasia 13,128 14,331 Asia 14,537 15,552 Africa America / Black 13,414 13,279 Hispaic / Latio 11,861 13,451 America or Alaska 11,025 Hawaiia / Pacific Islader 13,766 Other 12,497 More tha oe ethicity 13,262 13,439 Overall 13,124 14,366 6

7 Studets receive fiacial assistace from a variety of sources. The most traditioal of these, paretal support, is much less available to some ethic groups tha to others. As show i Table 6, atioally, a higher percetage of Hawaiia ad Asia studets received help from their parets to pay tuitio ad fees (58% ad 48%), while a smaller proportio of Africa America (29%) ad America (27%) studets received such help. Regioally, multiracial studets appeared to receive help from parets at the highest rate (56%), as compared with Asia (49%), Caucasia (46%), Hispaic (27%), ad Africa America (21%) studets. These patters held true for expeses other tha tuitio, as well. iv Table 6. Help from parets pay tuitio ad fees (percetage) Natio Regio Caucasia Asia Africa America / Black Hispaic/Latio America or Alaska 27 Hawaiia / Pacific Islader 58 Other 38 More tha oe ethicity Overall Other tha assistace from their parets, studets fuded their educatio through various forms of fiacial aid. This was defied by the NPSAS:04 as the total amout of fiacial assistace received by a studet durig the academic year from ay source except parets, relatives, or frieds. The ratio of total aid to studet budget provides the total aid received as a percetage of the total studet budget. As Table 7 shows, atioally ad regioally, Africa America studets have the highest fiacial aid applicatio rate (90%), the highest average amout of aid awarded ($10,015), ad the highest ratio of aid to budget (70%). The est applicatio rates were amog Caucasia (75%), Asia (75%), ad other studets (75%) ad these studets also had some of the est ratio of aid to studet budget (about 55%). Table 7. Fiacial aid Applied for ay aid Aid total amout ($) Ratio of total aid to studet budget Natio Regio Natio Regio Natio Regio Caucasia ,849 8, Asia ,190 8, Africa America / Black ,015 10, Hispaic/Latio ,571 8, America or Alaska 82 7, Hawaiia / Pacific Islader 86 6, Other 75 7, More tha oe ethicity ,966 5, Overall ,093 8,

8 The type of fiacial aid studets receive is importat. Grats are preferable to loas, of course, because grats do ot have to be paid back. Studets who take out loas to fiace their udergraduate educatios, eve iterest loas, ofte face staggerig debt whe they graduate. Natioally, the most commo overall aid package cosists of a combiatio of grats ad loas (32%), whereas, regioally, the most commo package cosists primarily of grats oly (see Tables 8 ad 9, respectively). Natioally, Asia studets are the most likely to receive oly grats (37%), foled closely by Hispaic studets (34%). Africa America studets are the least likely to receive oly grats (19%). Pacific Islader studets are more likely tha ay other group to receive loas oly (23%). Regioally, multi racial studets are more likely to receive oly grats (55%), while Africa America studets are the most likely to receive a combiatio of grats ad loas (42%). Table 8. Type of aid (percetage) Natio Grats Loas Work-study Other Grats ad loas Grats ad work-study Caucasia Asia Africa Amer. / Black Hispaic/Lati o America or Alaska Hawaiia / Pacific Islade r Other More tha oe ethicity Overall Grats ad other Loas ad work-study Loas ad other Work-study ad other Grats, loas, ad workstudy Grats, loas, ad other Grats, work-study, ad other Loas, work-study, ad other Grats, loas, work-study, ad other 8

9 Table 9. Type of aid (percetage) Regio Grats Loas Work-study Other Grats ad loas Grats ad work-study Caucasia Asia Africa Amer. / Black Hispaic/Lati o America or Alaska Hawaiia / Pacific Islader Other More tha oe ethicity Overall Grats ad other Loas ad work-study Loas ad other Work-study ad other Grats, loas, ad workstudy Grats, loas, ad other Grats, work-study, ad other Loas, work-study, ad other Grats, loas, work-study, ad other Regardig loas, the NPSAS:04 defied total loas as the amout of loas received by a studet durig the academic year, excludig federal Paret Loas for Udergraduate Studets (PLUS) loas to parets. Table 10 shows both the average amout of atioal ad regioal loas grated to studets by ethicity, as well as the average amout owed o udergraduate loas by ethicity i the academic year. As show, atioally, the average amout of total loas take out durig the academic year was higher amog Africa America ($6,147) ad Caucasia ($5,826) studets tha amog Hispaic ($5,049), Asia ($5,177), ad America studets ($5,331). The amout still owed o all udergraduate educatio loas was highest amog Africa America studets ($15,047), foled by multiracial studets ($13,637), ad Caucasia studets ($13,075). 9

10 Table 10. Loas (average) Total loas (excludig PLUS) Total usubsidized loas (all sources) Cumulative borrowed for udergraduate educatio Amout still owed o all udergraduate educatio loas ($) ($) ($) ($) Natio Regio Natio Regio Natio Regio Natio Regio Caucasia 5,826 5,248 5,716 6,469 13,275 13,536 13,075 13,146 Asia 5,177 4,340 5,467 6,375 12,840 11,677 12,158 11,641 Africa America / Black 6,147 5,154 14,944 14,239 15,047 13,353 Hispaic/Latio 5,049 4,550 4,978 5,103 12,078 11,480 12,050 11,681 America or Alaska 5,331 3,455 11,054 10,912 Hawaiia / Pacific Islader Low Other 5,620 4,437 12,084 11,953 More tha oe ethicity 5,550 6,223 13,175 13,637 Overall 5,756 4,958 5,548 6,065 13,337 12,720 13,186 12,659 While loas must be paid back with iterest, grats are awarded to studets free ad clear. The NPSAS:04 defied grats as the total amout of all eed ad merit based grats ad scholarships received by a studet durig the academic year, ad equal to the sum of all federal grats, state grats, istitutioal grats, as well as grats from employers or private sources. Tuitio waivers ad employer tuitio reimbursemets were icluded. As Table 11 shows, the atioal average amout of total grats received was sigificatly higher amog Asia ($5,613) ad Africa America studets ($5,178) tha for other groups (ragig from $3,584 to $4,455). Africa America studets also had a oticeably higher average of total merit oly grats ($4,589), tha did all other groups (ragig betwee $2,732 ad $3,438). Asia studets displayed the highest ratio of grats to loas at 222%. Table 11. Grats (average) Total merit-oly Total eed-based Ratio of grats to Total grats grats grat aid total loas ($) ($) ($) Natio Regio Natio Regio Natio Regio Natio Regio Caucasia 3,892 4,438 3,438 3,263 3,404 4, Asia 5,613 6,356 3,185 1,600 5,605 6, Africa America / Black 5,178 6,602 4,589 4,188 5, Hispaic/Latio 4,455 6,387 2,997 4,466 6, America or Alaska 4,112 3, Hawaiia / Pacific Islader Other 3,584 3, More tha oe ethicity 3,660 3,132 2,732 3, Overall 4,228 5,315 3,494 2,738 3,828 5,

11 I calculatig the ratio of grats to loas, the NPSAS:04 icluded the total studet loas (excludig PLUS) received as a percetage of the total studet budget as well as the total studet grat aid received as a percetage of the total studet budget. Table 12 shows that atioally, studets fiacial aid packages more ofte cosisted of loas tha grats. Regioally, the overall average is similar, but due to the umbers of studets i may of the categories, it is difficult see the regioal picture clearly. Table 12. Ratio of loas ad grat aid to total studet budget Loas Natio Grats Loas Regio Grats Caucasia Asia Africa America / Black Hispaic / Latio America or Alaska Hawaiia / Pacific Islader Other More tha oe ethicity Overall I additio to loas ad grats, may studets use credit cards to help fiace their educatios. Natioally ad regioally, as Table 13 idicates, the majority of studets (59% ad 67% respectively) pay off their credit card balaces each moth. Moreover, as the table shows, a higher percetage of Asia ad Caucasia studets tha other groups pay off their credit cards each moth. Natioally, Africa America studets are more likely to carry a mothly balace forward, but they have the est average balaces due o their cards amog all groups. Because of the umber of Africa America studets regioally who provided iformatio about their credit cards, regioal patters are ot clear. Table 13. Credit card balace (percetage) Pay off their balace Carry a balace Credit cards - balace due ($) Natio Regio Natio Regio Natio Regio Caucasia ,225 2,149 Asia ,321 Africa America / Black ,758 Hispaic / Latio ,657 2,582 America or Alaska Hawaiia / Pacific Islader Other More tha oe ethicity ,088 Overall ,202 2,177 11

12 Studet Backgroud Parets marital status ofte iflueces whether they ca cotribute to the costs of studets educatios, ad their levels of educatio ofte iflueces how much they ca help their collegeboud childre avigate the sometimes cofusig waters of higher educatio. As Table 14 shows, at the atioal level, Africa America studets have the est percetage of parets who are married (50%) ad the highest percetage of sigle paret households (19%), compared with all other ethicities. The percetage of studets i divorced/separated households is much higher amog Africa America (27%), Hispaic (22%), America (25%), ad multi racial studets (22%), tha amog Caucasia (17%), Asia (11%), ad Pacific Islader studets (8%). Iterestigly, multi racial studets had the secod highest rates of married parets at the regioal level, folig closely behid that of Asia studets. Table 14. Parets marital status (percetage) Married/remarried Sigle Divorced/separated Widowed Natio Regio Natio Regio Natio Regio Natio Regio Caucasia Asia Africa Amer. / Black Hispaic / Latio America or Alaska Hawaiia / Pacific Islader Other More tha oe ethicity Overall A similar picture emerged i the NPSAS:04 data o educatioal levels, as Table 15 shows. At both atioal ad regioal levels, most studets parets had bachelor s degrees, with the exceptio of Africa America ad Hispaic studets whose parets were more likely to have a high school diploma or equivalet. At the regioal level, multi racial studets had the highest percetage of parets with a bachelor s degree, foled by Caucasia ad Asia studets. 12

13 Table 15. Paret s educatioal status (percetage) Natio (Nat) ad Regio (Reg) Caucasia Asia Africa Amer. / Black Hispaic / Latio America or Alaska Hawaiia / Pacific Islader Did ot complete high school High school diploma or equivalet Vocatioal or techical traiig some college o degree Associate's degree Bachelor's degree Master's, first professioal or higher Nat Reg Nat Reg Nat Reg Nat Reg Nat Reg Nat Reg Nat Reg Other More tha Studet Academics Academic performace is ofte a factor i studets decisios to leave college, ad ca play a sigificat role i studets ability to pursue advaced degrees. As Table 16 shows, overall college GPAs were slightly higher at the regioal level (3.0) tha at the atioal level (2.9), with uder represeted miority studets earig slightly er GPAs (ragig from ) tha did Caucasia ad Asia studets (about 3.0). These mior GPA differeces do ot appear to be traslatig ito degree expectatios. Overall, the majority of studets both atioally ad regioally, expect to ear a master s degree or higher. Table 16. Grades (stadardized to a 4.0 scale) Natio (GPA) Regio (GPA) Caucasia Asia Africa America / Black Hispaic / Latio America or Alaska 2.8 Hawaiia / Pacific Islader 2.8 Other 3.0 More tha oe ethicity Overall

14 Aother factor related both to attritio ad time to degree is whether studets take remedial coursework. Table 17 shows the percetage of studets by ethicity who took remedial or developmetal college courses to improve basic skills after they completed high school. As the table shows, at the atioal level, the percetage of studets who had take remedial courses was higher amog Pacific Islader (39%), Africa America (35%), ad America studets (33%) tha amog White (24%), Asia (26%), ad multi racial studets (23%). Overall, the percetage of studets who took remedial courses was slightly higher at the far west regioal level (30%) tha at the atioal level (26%), ad Africa America studets represeted the highest percetages of studets who took remedial courses at both the atioal ad the regioal levels (35% ad 36%, respectively). Table 17. Remedial courses ever take (percetage) Natio Regio Caucasia Asia Africa America / Black Hispaic / Latio America or Alaska 33 Hawaiia / Pacific Islader 39 Other 27 More tha oe ethicity Overall Oe of the risk factors idetified by the NPSAS:04 was erollmet itesity. This is geerally cosidered to be liked with studet persistece, such that studets who are erolled full time are more likely tha part time studets to complete their post secodary degrees. As Table 18 shows, most studets atteded school exclusively full time (64% at the atioal level ad 75% at the regioal level). However, at both the atioal ad the far west regioal level, a higher percetage of Africa America ad Hispaic studets, as well as multi racial studets atteded school exclusively part time, compared with Caucasia ad Asia studets. v Both atioally ad regioally, the, a larger percetage of uder represeted studets atteded school exclusively part time tha did other studets. I additio, at the regioal level, a slightly higher percetage of Caucasia studets had a mixed part time/full time attedace patter tha did uder represeted studets. 14

15 Table 18. Attedace (percetage) Exclusively full-time Exclusively part-time Mixed full-time ad part-time Natio Regio Natio Regio Natio Regio Caucasia Asia Africa America / Black Hispaic / Latio America or Alaska Hawaiia / Pacific Islader Other More tha oe ethicity Overall Furthermore ad cosistet with their more part time status tha Caucasia ad Asia studets, at both the atioal ad regioal levels, uder represeted miority studets completed a slightly er average umber of credits tha others. At the atioal level, America studets appeared to take the est average umber of credits aually, ad Asia ad Hawaiia studets had the highest aual average. At the regioal level, Asia studets had the highest average aual credits, as Table 19 idicates. Table 19. Total credit hours per year (average) Natio (credits) Regio (credits) Caucasia Asia Africa America / Black Hispaic / Latio America or Alaska 20 Hawaiia / Pacific Islader 31 Other 25 More tha oe ethicity Overall Other Factors Aother factor ofte associated with retetio is the frequecy of ad amout of time studets sped o extracurricular activities, icludig work they do to help support themselves i college. As Table 20 shows, Hispaic ad America studets teded to sped more time at work tha do Caucasia ad Asia studets. Natioally, Asia ad Africa America studets appear spet more time tha ay other group o commuity service. 15

16 Table 20. Hours of work per week (average) Hours worked per week excludig work-study/ Hours worked per week icludig work-study/ Commuity service average hours per moth Natio Regio Natio Regio Natio Regio Caucasia Asia Africa America / Black Hispaic / Latio America or Alaska Hawaiia / Pacific Islader Other More tha oe ethicity Overall I additio, the NPSAS:04 foud that Africa America ad Hispaic studets were more likely tha other groups to cite college tuitio, fees, or livig expeses as their primary reaso for workig ad to deem their work as havig a egative impact o their grades, as show i Table 21. Table 21. Impact of work o grades (percetage) Positive effect Negative effect No effect Natio Regio Natio Regio Natio Regio Caucasia Asia Africa America / Black Hispaic / Latio America or Alaska Hawaiia / Pacific Islader Low Other More tha oe ethicity Overall Payig for housig icreases the costs of college, but livig at home ca make it difficult for studets to separate from family demads ad focus o academic work. The NPSAS:04 foud that both atioally ad regioally a higher percetage of studets reported livig off campus tha o campus or with parets, as Table 22 idicates. Natioally, a higher percetage of Africa America studets reported livig o campus ad a higher percetage of America studets lived off campus. Hispaic ad other studets more ofte reported livig with their parets tha did other studets. Regioally, Asia studets more ofte reported livig o campus or with their parets, ad Africa America studets most ofte reported livig off campus. Regioal data were ot available for America studets ad Hawaiia studets. 16

17 Table 22. Housig (percetage) O campus Off campus Livig with parets Natio Regio Natio Regio Natio Regio Caucasia Asia Africa America / Black Hispaic / Latio America or Alaska Hawaiia / Pacific Islader Other More tha oe ethicity Overall SUMMARY AND CONCLUSIONS The atioal ad regioal data shed light o curret treds i higher educatio with regard to studets academic backgrouds, performace, ad expectatios; socio ecoomic backgrouds ad studet aid status; as well as activities ad obligatios outside class. Cosistet with previous research o factors that affect uder represeted miority studet retetio, the fidigs i this report suggest that uder represeted miority studets are ofte more likely tha White ad Asia studets to be at risk for attritio. Thus, the Uiversity of Washigto ca us this iformatio to explore how the atioal ad regioal treds compare to UW treds ad use this iformatio as baselie iformatio that might help develop programs ad services that mitigate the impact of these treds o uder represeted miority studet attritio. 17

18 REFERENCES Tito, V. (1993). Leavig college: Rethikig the causes ad cures of studet attritio. Chicago: Uiversity of Chicago Press. Hurtado, S., Milem, J. F., Clayto-Pederso, A. R., ad Alle, W. R. (1998). Ehacig campus climates for racial/ethic diversity: Educatioal policy ad practice. The Review of Higher Educatio, 21(3), Mohammadi, J. (1994). Explorig Retetio ad Attritio i a Two-Year Public Commuity College. Report. Martisville, VA: Patrick Hery Commuity College Istitutioal Plaig ad Research Iformatio Services. Redo, L. I. (1994) Validatig culturally diverse studets: Toward a ew model of studet developmet. Iovatig Higher Educatio 19(1), Redo, L. I., Jalomo, R. E., & Nora, A. (2000). Theoretical cosideratios i the study of miority studet retetio i higher educatio. I J. Braxto (Ed.) Rethikig the departure puzzle: New theory ad research o college studet retetio (pp ).nashville, TN: Vaderbilt Uiversity Press. i ii iii iv v The sample was limited to istitutios that were eligible to participate i the federal fiacial aid programs icluded i Title IV of the Higher Educatio Act. Oly approximate totals are provided to secodary researchers, such as OEA. Data were available to us as percetages ad are reported by us i the same form. NCES defies selectivity from a combiatio of variables from the Istitutioal Characteristics survey. Ope admissio 4-year istitutios were formed ito a separate category. For o-ope admissio istitutios, a idex was created from two variables: a) the cetile distributio of the percetage of studets who were admitted to each istitutio (of those who applied); ad b) the cetile distributio of the midpoit betwee the 25 th ad 75 th percetile SAT/ACT combied scores reported by each istitutio. See NPSAS:04 report: Attedace is based o the umber of moths erolled full-time or part-time at all istitutios erolled at durig the academic year. Accordigly, exclusively full-time is defied as erolled full-time for ie moths or more. 18