SOECover 11/2/09 4:07 PM Page 1 SOCIOLOGY EDUCATION · Jeff Keesler Nathan D. Jones and Timothy G....

SOCIO

LOGY

OFED

UCATIO

N(ISSN

0038-0407)1430

KStreet,N

WW

ashin

gton

,DC

20005-2529

Periodicals

postag

epaid

atW

ashin

gton

,DC

and

addition

almailin

goffices A Journal of the American Sociological Association

SOCIOLOGYOF

EDUCATIONVolume 82 October 2009 Number 4

SOCIO

LOGY

OFED

UCATIO

NVolum

e82,N

umber

4,Octob

er2009

HOUSE AD HERE

High School Classmates and College SuccessJASON M. FLETCHER AND MARTA TIENDA

Organization by Design:Supply- and Demand-side Models of

Mathematics Course TakingDANIEL A. MCFARLAND AND SIMON RODAN

Informal Mentors and Education: Complementary or Compensatory Resources?LANCE D. ERICKSON, STEVE MCDONALD, AND

GLEN H. ELDER, JR.

Another Way Out: The Impact of Juvenile Arrests on

High School DropoutPAUL HIRSCHFIELD

SOECover 11/2/09 4:07 PM

CONTEXTS

EDITORS //

Christopher UggenDouglas Hartman

DETAILS //

ISSN: 1536-5042eISSN: 1537-60522009, Vol 8February, May, August,November

An award-winning quarterly magazine of the American SociologicalAssociation, Contexts presents cutting-edge perspectives on themost provocative issues facing contemporary society. This pioneeringjournal brings accessible, incisive writing and the best of sociologicalinquiry to bear on crucial concerns such as poverty, education, popculture, immigration, religion, environmental justice, and much more.

W W W. U C P R E S S J O U R N A L S . C O M

Understanding People in their Social Worlds

Job Name:94524 PDF Page:94524pbc.p1.pdfColor:

Date:09-11-03Cyan Magenta Yellow Black PANTONE 274 C

MISSION STATEMENT: The journal provides a forum for studies in sociology of education and human social develop-ment throughout the life cycle. It publishes research from all methodologies that examines how social institutions andindividuals’ experiences in these institutions affect educational processes and social development. Such research mayspan various levels of analysis, from the individual to the structure of relations among social and educational institutions,and may encompass all stages and types of education at the individual, institutional, and organizational levels.

Hanna AyalonTel Aviv University

Pamela R. BennettJohns Hopkins University

William J. CarbonaroUniversity of Notre Dame

Wade M. ColeMontana State University

Elizabeth C. CookseyOhio State University

Susan A. DumaisLouisiana State University

Cynthia FelicianoUniversity of California–Irvine

Sara Goldrick-RabUniversity of Wisconsin-Madison

Eric GrodskyUniversity of California–Davis

Angel Lou HarrisPrinceton University

Joseph C. HermanowiczUniversity of Georgia

Charles HirschmanUniversity of Washington

Sylvia HurtadoUCLA

Douglas Lee LauenUniversity of North Carolina-Chapel Hill

Kevin T. LeichtUniversity of Iowa

Samuel R. LucasUniversity of California–Berkeley

Daniel A. McFarlandStanford University

Lynn M. MulkeyUniversity of South Carolina–Beaufort

Brian PowellIndiana University

Kelly RaleyUniversity of Texas

Sean F. ReardonStanford University

John R. SchwilleMichigan State University

Salvatore SaporitoCollege of William and Mary

Christopher B. SwansonEditorial Projects in Education

Tony TamChinese University of Hong Kong

Edward E. TellesUniversity of California–Los Angeles

Marta TiendaPrinceton University

Sarah TurnerUniversity of Virginia

Karolyn TysonUniversity of North Carolina-Chapel Hill

Julia WrigleyCUNY Graduate Center

MANAGING EDITORWendy Almeleh

EDITORIAL ASSISTANT GRADUATE STUDENT EDITORSJeff Keesler Nathan D. Jones and Timothy G. Ford

EXECUTIVE OFFICERSally T. Hillsman

SOCIOLOGY OF EDUCATION (ISSN 0038-0407) is published quarterly in January, April, July, and October by theAmerican Sociological Association, 1430 K Street, N.W., Suite 600, Washington, DC 20005-2529, and is printed byBoyd Printing Company, Albany, New York. Periodicals postage is paid at Washington, DC, and additional mailingoffices. POSTMASTER: Send address changes to Sociology of Education, 1430 K Street, N.W., Suite 600, Washington,DC 20005-2529.

Address manuscripts and communications for the editors to Barbara Schneider, Editor, SOCIOLOGY OF EDUCATION,Department of Education, Michigan State University, 516B Erickson Hall, East Lansing, MI 48824; [email protected].

Send advertisements, changes of address, and subscriptions to the Executive Office, American Sociological Association,1430 K Street, N.W., Suite 600, Washington, DC 20005-2529. Subscription rates for members, $35 ($25 studentmembers); institutions, $164. Rates include postage in the United States and Canada; elsewhere, add $20 per journalsubscription for international postage. Single issues available: $7 to members and $20 to institutions. New subscriptionsand renewals will be entered on a calendar-year basis only. Change of address: Send old and new addresses to the ASAExecutive Office six weeks in advance. Claims for undelivered copies must be made within the month following the reg-ular month of publication. The publishers will supply missing copies when losses have been sustained in transit and thereserve stock will permit.

Copyright ©2009, American Sociological Association. Copying beyond fair use: Copies of articles in this journal may bemade for teaching and research purposes free of charge and without obtaining permission, as permitted under Sections107 and 108 of the U.S. Copyright Law. For all other purposes, permission must be obtained from the publisher.

The American Sociological Association acknowledges, with appreciation, the facilities and assistance provided byMichigan State University.

EDITORBarbara Schneider, Michigan State University

DEPUTY EDITORJohn Robert Warren, University of Minnesota Ainsworth, James W.: See Downey, Douglas B.

Bennett, Pamela R.: How African American Is theNet Black Advantage? Differences in CollegeAttendance Among Immigrant Blacks, NativeBlacks, and Whites 70

Bonneau, Kara: See Stearns, Elizabeth

Buchmann, Claudia: See Stearns, Elizabeth

Downey, Douglas B.: Rethinking the Attitude-Achievement Paradox Among Blacks 1

Elder, Glen H., Jr.: See Erickson, Lance D.

Erickson, Lance D.: Informal Mentors andEducation: Complementary or CompensatoryResources? 344

Elder, Glen H., Jr.: See Erickson, Lance D.

Fletcher, Jason M.: High School Classmates andCollege Success 287

Goldrick-Rab, Sarah: Beyond Access: ExplainingSocioeconomic Differences in College Transfer 101

Herman, Melissa R.: The Black-White-OtherAchievement Gap: Testing Theories ofAcademic Performance Among Multiracialand Monoracial Adolescents 20

Hirschfield, Paul: Another Way Out: The Impact ofJuvenile Arrests on High School Dropout 368

Kelly, Sean: The Black-White Gap in MathematicsCourse Taking 47

Koo, Jeong-Woo: See Suárez, David F.

Landale, Nancy S.: See Oropesa, R. S.

López Turley, Ruth N.: College Proximity: MappingAccess to Opportunity 126

Mangino, William: The Downside of Social Closure:Brokerage, Parental Influence and DelinquencyAmong African American Boys 147

McDonald, Steve: See Erickson, Lance D.

McFarland, Daniel A.: Organization by Design:Supply- and Demand-Side Models of Mathematics Course Taking 315

Morgan, Stephen L.: Intergenerational Closure andAcademic Achievement in High School: A NewEvaluation of Coleman’s Conjecture 267

Oropesa, R. S.: Why Do Immigrant Youths WhoNever Enroll in U.S. Schools Matter? SchoolEnrollment Among Mexicans and Non-Hispanic Whites 240

Pfeffer, Fabian T.: See Goldrick-Rab, Sarah

Qian, Zenchao: See Downey, Douglas B.

Ramirez, Francisco O.: See Suárez, David F.

Rodan, Simon: See McFarland, Daniel A.

Stearns, Elizabeth: Interracial Friendships in theTransition to College: Do Birds of a Feather FlockTogether Once They Leave the Nest? 175

Stevens, Peter A. J.: See Van Houtte, Mieke

Suárez, David F.: UNESCO and the AssociatedSchools Project: Symbolic Affirmation ofWorld Community, International Understand-ing, and Human Rights 197

Tienda, Marta: See Fletcher, Jason M.

Todd, Jennifer J.: See Morgan, Stephen L.

Van Houtte, Mieke: School Ethnic Compositionand Students’ Integration Outside and InsideSchools in Belgium 217

INDEX TO VOLUME 82

SOECover 11/2/09 4:07 PM

Job Name:94524 PDF Page:94524pbc.p2.pdfColor:

Date:09-11-03Black

SOCIOLOGYOF

EDUCATIONVolume 82 October 2009 Number 4

Contents

High School Classmates and College SuccessJASON M. FLETCHER AND MARTA TIENDA 287

Organization by Design: Supply- and Demand-side Models of Mathematics Course Taking

DANIEL A. MCFARLAND AND SIMON RODAN 315

Informal Mentors and Education: Complementary or Compensatory Resources?

LANCE D. ERICKSON, STEVE MCDONALD, AND GLEN H. ELDER, JR. 344

Another Way Out: The Impact of Juvenile Arrests on High School DropoutPAUL HIRSCHFIELD 368

NOTICE TO CONTRIBUTORS

Editorial ProceduresAll papers considered appropriate for this journal are reviewed anonymously. To ensure anonymity, authors’names, institutional affiliations, and other identifying material should be placed on the title page only. Papersare accepted for publication subject to nonsubstantive, stylistic editing. A copy of the edited paper is sent tothe author for final review. Proofs of articles are sent only to authors who reside in the United States.Submission of a paper to a professional journal is considered an indication of the author’s commitment topublish in that journal. A paper submitted to this journal while it is under review for another journal will notbe accepted for review.

Preparation and Submission of Manuscripts.1. Type all copy (including indented material, references, and footnotes) double-spaced in no smaller than

12-point type using 11/2-inch margins on all sides.2. Type each table on a separate page. Insert a note in the text indicating where the table should appear.3. On an article’s acceptance, submit camera-ready art for all figures, rendered on a laser printer or as

glossy prints or electronic format.4. Include an abstract of no more than 100 words.5. Submit four copies of the paper and retain the original for your files. Enclose a stamped, self-addressed

postcard so we can acknowledge receipt of your paper.6. A check or money order for $25.00, payable to the American Sociological Association, must accompa-

ny each submission. This fee is waived for papers written by student members of the ASA. The submis-sion fee reflects a policy of the ASA Council and Committee on Publications, which affects all ASA jour-nals. It is a reluctant response to the accelerating costs of manuscript processing.

Reference Format1. In the text: All references to books, articles, and other works should be identified at the appropriate point

in the text by the surname of the author and year of publication; add page numbers only when citingstatistics or direct quotes. Endnotes should be used only for substantive observations and explanations.Subsequent citations of a source should be identified in the same way; do not use “ibid.,” “op. cit.,” or“loc. cit.”a. If the author’s name is part of the narrative, place only the year of publication in parentheses:

Duncan (1959). Otherwise, place both the name and the year, with no intervening punctuation,in parentheses: (Duncan 1959).

b. Insert page numbers, preceded by a colon after the year of publication: (Kuhn 1970:120–45).c. If the work cited has three or fewer authors, list all authors in the first citation; thereafter, include

only the name of the first author followed by “et al.” If the work has four or more authors, includeonly the name of the first author followed by “et al.” in all citations.

d. Abbreviate or shorten the names of institutional or corporate authors, making sure that the textcitation and the entry in the reference list begin with the same element.

e. Distinguish two or more works by the same author with the same publication date by appendingletters (a, b, c) to the date: (Levy 1965a).

f. Separate a series of references with semicolons and enclose them in a single pair of parentheses:(Featherman and Hauser 1979; Coleman et al. 1982; U.S. Bureau of Census 1981).

2. In the Reference List: List all entries alphabetically by author and, within author, by year of publication.List all authors in citations of multiauthor works; do not use “et al.” in the reference list.

Examples follow:Bourdieu, Pierre, 1977. “Cultural Reproduction and Social Reproduction.” Pp. 487–511 in Power and Ideology

in Education, edited by J. Karabel and A.H. Halsey. New York: Oxford University Press.Coleman, James S., Thomas Hoffer, and Sally B. Kilgore. 1982a. “Cognitive Outcomes in Public and Private

Schools.” Sociology of Education 55:65–76.——. 1982b. High School Achievement: Public, Catholic and Other Private Schools Compared. New York: Basic.Mare, Robert D. 1979. “Change and Stability in Educational Stratification.” Paper presented at the annual

meeting of the American Sociological Association, Boston.Marx, Karl (1867) 1976. Capital. Vol. 1. Translated by S. Moore and E. Aveling. New York: International.U.S. Bureau of the Census. 1979. 1970 Census Population and Housing. Fourth Count Population Summary

Tape. Machine-readable data file. Washington, DC: U.S. Bureau of the Census (producer). Rosslyn, VA:DUALabs (distributor).

Informal Mentors and Education: Complementary or Compensatory

Resources?

Lance D. EricksonBrigham Young University

Steve McDonaldNorth Carolina State University

Glen H. Elder, Jr.University of North Carolina at Chapel Hill

Few studies have examined the impact of mentoring (developing a special relationship with a

nonparental adult) on educational achievement and attainment in the general population. In

addition, prior research has yet to clarify the extent to which mentoring relationships reduce

inequality by enabling disadvantaged youths to compensate for the lack of social resources or

to promote inequality by serving as a complementary resource for advantaged youths. The

results of a nationally representative sample of youths show (1) a powerful net influence of

mentors on the educational success of youths and (2) how social background and parental,

peer, and personal resources condition the formation and effectiveness of mentoring relation-

ships. The findings uncover an interesting paradox—that informal mentors may simultaneous-

ly represent compensatory and complementary resources. Youths with many resources are

more likely than are other young people to have mentors, but those with few resources are

likely to benefit more from having a mentor—particularly a teacher mentor—in their lives.

344 Sociology of Education 2009, Vol. 82 (October): 344–367

Sociological research has long recognizedthe contributions of significant others inthe educational process (Sewell, Haller,

and Ohlendorf 1970), but the research hastended to focus on the contributions of par-ents, peers, and teachers (e.g., Conley 2001;Crosnoe, Johnson, and Elder 2004; Downeyand Pribesh 2004) and paid little attention toother individuals in the social networks ofyouths (but see Cheng and Starks 2002;Hofferth, Boisjoly, and Duncan 1998).However, an emerging line of work hasbegun to study the role of mentors in the lives

of young people (for reviews, see DuBois etal. 2002; Jacobi 1991). Mentors are non-parental adults who take a special interest inthe lives of youths. They step outside theirnormal social roles as teachers, relatives,youth workers, ministers, and employers byhelping to guide young people in the transi-tion to adulthood with advice and emotionalsupport and by serving as role models.

Much of the work on mentoring has exam-ined relationships with formal mentors, oradults who volunteer in organized mentoringprograms like Big Brothers/Big Sisters that aredesigned as interventions for youths who are

“at risk” or who experience substantial disad-vantage. Fewer studies have investigatedinformal mentoring—or naturally occurringmentoring—even though most young peoplereport a relationship with an informal mentor(Beam, Chen, and Greenberger 2002;McDonald et al. 2007). Scholarship on infor-mal mentoring has been limited because ofthe almost-exclusive focus on the role of men-toring in the lives of disadvantaged youths.Consequently, little is known about theimpact of mentoring on the educational per-formance and attainment of America’syouths.

Furthermore, most research has failed toexamine the interdependence of informalmentors and other social relationships.Studies have typically held constant theresources available to youths by sampling nar-rowly defined disadvantaged populations.When broader populations have been exam-ined, the effects of mentoring have tended tobe assessed independently of other potentialresources. These strategies mask issues ofselection and the extent to which existingresources condition the effectiveness of men-toring relationships. Without examining men-toring as a component of a larger constella-tion of social relationships among youths,past studies have been unable to identifywhether informal mentors enable youths tocompensate for the lack of available socialresources or complement the wealth ofresources maintained by the advantaged(Hamilton and Hamilton 2004; Rhodes 2002).The focus on mentors of disadvantagedyouths implies that mentoring helps at-riskyouths “catch up” to their more fortunatecontemporaries. Yet, informal mentoring mayactually be more prevalent among and moreeffective for advantaged youths.

In this article, we address these issuesthrough an empirical examination of theNational Longitudinal Study of AdolescentHealth (Add Health). First, we assess theimpact of informal mentoring on the educa-tional success of young people. In doing so,we draw from the Add Health AcademicAchievement study (AHAA), which links infor-mation on Add Health respondents to datafrom high school transcripts to examine stu-dents’ scholastic performance in high school.We also investigate educational attainmentthrough self-reported survey information in

Wave 3 of Add Health. These outcomes referto distinct, although related, processes, withachievement referring to the educational per-formance of students during the high schoolyears and educational attainment generatinginsights into longer-term educational trajecto-ries and prospects for socioeconomic careers.In combination, these perspectives provide amore nuanced understanding of the specificcontexts in which mentors are likely to matterthe most. Second, we investigate the specifictypes of mentors (relatives, friends, teachers,or community members) who have the great-est influence on educational performance andattainment. Third, we situate the relationshipbetween informal mentoring and educationalsuccess within the context of a broader set ofpotential resources (including those that arelinked to social background, parents andpeers, school, and the individual). In otherwords, we consider whether the availability ofthese various social resources conditions theformation and effectiveness of mentoringrelationships.

The present study shows that mentorshave a strong positive impact on both per-formance in high school and educationalattainment overall, even after other resourceson which youths may draw are controlled for.Relatives, friends, teachers, and community-based mentors all contribute to educationalsuccess. The findings also show that mentor-ing can serve as both a compensatory and acomplementary resource for young people.Mentoring relationships are more likely toform among youths with an abundance ofother resources to draw upon, thereby high-lighting the complementary role that mentor-ing plays for the socially advantaged. Theresults on the effectiveness of mentoring aremore mixed. Mentoring relationships with rel-atives result in more positive educationalattainment for advantaged than disadvan-taged youths. Relatives of advantaged adoles-cents most likely have valuable expertise oneducation. However, teacher mentors have asubstantial impact on the educational attain-ment of disadvantaged youths. Consequently,this article shows (1) how mentoring relation-ships can contribute to individual educationalsuccess and (2) how mentoring can promoteboth greater equality and greater inequality insociety.

Informal Mentors and Education 345

IMPACT OF MENTORING RELATIONSHIPS ON EDUCATION

By focusing on the influence of social relation-ships on educational success, our study fol-lows a tradition of research on social capital.Much of the initial emphasis on social capitalwas on the controlling and constrainingaspects of relationships (e.g., Coleman 1988).However, our research is more squarelyaligned with recent conceptualizations ofsocial capital that have highlighted the impor-tance of resources embedded in social net-works (Kim and Schneider 2005; Lin 2001;Maeroff 1998; Stanton-Salazar 2001;Stanton-Salazar and Dornbusch 1995). In thisway, we note that mentors serve as vitalresources that youths may draw on to achieveacademic success.

Mentors are nonparental adults in the livesof young people, typically several years olderthan their protégés, and may come from dif-ferent relationships: relatives, older siblingsand friends, teachers, coaches, clergy,employers, or coworkers (Rhodes 1994).What distinguishes typical nonparental adultrelationships from mentoring relationships—for example, teachers from teacher mentors—is that a mentor steps outside the boundariesof his or her typical role to take a special inter-est in a young person, offering advice andsupport to help the young person find his orher way in the social environment (Jacobi1991). Consequently, mentors represent apotentially important resource for youths dur-ing the transition to adulthood and beyond.

Compared to mentoring, other predictorsof educational success, such as social back-ground and the resources indexed by parents,peers, schools, and individual attributes, havereceived far greater attention fromresearchers. First, a number of social back-ground factors are associated with education-al performance, including race (Wiggan2007) and neighborhood characteristics(Garner and Raudenbush 1991). Second,numerous studies have documented how par-ents can positively influence their children’seducation through effective surveillance andsanctioning (Coleman and Hoffer 1987) andby providing information and encouragementabout educational opportunities (Kim andSchneider 2005; Stanton-Salazar and

Dornbusch 1995). Third, although manystudies have examined the negative influenceof peers (e.g., Woodward and Fergusson2000), peers can also serve as a positiveresource for young people. Friendship net-works can influence educational achievementby promoting engagement with school, pro-viding social support, and offering a set ofbehavioral models (Crosnoe, Cavanagh, andElder 2003; Nora et al. 1996; Robertson andSymons 2003). Fourth, youths also draw fromresources that are available in the school envi-ronment. Prior research has shown that small-er class sizes (Mosteller 1995) and schoolswith fewer enrollees (Kuziemko 2006) con-tribute substantially to students’ learning.Furthermore, student-teacher relationshipscreate school environments that can facilitatestudents’ engagement in school and reducerates of dropout (Lee and Burkam 2003). Last,young people also bring their own personalresources to social interactions and environ-ments that can influence their overall educa-tional trajectories. For example, researchershave long recognized that aspirations andaptitude have a strong impact on educationalattainment (e.g., Sewell et al. 1970).

Mentoring is distinct from these otherresources, yet few have examined its impacton education. Most research on the mentor-ing of adolescents has come from the “riskand resilience” tradition in developmentalpsychology. This work has shown that men-tors enable “at-risk” youths to adapt to signif-icant adversity in their lives (e.g., Masten andCoatsworth 1998). Even in the face of seriousthreats to their developmental well-being(e.g., neighborhood disadvantage, childhoodpoverty, and abuse), young people are oftenable to become competent, well-adjustedadults (Luthar, Cicchetti, and Becker 2000).Resilience depends, in large part, on access toand participation in nurturing relationshipswith nonparental adults (Werner and Smith1982, 2001).

Most of the empirical research on mentor-ing has drawn from samples of disadvantagedyouths (e.g., Rhodes, Ebert, and Fischer 1992)or has examined formal mentoring programsthat match at-risk youths to adult volunteersor mentors (e.g., Rhodes, Grossman, andResch 2000). The results from these studieshave shown that these relationships generallyhave a positive impact on the lives of disad-

346 Erickson, McDonald, and Elder

vantaged youths (DuBois et al. 2002).However, little has been said about the rolethat mentoring plays in the larger populationof youths. Mentors may (or may not) play aninstrumental role in the lives of youths whodo not experience similar adversity.Furthermore, the emphasis on mentoring informal programs fails to address the effective-ness of mentoring that occurs in natural envi-ronments.

The few studies that have examined natu-rally occurring mentoring relationships havereported that 75 percent to 80 percent ofyouths have an informal mentor (Beam et al.2002; McDonald et al. 2007). The results ofboth quantitative and qualitative researchhave shown that young people tend to bene-fit from having informal mentors in their lives(Ianni 1989; Maeroff 1998; Rhodes,Contreras, and Mangelsdorf 1994; Stanton-Salazar 2001; Williams and Kornblum 1985;Zimmerman, Bingenheimer, and Notaro2002). For example, Stanton-Salazar andSpina (2003) presented numerous examplesof how nonparental adults provide advice,emotional support, and role modeling to helpMexican American youths turn their livesaround. In particular, they emphasized theemotional support provided by mentors, asexpressed by the following quote from arespondent: “[Mentors] make you feel likeyou’re not the only one that’s going throughthis, that there are other people like youthat’ve succeeded” (p. 248).

Studies of informal mentoring are valuable,but they have been based primarily on smallregional samples of at-risk or disadvantagedpopulations, such as pregnant teenagers,lower-income African Americans, Hispanics,or urban youths (e.g., Klaw and Rhodes 1995;Sanchez and Reyes 1999; Zimmerman et al.2002). Consequently, much remainsunknown about informal mentoring in thelives of youths in the general population. Oneexception, though, is the research of DuBoisand Silverthorn (2005), who analyzed thepublic-use version of the Add Health data andfound that youths who identified informalmentors were more likely to complete highschool and to attend college.

We extend DuBois and Silverthorn’s (2005)research in a number of ways. First, we analyzethe relationship between mentoring and educa-tional outcomes using the full set of data from

the Add Health study, rather than the morerestrictive public-use sample. We also drawfrom a richer set of information from AddHealth. That is, we conducted analyses on mul-tiple waves of survey data, including the youthin-home interviews, in-school surveys, parentinterviews, peer network data, school adminis-tration data, and high school transcripts. Thisapproach allows us to analyze a more extensiveset of control variables and to expand the set ofdependent variables to examine educationalperformance in high school and overall educa-tional attainment. On the basis of evidencefrom prior studies, we anticipate that informalmentoring relationships have a positive effecton educational performance (grades in highschool) and on educational attainment (highestdegree received). We also expect this relation-ship to be independent of the social back-ground and other resources from which youthscan draw. With these issues in mind, we pro-pose the following hypothesis:

Hypothesis 1: Mentoring will be positivelyrelated to educational performance andattainment.

As we noted earlier, mentors vary in thesocial roles that they fill for young people.Consequently, we examine variation in theeffect of mentoring on educational outcomesacross different social roles. We expect that alltypes of mentors—relatives, friends, teachers,and community members—will have a posi-tive influence on educational performanceand attainment. However, teacher mentorsshould have a greater impact than other men-tors in view of their central role in the educa-tional process and the kinds of knowledge,skills, and ethics they impart to young people,which leads to Hypothesis 2:

Hypothesis 2: Relative, friend, teacher, and com-munity mentors will positively influence the edu-cational performance and attainment, butteachers are likely to have the greatest impact.

INFORMAL MENTORING IN CONTEXT

Mentors are merely one resource in a constel-lation of other potential resources (Higginsand Thomas 2001). While studies have exam-


ined the independent influence of mentoringon various outcomes, few have examinedmentoring within the broader context of ado-lescent relationships and environments(Hamilton and Hamilton 2004). The kinds ofresources and environments (e.g., coaches,friends’ parents, and religious youth groups)that are available to youths should affect thelikelihood of mentoring relationships and con-dition their effectiveness.

The second part of our analysis thereforeexplores how mentoring interacts with theseother resources to produce unique relation-ship configurations and educational out-comes. Understanding these patterns of inter-action is useful because the patterns help tospecify the role that mentoring plays in eitherpromoting equality in society or in perpetuat-ing existing inequities. In the literature on riskand resilience, the focus on at-risk youthsimplies that mentoring relationships help dis-advantaged youths to catch up with theirmore advantaged peers. This is a plausibleexpectation, especially in the context of for-mal mentoring programs that target at-riskyouths, but informal mentoring relationshipsthat develop naturally may not follow thispattern. These relationships may onlyenhance existing inequalities by enablingadvantaged youths to achieve even greatersuccess, while leaving disadvantaged youthsto fend for themselves. Theories of cumulativeadvantage or disadvantage imply that socialresources tend to accumulate for individualswho start out with many advantages in life(DiPrete and Eirich 2006; O’Rand 2006).Consequently, one of the more pressing ques-tions in research on mentoring is whethermentoring resources compensate for the lackof other social resources or complement theresources of those who are rich in social capi-tal (see Darling, Hamilton, and Shaver 2003;McDonald et al. 2007; Rhodes 1994).

Answering this question requires an under-standing of how existing resources both facil-itate selection into and condition the effec-tiveness of mentoring. Selection processes aresubstantively important because they leadyoung people either into or away from suchrelationships, ultimately influencing theirchances for educational success or failure(Caspi 2004). Personal resources, for exam-ple, should be considered important determi-nants of selection into mentoring relation-

ships, since they are not formed at random.Competent, intelligent, goal-oriented adoles-cents are likely to be more able to engage inrelationships with nonparental adults (Sewell,Haller, and Portes 1969). At the same time,these skills tend to make youths more attrac-tive as relationship partners to adults who areinclined to mentor young people. Therefore,we anticipate that mentoring relationships aremore likely to form among youths who areacademically gifted, physically attractive,and/or gregarious and easy to get along with.

Furthermore, access to a variety of socialenvironments (e.g., athletic teams and employ-ment) facilitates the development of mentoringrelationships (Barajas and Pierce 2001;Hamilton and Hamilton 2005; Mortimer 2003).Significant others are also more likely to formrelationships with young people with whomthey share similar cultural values (Stanton-Salazar and Dornbusch 1995). In this way,race/ethnicity and neighborhood disadvantagestructure access to informal mentors. Priorresearch has shown that young blacks in disad-vantaged inner cities have relatively few adultsin their communities whom they can dependon for guidance and positive role models(Newman 2000; Wilson 1987). Finally, youthswith many of the social resources described ear-lier would seemingly be more likely than disad-vantaged youths to develop useful relationshipswith nonparental adults. These resources facili-tate access to environments that are rich insocial capital and help to provide young peoplewith the skills that are necessary to engage inthese relationships and to identify their useful-ness. Some evidence indicates that advantagedyouths are more likely to develop informal men-toring relationships (Zimmerman,Bingenheimer, and Behrendt 2005), suggestingthat access to mentoring is indeed complemen-tary rather than compensatory, which leads toHypothesis 3:

Hypothesis 3: Youths with extensive socialresources (in terms of social background andparental, peer, school, and personal resources)will be more likely to form mentoring relation-ships.

The effectiveness of mentoring relation-ships also depends on existing resources. Forexample, studies of resilience have consistent-ly shown that intelligent youths are more like-ly to show resilience when faced with a variety


of personal and environmental risks (Sameroff,Gutman, and Peck 2003). This evidence sug-gests that personal resources are importantfactors that condition the effects of the largerenvironment. The same has been shown forparental, peer, and community resources(Rhodes et al. 1994). Mentoring may unlockbenefits that are latent within the existingresources of young people (Anderson 1990).In other words, advantaged youths may bebetter equipped to benefit from the opportu-nities afforded by mentoring relationships.However, other research has found the exactopposite: that mentoring may be more bene-ficial for at-risk youths. For example, evidencefrom qualitative interviews has revealed thatsupport from teachers, coaches, and otherschool personnel has a greater impact on thelives of youths from lower- and working-classbackgrounds than those from the middle- andupper-class backgrounds (Stanton-Salazar2001). This evidence is consistent with find-ings from quantitative analyses that havedemonstrated that contact with teachers out-side the classroom is negatively and signifi-cantly associated with dropping out amongat-risk youths, but not among socially advan-taged youths (Croninger and Lee 2001).Disadvantaged youths therefore have more togain from mentoring relationships and moreto lose from not having them. Consequently,we expect that while at-risk youths are leastlikely to develop informal mentoring relation-ships, they are more likely to benefit fromthem than are socially advantaged youths(Rhodes 2005), as suggested in Hypothesis 4:

Hypothesis 4: The effectiveness of mentoringon educational performance and attainmentwill be greatest for youths with few socialresources (in terms of social background andparental, peer, school, and personal re-sources).

METHODOLOGY

Data

Our study was based on data from theNational Longitudinal Study of AdolescentHealth (Add Health) and the AdolescentHealth and Academic Achievement (AHAA)study, which is an extension of Add Health.

Add Health is a nationally representative studyof adolescents in Grades 7–12 in the UnitedStates in 1994. The data included in-depthinterviews with adolescents and their parents,which provided detailed information on childoutcomes, family and peer relationships, andschool and neighborhood characteristics. AddHealth used a multistage, stratified, school-based cluster sampling design. Included inthe sample were students from 80 highschools (both public and private) and a corre-sponding feeder junior high or middle school.While some minority racial/ethnic groupswere sampled in proportion to their size with-in the U.S. population, smaller racial/ethnicgroups were oversampled. Additional infor-mation on the study can be found in Harris etal. (2003). The Add Health data in our studyincluded the in-school survey (1994), threewaves of in-home interviews (1995, 1996,and 2001), the parent survey (1994), and theschool administrator survey (1994).

As part of Wave 3 of Add Health, respon-dents were asked to sign a high school tran-script release form authorizing the collectionof their official transcripts from the last highschool they attended (see www.cpc.unc.edu/projects/addhealth). Approximately 91 per-cent of the respondents complied, resulting inthe collection of more than 12,000 tran-scripts. The AHAA team subsequently devel-oped a variety of indicators of students’course taking and performance throughouttheir high school experience (seewww.prc.utexas.edu/ahaa).

The nationally representative componentof Add Health (with valid sample weights) atWave 3 included 14,322 participants, for aresponse rate of 76 percent. This was thebaseline sample for this study because men-tors were assessed at Wave 3. We removed170 respondents from the sample who didnot have valid data on whether they had amentor or did not describe the relationshipwith this person. A small number of variablesin our analyses had relatively large amounts ofmissing data, partly because of the inclusionof data from various sources within the AddHealth data archive and the AHAA (rangingfrom 5 percent to 39 percent of cases).Following previous studies using Add Healthand AHAA (e.g., Riegle-Crumb, Farkas, andMuller 2006), we treated these cases usingembedded dummy variables (Hardy and


Reynolds 2004). This procedure creates adichotomous indicator for missing data andrecodes the substantive variable to its samplemean. The remaining variables had negligibleamounts of missing data and were treatedwith listwise deletion. This procedure resultedin 9,216 cases with valid data for 12th-gradeGPA (from AHAA) and 12,621 cases for thehighest degree achieved. To ensure temporalordering for the analysis of the contexts thatcondition the formation of mentoring rela-tionships (Hypothesis 3), we restricted analy-ses of mentors’ prevalence (Table 3) to only6,819 cases; 4,433 respondents reported thattheir mentor became important before AddHealth’s initial data collection. In other words,by limiting the analysis predicting mentoringrelationships in this way, we attempted tomake certain that the analyses represent howthe characteristics of adolescents conditionedthe formation of mentoring relationships andnot the effect of mentoring on the character-istics of adolescents.1 Those who were exclud-ed here were included in all the other analy-ses.

Because Add Health is clustered by schooland cases had an unequal probability of beingselected into the sample, biased coefficientsand underestimated standard errors are likely.To correct for this problem, all models wereestimated using the SURVEY commands inSTATA, which provides correct estimates ofcoefficients and standard errors (Chantala2002).

Measures

Dependent Variables Twelfth-grade GPAwas taken from the AHAA and represents theaverage of all courses taken for the entire sen-ior year. Highest degree achieved came fromthe Wave 3 in-home interviews; it is an ordi-nal measure with five categories—less thanhigh school degree, high school degree orgeneral equivalency diploma, some collegebut no degree, a two-year college degree,and a four-year college degree or more.

Mentoring All data on mentoring camefrom the Wave 3 in-home interviews and werereported by the young person. Informal men-tors were identified with the following ques-tion: “Some young people know adults, otherthan their parents, who make an important

positive difference in their lives. Some do not.Has an adult, other than your parents or step-parents, made an important positive differ-ence in your life at any time since you were 14years old?” If there was more than one influ-ential adult, the respondents were asked toreport only on the most important.

The mentor’s social role, or relationship tothe young person, was identified by therespondent as an adult relative (brother, sis-ter, grandparent, aunt, uncle, spouse, or part-ner), friend, teacher (teacher or guidancecounselor, coach or athletic director), or com-munity member (minister, priest, rabbi, orother religious leader; employer; coworker;neighbor; friend’s parent; doctor, therapist, orsocial worker; or other). It is important to notethat most of these roles are formal roles, andsome place adults in positions in which theyare expected to contribute to the develop-ment of young people (e.g., teacher and min-ister). However, the mentoring behaviorsthemselves, or the perception that someonein one of these roles is a mentor, occurs infor-mally. We do not consider that any of theroles just listed occur within formal mentoringorganizations, which is a requirement foridentifying formal mentors.

Social Background Race-ethnicity was codedwhite, black or African American, Hispanic,American Indian or Native American, andAsian or Pacific Islander. In Add Health,Hispanic ethnicity is assessed using a separatequestion than race. In our variable, those whoreported Hispanic ethnicity were coded asHispanic regardless of their racial classifica-tion. The neighborhood disadvantage variablewas developed to identify the concentrateddisadvantage that is associated with raciallysegregated urban neighborhoods (Sampson,Morenoff, and Earls 1999). The scale useddata from the 1990 U.S. census and wasmeasured at the block-group level. Five itemswere included: percentage below the povertyline, percentage receiving public assistance,percentage unemployed, percentage offemale-headed families with children, andpercentage black. The items were submittedto a principle-components factor analysis, andonly one factor was extracted with factorloadings ranging from .72 to .89. In creatingthe summed scale, individual items wereweighted by their factor scores.


Parental Resources In the parent survey, theresponding parent reported the total house-hold before-tax income. We also includedwhether or not the responding parent was amember of the PTA as a measure of intergen-erational closure. Living in a two-parent biolog-ical family was coded 1 if the young personlived with both biological parents and 0 for allother family types. Highest parent’s educationwas measured using the parent report. Theresponding parent was asked: “How far didyou go in school?” and “How far did your cur-rent spouse or partner go in school?” Parentaleducation was the higher of the two reports.If a parent report was not available, we usedthe Wave 1 child in-home data (when avail-able) or the in-school data. Relationship withparents consisted of the mean score of fouritems concerning the quality of the parent-child relationship. The youths said whetherthey (1) felt close to their mothers or fathers,(2) felt that their mothers or fathers werewarm, (3) felt that they communicated wellwith their mothers or fathers, and (4) weresatisfied with their relationships with mothersor fathers. The responses were rated on a 5-point scale, which ranged from 1 (“not at all”for the first item and “strongly disagree” forthe remaining three) to 5 (“all the time” forthe first item and “strongly agree” for theremaining three). Mean scores were calculat-ed for these items (a Cronbach’s alpha of .86for fathers and .90 for mothers). The mean ofthe mothers’ and fathers’ scores was used.

Peer Resources The respondents were askedto nominate up to 10 friends (5 male and 5female). Number of friends is the sum of therespondents’ total nominations. When one ofthe nominated friends was also a member ofthe Add Health sample, it was possible tomatch the nominations with the data fromtranscripts. Mean friends’ 9th-grade GPA wasthe mean of the average grade for the entireschool year for all nominated friends whocould be matched with valid data from thetranscript. This variable was used as a predic-tor of 12th-grade GPA and educational attain-ment (see Table 2). For the analysis of thementoring antecedents (see Table 3 andAppendix Table A1), we used mean friends’Wave 1 GPA, which was the mean of the aver-age grade for all nominated friends who hadvalid Wave 1 data. Peer network centrality was

measured using the Bonacich centrality meas-ure (Bonacich 1987), which is an ego-basedmeasure of centrality that is weighted by thecentrality of those to whom ego is tied.

School Resources We used two questionsfrom the in-school survey to assess the stu-dent-teacher environment at school. Therespondents reported how often they hadtrouble getting along with teachers at theirschool and how much they agreed that teach-ers at their school treat students fairly. Themean of these two items was calculated foreach school. School administrators estimatedthe average class size in their school. They alsoreported on the overall school size. Theresponse categories presented to them were(1) small (1–400), (2) medium (401–1,000),and (3) large (1,001–4,000).

Personal Resources The interviewers wereasked to rate each respondent’s attractivenessin terms of physical appearance and personali-ty. Responses ranged from 1 “very unattrac-tive” to 5 “very attractive.” The respondents’college aspirations were assessed using the fol-lowing question: “How much do you want togo to college?” Responses ranged from 1(low) to 5 (high). The Add Health PictureVocabulary Test (PVT) is an abridged versionof the Peabody Picture Vocabulary Test,Revised (Dunn 1981), which was designed tomeasure hearing vocabulary for StandardAmerican English. Scores were standardizedby age to have a mean of 100 and a standarddeviation of 15. To account for some of thedifficulties with the timing of the formation ofmentoring relationships and Add Health’smeasurement of mentoring, we used two dif-ferent measures of GPA. The analysis of 12th-grade GPA and highest degree achieved (seeTable 2) used 9th-grade GPA, which was takenfrom the AHAA and is the average of all cours-es taken for the entire 9th-grade year. Weused the self-report of grades from Wave 1 tocreate a measure of GPA in the analysis pre-dicting mentoring relationships (see Table 3and Appendix Table A1).

Additional Variables We controlled for anumber of additional variables. Age was meas-ured in years at the time of the Wave 3 inter-view. Gender was indicated with a dummyvariable for female (male = reference catego-


ry). Private school was also a dummy variable(private school = 1, public school = 0).Extracurricular activities were measured bysumming the affirmative responses to a list ofactivities that the students reported beinginvolved in, including school academic clubs(e.g., math or French), sports teams, and stu-dent government. The respondents reportedthe number of hours they worked for pay dur-ing a typical nonsummer week. We createdcategories representing 0 hours, 1 to 20hours, and 20+ hours. Descriptive statistics forall the variables are presented in Table 1.

Analysis

The analysis proceeds in three stages. First, wemodel 12th-grade GPA and highest degreeachieved using ordinary least-squares (OLS)regression and the ordinal logit model (Long1997), respectively. This set of analyses con-trols for resources that are related to educa-tion to assess the independent effects of men-toring (Hypothesis 1) and whether a mentor’ssocial role has an impact on the educationalprocess (Hypothesis 2). Second, we investi-gate whether the prevalence of mentoringrelationships for youths depends on theiradvantaged or disadvantaged status, as indi-cated by social background and parental,peer, school, and personal resources(Hypothesis 3). To ensure that our measuresof advantage/disadvantage condition the for-mation of mentoring relationships as opposedto being their result, we include only respon-dents in this portion of the analysis whoreported that their mentor became importantafter Wave 1. Third, we test whether the effectof mentoring relationships on educationaloutcomes depends on resources available toyouths (Hypothesis 4). Specifically, we exam-ine interactions between having a mentor (inparticular, the mentor’s social role) and otherresources, such as parents’ socioeconomicstatus, friends’ grades, a positive student-teacher environment, and educational aspira-tions. Positive interactions between youthresources and mentoring would indicate thatmentors are more effective for youths withmany resources; negative interactions suggestthat mentors are more effective for disadvan-taged youths who lack resources.

RESULTS

First, we report the impact that informal men-tors have on the educational achievementand attainment of young people. Models 1–3in Table 2 present the slope coefficients fromthe OLS regression on 12th-grade GPA. It isimportant to note that because we controlledfor 9th-grade GPA, the results for the firstthree regression models report the relation-ship between mentoring and a change in ayoung person’s GPA between the 9th and12th grades. Model 1 estimates the relation-ships between GPA and various resources andserves as a baseline for testing the independ-ent effect of mentoring on grades.

On average, the young women displayed agreater increase in grades over the course ofhigh school than did the young men.Students who attended private schools hadsmaller increases in grades than did those inpublic schools, perhaps because privateschool students had higher GPAs beginning inthe 9th grade (the mean GPA for students inprivate schools was 2.94 compared to 2.57 inpublic schools). Black and Asian youthsshowed less change in GPA relative to whites.Parental resources (living with biological par-ents, highest parent’s education, and qualityof the relationship with parents) are all posi-tively related to change in grades. Amongpeer resources, only friends’ average GPA issignificantly related to grades—young peoplehave higher grades when their friends arehigh achievers. Having a positive student-teacher environment in the school is associat-ed with students’ higher achievement. Beingmore physically attractive is also positivelyrelated to an increase in grades from thefreshman to the senior year. In Model 2, hav-ing a mentor is highly significant and positive-ly related to higher grade achievement, evenafter all the other independent variables arecontrolled. The control variables changed lit-tle with mentoring included in the regressionmodel. The results in Model 3 show that allthe social roles of mentors except for friendscontribute to significant increases in grades.

Models 4–6 in Table 2 focus on the highestdegree achieved and report the odds ratiosfrom ordinal logistic regression on this out-come. Odds ratios represent the factorchange in the odds of an event occurring fora unit change in the independent variable.



Table 1. Informal Mentoring and Education: Descriptive Statistics (N = 12,621)

Variable Mean SD Minimum Maximum

12th-Grade GPAa 2.75 .73 0 4Highest degree achieved 2.71 1.08 1 5Age 21.72 1.82 18 27Female .49 .50 0 1Private school .07 .26 0 1Extracurricular activities 1.64 2.12 0 15Work

0 hours .47 .50 0 11 - 20 hours .41 .49 0 121+ hours .12 .32 0 1

Social BackgroundRace-ethnicity

White .65 .48 0 1Black .16 .36 0 1Asian .04 .20 0 1Hispanic .12 .32 0 1Native American .03 .17 0 1

Neighborhood disadvantage .55 .43 0 3Neighborhood disadvantage—Missing .14 .35 0 1

Parental ResourcesFamily income (in $1,000) 45.82 40.74 0 999Family income—Missing .21 .41 0 1Parent in PTA .33 .44 0 1Parent in PTA—Missing .12 .32 0 1Two-parent biological family .57 .49 0 1Highest parent’s education 2.80 1.01 0 4Relationship with parents 3.26 .66 1 4

Peer ResourcesNumber of friends 6.43 3.46 0 10Friends’ mean 9th-grade GPA 2.86 .43 0 4Friends’ mean 9th-grade GPA—Missing .39 .49 0 1Friends’ mean Wave 1 GPA 2.89 .53 1 4Peer network centrality .83 .52 0 4Peer network centrality—Missing .32 .47 0 1

School ResourcesStudent-teacher environment 3.60 .14 3 4Student-teacher environment—Missing .28 .45 0 1Average class size 25.60 4.94 10 39School size 2.22 .71 1 3

Personal ResourcesPhysical attractiveness 3.57 .86 1 5Personality attractiveness 3.60 .83 1 5College aspirations 4.44 1.02 1 5Picture Vocabulary Test 100.55 14.09 10 137Picture Vocabulary Test—Missing .05 .22 0 1

Odds ratios higher than 1 represent anincrease, and those lower than 1 indicate adecrease in the odds of the outcome. Forexample, the odds ratio for age in Model 4 is1.415, indicating that a year increase in ageincreases the odds of obtaining more educa-tion by a factor of 1.415, or about 42 percent.Alternatively, the odds ratio for working 21hours or more per week, .892 (though notstatistically significant), means that intensivework experience decreases the odds of highereducational achievement by a factor of .892(or by about 11 percent) relative to not work-ing.

Model 4 reports the baseline results beforeexamining whether having a mentor affectseducational attainment. Older youths com-pleted more schooling, but they also hadmore time to advance through higher educa-tion.2 Young women achieved higher levels ofeducation, as did students who participatedin extracurricular activities and attended pri-vate school. Black, Hispanic, and Asian youthstend to rank higher on educational attain-ment when other resources are controlled,while youths from disadvantaged neighbor-hoods rank lower. Higher incomes and havinga parent in the PTA are both associated withincreases in the odds of further educationaladvancement. Parents’ education and livingwith both biological parents are positivelyrelated to attainment, but the quality of theparent relationship is not. Number of friend-ships, friends’ GPA, and being more central inone’s peer network are all positively related toattainment. School resources, however, are

not significantly related to the highest degreeattained. Being more personally attractive,having higher educational aspirations andhigher grades as a freshman, and scoringhigher on the PVT are all significantly predic-tive of advancement in educational attain-ment. Having a mentor is also beneficial foreducational attainment (see Model 5). Youthswith a mentor are 53 percent more likely toadvance to the next level of education thanare youths who do not have a mentor.Teacher mentors are most strongly related toattainment, but relative, friend, and commu-nity mentors also have positive and significanteffects (see Model 6).

To summarize the analysis thus far, infor-mal mentors have a positive impact on youngpeople’s education, lending strong support toHypothesis 1. These findings are consistentwith those in the literature that mentors aresignificant influences on the education ofyoung people. The effect of mentoringremains strong and statistically significanteven after the effects of other resources(including social background and parent,peer, teacher, and personal resources) that areknown to influence education are controlled.

We turn next to whether the resources ofstudents facilitate the formation of mentoringrelationships. Table 3 reports the predictedprobabilities of having a mentor with a partic-ular social role for advantaged and disadvan-taged youths. These results are based on themultinomial logistic regression reported inAppendix Table A1.3 We present our findingsas predicted probabilities because this


Table 1. Continued

Variable Mean SD Minimum Maximum

9th-grade GPA 2.60 .89 0 4Wave 1 GPA 2.89 .75 1 4

MentoringMentor .75 .44 0 1

Social roleRelative .26 .44 0 1Teacher .19 .39 0 1Friend .13 .34 0 1Community .16 .37 0 1

aValid N = 9,216.


Table 2. Influences on Educational Achievement and Attainment

12th-Grade GPAa Highest Degree Achievedb

Variable (1) (2) (3) (4) (5) (6)

Age -.007 -.005 -.004 1.373*** 1.381*** 1.387***Female .204*** .203*** .204*** 1.312*** 1.298*** 1.301***Private school -.113* -.110 -.110 1.842*** 1.860*** 1.842***Extracurricular activities .008 .007 .007 1.058*** 1.055*** 1.052***Work

0 hours — — — — — —1–20 hours -.023 -.028 -.026 1.052 1.036 1.04421+ hours -.043 -.049 -.046 .891 .879 .889

Social BackgroundRace

White — — — — — —Black -.076* -.078* -.077* 1.760*** 1.755*** 1.754***Asian -.116* -.114* -.115* 1.373 1.387 1.376Hispanic -.037 -.035 -.035 1.225 1.247 1.250Native American -.031 -.031 -.028 .814 .818 .823

Neighborhood disadvantage -.028 -.029 -.031 .826 .823* .824*Neighborhood disadvantage—Missing .010 .008 .006 .926 .919 .912

Parental ResourcesFamily income (in $1,000) .001** .001** .001** 1.003*** 1.003*** 1.003***Family income—Missing -.021 -.020 -.019 .970 .973 .973Parent in PTA .017 .019 .018 1.217*** 1.226*** 1.222***Parent in PTA—Missing .041 .043 .042 .986 .990 .982Two-parent biological family .058** .058** .057** 1.579*** 1.579*** 1.573***Highest parent’s education .044*** .041*** .041*** 1.443*** 1.431*** 1.435***Relationship with parents .035* .035* .035* 1.045 1.042 1.042

Peer ResourcesNumber of friends .004 .003 .003 1.020 1.015 1.015Friends’ mean 9th-grade GPA .102*** .101*** .100*** 1.174** 1.167* 1.163*Friends’ mean 9th-grade GPA—Missing .004 -.001 .000 1.155 1.126 1.125Peer network centrality .008 .010 .011 1.152* 1.157* 1.159*Peer network centrality—Missing .119 .116 .119 .824 .813 .820

School ResourcesStudent-teacher environment .255* .250* .246* 1.381 1.343 1.331Student-teacher environment—Missing -.090 -.085 -.089 1.021 1.045 1.029Average class size -.002 -.002 -.002 .990 .990 .990School size -.084** -.085** -.086** 1.283** 1.278** 1.269**

Personal ResourcesPhysical attractiveness .035** .036** .037** 1.028 1.035 1.038Personality attractiveness .015 .013 .012 1.089* 1.079* 1.079*College aspirations .023 .021 .020 1.489*** 1.482*** 1.477***Picture Vocabulary Test .002* .002 .002 1.026*** 1.025*** 1.024***Picture Vocabulary Test—Missing -.008 -.009 -.010 1.038 1.033 1.0289th-grade GPA .556*** .553*** .551*** 2.301*** 2.290*** 2.268***

approach allowed us to create hypotheticalprofiles of youths that have substantive mean-ing (see Table 3 for a description of the pro-files). To avoid extreme and unrealistic esti-mates, we use the 20th and 80th percentilesto represent youths who are advantaged anddisadvantaged in terms of resources. Forexample, we compare the probabilities ofhaving mentors for youths from advantagedsocial backgrounds (whites and Asians plusthose in the 20th percentile for neighborhooddisadvantage) to youths from disadvantagedsocial backgrounds (blacks, Hispanics, andNative Americans plus those in the 80th per-centile for neighborhood disadvantage).4

Overall, disadvantaged youths are signifi-cantly less likely to report having a mentorthan are advantaged youths across each ofthe categories that we modeled. Youths fromadvantaged social backgrounds identify amentor 66 percent of the time compared to62 percent for youths from disadvantagedbackgrounds, but this difference is not statis-tically significant. Parental resources, howev-er, significantly predict having a mentor.Having advantaged peer resources, such ashaving more friends, having friends withhigher GPAs, and being central in one’s peernetwork, increases the chances of having amentor—68 percent for advantaged youthsversus 61 for disadvantaged youths. Schoolresources are unrelated to identifying a men-tor. Personal resources, such as having a moreattractive personality and physical appear-

ance, higher educational aspirations, andintelligence, display the largest gap in theprobability of having a mentor. Seventy-fourpercent of these advantaged youths havementors compared to only 53 percent of thedisadvantaged. A young person who faces allthese disadvantages has only a 44 percentlikelihood of having a mentor, while anadvantaged youth has an 82 percent chance.Although no respondents in the sample weredisadvantaged or advantaged in all theseways, this difference provides a sense of thepower that resources exert on the possibilityof having a mentoring relationship.

The overall pattern of advantaged youthsbeing more likely to have a mentor persists,regardless of the mentor’s social role,although in some cases there are no differ-ences in the predicted probabilities or the dif-ferences were not statistically significant. Thelevel of parental resources significantly pre-dicts the probability of having a teacher men-tor: 20 percent of youths with advantagedparental resources identified a teacher men-tor, compared to 14 percent of youths withfew parental resources. Personal resourcescontribute significantly to the likelihood ofhaving a teacher mentor or a mentor in thecommunity. The combined advantage versusdisadvantage comparison suggests that socialresources play an important role in channel-ing youths into mentoring relationships, espe-cially those with teachers and communitymembers.


Table 2. Influences on Educational Achievement and Attainment

12th-Grade GPAa Highest Degree Achievedb

Variable (1) (2) (3) (4) (5) (6)

MentoringMentor .103*** 1.529***Social role

Relative .100*** 1.501***Friend .059 1.400***Teacher .151*** 1.987***Community .088** 1.303***

Constant -.483 -.514 -.491N [9,216] [9,216] [9,216] [12,621] [12,621] [12,621]

aUnstandardized coefficients from the OLS regression.bOdds ratios from the ordered logistic regression. *p < .05, **p < .01, ***p < .001; two-tailed tests.

Finally, we examined the interactionsbetween mentors’ social role and theresources available to young people to deter-mine the extent to which mentoring relation-ships unlock existing resources or compensatefor the lack thereof. We used Model 3 (for

GPA) and Model 6 (for the highest degreeachieved) from Table 2 as baselines and esti-mated a series of models that included a sin-gle interaction term between the mentor’ssocial role and a youth resource. In Table 4,we report only the significant interaction


Table 3. Predicted Probabilities of Having a Mentor from the Multinomial LogisticRegression for Advantageda and Disadvantagedb Youths

Mentor’s Social Role

Variable Mentor Relative Friend Teacher Community

Social BackgroundAdvantage 66 11 15 18 22Disadvantage 62 13 16 15 18

Parental ResourcesAdvantage 68 14 14 20 20Disadvantage 60 10 16 14 20

Peer ResourcesAdvantage 68 12 17 17 21Disadvantage 61 12 14 15 20

School ResourcesAdvantage 64 11 19 16 18Disadvantage 65 12 14 18 21

Personal ResourcesAdvantage 73 12 13 23 25Disadvantage 53 11 17 10 15

Full ModelAdvantage 82 11 15 32 24Disadvantage 44 9 15 7 13

Note: N = 7,840—Includes only respondents whose mentors became important after Wave1 to ensure temporal ordering. Bolded predicted probabilities represent statistically significantdifferences (i.e., did not have overlapping 95% confidence intervals) within type of resourceand social role. Within a row of the table, the sum of the mentor’s social roles may not equalthe report of mentor because of rounding.

aAdvantage: social background—white or Asian,.24 neighborhood disadvantage, parentalresources—$58,000 parental income; participates in the PTA, lives with both biological par-ents, has a parent with a college degree, 3.75 connection to parents; peer resources—nomi-nated 10 friends, 3.3 mean friends’ Wave I GPA, 1.16 peer network centrality; schoolresources—4 teacher-student environment, 20 students per class at school, 2 (401–1,000 stu-dents) school size; personal resources—very attractive personality, very attractive physically,high aspirations to attend college, 113 PVT score, 3.5 GPA.

b Disadvantage: social background—black, Hispanic, or Native American, .78 neighborhooddisadvantage; parental resources—$23k parental income; Does not participate in the PTA, doesnot live with both biological parents, highest parent education is high school diploma, 2.5 con-nection to parents; peer resources—nominated 4 friends, 2.5 mean friends’ Wave 1 GPA, .34peer network centrality; school resources—3.5 student-teacher environment, 30 students perclass at school, 3 (1,001–4,000 students) school size; personal resources—unattractive person-ality, physically unattractive, low aspirations for college, 87 PVT score, 2.0 GPA.

coefficients from these models. To decreasethe likelihood of chance results, we presentonly interactions that were significant at p <.01. For the most part, negative interactionsindicate that mentors are compensatory,while positive interactions involve comple-mentary resources.

Few of the interactions proved to be statis-tically significant for the models predictingeducational performance in high school. Noclear patterns of association emerged, sug-gesting that mentoring has a similar impacton the educational achievement of those withboth few and many social resources.However, the educational attainment interac-tions reveal two notable patterns. First, men-toring from relatives interacts with several ofthe personal resource variables in a positivedirection, suggesting that relative mentorscomplement existing personal resources ineducational attainment. In other words, rela-tives in the mentor role are more effectiveamong youths who already have substantialpersonal resources. Second, teacher mentor-ing interacts negatively with indicators fromeach of the resource domains (with theexception of school resources). Note that theonly positive interaction for teacher mentorsis with neighborhood disadvantage, which iscoded in the opposite direction (higher valuesindicate a greater disadvantage) compared to

the other variables (where higher values referto a greater advantage).

It is important to keep in mind that thesecoefficients represent interaction effects andcannot be interpreted in isolation from theirmain effects. For example, the interactioncoefficient for teacher and white or Asian is -.427 (see Table 4). The negative valence ofthe coefficient indicates that the change inthe odds of having a mentor is smaller forwhite or Asian youths than for youths fromother racial-ethnic backgrounds. Whether ornot this finding ultimately means that teachermentors harm white or Asian youths dependson the magnitude of the interaction and maineffects. To clarify the interpretation of interac-tion effects and provide a sense of their sub-stantive importance, we plotted the changein probability of attending college across dif-ferent levels of resources for youths whoreported a relative and teacher mentor inFigures 1 and 2. These figures incorporateboth interaction and main effects. The valuesof resources measured as continuous variables(presented on the category axis) represent the20th, 40th, 60th, and 80th percentiles.

Figure 1 reveals the complementary role ofrelatives as mentors. Youths with these mentorsare more likely to attend college than are thosewithout, but the gap between the two is great-est among youths with the greatest amount of


Table 4. Interactions Between Mentoring and Other Resources

Variable Relative Friend Teacher Community

12th Grade GPAWhite or Asian -.109Student-teacher environment .351Parent in PTA .138Number of friends -.020

Highest Degree AchievedWhite or Asian -.427Neighborhood disadvantage .409Parent in PTA -.505Highest parent’s education -.268Friends’ average GPA -.433Personality attractiveness .1739th-grade GPA .187 -.288Picture Vocabulary Test .013 -.019

Note: To reduce the potential of reporting chance results, we present only coefficients thatare significant at p < .01, two-tailed tests.

personal resources. In contrast, the compensa-tory nature of mentoring relationships is shownin Figure 2 by large differences in the probabil-ity of attending college when resources are lowand small differences when resources are high.For example, youths with parents whose educa-tion is limited have only a 35 percent probabil-ity of attending college. But if they have ateacher as a mentor, their chances increase to65 percent. Children of highly educated par-ents are very likely to go on to college, regard-less of whether they have a teacher as mentor(75%) or not (67%). In other words, the nega-tive interaction indicates that the help ofteacher mentors is less consequential for theadvantaged than it is for the disadvantaged.Thus, relative mentors serve as complementaryresources for educational attainment, whereasteacher mentors have a compensatory impacton educational attainment.

DISCUSSION AND CONCLUSION

Informal mentoring is an important andunderstudied resource for youths in their edu-

cational careers. In a nationally representativesample of youths, we find that exposure tothe mentoring of an adult is associated withgreater educational success, in terms of botheducational performance in high school andoverall educational attainment. The effect ofmentoring on education remains strong evenafter social background and parental, peer,school, and personal resources are controlled.Mentors vary by social role in their effect onyouths’ educational outcomes. It is not sur-prising that young people with teachers asmentors tend to have greater educational suc-cess, whereas having a friend as a mentor isunrelated to an improvement in grades inhigh school.

Contrary to usual expectations, much ofthe evidence shows that mentoring relation-ships that develop naturally have the potentialfor contributing to—rather than reducing—social inequality. First, youths have unequalaccess to the benefits of informal mentoring.Overall, those with an advantaged back-ground are more likely than the disadvan-taged to have an informal mentoring relation-ship. In other words, mentoring is most com-


Figure 1. Interaction Between Having a Relative as a Mentor and Youth Resources on theHighest Degree Achieved: Probability of Attending College. Note: The category axis representsthe 20th, 40th, 60th, and 80th percentiles.

mon among youths who already possess awealth of social resources. Second, relativesappear to be more effective as mentors for theeducational attainment of youths who havemany personal resources on which to draw.This finding suggests that young people witha great many personal resources are in a bet-ter position than are others to take advantageof the guidance, advice, and support provid-ed by relatives in the mentor role. In theseways, mentoring serves primarily as a comple-mentary resource for advantaged young peo-ple.

These findings are consistent with priorresearch, which indicated that middle- andupper-class students develop a broader net-work of support relationships than do lower-and working-class students (e.g., Ianni 1989).Future research should investigate in greaterdetail the microprocesses that link youths toinformal mentors. For example, recentresearch suggested that some individuals maypossess a “social intelligence” that enablesthem to establish rapport and manage social

interactions with relative ease (Goleman2006). Young people with these abilitieswould be more likely than others to establishand maintain relationships with nonparentaladults. At the same time, other researchershave noted that a youth’s failure to engage inrelationships with adults is often a purposiveresponse to disadvantage. That is, someyoung people in disadvantaged social envi-ronments tend to develop a “defiant individ-ualist character,” which involves extremecompetitiveness, mistrust, and self-reliance(Sánchez-Jankowski 1991). Such attitudeslimit help-seeking efforts, produce social isola-tion, and are likely to impede both academicand personal development (Stanton-Salazar2001).5

At the same time, disadvantaged youthsbenefit significantly when they develop rela-tionships with mentors. The presence of men-tors is related to substantial improvements inthe educational fortunes of disadvantagedyouths, even in those instances when mentor-ing does not provide benefits as profound as


Figure 2. Interaction Between Having a Teacher as a Mentor and Youth Resources on theHighest Degree Achieved: Probability of Attending College. Note: The category axis for con-tinuous variables represents the 20th, 40th, 60th, and 80th percentiles.

those for more advantaged youths.Furthermore, compelling evidence on the com-pensatory role of teacher mentoring for educa-tional attainment is encouraging. Consistentwith prior research (Stanton-Salazar 2001), wefound that the influence of teacher mentors oneducational attainment is the greatest amongthe most disadvantaged youths. This findinghighlights an interesting paradox. Teachermentors simultaneously serve as complementa-ry and compensatory resources for young peo-ple. On the one hand, they are complementaryin prevalence, with advantaged youths beingmost likely to identify a teacher as a mentor. Onthe other hand, they are compensatory in effec-tiveness, with disadvantaged youths gainingmore from having a teacher as a mentor.Disadvantaged youths are the least likely tohave teacher mentors, but they are the mostlikely to benefit from them.

Of course, the mentoring module availablein the Add Health data set has a number oflimitations.6 First, the fact that the mentoringmodule was measured concurrently with edu-cational attainment leaves open the possibili-ty that the responses may have been influ-enced by educational attainment. Second,the youths were asked to identify only the“most influential” mentor. Some studies havefound that youths identify a variety of individ-uals in their social networks who providementoring functions (e.g., Darling et al.,2002). It is possible that youths who havemultiple mentors, either simultaneously orserially, experience even greater gains in theiracademic careers. Moreover, when the youthshad multiple mentors in their lives, the word-ing of the question may have led them torecall long-lasting mentors more often thanmentoring relationships that were moreephemeral in nature. Therefore, when com-pared to the full population of mentoringrelationships, it is reasonable to assume thatthis measure may be biased toward long-termmentoring relationships, although we cannotbe certain on this point, given the availabledata. However, this potential bias is unlikely toalter the findings in a substantial way. Priorresearch showed that even temporary assis-tance from mentors often has long-term con-sequences for developmental trajectories(Stanton-Salazar and Spina 2003).

Third, we are limited in our ability to assessthe influence of the characteristics of mentors

on the educational outcomes of youths. Forexample, these outcomes are likely to bedependent as much on the quality of mentorsas they are on the resources of youths. Youthswith higher levels of personal resources couldbenefit more from having a relative as a men-tor because their relatives come from advan-taged social backgrounds themselves andconsequently have skills that the relatives ofdisadvantaged youths are less likely to have(Anderson 1999). Similarly, teacher mentorsmay help disadvantaged youths not justbecause teachers are key in the educationalcareers of young people, but because theycome from more advantaged backgroundsand thus have more to contribute to adoles-cents’ success. The Add Health data, however,do not contain measures of the social class orstatus of mentors that would be necessary totest these ideas empirically. Therefore, weregard these issues as problems for furtherinvestigation.

Overall, the findings of this study validatethe efforts of formal mentoring programs tomatch disadvantaged youths with non-parental adults. Not only are these youths atrisk of countless negative life events and dis-appointments during the transition to adult-hood, they are also (not coincidentally) at riskof failing to develop an important relationshipwith a nonparental adult. With this point inmind, additional programs are needed to linkdisadvantaged youths to adults outside thefamily. Such initiatives would supplement thegaps in informal mentoring relationships andhelp young people succeed in educationalsettings. On the basis of the findings present-ed here, advances in the area of teacher men-toring offer the greatest promise for reducingsocial inequities. These relationships providethe greatest compensatory benefits for disad-vantaged youths, while their relative absenceamong the disadvantaged highlights a poten-tial target for social intervention programs.Programs that facilitate the development ofmentoring relationships between at-riskyouths and teachers therefore deserve to be apriority in policy considerations.

The study also contributes to the researchliterature on educational achievement andattainment. Social and behavioral scientistshave long emphasized the constraints, socialcontrol mechanisms, and norm-enforcementactivities that are designed to keep youths on



the “right track” (Dika and Singh 2002). Morerecently, scholars have convincingly arguedfor the importance of supplementing this per-spective with one that views social relation-ships as providing valuable resources foryouths to draw upon when making the tran-sition to adulthood (Kim and Schneider2005). Our research is aligned with the latterstrategy. Success in the educational process isoften accompanied by help from caringadults in the lives of young people. Mentorsplay an important role in this process.Research on formal mentoring relationshipshas identified a variety of mechanisms bywhich mentors influence the lives of youngpeople (Rhodes et al. 2000; Stanton-Salazarand Dornbusch 1995), yet little is knownabout the mechanisms that operate in theseinformal relationships. Future research wouldprofit from in-depth investigations of thosemechanisms.

The study also demonstrates the impor-tance of the social context surrounding infor-mal mentoring relationships. More researchattention is needed on how the timing ofmentoring is related to its character and effec-tiveness, since the timing of life transitionscan have substantial developmental effects(Elder and Shanahan 2006). The establish-ment of mentoring relationships is a potentiallife-altering event when experienced at a piv-otal time in a person’s life. Thus, we need togain a better understanding not only of theshort-term gains from social relationships, butof how these experiences contribute tolonger-term patterns of attainment andresource accumulation across the life course.

NOTES

1. An additional sample weight was createdso that the ratio of respondents with mentors

was the same in this reduced sample as in theoriginal sample. A disproportionate number ofthose who were removed reported having anadult relative as mentor. Analyses (not reportedhere) suggested that respondents who wereexcluded for this reason were older and morelikely to live with both biological parents andhad lower PVT scores. Furthermore, the resultsin Table 4 do not differ substantially from analy-ses (not reported here) in which these respon-dents were included.

2. To examine whether the relationshipbetween age and educational attainmentinfluenced our results, we compared analysesof highest degree achieved in Table 3 for ayounger (age 18–21) and an older (age22–27) cohort (results not shown). We foundno substantive differences in the effects acrossthe two groups.

3. To calculate predicted probabilities for aparticular social role, that response categorymust be set as the reference category in themultinomial logistic equation. For instance,Appendix Table A1 reports odds ratios for theanalysis that was used to calculate the predict-ed probabilities of not having a mentor(which we report as having any kind of men-tor for a more straightforward interpretation).Thus, the results in Appendix Table A1 areactually the results of only one of five equa-tions that were used to calculate the entire setof predicted probabilities.

4. We conceptualize private school,extracurricular activities, and employment asenvironments that may be conducive to theformation of mentoring relationships but notresources per se. We therefore included thesevariables in the logistic regression, but heldthem constant in the calculation of predictedprobabilities.

5. We thank one of the reviewers for direct-ing our attention to this line of research.

6. We thank the anonymous reviewers forpointing out these limitations.


APPENDIX

Table A1. Antecedents of Informal Mentor Relationships: Odds Ratios from theMultinomial Logistic Regression

Relative Friend Teacher Community

Age .820*** .852*** .713*** .909**Female 1.554*** 1.240* 1.184 .969Private school 1.225 .824 1.035 .696Extracurricular activities 1.015 1.033 1.057* .995Work

0 hours — — — —1–20 hours 1.239 1.325* 1.122 1.20721+ hours .789 1.479* .877 1.040

Social BackgroundRace-ethnicity

White — — — —Black 1.041 .932 .944 .707*Asian .846 1.209 1.041 .906Hispanic .934 .901 .699* .851Native American 1.120 1.638* .961 .797

Neighborhood disadvantage .982 1.003 .754 .952Neighborhood disadvantage—Missing 1.442* .887 1.295* 1.042

Parental ResourcesFamily income (in $1,000) 1.001 1.000 .999 1.000Family income—Missing .923 .998 .895 .911Parent in the PTA .892 .827 .989 .888Parent in the PTA—Missing 1.171 .920 .901 .729Two-parent biological family 1.384** .957 1.260* 1.122Highest parent’s education 1.067 1.127* 1.221*** 1.164**Relationship with parents 1.142 1.040 .988 .947

Peer ResourcesNumber of friends 1.045* 1.068** 1.063** 1.062***Friends’ mean Wave 1 GPA .877 1.115 1.096 1.030Peer network centrality 1.089 .932 .861 .822Peer network centrality—Missing 1.399 1.408 .989 1.026

School ResourcesStudent-teacher environment .952 1.983 1.230 .821Student-teacher environment—Missing .633 .508* 1.104 .793Average class size 1.000 1.004 1.003 1.001School size 1.105 .992 1.245* 1.026

Personal ResourcesPhysical attractiveness .961 .870 .911 .985Personality attractiveness 1.142 1.094 1.073 1.080College aspirations 1.095 1.038 1.262*** 1.164**Wave 1 GPA 1.235* .987 1.426*** 1.039Picture Vocabulary Test .998 1.016*** 1.026*** 1.025***Picture Vocabulary Test—Missing 1.447 1.167 1.190 1.132

Note: No mentor is the reference category. N = 7,840. *p < .05, **p < .01, ***p < .001; two-tailed tests.

REFERENCES

Anderson, Elijah. 1990. Race, Class, and Change:Streetwise in an Urban Community. Chicago:University of Chicago Press.

—-. 1999. Code of the Street: Decency, Violence andthe Moral Life of the Inner City. New York: W. W.Norton.

Barajas, Heidi Lasley, and Jennifer L. Pierce. 2001.“The Significance of Race and Gender inSchool Success Among Latinas and Latinos inCollege.” Gender & Society 15:859–78.

Beam, Margaret, Chuansheng Chen, and EllenGreenberger. 2002. “The Nature ofAdolescents’ Relationships with Their “VeryImportant” Nonparental Adults.” AmericanJournal of Community Psychology 30:305–25.

Bonacich, Phillip. 1987. “Power and Centrality: AFamily of Measures.” American Journal ofSociology 92:1170–82.

Caspi, Avshalom. 2004. “Life-Course Development:The Interplay of Social Selection and SocialCausation Within and Across Generations.”Pp. 8–27 in Human Development Across Livesand Generations: The Potential for Change, edit-ed by P. Lindsay Chase-Lansdale, KathleenKiernan, and Ruth J. Friedman. New York:Cambridge University Press.

Chantala, Kim. 2002. “Introduction to AnalyzingAdd Health Data.” Paper presented at the AddHealth Users Workshop, National Institutes ofHealth, Bethesda, MD.

Cheng, Simon, and Brian Starks. 2002. “RacialDifferences in the Effects of Significant Otherson Students’ Educational Expectations.”Sociology of Education 75:306–27.

Coleman, James S. 1988. “Social Capital in theCreation of Human Capital.” American Journalof Sociology 94:S95–S120.

Coleman, James S., and Thomas Hoffer. 1987.Public and Private High Schools: The Impact ofCommunities. New York: Basic Books.

Conley, Dalton. 2001. “Capital for College:Parental Assets and Postsecondary Schooling.”Sociology of Education 74:59–72.

Croninger, Robert G., and Valerie E. Lee. 2001.“Social Capital and Dropping Out of HighSchool: Benefits to At-Risk Students ofTeachers’ Support and Guidance.” TeachersCollege Record 103:548–81.

Crosnoe, Robert, Shannon Cavanagh, and Glen H.Elder, Jr. 2003. “Adolescent Friendships asAcademic Resources: The Intersection ofFriendship, Race, and School Disadvantage.”Sociological Perspectives 46:331–52.

Crosnoe, Robert, Monica Kirkpatrick Johnson, andGlen H. Elder, Jr. 2004. “IntergenerationalBonding in School: The Behavioral andContextual Correlates of Student-Teacher

Relationships.” Sociology of Education77:60–81.

Darling, Nancy, Stephen F. Hamilton, andKatherine Hames Shaver. 2003. “RelationshipsOutside the Family: Unrelated Adults.” Pp.349–70 in Blackwell Handbook of Adolescence,edited by Gerald R. Adams and Michael D.Berzonsky. Malden, MA: Blackwell.

Darling, Nancy, Stephen F. Hamilton, TeruToyokawa, and Sei Matsuda. 2002. “NaturallyOccurring Mentoring in Japan and the UnitedStates: Social Roles and Correlates.” AmericanJournal of Community Psychology 30:245–70.

Dika, Sandra L., and Kusum Singh. 2002.“Applications of Social Capital in EducationalLiterature: A Critical Synthesis.” Review ofEducational Research 72:31–60.

DiPrete, Thomas A., and Gregory M. Eirich. 2006.“Cumulative Advantage as a Mechanism forInequality: A Review of Theoretical andEmpirical Developments.” Annual Review ofSociology 32:271–97.

Downey, Douglas B., and Shana Pribesh. 2004.“When Race Matters: Teachers’ Evaluations ofStudents’ Classroom Behavior.” Sociology ofEducation 77:267–82.

DuBois, David L., Bruce E. Holloway, Jeffrey C.Valentine, and Harris Cooper. 2002.“Effectiveness of Mentoring Programs forYouth: A Meta-Analytic Review.” AmericanJournal of Community Psychology 30:157–97.

DuBois, David L., and Naida Silverthorn. 2005.“Natural Mentoring Relationships andAdolescent Health: Evidence from a NationalStudy.” American Journal of Public Health95:518–24.

Dunn, Lloyd M. 1981. Peabody Picture VocabularyTest, Revised. Circle Pines, MN: AmericanGuidance Service.

Elder, Glen H., Jr., and Michael J. Shanahan. 2006.“The Life Course and Human Development.”Pp. 665–715 in The Handbook of ChildPsychology, Vol. 1, edited by William Damonand Richard M. Lerner. New York: Wiley.

Garner, Catherine L., and Stephen W. Raudenbush.1991. “Neighborhood Effects on EducationalAttainment—A Multilevel Analysis.” Sociologyof Education 64:251–62.

Goleman, Daniel. 2006. Social Intelligence: The NewScience of Human Relationships. New York:Bantam Books.

Hamilton, Mary Agnes, and Stephen F. Hamilton.2005. “Work and Service-Learning.” Pp.348–63 in Handbook of Youth Mentoring, edit-ed by David L. DuBois and Michael J. Karcher.Thousand Oaks, CA: Sage.

Hamilton, Stephen F., and Mary Agnes Hamilton.2004. “Contexts for Mentoring: Adolescent-Adult Relationships in Workplaces and


Communities.” Pp. 395–428 in Handbook ofAdolescent Psychology, edited by Richard M.Lerner and Laurence Steinberg. New York: Wiley.

Hardy, Melissa A., and John Reynolds. 2004.“Incorporating Qualitative Information intoRegression Models: The Utility of DummyVariables.” Pp. 209–36 in Handbook of DataAnalysis, edited by Melissa A. Hardy and AlanE. Bryman. Thousand Oaks, CA: Sage.

Harris, Kathleen Mullan, Francesca Florey, JoyceTabor, Peter S. Bearman, Jo Jones, and J.Richard Udry. 2003. “The NationalLongitudinal Study of Adolescent Health:Research Design.” Available online: http://www.cpc.unc.edu/projects/addhealth/design

Higgins, Monica C., and David A. Thomas. 2001.“Constellations and Careers: TowardUnderstanding the Effects of MultipleDevelopmental Relationships.” Journal ofOrganizational Behavior 22:223–47.

Hofferth, Sandra L., Johanne Boisjoly, and Greg J.Duncan. 1998. “Parents’ ExtrafamilialResources and Children’s School Attainment.”Sociology of Education 71:246–68.

Ianni, Francis A. J. 1989. The Search for Structure : AReport on American Youth Today. New York:Free Press.

Jacobi, Maryann. 1991. “Mentoring andUndergraduate Academic Success: ALiterature-Review.” Review of EducationalResearch 61:505–32.

Kim, Doo Hwan, and Barbara Schneider. 2005.“Social Capital in Action: Alignment ofParental Support in Adolescents’ Transition toPostsecondary Education.” Social Forces84:1181–1206.

Klaw, Elena L., and Jean E. Rhodes. 1995. “MentorRelationships and the Career Development ofPregnant and Parenting African-AmericanTeenagers.” Psychology of Women Quarterly19:551–62.

Kuziemko, Ilyana. 2006. “Using Shocks to SchoolEnrollment to Estimate the Effect of SchoolSize on Student Achievement.” Economics ofEducation Review 25:63–75.

Lee, Valerie E., and David T. Burkam. 2003.“Dropping Out of High School: The Role ofSchool Organization and Structure.” AmericanEducational Research Journal 40:353–93.

Lin, Nan. 2001. Social Capital: A Theory of SocialStructure and Action. Cambridge, England:Cambridge University Press.

Long, J. Scott. 1997. Regression Models forCategorical and Limited Dependent Variables.Thousand Oaks, CA: Sage.

Luthar, Suniya S., Dante Cicchetti, and BronwynBecker. 2000. “The Construct of Resilience: ACritical Evaluation and Guidelines for FutureWork.” Child Development 71:543–62.

Maeroff, Gene I. 1998. Altered Destinies: Making LifeBetter for Schoolchildren in Need. New York: St.Martin’s Press.

Masten, Ann S., and J. Douglas Coatsworth. 1998.“The Development of Competence inFavorable and Unfavorable Environments:Lessons Learned from Research on SuccessfulChildren.” American Psychologist 53:205–20.

McDonald, Steve, Lance D. Erickson, MonicaKirkpatrick Johnson, and Glen H. Elder, Jr.2007. “Informal Mentoring and Young AdultEmployment.” Social Science Research36:1328–47.

Mortimer, Jeylan T. 2003. Working and Growing Upin America. Cambridge, MA: HarvardUniversity Press.

Mosteller, Frederick. 1995. “The Tennessee Studyof Class Size in the Early School Grades.”Future of Children 5:113–27.

Newman, Katherine. 2000. No Shame in My Game:The Working Poor in the Inner City. New York:Vintage Books.

Nora, Amaury, Alberto Cabrera, Linda SerraHagedorn, and Ernest Pascarella. 1996.“Differential Impacts of Academic and SocialExperiences on College-Related BehavioralOutcomes Across Different Ethnic and GenderGroups at Four-Year Institutions.” Research inHigher Education 37:427–51.

O’Rand, Angela M. 2006. “Stratification and the LifeCourse: Life Course Capital, Risks and SocialInequality.” in Handbook of Aging and the SocialSciences (6th ed.), edited by Robert H. Binstock,Linda K. George, Stephen J. Cutler, JonHendricks, and James H. Schulz. Boston:Academic Press.

Rhodes, Jean E. 1994. “Older and Wiser: MentoringRelationships in Childhood and Adolescence.”Journal of Primary Prevention 14:187–96.

—-. 2002. Stand by Me: The Risks and Rewards ofMentoring Today’s Youth. Cambridge, MA:Harvard University Press.

—-. 2005. “A Model of Youth Mentoring.” Pp.30–43 in Handbook of Youth Mentoring, editedby David L. DuBois and Michael J. Karcher.Thousand Oaks, CA: Sage.

Rhodes, Jean E., Josefina M. Contreras, and Sarah C.Mangelsdorf. 1994. “Natural MentorRelationships Among Latina AdolescentMothers: Psychological Adjustment, ModeratingProcesses, and the Role of Early ParentalAcceptance.” American Journal of CommunityPsychology 22:211–27.

Rhodes, Jean E., Lori Ebert, and Karla Fischer. 1992.“Natural Mentors: An Overlooked Resources inthe Social Networks of Young African AmericanMothers.” American Journal of CommunityPsychology 20:445–61.

Rhodes, Jean E., Jean B. Grossman, and Nancy L.



Resch. 2000. “Agents of Change: PathwaysThrough Which Mentoring RelationshipsInfluence Adolescents’ Academic Adjustment.”Child Development 71:1662–71.

Riegle-Crumb, Catherine, George Farkas, andChandra Muller. 2006. “The Role of Genderand Friendship in Advanced Course Taking.”Sociology of Education 79:206–28.

Robertson, Donald, and James Symons. 2003. “DoPeer Groups Matter? Peer Group versusSchooling Effects on Academic Attainment.”Economica 70:31–53.

Sameroff, Arnold, Leslie Morrison Gutman, andStephen C. Peck. 2003. “Adaptation AmongYouth Facing Multiple Risks: ProspectiveResearch Findings.” Pp. 364–91 in Resilienceand Vulnerability: Adaptation in the Context ofChildhood Adversities, edited by Suniya S.Luthar. New York: Cambridge University Press.

Sampson, Robert J., Jeffrey D. Morenoff, and FeltonEarls. 1999. “Beyond Social Capital: SpatialDynamics of Collective Efficacy for Children.”American Sociological Review 64:633–60.

Sánchez-Jankowski, Martín. 1991. Islands in theStreet: Gangs and American Urban Society.Berkeley: University of California Press.

Sanchez, Bernadette, and Olga Reyes. 1999.“Descriptive Profile of the MentorshipRelationships of Latino Adolescents.” Journal ofCommunity Psychology 27:299–302.

Sewell, William H., Archibald O. Haller, and GeorgeW. Ohlendorf. 1970. “The Educational andEarly Occupational Status Attainment Process:Replication and Revision.” AmericanSociological Review 35:1014–27.

Sewell, William H., Archibald W. Haller, andAlejandro Portes. 1969. “The Educational andEarly Occupational Attainment Process.”American Sociological Review 34:82–92.

Stanton-Salazar, Ricardo D. 2001. ManufacturingHope and Despair: The School and Kin SupportNetworks of U.S.-Mexican Youth. New York:Teachers College Press.

Stanton-Salazar, Ricardo D., and Sanford M.

Dornbusch. 1995. “Social Capital and theReproduction of Inequality: InformationNetworks among Mexican-Origin High SchoolStudents.” Sociology of Education 68:116–35.

Stanton-Salazar, Ricardo D., and Stephanie UrsoSpina. 2003. “Informal Mentors and RoleModels in the Lives of Urban Mexican-OriginAdolescents.” Anthropology & EducationQuarterly 34:231–54.

Werner, Emmy E., and Ruth S. Smith. 1982.Vulnerable but Invincible: A Longitudinal Study ofResilient Children and Youth. New York:McGraw-Hill.

—-. 2001. Journeys from Childhood to Midlife: Risk,Resilience, and Recovery. Ithaca, NY: CornellUniversity Press.

Wiggan, Greg. 2007. “Race, School Achievement,and Educational Inequality: Toward a Student-Based Inquiry Perspective.” Review ofEducational Research 77:310–33.

Williams, Terry M., and William Kornblum. 1985.Growing Up Poor. Lexington, MA: LexingtonBooks.

Wilson, William Julius. 1987. The TrulyDisadvantaged: The Inner City, the Underclass,and Public Policy. Chicago: University ofChicago Press.

Woodward, Lianne J., and David M. Fergusson.2000. “Childhood Peer Relationship Problemsand Later Risks of Educational Under-achieve-ment and Unemployment.” Journal of ChildPsychology and Psychiatry 41:191–201.

Zimmerman, Marc A., Jeffrey B. Bingenheimer, andDiana E. Behrendt. 2005. “Natural MentoringRelationships.” Pp. 143–57 in Handbook ofYouth Mentoring, edited by David L. DuBoisand Michael J. Karcher. Thousand Oaks, CA:Sage.

Zimmerman, Marc A., Jeffrey B. Bingenheimer, andPaul C. Notaro. 2002. “Natural Mentors andAdolescent Resiliency: A Study with UrbanYouth.” American Journal of CommunityPsychology 30:221–43.

Lance D. Erickson, Ph.D., is Assistant Professor, Department of Sociology, Brigham YoungUniversity. His main fields of interest are the life course, adolescence, and family. His current proj-ects include examinations of mentoring and trajectories of delinquency, whether marriage facilitatesor inhibits success among graduate students, and the causal relationship between divorce and well-being.

Steve McDonald, Ph.D., is Assistant Professor, Department of Sociology, North Carolina StateUniversity. His main fields of interest are social capital, social networks, and inequality across the lifecourse. Dr. McDonald is conducting studies on the influence of informal mentoring relationships onthe transition to adulthood and beyond and on informal job-matching processes and their influenceon career attainment.


Glen H. Elder, Jr., Ph.D., is University Research Professor, Department of Sociology and the CarolinaPopulation Center, University of North Carolina at Chapel Hill. His main field of interest is life-coursestudies. Dr. Elder is currently conducting longitudinal studies of the transitions from childhood to theadult years, research on significant others beyond the family, and studies of pathways from disad-vantage to greater life opportunity.

The first two authors made equivalent contributions to this research. This research used data fromAdd Health, a program project designed by J. Richard Udry, Peter S. Bearman, and Kathleen MullanHarris and funded by Grant P01-HD31921 from the Eunice Kennedy Shriver National Institute ofChild Health and Human Development, with cooperative funding from 17 other agencies. Specialacknowledgment is due to Ronald R. Rindfuss and Barbara Entwisle for assistance in the originaldesign. Persons who are interested in obtaining data files from Add Health should contact AddHealth, Carolina Population Center, 123 West Franklin Street, Chapel Hill, NC 27516-2524([email protected]). No direct support was received from grant P01-HD31921 for this analysis.We gratefully acknowledge support from NICHD to Glen H. Elder, Jr., and Michael J. Shanahanthrough their subproject to the Add Health Wave IV Program Project (Grant 3P01 HD031921).Acknowledgement is also given to NIH/NIA for fellowship support from the Demography of Agingtraining grant (5 T32 AGOO155-14). The Add Health Academic Achievement study was funded bya grant from the National Institute of Child Health and Human Development under Grant 01HD40428-02 to the Population Research Center, University of Texas at Austin; Chandra Muller(principal investigator) and from the National Science Foundation under Grant REC-0126167 to thePopulation Research Center, University of Texas at Austin; Chandra Muller and Pedro Reyes (co-prin-cipal investigators). Address correspondence to Lance D. Erickson, Department of Sociology,Brigham Young University, 2008 JFSB, Provo UT, 94602; e-mail: [email protected].

SOECover 11/2/09 4:07 PM Page 1 SOCIOLOGY EDUCATION · Jeff Keesler Nathan D. Jones and Timothy G....

Documents

Transcript of SOECover 11/2/09 4:07 PM Page 1 SOCIOLOGY EDUCATION · Jeff Keesler Nathan D. Jones and Timothy G....