S P E C I A L T H E M E R E S E A R C H R E P O R T

Educational Benefits of Diversity in Medical School:

A Survey of Students
Dean K. Whitla, PhD, Gary Orfield, PhD, William Silen, MD, Carole Teperow, Carolyn Howard, MEd, and
Joan Reede, MD, MPH

ABSTRACT

Purpose. Many U.S. medical schools have abandoned Results. Of the 55% of students who could be located,
affirmative action, limiting the recruitment and reducing 97% responded to the survey. Students reported having
the admission of underrepresented minority (URM) stu- little intercultural contact during their formative years but
dents even though research supports the premise that the significantly more interactions during higher education
public benefits from an increase in URM physicians and years, especially in medical school. Students reported
that URM physicians are likely to serve minority, poor, contacts with diverse peers greatly enhanced their edu-
and Medicaid populations. Faculty and students com- cational experience. They strongly supported strength-
monly assume they benefit from peer cultural exchange, ening or maintaining current affirmative action policies
and the published evidence for the past two decades in admissions. The responses and demography of the
supports this notion. This research examined the students’ Harvard and UCSF students did not differ significantly,
perceptions of the educational merits of a diverse student nor did they differ for majority students and URM
body by surveying medical students at two schools. students—all groups overwhelmingly thought that racial
Method. In 2000, medical students from all four years at and ethnic diversity among their peers enhanced their
Harvard Medical School and the University of California, education.
San Francisco, School of Medicine were enrolled in Conclusions. Diversity in the student body enhanced
a telephone survey about the relevance of racial diversity the educational experiences of students in two U.S.
(among students) in their medical education. Students medical schools.
responded to the interviewer’s questions on a five-point Acad. Med. 2003;78:460–466.
Likert-type scale.

The Bakke case has influenced admis- and graduate schools for the past three ton) and in the 5th District Court area
sions of minority students to college decades.1 In its 1978 ruling, the (Louisiana, Mississippi, and Texas)
Supreme Court rested its decision on both ballot initiatives and lower court
Dr. Whitla is director, Counseling and Psychology the importance of a diverse student decisions have placed restrictions on
Program, and lecturer on education, Harvard Grad- body for the educational experience of using race as a factor in higher
uate School of Education; Dr. Orfield is professor of
education and social policy, Harvard Law School, all students. The Court stated that education admission decisions. Critics
Harvard Graduate School of Education, and Harvard race could legally be considered only of affirmative action argue not only
University Kennedy School of Government; Dr. Silen as one of a number of factors in that affirmative action is unfair to
is Johnson and Johnson Distinguished Professor of
Surgery, Emeritus, Harvard Medical School; Ms. selecting a class but forbade the use of whites but also that such polices have
Teperow is administrative director, Office of Faculty quotas. However, in some states (Cali- not produced the educational gains for
Development, Children’s Hospital Boston; Ms. fornia, Florida, Georgia, and Washing- students that were anticipated.2,3
Howard is research assistant, Harvard Graduate
School of Education; and Dr. Reede is dean of This study represents an effort to
diversity and community partnership, Harvard Medi- Correspondence and requests for reprints should be add a new level of understanding
cal School, Boston, Massachusetts. At the time of this addressed to Dr. Whitla, Harvard Graduate School to the educational effects of diver-
research, Dr. Silen and Ms. Teperow were with the of Education, 409 Larsen Hall, 15 Appian
Harvard Medical School Office of Faculty Develop- Way, Cambridge, MA 02138; e-mail: hwhitla@ sity, especially in medical education.
ment and Diversity, where Dr. Silen was the dean. fas.harvard.edui. The December 2002 decision of the

460 ACADEMIC MEDICINE, VOL. 78, NO. 5 / MAY 2003

Supreme Court to review affirmative surveys of law students at eight U.S. In May and June of 2000, Gallup
action admission policies enhances the law schools14 and the Bowen and Bok interviewers phoned students enrolled
importance of these research efforts. research effort,15 that telephone inter- in all four years of the Harvard and
A diverse student body enables viewing was the most effective method UCSF medical schools. Interviewers
students to exchange information and of collecting responses. The deans of made up to five calls per student, and
share value systems of different cul- the two participating medical schools, if no contact occurred, that instance
tures as a basic foundation for cultural Dean Debas of the University of was deleted from the total number.
sensitivity.4 A major benefit of affirma- California, San Francisco, School of The response rate, taking into account
tive action in medical school admis- Medicine and Dean Martin of Harvard these deletions, was 97%. However,
sions is the ability to expand health Medical School, approved of the pro- due to the infrequency of actual stu-
care delivery to traditionally under- ject. Their representatives provided dent contact, only 55% of the total
served communities, generating social telephone numbers of each school’s enrolled student body at both schools
benefits that go beyond the individual enrolled undergraduates. The Harvard could be sampled. Interviewers also
physician.5 Research indicates that Committee on the Use of Human recorded students’ explanatory remarks
underrepresented minority (URM) Subjects approved the project. We in response to the questions.
physicians are more likely to serve employed The Gallup Organization to Our data represent the views of 639
minority, poor, and Medicaid popula- the complete phone interviews. Al- students, 338 from Harvard and 301
tions than are their majority counter- though a phone call even from a pro- from UCSF. The responders consisted
parts.6–8 Moreover, minorities in North fessional polling organization does not of roughly equal numbers of students in
America tend to choose physicians guarantee anonymity, research con- each of the four years of medical school
of their own races, due not only to ducted using this method has normally study. The response patterns and the
geographic location but also to the been sanctioned as meeting this qual- demographics of the Harvard and UCSF
nature of the care they receive—care ification. As such, ‘‘implicit informed medical students were not found to
based on mutual understanding and consent’’ meets the review standards of be significantly different. Therefore, the
trust.9,10 African American patients the two medical schools. responses from the two samples were
who see physicians of their own race A committee with expertise in combined in the analyses. The racial
tend to rate their physicians’ decision- questionnaires and medical educa- and ethnic characteristics of the UCSF
making styles as more participatory.11 tion constructed the survey instru- and Harvard samples were also typical of
Because satisfaction with health care ment, drawing on previous work in the total enrolled student populations at
is positively associated with patients’ this area. Previous questionnaires by the two schools (chi-square test p ¼ .87).
treatment compliance, researchers be- the National Science Foundation, the Furthermore, the composition of the
lieve that increasing the pool of URM American Medical Association, the combined sample did not differ from the
physicians, and improving cultural Canadian Federation of Medical Stu- U.S. population of enrolled medical
competence among all physicians, dents, and the Institute of Ethics were school students (chi-square test p ¼
may lead to better health outcomes examined. The instrument, a series of .71). There were 2% more African
for minority populations.12,13 To para- five-point Likert-type questions asking Americans in the study sample than
phrase Dr. Jordan Cohen, current students to rate the importance of were enrolled nationally (9% versus
president of the Association of Amer- diversity in the student body in a 7% nationally); 6% more Asians (26%
ican Medical Colleges, anti-affirmative number of areas, was pilot tested with versus 20% nationally); 3% more Lati-
action would be bad for our collective a small group of graduate students in nos (9% versus 6% nationally); 0.3%
national health.13 the medical sciences. ‘‘Diversity’’ was fewer Native Americans (0.5% versus
defined for students as being limited to 0.8% nationally); and 10% fewer others
racial and ethnic diversity. The con- (56% versus 66% nationally). Approx-
M ETHOD struct validity of the instrument was imately 93% of those surveyed (597
deemed appropriate and adequate from students) were U.S. citizens, and just
Data collection from medical students, the pretest results and by the oversight over 6% (42) were foreign nationals.
because of their complicated and over- of a team of psychometricians and Because the sample was representative
loaded schedules, is very difficult. Of medical educators. The internal con- of the enrolled students at UCSF and
the various methods of data collec- sistency of the series of items focusing Harvard and the U.S. medical school
tion—e-mails, personal interviews, on attitudes toward diversity was found population, there may be some infer-
questionnaires, telephone interviews— to be substantial (Cronbach’s alpha ¼ ences that can be drawn from the find-
we decided from prior experience with .87). ings that have national implications.

ACADEMIC MEDICINE, VOL. 78, NO. 5 / MAY 2003 461

 R ESULTS Table 1

Interactions with Those of Reports of Frequencies of Contacts with Different Races or Ethnicities at Various Stages of
Different Race or Ethnicity Development by 639 Students at Two Medical Schools, 2000*

The first set of questions surveyed Percentage

the frequency of students’ contact with Often None
people of different races and ethnicities Stage 5 4 3 2 1
during their formative years, in sec- Growing up 27 21 22 21 7
ondary school, in college, and finally in Secondary school 28 21 25 20 6
medical school. As Table 1 indicates, College 41 25 23 9 1
the percentage of students who had Medical school 54 31 13 2 1
contact with those of other groups *Percentages may exceed 100% due to rounding.
increased from 50% (combining cate-
gories 4 and 5) in their early years and
in secondary school, to 67% while in
college and 85% in medical school dents, interviewers asked students how of medical conditions and treatments
(chi-square test p ¼ .01). These trends much a diverse student population concentrated among racial and ethnic
were true for minority groups as well as helped, or did not help, them to work groups (see Table 2).
for the majority whites. For example, more effectively with those from dif- Eighty-six percent of the students
64% of African American students had ferent racial or ethnic backgrounds. felt that diversity in the classroom was
contact with other groups while grow- Seventy-six percent of students felt more likely to foster serious discussions
ing up, and that number increased to that a diverse student body helped of alternative viewpoints (chi-square
91% while these students were in them work more effectively with those test p ¼ .04), and 77% indicated that
medical school. of different backgrounds, compared a greater understanding of medical
To gauge the amount of collegial with 4% who said such diversity was conditions and treatments was more
interaction among students beyond the of little or no help (F test p ¼ .01). At likely when the student body was
classroom, interviewers asked students the time of the study, neither medical diverse (chi-square test p ¼ .05).
about their frequency of study time school offered any official cultural Sixty-two percent of the students also
with those of other races and ethnic- sensitivity training to its students, found that professors and peers offered
ities. Fifty percent of the responding which made each student’s personal a broader range of examples in a di-
students indicated that they often diversity experiences even more rele- versified classroom (chi-square test p ¼
studied with persons of a different race vant. .05). As shown in Table 2, those who
or ethnicity, while 36% and 14% stated did not feel that diversity afforded
that this occurred sometimes or never, Classroom Dynamics a great advantage were largely neutral.
respectively. Of the 88 students who Students did not feel that diversity was
reported never studying with those Students were asked about the impact especially important in giving rise to
from different racial and ethnic groups, of diversity of students on the way higher levels of intellectual challenge
80 indicated that they always chose to topics were discussed in class. Only 16 or conflict. It is possible that the terms
study alone. Removing these 80 stu- students (3%) felt that diversity de- ‘‘challenge’’ and ‘‘conflict’’ in this ques-
dents from the calculations lowers the tracted from the discussions, while the tion might have been somewhat emo-
percentage of those who never studied majority (84 %) thought that diversity tionally loaded for students and, hence,
with another student of a different race enhanced discussion. The diversity of skewed their responses.
or ethnicity to 1%. Fifty-eight percent students appears to improve classroom Most students (77%) found that
of the students indicated they often discussion—a fundamental educational they felt challenged to rethink their
studied with others whose backgrounds benefit. The next series of questions values when racial conflicts occurred,
differed from their own. Clearly, many probed whether diversity in the class- 68% thought such occurrences were
students had experience in working rooms was more or less likely to change learning experiences, and 23% thought
across racial and ethnic lines (F test the nature of the discussions through that the conflicts they encountered
p ¼ .04). examples used, viewpoints seriously reinforced stereotypical positions. The
To further assess the impact of discussed, the level of intellectual students’ responses suggest that both
multicultural interaction among stu- challenge, and greater understanding conflict and difference of opinion add

462 ACADEMIC MEDICINE, VOL. 78, NO. 5 / MAY 2003

Table 2 and only 0.3% felt campus diversity
was a negative element of their educa-
Resultant Dynamics of a Diverse Classroom Reported by 639 Students at Two Medical tional experience. Although there is
Schools, 2000* some variation in the response pattern
between the racial and ethnic groups,
Percentage these differences are not statistically
Definitely More No Difference Less Definitely significant (F test p ¼ .87).
Dynamic More Likely Likely Based on Diversity Likely Less Likely In fact, an interesting finding of this
More examples used 31 30 36 2 1 research was that the distribution of
Alternative viewpoints responses by racial groups did not differ
seriously discussed 51 36 11 2 1 significantly. The pattern of African
High levels of intellectual American students’ responses, for ex-
conflict or challenge 20 25 49 3 2 ample, showed that they were typically
Greater understanding of
a bit more supportive of the educa-
medical conditions and
treatments for disease/ tional benefits of diversity than were
problems 42 35 19 2 1 the white students—but not to a degree
*Percentages may exceed 100% due to rounding. that reached significance. Chi-square
results comparing the response patterns
of African American, Asian, White,
Latino, Native American, Other, and
to a richness and breadth of under- positive) of the students indicated that Mixed students were all not significant.
standing of people and their differ- a diverse student body was a positive Of the seven groups, very few students
ences. element in their educational experi- chose ‘‘Moderately detracts’’ or
As shown in Figure 1, 94% (78% ence (chi-square test p ¼ .001). Six ‘‘Clearly detracts’’ as a response when
clearly positive and 16% moderately percent felt diversity had no impact asked about the quality of their expe-

Figure 1. Percentages of 639 students from two medical schools (by ethnicity or race) answering ‘‘clearly positive’’ to the question: ‘‘Do you consider having
students of different races and ethnicities to be a positive or negative element of your educational experience?’’

ACADEMIC MEDICINE, VOL. 78, NO. 5 / MAY 2003 463

riences with diversity. Essentially, the groups by 58% of students (chi-square medical school admissions, and whe-
differences, when they existed among test p ¼ .01) but also was considered ther schools should admit more URM
groups, were among the degrees of important. students. Over 90% of respondents
satisfaction: for example, between cat- Researchers sampled the students’ indicated that admission policies seek-
egories such as—‘‘Enormously Gratify- opinions on the extent to which dis- ing URM students should be strength-
ing’’ and ‘‘Gratifying,’’ or between cussions with students from different ened (43%) or maintained (47%)—
‘‘Clearly Positive’’ and ‘‘Moderately racial and ethnic backgrounds had evidence of very strong student sup-
Positive.’’ affected their thinking about the equity port for continuing affirmative action in
of the health care delivery system, admissions. Less than 3% of students
Impact of Diversity on access to medical care for the un- said such policies should be discon-
Policy Matters derserved, cultural competency issues tinued. Again, student support of affir-
in treating a diverse population, and mative action policies was statistically
Researchers explored the impact and priority areas for future research. Stu- very significant (chi-square test p ¼
importance of diversity for students in dents stated that their concerns in .001).
three distinct segments of the medical these areas were significantly enhanced Consistent with the published re-
school community: the student body, by the diversity of students (F test p ¼ ports of many investigators,4,13 84%
the basic science faculty, and the .001). Indeed, students reported that of the polled students felt that the
clinical faculty. Students indicated that having a diverse student body in- medical profession should represent
their most significant educational ex- creased students’ concern for treating the country’s racial and ethnic com-
periences influenced by peers of differ- a diverse population (84%, chi-square position to a large degree (59%) or
ent races or ethnicities came through test p ¼ .001), the equity of the a significant degree (24%). Only 3% of
two major avenues: regular exchanges delivery system (78%, chi-square test the students indicated that the racial
with students from different back- p ¼ .001), and the access to care for and ethnic composition of the medical
grounds and in the clinic while learning the underserved (76%, chi-square test profession should reflect such compo-
how to actually treat patients (see p ¼ .001) (see Table 4). The lesser sition of society either slightly (2%) or
Table 3). Thus, among the three concern for priority in research areas not at all (1%). These findings are
segments of the community (i.e., peers, (46%) was also significant (chi-square statistically significant (F test p ¼ .05).
basic science faculty, and clinical test p ¼ .01) but received less support The researchers probed the idea of
faculty), 88% of students felt that the from students; this is surprising consid- cultural competency and extending
diversity of their peers was the most ering that both UCSF and Harvard oneself in the physician–patient re-
important element in their education train large numbers of research profes- lationship by having the interviewers
(chi-square test p ¼ .001). The impor- sionals. It may be that many students ask the students about their confidence
tance of a diverse clinical faculty was have not engaged in sufficient research in their ability to establish positive
considered a close second by 83% of to formulate ideas and gauge future relationships with patients of different
the students (chi-square test p ¼ .001). priorities. racial and ethnic backgrounds. Eighty-
The need for a diverse basic science We asked this very competitive six percent of students felt confident
faculty was ranked third among the group of students how they felt about they could establish a positive rapport
with patients from different racial or
ethnic groups, while only 1% felt that
Table 3 they were not confident in their ability
to do so (chi-square test p ¼ .001). We
Reports of Importance of Diversity in Groups of the Medical Community in Improving clearly need physicians who will treat
Medical Education by 639 Medical Students from Two Medical Schools, 2000* patients from different backgrounds
and who will undertake research agen-
Percentage das that represent a broader spectrum
Very Significantly Moderately Slightly Not at All of diseases, as well as effective treat-
Group Important 5 Important 4 Important 3 Important 2 Important 1 ment delivery, across communities.16,17
Student body 62 26 9 2 2 Finally, interviewers asked students
Basic science faculty 30 28 29 6 6 to rate their experiences in medical
Clinical faculty 55 28 11 2 2 school overall and, specifically, during
*Percentages may exceed 100% due to rounding. the preclinical and clinical years. Se-
venty percent of the students found

464 ACADEMIC MEDICINE, VOL. 78, NO. 5 / MAY 2003

Table 4 the cultures, beliefs, and values of their
communities.4,16 Indeed, the core cur-
Reports of Extents to Which Discussions with Students of Different Racial and Ethnic riculum guidelines of the Society of
Backgrounds Affected Concern about Policy-related Issues by 639 Students Teachers of Family Medicine, approved
from Two Medical Schools, 2000 by the Academy of Family Physicians,
recognize the need to teach respect
Percentage and tolerance for cultural and social
Great Significant Moderate Slight class differences in a pluralistic society
Increase in Increase in Increase in Increase in by setting forth a three-tier approach:
Concern Concern Concern Concern No Effect attitude, knowledge, and skills devel-
Policy-related Issue 5 4 3 2 1
opment.19 Diversity among students
Equity of the health care clearly improves the breadth of class
delivery system 47 32 14 4 3 discussion, a fundamental educational
Access to medical care for
benefit and a basis for learning cultu-
underserved populations 46 30 17 4 3
Cultural competency when rally competent health care.
treating a diverse That students gave high ratings for
population 52 33 9 4 2 a diverse student body supports the
Priority areas for future hypothesis that students regularly edu-
research 19 28 35 9 9 cate each other on important issues,
such as differences among the cultures
and how to best respond to those
the basic science years gratifying. This greatly enhanced their educational differences. The teaching dynamic in
number increased to 79% during the experiences in medical school. These a biochemistry or anatomy class may
clinical years. These responses, while students strongly supported maintain- be less affected by the racial and
somewhat lower than might be ex- ing or strengthening current affirmative ethnic diversity of students. However,
pected, compare very favorably with action policies in admissions at their students’ understanding of patients
those of students in the professions of respective schools. and colleagues is likely to be affected
law and business. The frequency with which the ma- when, for example, an Asian student
jority of students study with those from learns from a Native American student
different racial groups suggests that about tribal views of healing. Further-
D ISCUSSION students attempt to take advantage more, treatment compliance may be
of the diverse student body medical positively affected if, for example, a
There are two important findings in school provides. The consistently low Caucasian student from an affluent,
this study. First, students typically had numbers of minority faculty in medical predominantly Caucasian suburb learns
less contact during their formative school compounded with the dearth of from an African American inner-city
years with those of different races and cultural sensitivity training suggests colleague how to better engage African
ethnicities than they did during their that students’ interactions—both in- American inner-city patients in follow-
college years. Student cross-cultural side and outside the classroom—pro- ing a course of treatment through the
and cross-racial interactions increased vide one of the few arenas in which public health clinic.
even more during medical school. students can gain cultural awareness For medical schools to accomplish
When interviewed, over 60% of the before they mingle with a multicultural the goal of increasing the diversity of
students stated that they had three or patient population. In a recent poll the physician population to mirror that
more close friends who differed racially of 98 medical schools, many school of the general population, the aca-
and ethnically from themselves. Such deans felt that their recent graduates demic community will need to recon-
collegial relationships and friendships were only ‘‘somewhat prepared’’ to pro- sider the current stand on affirmative
are critical given the multicultural vide culturally sensitive clinical care.18 action in admissions.20 In a recent
society in which they will later practice Although cultural competence is in- survey of 15 medical schools, research-
medicine. cluded in some medical curricula, it is ers found that the weights given to
The second and perhaps even more too often a rather sterile course taught qualitative factors such as URM status
important finding is that both Harvard from a syllabus. Medical students and in the admission process vary widely
and UCSF students reported that the faculty from diverse racial and ethnic from school to school.20 However, the
interaction with a diverse student body backgrounds teach each other about transition from a predominantly male

ACADEMIC MEDICINE, VOL. 78, NO. 5 / MAY 2003 465

profession to one today in which it was absent in the responses in this 6. Cantor JC, Miles EL, Baker LC, Barker DC.
women make up a majority of medical medical school survey. Physician service to the underserved: impli-
cations for affirmative action in medical
students has been accomplished with- In summary, students enrolled in
education. Inquiry. 1996;33:167–80.
out compromising medical education in Harvard and University of California, 7. Thurmond VB, Kirch DG. Impact of minor-
any way. Thus, it should be possible to San Francisco, medical schools over- ity physicians on health care. South Med J.
make a similar shift in the racial and whelmingly supported affirmative ac- 1998;91:1009–13.
ethnic composition of students as well. tion in admissions. They strongly 8. Rabinowitz HK, Diamond JJ, Veloski JJ,
Students in the present survey ex- believed that diversity enhanced their Gayle JA. The impact of multiple predictors
on generalist physicians’ care of underserved
pressed in parenthetic remarks that educational experiences and provided
populations. Am J Public Health. 2000;
there should be more socio-economic them with culturally rich opportunities. 90:1225–8.
as well as racial and ethnic diversity in They had established close collegial 9. Saha S, Taggart SH, Komaromy M, Bindman
the student body. Looking at national and personal friendships with students AB. Do patients choose physicians of their
demographics, one can see the oppor- of different races and ethnicities. These own race? Health Aff. 2000;19 (4):76–83.
tunities to broaden the student base students stated that such ties contrib- 10. Saha S, Komaromy M, Koepsall TD, Bind-
man AB. Patient–physician racial concor-
and, certainly, the need for physicians uted greatly to their understanding of
dance and the perceived quality and use of
to become culturally competent.11,16 medical practice and, ultimately, would health care. Arch Intern Med. 1999;
Despite the support for a diverse better train them for service in a mul- 159:997–1004.
student body and affirmative action in ticultural society. 11. Cooper-Patrick L, Gallo JJ, Gonzales JJ. et al.
admissions, we should mention that Race, gender, and partnership in the patient–
a number of students responded to The authors thank Joseph B. Martin, Dean of the physician relationship. JAMA. 1999;282:
the open-ended question about affirma- Harvard University Faculty of Medicine, Harvard 583–9.
Medical School, and Haile T. Debas, Dean of the 12. Flores G, Gee D, Kastner B. The teaching of
tive action with statements about the School of Medicine, University of California, San cultural issues in U.S. and Canadian Medical
importance of merit in the selection Francisco for giving them permission to interview Schools. Acad Med. 2000;75:451–5.
process (8% of total responses), and a students enrolled at Harvard and UCSF medical 13. Cohen JJ. Statistics don’t lie: anti-affirmative
few were concerned about standards. schools. They are grateful to Warren C. Reed and action is bad for our health. Acad Med.
Terrence Tivnan for aid in the experimental 1997;72:1084.
However, 57% of the students re-
design and data analysis, and to Jody E. Clarke 14. Orfield G, Whitla D. Diversity and legal edu-
sponding to the open-ended section for assisting in data analysis. They also thank The cation: student experiences in leading law
gave responses that were overwhelm- Gallup Organization for conducting telephone schools. In: Orfield G (ed). Diversity Chal-
ingly in favor of affirmative action in interviews. This research was supported by The lenged: Evidence on the Impact of Affirma-
admissions, and these students further Andrew W. Mellon Foundation and The Atlan- tive Action. Cambridge, MA: Harvard
commented upon the need to continue tic Philanthropies. Education Publishing Group, 2001: chapter 6.
using such measures. Many of the ma- 15. Bowen WG, Bok D. The Shape of the River:
jority students mentioned that the di- Long-Term Consequences of Considering
versity of students was one of the R EFERENCES Race in College and University Admissions.
Princeton, NJ: Princeton University Press,
more important reasons in their choice 1. Regents of the University of California v. Bakke, 1998.
of a school. They encouraged other 438 U.S. 265 (1978). 16. Cohen JJ. Finishing the bridge to diversity.
schools that have not achieved such 2. Cohen C. Race preference and the univer- Acad Med. 1997;72:103–9.
diversity to be more aggressive in sities—a final reckoning? Commentary. 17. Zweifler J, Gonzalez AM. Teaching residents
2001;112(2):31–9. to care for culturally diverse populations.
recruiting URM students and ex-
3. Wood TE, Sherman MJ. Is Campus Ra- Acad Med. 1998;73:1056.
pressed that it was a privilege to have cial Diversity Correlated with Educational 18. Lum CK, Korenman SG. Cultural-sensitivity
been admitted to a school known for Benefits? hhttp://www.nas.org/reports/umich_ training in U.S. medical schools. Acad Med.
such efforts. diversity/umich_execsum.htmi. Accessed 1/ 1994;69:239–41.
None of the URM students ex- 10/03. National Association of Scholars, 19. Like RC, Steiner RP, Rubel RJ. Recommen-
pressed concern about being burdened Race and Higher Education, Princeton, NJ, ded core curriculum guidelines on culturally
with the mantle of ‘‘spokesperson’’ for 2001. sensitive and competent health care. Fam
4. Burrows GN. Medical student diversity— Med. 1996;28:291–7.
their racial or ethnic groups. In our 20. Tekian A. Minority students, affirmative
elective or required? Acad Med. 1998;73:
work with undergraduates, that reac- 1052–3. action, and the admissions process: a survey
tion frequently occurs—more in re- 5. Holzer H, Neumark D. Assessing affirmative of 15 medical schools. Acad Med. 1998;73:
sponse to classroom interaction—but action. J Econ Lit. 2000;38:483–568. 986–92.

466 ACADEMIC MEDICINE, VOL. 78, NO. 5 / MAY 2003