Response to Chetly Zarko Critique of U-M Diversity Research

By Gerald Gurin, Professor Emeritus of Higher Education and Research Scientist Emeritus, Center for Survey Research, Institute for Social Research; and John Matlock, Associate Vice Provost and Director of the Office of Academic Multicultural Initiatives, University of Michigan

May 16, 2003

The following is in response to the charge by Chetly Zarko that the expert testimony of Patricia Gurin, based in part on the analyses of data of the Michigan Student Study, essentially “lied to the Supreme Court,” because “no one at the University of Michigan has revealed that U-M policy makers used the same data set to arrive at exactly the opposite conclusions.” This charge is completely false and misrepresents the research that has been done using the data from the Michigan Student Study.

Zarko’s General Charge

Zarko’s charges derive from a short, preliminary summary of data that included just the first two years of the Michigan Student Study, prepared in May, 1994, for a retreat of a University-wide “committee on a multicultural university.” This is only a small part of the total study, as it was a longitudinal series of surveys given to a large, representative sample of the undergraduate class that entered the University in 1990. Students were given surveys at the time they entered the University and follow-up surveys at the end of their first, second, and fourth years. A summary of the results of the study, covering the four years of data, appears on the U-M lawsuits website.

Zarko refers to the 1994 document as a “secret” version of the study, that he found it in a “restricted” area of the library, that it contradicts the “cleansed” version that is now on the U-M website, and since Patricia Gurin’s husband is one of the two principal investigators of the Michigan Student Study, Patricia Gurin must have been aware of his “secret analysis” and “contrary research” that refutes the “data deception” of her expert testimony.

No Reports Have Been Hidden

As we detail point by point, every supposedly “hidden” finding that Zarko quotes from the earlier 1994 report on the two-year data is also presented and brought up to the date in the summary of the four-year data that appears on the website. The charge that we “hid” the first report and “cleansed” the final report and that we lied to the Supreme Court is outrageous.

The Research Task for the Litigation

The task that Patricia Gurin undertook was to examine the relationship (or lack of relationship) between students’ own experiences with diversity — in the classroom and in the informal peer environment — and their educational outcomes over four years of college. The question was: Does experience with diverse peers have educational benefits for students?

None of the findings that Zarko focuses on from the 1994 report on the first two years’ of data from the Michigan Student Study is relevant to this question. None of them refer to students’ own experiences with diversity. Instead, all of them refer either to perceptions of racial climate on the Michigan campus or to attitudes about specific university affirmative action and multicultural education policies.

Why Are Findings on Perceptions and Attitudes Irrelevant to
Both Experiences with Diversity and Educational Outcomes?

Dr. Gurin’s testimony did not cover perceptions of racial climate or policy attitudes because they do not tell anything about students’ own experiences or about their educational outcomes. Moreover, perceptions of racial climate cannot be used as proxies for students’ own experiences. The Michigan Student Study shows this dramatically with two sets of questions asked of students in their senior year. One asks them the extent to which a set of racial climate phrases applies to the University of Michigan. Over one-quarter (27%) of the white students responded that there was “quite a bit” or “a great deal” of “interracial tension on campus.” Yet, when asked to characterize their own relationships with students of other racial backgrounds only 1% to 7% of the white students said that their own relationships with Latino(a), Asian American, and African American students were “tense, somewhat hostile interactions,” or “guarded and cautious.”

• To assess students’ experiences with diversity, you have to ask them about their own relationships, not their perceptions of interracial relationships in general.
  
  
• Professor Gurin did not include data on perceptions of racial climate because they tell nothing about students’ own experiences with diversity, and it is those actual experiences that are the causal mechanism by which diversity has educational impact.

The Six Charges in Zarko’s Original Press Release

We will now examine the particular findings from the 1994 report that Zarko refers to, making two points for each finding: 1) that far from being “hidden” or “cleansed,” the finding is presented in the current website summary report of the Michigan Student Study, and 2) that it does not contradict the testimony that the U-M presented to the Courts.

First Charge: Zarko quotes from the 1994 the study’s executive summary that “increasing the numbers of students who attend the institution from different racial/ethnic backgrounds does not in itself lead to a more informed, educated population.”

This quote does not refer to any particular empirical finding from the Michigan Student Study. It is the second part of the last sentence of the 1994 report. The beginning of that sentence, left out of Zarko’s quote, is: “Quite simply, access is not enough.” This quote does not reflect a conclusion that diversity has failed, but that diversity means more than just bringing increased proportions of minority students into an institution. Far from being a “hidden” conclusion, the importance of doing more than simply bringing students together is a central tenet of all proponents of diversity.

This conclusion is completely consistent with the current Michigan Student Study summary on the website and with Professor Gurin’s expert report for the litigation.

• The Michigan Student Study summary also ends with a statement of the need for all members of the University community to work to maximize the potential benefits of diversity.
  
  
• Professor Gurin explicitly notes in her expert testimony that: “Structural diversity (i.e., increasing the numerical representation of various racial/ethnic groups) is essential but by itself usually not sufficient to produce substantial benefits; in addition to being together on the same campus, students from diverse backgrounds must also learn about each other in the courses they take and in informal interaction outside the classroom. For new learning to occur, institutions of higher learning have to make appropriate use of structural diversity.”

Second Charge: Zarko quotes the statement from our 1994 report that our study did “not support the idea that attending a university with an increasing racial/ethnic student body has much of an impact in educating large numbers of students on multicultural matters.”

This comment from our 1994 report referred to our findings on students’ responses to a few questions on specific affirmative action policies. Those findings indicated no changes from the time the students entered the university to the end of their second year. The findings on attitudes toward affirmative action are prominently discussed in the current summary on the website.

• As already indicated, attitudes toward affirmative action were not relevant to the basic question to which Professor Gurin’s testimony was directed — Does actual experience with diversity have educational benefits?
  
  
• It is noteworthy, however, that the Michigan Student Study summary on the Michigan website, covering four years of data, indicates that by the end of the 4th year there was some change in attitudes toward affirmative action after four years in college. White students and students in the three groups of color were all more supportive (increases of support from 10 to 15%) of policies involving standard test scores and financial aid that foster the access of students of color.

Third Charge: Zarko uses the 1994 report to claim that there is increasing polarization of racial attitudes over the first two years of college. He quotes our finding that “at entrance 46% of African American students perceive a university commitment to students of color, and at the end of the second year this number decreases to 19%. For white students, the number perceiving a university commitment to students of color increases from 57% at entrance to 70% in the sophomore year.”

The point that over the years at Michigan there is a great increase in the difference between white and African American student views about Michigan’s commitment to students of color is again not “hidden” but reported in the current summary on the website.

• This finding, however, does not indicate a negative impact of diversity that contradicts the conclusions of Patricia Gurin’s expert testimony. The fact that African American students increase in their feeling that Michigan is only minimally committed to diversity suggests that they would like more diversity than they find at Michigan but says nothing about its impact on their educations — neither negative or positive.
  
  
• In fact, Patricia Gurin’s testimony shows that actual experiences with diversity had very similar educational effects for white and African American students. Moreover, that white and African American students differ in their perceptions of institutional commitment is hardly reflective of the supposed racial confrontations on college campuses that critics of affirmative action are referring to when they speak of the “polarization” that is a supposed consequence of affirmative action.

Fourth Charge: Zarko quotes from the 1994 report that our findings “supported the concern in the literature on minority students that they feel stigmatized and academically disregarded by faculty at predominantly white institutions.” Zarko goes on to suggest that “although reflecting the authors’ bias that the predominantly white nature of the university was not the cause, this conclusion equally supports a counter argument to racial preference programs that they create stigmatization and lower self-confidence.”

Again, this is an issue discussed in the current report on the Michigan Student Study. The current website report contains a section on “feelings of belonging and alienation,” which includes a discussion of African American students somewhat more often feeling than other students that “the basic legitimacy of their presence on campus was questioned.”

Neither the 1994 nor the current report supports Zarko’s suggestion that this stigmatization is caused by affirmative action policies. As a matter of fact, Figure 8 in the 1994 report clearly contradicts such an argument. Figure 8 compares the responses of white students and students of color to a question asked when they entered and two years later. The first question at entrance in 1990 asked how much they expected the racial climate at Michigan to be characterized by “respect by white faculty for students of color.” The question two years later asked what they had found at Michigan in terms of “respect by white faculty for students of color.” The findings in Figure 8 show that African American students differ greatly from other students, but that these differences are completely a reflection of their feelings at the time they entered Michigan, not of their Michigan experiences. For example, at the time of entrance 33% of African American students compared to 86% of white students expected that they would find “quite a bit” or “a great deal” of “respect by white faculty for students of color.” At the end of two years when asked about what they had found at Michigan, the comparative figures were virtually unchanged: 32% for African American and 88% for white students. African American students’ feelings of stigmatization come from the history of devaluation they have experienced before college, not from affirmative action policies at Michigan.

• These findings on stigmatization were not discussed in Patricia Gurin’s expert testimony that utilized the data from the Michigan Student Study. As we indicated above, Patricia Gurin’s expert testimony was concerned with educational outcomes and did not present any finding on stigmatization from the Michigan Student Study.
  
  
• However, in the first paragraph of his press release, Zarko states that in oral arguments to the Supreme Court, John Payton, the U-M lead counsel in the undergraduate law suit, cited Patricia Gurin’s expert testimony “that diversity decreases stigmatization.” John Payton’s reference was to a supplemental expert report (July 13, 2000) by Patricia Gurin that addressed the issue of critical mass. In the supplemental report she buttressed her argument of the need for more than “token” minority students in classes and other educational settings with findings from the general social psychological literature, not with any analysis of the Michigan data.
  
  
• This distinction between Patricia Gurin’s discussion of general social psychological literature and the Michigan data is important because in his opening paragraph Zarko states that “Gurin’s testimony that diversity decreases stigmatization…. relied largely on her statistical analyses of data from the Michigan Student Study” and that “no one at U-M has revealed that U-M policy makers used the same data set to arrive at exactly the opposite conclusion.” By erroneously attributing the basis of John Payton’s oral argument to analysis of data of the Michigan Student Study and by erroneously stating that these data suggest that diversity at Michigan is the cause of stigmatization, Zarko falsely accuses us of “manipulating” the dataset of the Michigan Student Study to “tailor” our testimony and “lie” to the Supreme Court.

Fifth Charge: Zarko quotes from our 1994 report that “few group differences were found when (the study) examines students’ general expectations…. not directly focused on racial/ethnic diversity.” Zarko claims that this quote supports “the counter argument to diversity that blacks and whites don’t think differently.”

Again, the point about minimal group differences in a number of areas is not “hidden” but made in the current executive summary on the website. For example, the current summary notes that “despite differences in parental backgrounds and racial/ethnic experiences students of all origins showed great commonality in their goals for college and their high academic investment and motivation.”

The Michigan Student Study has consistently pointed out that there are both differences and similarities between students of different racial/ethnic backgrounds, depending upon the specific arena being measured. Students of color and white students hold very much the same goals and aspirations for college and for their lives. They frequently express very different perspectives about how various groups are treated in our society and how various institutions (including the University of Michigan) operate.

It is in fact these commonalities about their own lives and different perspectives on society and its institutions that students learn in intergroup encounters on the Michigan campus.

In both the 1994 and current website summary (as well as in Patricia Gurin’s expert report), We stress the point that increased interactions with diverse peers lead to greater appreciation of both differences and similarities of students of varying racial/ethnic backgrounds. We find it puzzling that critics of affirmative action, like Zarko, argue that it is inconsistent for proponents of affirmative action to say that interracial contact produces an appreciation of both similarities and differences from others. They apparently feel that it has to be one or the other. But this portrays a remarkably simplistic understanding of human relationships.

Positive interactions with people from different racial/ethnic groups produce a more complex and nuanced view of members of one’s own group and members of other groups. It is not inconsistent to say that this complexity involves an understanding of the differences that come from the different life experiences of racial and ethnic groups in the United States as well as the similarities that reflect their many shared experiences. This complexity further involves an appreciation of the individual variability of members of both one’s own and other racial/ethnic groups. We learn how we differ in ways we have not thought about, and we learn how we are similar in ways we had previously stereotypically assumed that groups differ.

Sixth Charge: Zarko incorrectly states that our 1994 report “concluded that most of the differences in educational performances between whites and minorities were related to real socio-economic disparities.” He says that this statement (which is not in the 1994 report) implies that “a race-neutral policy of socio-economic preference alone could be designed that would statistically benefit minorities.”

The 1994 report did not relate differences in socio-economic status to differences in “educational performance.” It related the lower socio-economic backgrounds of African Americans and to a lesser extent of Latino(a) students to less financial resources available for financing their college educations, to their greater reliance on financial aid, and to their greater concern about whether or not they would have the finances to complete their educations. These findings are also reported in the current executive summary on the U-M website.

The relationship of racial/ethnic background to socio-economic status and economic resources is one of the most well-documented findings in the social science literature. It not exactly the surprising or damaging finding that would have to be “secreted” in a “restricted” area of the Bentley library. Moreover, our data certainly do not suggest that the relationship between ethnicity and economic resources is great enough to justify Zarko’s conclusion that a race-neutral policy of socio-economic preference alone could be designed to limit the need for Michigan’s current use of race in its admission policies (which of course already assign points in the admission process to students from economically disadvantaged backgrounds).

A New Charge in Zarko’s Second Press Release

Zarko made one charge in his second press release that was not made in the first press release. The new charge is particularly significant and insidious because it claims “a sharp contrast” between the 1994 report and the Patricia Gurin testimony before the Supreme Court. Therefore, this charge supposedly supports Zarko’s title: “Lying to the Supreme Court?”

Specifically, Zarko quotes from the 1994 report that “interaction is generally not occurring in close social networks.” He then continues, stating that “this is in sharp contrast to the Gurin expert testimony before the Court that ‘Michigan students indicate a considerable degree of interracial contact in their general relationships on the Michigan campus’ and that ‘the quality of these interactions is predominantly positive, particularly between white students and Asian Americans and Latinos.’ The prevarication in the Court testimony is found by the Gurins’ focus on the qualifier ‘particularly between white students and Asian Americans and Latinos.’ The absence of quality relationships between African Americans and whites is almost implied.”

In response, we note the following three points:

• First, the Gurin testimony does not “hide” the findings of our “secret report of 1994.” The statement from our 1994 report that “interaction is generally not occurring in close social networks” comes from the students’ responses to a question that asks them to think of their six best friends and then indicate the race/ethnicity of each friend. Our 1994 report indicated that “at the end of the first year most African American and white students indicate that most of their close friendships at Michigan remained with people from their own group.”
  
  In her expert testimony, Patricia Gurin includes a discussion of students’ responses to the “six best friends” question. She notes, as we did in our 1994 report, that “close friendship circles of African American and white students are predominantly with peers of their own backgrounds, both at entrance and after four years at the University of Michigan.”
  
  With the benefit of the senior data, not available for the 1994 report on the first and second years in college, Patricia Gurin does indicate that there is a significant increase in the racial/ethnic diversity of the six close friendships of white and African American students over the four years at Michigan. But there is no attempt to deny that they are still “predominantly with peers of their own background.”
  
  
• Second, in 1993 when we were preparing our senior survey, we realized that while we had a question about the six best friends, we had no detailed information about the quality of their overall interracial interactions on campus. Quality and not just quantity of their overall interactions is crucially important. Therefore, we added some new questions about quality of interactions.
  
  We asked students which racial/ethnic group they interacted most with on the Michigan campus, and then asked how much their relationships with students in that group involved a set of positive and negative characteristics (for example, “shared personal feelings and problems,” “had tense, somewhat hostile interactions,” etc.).
  
  Patricia Gurin’s conclusion that the students’ general interracial interactions are predominantly positive comes from the answers to the question about positive and negative characteristics of those relationships. As we have indicated, that question was not asked in the first two years of the surveys that the 1994 report covered.
  
  It is interesting that Zarko never considered the possibility that differences between the conclusions in a very preliminary statement of the first two years of data and those of a final four-year report might come from new data, rather than “lying” and “cheating.” The whole point of longitudinal studies is to be able to assess new aspects of experience as they happen over time.
  
  
• Zarko states that there is some “prevarication” in the use of the “qualifier” that the positive interracial relationships were “particularly between white students and Asian Americans and Latinos” and that “the absence of quality relationships between African Americans and whites is almost implied.”
  
  Patricia Gurin’s expert testimony devotes as much space to relationships between whites and African Americans as to relationships between whites and the other two groups. The testimony concludes that “relationships that white students have with African American students were somewhat less personal than their relationships with other students of color, but very few white students felt that their interactions with African Americans were negative.” The testimony further states that “from the perspectives of African American students their relationships with white students were somewhat ambivalent, reflecting negative as well as positive interactions.” These conclusions in the testimony hardly merit the epithet “prevarication” and “almost implied.”

Conclusion

Repeatedly we have shown that every charge of “secretive, hidden” findings is entirely false since these findings also appear in the current summary on the U-M website. We have also shown that none of these findings contradicts Patricia Gurin’s conclusions about the beneficial effects of experiences with diversity on students’ educational outcomes.

Top of page