The Journal of Negro Education, 1974, 43 (No. 4).
Issues in Predicting Black
William E. Sedlacek, Cultural Study Center, University o f
The topic of predicting black student success in higher education has received a great deal of attention in the last several years. I would like to summarize where I feel the area is at the moment (1972) and point to some unresolved issues which might give some direction for the future.
Stanley (1971) in summarizing the work on predicting the success of what he calls "disadvantaged" students, has concluded that admission to selective colleges and universities should be based substantially on test scores and high school grades, irrespective of whether the applicant is from a minority racial, ethnic or socioeconomic group. Stanley feels pessimistic about the possibility of remediation for disadvantaged students and states (1971, p. 642) "an admissions officer ignores test scores at his institution's peril." While there have been an increasing number of studies showing that the same predictors work about as well for Blacks or whites (e.g., Thomas and Stanley, 1969; Pfeifer and Sedlacek, 1971), there also exist studies with contrary or unexplained findings (e.g., Clark and Plotkin, 1964; Green and Farquhar, 1965; Cleary, 1968; Pfeifer and Sedlacek, 1970, 1971, 1974).
There are many potential reasons why there seems to the writer to be at least a reasonable doubt that commonly employed predictors (e.g., tests and high school grades) can be universally applied to all racial groups. First, there never has been an adequate study o f the issue. If we look closely at the literature we see that nearly every study done has repeated the same basic methodology: sample a group of intact Blacks at one or more institutions, correlate their college grades with SAT or ACT scores and high school grades and reach a conclusion. There are many problems with such a procedure, some of which are obvious and some perhaps less
obvious. Many of the problems center on restriction of range and related measurement phenomena. For instance, the types of Blacks attracted to higher education in general or to a specific school tend to be relatively homogeneous and are not very representative of the Blacks with the potential to do college work (Pfeifer and Sedlacek, 1970). Several researchers have reported that Blacks entering different types of institutions tend to have small within school variability and large between school variability (Hartnett, 1969; Borgen, 1970). A second, and related, problem is that typically freshman grades are used as a criterion and therefore only the Blacks who stayed in school for a full year are examined, causing further restriction of range. But many of you are probably saying "true but don't the same problems apply to whites?" The question is a good one and the best answer I have is that there is evidence to suggest that the whole sociocultural process involved in a Black attending college is different from that of a white. For instance, a prospective black student often has no people in his past or present environment who are closely or at all associated with higher education. He must decide (or be surprised) about whether to attend a primarily white institution and face racism and prejudice or go to a primarily black school which may not prepare him for living in the larger society. A Black will often require a longer period to adjust to higher education. As Brooks, Sedlacek and Farver (1972) pointed out, the relatively great increase in mean black sophomore grades compared to freshman grades may be related to this adjustment process. These are just a few conclusions based on research which point to some uniqueness of black experience in higher education. Thus, the way black subjects (Ss) are lost from the potential prediction elipse and the effects on validity coefficients could well be different from the variables that operate to eliminate white Ss from samples studied. Pfeifer and Sedlacek (1970) have made a case suggesting that the elimination of Blacks from potential research samples tends to retain those for whom traditional predictors are related to performance and exclude those who show a discrepancy.
Another point worth noting is that in any dynamic system, that is one changing or evolving such as higher education for Blacks, it is particularly important that we predict for the future, not for the past. There is evidence from the studies cited above, as well as others (Sedlacek, Brooks and Mindus, 1972), that says if the society is to get serious about providing higher education for more Blacks it must tap great resources of black citizenry heretofore untouched. The last few years have seen us reach the black elite or possibly the Blacks who had most adapted to the larger
white society for entry into our colleges and universities. But there are likely many Blacks beyond that group who need and indeed deserve higher education but their culture, background and lifestyles are different than what we typically measure with our predictors and criteria. We, therefore, must be very sure these variables have been adequately explored before we give up on the research area. But let us take some new directions. Let us not simply repeat our favorite study over and over again, reaching the same conclusion. No doubt, it will take some of the most creative research ever done by social scientists to come up with designs which get at these new and different variables. One should also keep in mind that training in research and measurement methods at most institutions does not provide enough content or cultural background to conduct such research. Having a person highly knowledgeable of the kinds of variables you might be dealing with is imperative in designing and conducting research on prediction of black student success. This person will typically be black and from a similar cultural background to the group studied. Having such a person do the research or collaborate with others is no guarantee but it appears to be a prerequisite to good research. I obviously do not belong to such a cultural group and have tried to limit my remarks and interpretations to those developed with others in such a position.
The Cultural Study Center at the University of Maryland has begun a three-phase research program aimed at answering the broad research question "Is there anything that could be called a unique black experience or experiences which could be measured and translated into practical terms?" The Cultural Study Center is interdisciplinary and interracial. The program was begun because of the feeling that it was inappropriate to conduct one shot studies and that only a sustained and systematic effort would have a reasonable chance of success. The first phase of the research was to examine the utility of currently used predictors and criteria. The second phase involves examining currently available variables for use as potential predictors and the third phase will be taking what is learned from the first two phases and working with a variety of Blacks on and off campus to develop predictors and criteria that reflect their experiences. Obviously, the third phase is the most difficult and the "put up or shut up" phase. Studies have been conducted in the first phases at Maryland with the following results:
(1) Multiple R's in the .60's (using freshman grades as a criterion) are possible for Blacks and whites using SAT, high school grades and high school rank, although optimal weights vary by
race and sex (e.g., high school grades do not predict well for black males) (Pfeifer and Sedlacek 1970, 1971).
(2) Attrition and second year grades may be more relevant criteria for Blacks (DiCesare, Sedlacek and Brooks, 1972; Brooks, Sedlacek and Farver, 1973).
(3) Several empirically developed predictors have been identified which bear careful scrutiny and should be restudied and tried out in some form in phase III (DiCesare, Sedlacek and Brooks, 1972; Pfeifer and Sedlacek, 1974).
These variables come from a variety of locally developed and commercial measures such as the California Psychological Inventory and Vocational Preference Inventory. The variables apparently measure such concepts as being independent, self assured, realistic and able to handle difficult adjustments. The term "Social Maturity" is currently being used to describe these variables. The University of Maryland, College Park, planned to admit 104 students for Fall 1972 using these predictors and also to gather data from students admitted on regular bases. Several additional studies are underway in the first two phases.
There are a number of other issues in predicting black student success that were not discussed here, but are also important. These include the ethics of examining racial subgroups and developing separate predictors and the appropriateness of any paper and pencil measure in assessing the ability of black students. The writer feels that much research needs to be done before we can say we have the answers and have given it a reasonable try. If we examine Type I and Type II errors in regard to this whole area of research, we find the following: The Type I error (saying racial differences in prediction exist when, in fact, none do) is the probability that researchers are wasting their time or developing inappropriate measures that will hurt higher education and society. The Type II error (saying there are no racial differences in prediction when they, in fact, exist) is the probability that we systematically deny opportunities to one group (Blacks) in the society. This writer feels the Type II error is much worse and has much broader implications for the society than the Type I error. Enough evidence for a reasonable doubt exists regarding the null hypothesis of no racial differences to compel further and, indeed, better research on predictors of success for black students in higher education.
Borgen, F. H. Able Black Americans in College: Entry and Freshman Experiences. National Merit Scholarship Corporation Research Reports, 6, No. 2, (1970).
Brooks, G. C., Jr., Sedlacek, W. E., & Farver, A. S., Jr. Longitudinal and Cross-sectional Study of
The Journal of Negro Education 515
Academic Predictors for Black and White University Students. Presented at National Council on Measurement in Education Convention, Chicago, April 4, 1972.
Clark, K. B., & Plotkin, L. The Negro Student at Integrated Colleges. New York: National Scholarship Service and Fund for Negro Students, 1964.
Cleary, T. A. Test Bias; Prediction of Grades of Negro and White Students in Integrated Colleges. Journal o f Educational Measurement, 5, (1968) 115-124.
DiCesare, A. C., Sedlacek, W. E., & Brooks, G. C., Jr. Non-intellectual correlates of Black Student Attrition. Journal of College Student Personnel, 13, (1972) 319-324.
Green, R. L., & Farquhar, W. W. Negro Academic Motivation and Scholastic Achievement. journal o f Educational Psychology, 56, (1965) 241-243.
Hartnett, R. T. Differences in Selected Attitudes and College Orientations Between Black Students Attending Traditionally Negro and Traditionally White Institutions. Educational Testing Service Research Bulletin RB-69-48, Princeton, New Jersey, 1969.
Pfeifer, C. M., Jr., & Sedlacek, W. E. The Validity of Academic Predictors for Black and White Students at the University of Maryland. Cultural Study Center Research Report # 2-70, 1970, University of Maryland. Also presented at National Council on Measurement in Education Convention, New York, February 6, 1971.
Pfeifer, C. M., Jr., & Sedlacek, W. E. The Validity of Academic Predictors for Black and White Students at a Predominantly White University. Journal of Educational Measurement, 8, (1971) 253-261.
Pfeifer, C. M., Jr., & Sedlacek, W. E. Predicting Black Student Grades with Nonintellectual Measures. Journal of Negro Education, 43, (1974) 67-76.
Sedlacek, W. E., Brooks, G. C., Jr., & Mindus, L. A. Black and Other Minority Admissions to Large Universities: Three-Year National Trends. Journal of College Student Personnel, 14, (1973) 16-21.
Stanley, J. C. Predicting College Success of the Educationally Disadvantaged. Science, 171, (1971) 640-647.
Thomas, C. L., & Stanley, J. C. Effectiveness of High School Grades for Predicting College Grades of Black Students: A Review and Discussion. Journal o f Educational Measurement, 6, (1969) 203-215.