Alternative Admissions Models


This article is an excerpt from Brown, S.E. and Marenco, E., Jr., Law School Admissions Study. San Francisco: Mexican-American Legal Defense and Educational Fund, 1980.

There is no argument against the logic that Blacks and whites who seek to master the same professional discipline must each develop the same body of skills and understandings. What is argued is that the disposition to learn them (aptitude) may be measured differentially with respect to group membership, and this should be considered in selection.

Open affirmative action programs have no need to apologize for using tests as only one part of the selection process. In order to develop a program of fair selection, the weaknesses of employing a purely psychometric basis for selection must be recognized and dealt with squarely. No apologies need be made for including relevant dimensions in the selection process.

Selection instruments need not be discarded, because they are found wanting. The same argument presented here in assessing tests could be repeated for any arbitrary procedure based on comparing an individual to group performances. The solutions are not psychometric; they are social policy decisions, but they can be made more difficult by psychometric misunderstanding. [emphasis added] Psychometricians can give us useful tools, but they must not be misused. This does nor absolve test constructors of responsibility. They should be spending as much time and money assisting selectors for school and employment in using other predictors as they do in developing tests. In fairness to all, affirmative action has the potential to enrich our society with the contributions to many areas which have too long stood dry from healthy infusion of a diverse racial and sexual populace.

From Johnson, The Measurement Mystique, pp. 47-48.

Introduction

The alternative law school admissions criteria which follow focus on a deemphasis of the LSAT as an evaluator of law school potential with an accompanying emphasis on other factors which have been shown to be indicative of success in legal studies, as well as constitutionally permissible. The advocated models recognize the unique character of each law school and should be evaluated an implemented as appropriate in each individual institution while also satisfying the constitutionally permissible and morally compelling goal of increasing access for minority law students.[1]

Moreover, the following models are not mutually exclusive one from the other, but rather may be combined and tailored to the stated institutional objectives of each school. As discussed in Section III, supra, the alternative criteria are structured around the pedagogically sound premise that standardized tests , in general, and the LSAT, in particular, measure only one set of cognitive skills out of the many that are necessary to become a competent attorney.[2] As the Carnegie Council of Policy Studies in Higher Education stated:

"Grades and tests looked at together are more predictive of subsequent academic performance than grades alone or tests alone. More generally, Torsten Husen (1976) has noted that: “Extensive empirical research tells us that at most half of individual differences in educational attainments are attributable to purely intellectual factors. The rest may be attributed to motivation, interest, perseverance, health, and, of course, home background.”…Some of these other factors, however, have proven difficult to assess prior to admission. These other factors also, of course, affect earlier grades and entrance test scores.[3] "

For minority students it is questionable whether a combination of tests and grades is more predictive than grades alone.[4] Irrespective, minority test scores cannot be divorced from the economic and cultural background of the examinee.

The following alternative criteria are also presented within a sociological context which acknowledges the necessity of a critical mass of minority students in a given institution not only to approach population parity but to guarantee optimal performance of minority candidates.[5] For example, studies conducted on women in law school indicate that women law students when enrolled in proportionally small numbers within a predominately male student body either overachieve or settle for very mediocre academic performance. The same studies, however, demonstrate that women tend to distribute themselves evenly along the range of academic performance with other students when enrolled in more than token numbers.[6] Thus, it is suggested that for minority law students to reach their true potential in legal studies, their numbers must allay possible feelings of isolation and alienation.[7]

With these considerations in mind, the following alternative criteria are offered to law school deans, professors, admissions officers, students, and shapers of public policy for their serious consideration. It must be followed that the following admissions models, just as every existing admissions model now in use in law schools, contain an implicit or explicit policy decision. Since the LSAT is not a perfect instrument for selection, law schools consider a variety of other criteria. Thus, admissions decisions are currently made not only on the basis of minute yet seemingly significant differences in predicted first-year law grades, but also on a consideration of a diverse set of noncognitive characteristics.[8] Our models seek to effectuate a policy of greater diversity and integration within the legal profession by suggesting other indices for selection apart from GPA and LSAT.[9]

Although our models are not necessarily predictive of first-year success in law school in the same way as traditional admissions indices, they assist in a more complete assessment of law school candidates. By definition, the advocated models are not finished admissions procedures that can be used without further test-piloting and subsequent revision. Rather, our models are specific, concrete approaches with directions for implementation, that law schools will need to review closely and test before the models can be fully adopted.

A. A Cultural Diversity Model

Rationale of the Diversity Model

The cultural diversity model directly responds to and satisfies the constitutional of Bakke. As Justice Powell observed:

"Thus, in arguing that its universities must be accorded the right to select those students who will contribute the most to the “robust exchange of ideas,” petitioner invokes a countervailing constitutional interest, that of the first amendment. In this light, petitioner must be viewed as seeking to achieve a goal that is of paramount importance in the fulfillment of its mission.[10]"

Not only is diversity a compelling constitutional interest, but the diversity formula set forth in this model capitalizes on the established admissions procedures and student composition of each law school, thereby according utmost deference to traditional university autonomy while satisfying first amendment rights.

Further, the cultural diversity model recognizes and resolves the perplexing fact that diversity, in the sense of meaningful racial and ethnic diversity, will not exist in the absence of admissions models, which expand from the traditional cognitive criteria of GPA and LSAT. Franklin Evans of the Educational Testing Service documented that:

"If the nation’s law schools were to adopt and admissions policy taking no account of minority backgrounds of blacks and Chicanos, a majority of the students from those groups now admitted and enrolled would be excluded….If numerical predictors were employed exclusively for all applicants, the resulting reductions would be 76 to 78% for blacks and 45 to 48% for Chicanos.[11]"

Yet there is no statistical, constitutional, or moral reason to limit admissions criteria to strictly numerical indices. On the contrary, the studies cited in Section III, supra, indicate that the cognitive scores are likely to be misused against all applicants if isolated from other relevant candidate data.

This diversity model provides other relevant factors for applicant evaluation. It has the important feature of adjusting the weight accorded to an applicant’s cultural diversity of the basis of the racial/ethnic enrollment in that particular institution. Applicants who are underrepresented will automatically receive more weight on the cultural diversity part of the formula than will candidates who are already well represented at the law school. Thus, the “robust exchange of ideas” which Justice Powell found compelling will be achieved without quotas and within the framework of a formula which is relatively simple and administratively feasible. Moreover, this model, by virtue of its noncognitive component, encompasses diversity characteristics apart from race and ethnicity.

Description

The purpose of this model is to provide a systematic procedure focusing on noncognitive admissions characteristics which research demonstrates may be useful in evaluating candidates for graduate and professional schools.[12] Although these characteristics are particularly relevant to minority candidates, they can be used as criteria to gain additional information on all applicants.[13]

The elements of our cultural diversity model are expressed by the following formula:

[NC (noncognitive) score + C (cognitive) score] x CD (cultural diversity) score = AS (applicant score)[14]

This formula describes a procedure for law student selection which gives weight to cultural diversity based on the existing racial/ethnic composition of a particular law school, and, hence, increases the chance of minority student selection. Moreover, the criteria are keyed to the philosophy that excellence in education is promoted when a critical number of individuals with varying characteristics are recruited for professional school preparation.[15]

As designed, the cultural diversity formula may be implemented in one of two ways. Pursuant to a unitary admissions approach, the formula may be applied to every law school applicant to a particular institution. In the alternative, a predetermined percentage of students may be admitted under the existing criteria of a law school with the remaining applicants being evaluated on the basis of the noncognitive and diversity factors which follow.

Dr. William E. Sedlacek, developer of the model, suggests that, for administrative reasons, 50 percent of law applicants to a particular school be admitted under the school’s established criteria. This percentage, of course, could be adjusted by any school, in keeping with the HEW guidelines on Title VI, according to its own numerical targets and/or prior experience in minority enrollment.[16] If 50 percent were admitted traditionally, then the remaining 50 percent would be chosen based on the above formula which seeks to foster true diversity in entering law classes. Although the cultural diversity and noncognitive components of the formula may be adjusted by different schools based on their test-pilots of the model, the essential procedure for implementation is as follows:

Implementation Procedure

1. Select 50 percent of the entering class using traditional methods (GPA, LSAT, letters of recommendation, etc.). In the alternative, omit this step and evaluate all applicants as detailed in steps 2 through 5.

2. Develop a composite score for all applicants, or the remaining 50 percent, on the following noncognitive variables. These variables are scored on a scale of 1, 2, or 3 points, with 3 being the highest. Data to achieve scores may be obtained from letters of recommendation, personal statements, interviews, etc. They are variables which admissions personnel and committees must scan for, since they could be obtained anywhere in the applicant’s record. All are supported with research as to their utility, particularly for all racial/ethnic groups, but for whites as well.[17]

a. Noncognitive variables:

1. Self-concept.

2. Realistic self-appraisal.

3. Understanding racism.

4. Long-range goals.

5. Availability of a strong support person.

6. Leadership.

7. Community Service.

8. Demonstrated legal interest.[18]

b. The highest score attainable is 8 x 3 = 24, while the lowest is 8 x 1 = 8. Since there are eight traits with a maximum value of 3 points per trait, a score of 24 would be the maximum while 8 would be the minimum because each applicant receives at least 1 point per trait. Develop a distribution of these scores for all candidates, or the remaining 50 percent, and convert these scores into T scores which have a mean of 50 and a standard deviation of 10. A T score is merely a statistical method for equating scales which are not equivalent.[19] The resulting score, in this component of the formula, is called the NC or noncognitive score.

3. Develop a distribution of the remaining applicants based on the traditional cognitive variables (GPA and LSAT) used by an institution in ranking and selecting admittees. This distribution,as discussed previously, will contain either all candidates or the remaining 50 percent of the applicants after the first 50 percent were admitted pursuant to established criteria. The goal is to develop a distribution based on a single composite ranking of the cognitive variables for each applicant. This distribution, as the noncognitive distribution, will be converted into T scores and will represent the C or cognitive score. For hypothetical examples of conversions of student GPAs, LSATs, and nocognitive traits into T scores, please refer to Addendum B at the end of this part.

4. Depending on the admissions formula chosen by the institution, cultural diversity scores will be assigned to all students or to those remaining after a specified percentage were admitted by the existing criteria of a law school. For a cultural diversity score to be assigned, however, there must be some external norm against which the weight of the score is determined; this is so because the purpose of the diversity model is to automatically adjust the weight each racial/ethnic group receives in evaluation for admission based on the representation of that particular group in a specific law school. The model, therefore, facilitates law school access to those groups least represented in a given law school by assigning them a higher cultural diversity score.

5. The HEW “Nondiscrimination Policy Interpretation,” on Title VI as discussed in Section V, supra, permits a university to establish a numerical target for ethnic/racial minority admissions. Certainly a law school could use some numerical target or even last year’s actual enrollment, as broken down by race and ethnicity, to establish a benchmark against which to measure applicants for cultural diversity. Cultural diversity scores then, are calculated as follows:

Institutional Composition CD Score Multiplier
Less than 10 percent of the Applicant’s racial/ethnic group is represented (a) in the 50 percent of the class already admitted under established criteria or (b) in the student body of a particular law school or (c) by some other numerical target used as a benchmark for assigning cultural diversity scores.

1.5

Between 11 and 50 percent is represented.

1.25

More than 50 percent is represented.

1.0

6. Final selection is made pursuant to the following formula:

a. (NC score + C score) x CD score = AS

b. Those individuals with the highest applicant scores are selected for admission.

Conclusion

The cultural diversity model is inherently fait in that every applicant is compared against every other applicant on the basis of cognitive, noncognitive, and cultural diversity traits. If the model is not applied to all students, it is still fair since every applicant not admitted pursuant to the traditional criteria of the law school must compete individually with every other applicant not admitted by the established criteria. The criteria utilized in this model, moreover, is indisputably within the letter and spirit of Bakke, as articulated by Justice Powell, in that applicants are assessed on a multitude of traits in an effort to achieve true diversity within the law school. Cultural diversity is not assigned a fixed weight nor is it implemented by arbitrary quotas. Rather, applicants are individually evaluated for cultural diversity, among other traits, based on the specific cultural composition of the law school to which they are applying. Finally, the cultural diversity component of the model is adjusted automatically in the admissions process according to the percentage of particular racial/ethnic groups already represented in the institution.

Addendum A to Section VI, Part A

Below is a description of the weighting scale values for each of the components that make up the NC score.

I. Positive Self-Concept
(Strong self-feeling; strength of character. Determination, independence.)

Meaning Code (Points)
Initiates statements or Behaviors that indicate strong positive feelings about oneself, e.g., “I felt I could do well on a project so I took extra initiative.” Took Heavy course-loads in school. Willingness to try new things over a long period of time.

3

Some evidence of positive Feelings or behaviors but not strong. Some good evidence, some bad. Does not take initiative in trying new things or presenting evidence of self-worth; or only recent evidence of good self concept.

2

Shows no evidence of good self-concept or negative evidence. No evidence of trying new things; statements of unexpected failure made.

1

2. Realistic Self-Appraisal
(Especially academic. Recognizes and accepts any deficiencies and works hard at self-development. Recognizes need to broaden his/her individuality.)

Meaning Code (Points)
Presents clear evidence of assessing shortcomings in his/her background and has taken steps to overcome. Could be curricular or personal, e.g., “I knew that I was short in math so I took an extra course.” “I was not effective in dealing with colleagues so I sought them out for reasons why.”

3

Some recognition of some shortcomings but has generally not taken action to correct.

2

No evidence that shortcomings recognized; defensive or avoids questions concerning possible problem. Covers up and offers excuses.

1

3. Understands and Deals with Racism
(Realist based on personal experiences of racism. Is committed to fighting to improve existing system. Not submissive to existing wrongs, Hostile to society, or a “cop-out.” Able to handle racist system.)

 

Meaning Code (Points)
Initiates realistic explanations of how racism (particularly institutional racism) affects life. Not bitter. Understands that some of his/her life is controlled by the system based on race or sex and some is individually determined. Evidence of successfully handling interracial and/or intersexual situations, e.g., “I expect that some people may not understand modern women, but I had one supervisor who came around after I let him know what I could do.”

3

Some good evidence, some not so good or tentative. Not a full understanding. May be bitter or confused.

2

No understanding of racism, hostile, resentful. Blames everything on the system being against Hispanics, Blacks, etc., if a minority. Feels resentful of reverse discrimination if white. No demonstrated method of handling interracial or intersexual situations well.

1

4. Prefers Long-Range Goals to Short Term or Immediate Needs
(Able to respond to deferred gratification.)

Meaning Code (Points)
Consistent evidence of planning and future orientation over a long period, e.g., “As a freshman, I figured I had better study if I wanted to get into law.” “I realized I had to learn X procedure on the job before I could get promoted.”

3

Some recognition of long term goals but no long term evidence, or mixed evidence.

2

No evidence of long-term planning; looks at issues in immediate terms, unprepared for future.  

1

5. Availability of Strong Support Person
(To whom to turn in crises.)

Meaning Code (Points)
Someone has provided assistance in times of crisis. Generally same person or one at a time sequentially, e.g., grandmother, then teacher, then boss, etc. knows where to go in a crisis.

3

Sometimes has received  help but not consistently; somewhat unclear about where to go in crisis.

2

No evidence of turning to others, loner, tough it out, then says no problem.  

1

6. Successful Leadership Experience
(In any area pertinent to background, e.g., gang leader, sports.)

Meaning Code (Points)
Behavioral evidence of influencing others in the context of his/her cultural or socialized background (may not be traditional, e.g., gang leader, unusual hobby, or community work). Has shown evidence over a period of time.

3

Some evidence of leadership position. Not clear what his/her influence really was, may list offices held in student organizations or other organization.

2

No evidence of influencing others or holding office. May avoid or be uncomfortable in leadership role, e.g., “Let others do it—I’m too busy.”

1

7. Demonstrated Community Service

Meaning Code (Points)
Behavioral evidence of activity and identification with community. Long term involvement and interest. Community must be allowed to be cultural/racial as well as geographical.

3

Some contacts with community but may be just recent, or perhaps, more likely, in the past with an uncertain present and future.

2

No contact with community. Little or no evidence that he or she is aware of the concept or its importance. Alienated, separated from cultural/racial background.

1

8. Demonstrated Legal Interests

Meaning Code (Points)

Behavioral evidence of activity and interest in the law and legal issues for some time. Interests may be through one’s culture, bettering one’s culture through the law, etc. Allow for nontraditional views of legal interest.

3

Some behavioral evidence of legal interests but not strong or long term.

2

No evidence of interest in the law or legal issues, or perhaps avoidance of such issues.

1

Addendum B to Section VI, Part A

Applicant A

Description: White, high grades an LSATs but not involved in activities. Shows performance in traditional ways in classroom.

GPA = 3.6 = T score of 65

LSAT = 750 = T score of 75

Computation of C score:

The school evaluating Applicant A weights GPA 50% and LSAT 50 %. Thus we can simply get the mean of the two T scores (65 + 75) divided by 2 = 70. (C = 70)

Computation of NC score:

Applicant A scored as follows on the eight noncognative variables making up the NC score:

Self-concept = 2

Realistic self-appraisal = 2

Understands racism = 1

Long-range goals = 2

Strong support person = 1

Leadership = 1

Community = 1

Demonstrated legal interests = 3

The sum of these eight scores is 13. If we compare this to a distribution of these scores from all applicants to the school, we get a T score of 40 or 1 standard deviation below the mean. This person would be at the 16th percentile, or the lowest 16% of the applicants on NC. (NC = 40)

Computation of CD score:

Based on applicant’s race applicant receives a 1 for being in a group that represents more than 50% of the applicants. The reference group here could be the current year’s applicants, last year’s applicants, residents in the area, etc. The weights assigned to cultural/racial groups as of this date are: 1 = more than 50% represented; 1.25 = 11 to 50% represented; 1.5 = 10% or less represented. (CD = 1)

Computation of applicant score (AS):

AS = (70 + 40) x 1 = 110

Applicant B

Description: Chicano, average LSAT and grades, but shows good performance in many areas outside the traditional educational setting.

GPA = 2.9 = T score of 45

LSAT = 450 = T score of 38

Computation of C score:

The scores for GPA and LSAT are based on the applicant pool of the school involved. The school evaluating Applicant B weights GPA two-thirds and LSAT one-third. Thus, C = (45 + 45 + 38) divided by = 42.67.

Computation of NC score:

Applicant B scored as follows on the eight noncognitive variables making up the NC score:

Self-concept = 2

Realistic self-appraisal = 2

Understands racism = 3

Long-range goals = 3

Strong support person = 2

Leadership = 2

Community = 3

Demonstrated legal interests = 2

The sum of the noncognitive variables is 19. If we compare this score to a distribution of these scores form all applicants we get a T score of 66 for the NC component. (NC = 66)

Computation of CD score:

There were 11 – 50% Chicanos in the reference group employed by the school. (CD = 1.25)

Computation of applicant score (AS):

AS = (42.67 + 66) x 1.25 = 135.83

Applicant C

Description: Black, low grads and SATs, few activities and performance in areas outside education.

GPA = 2.6

LSAT = 370

Computation of C score:

The school evaluating Applicant C does not specifically weight GPA and LSAT, but makes an overall assessment of academic qualifications and ranks all the applicants to this school. Applicant C was in the lowest 20%, the T score equivalent of 30. (C = 30)

Computation of NC score:

Applicant C scored as follows on the eight noncognitive variables making up the NC score:

Self-concept = 2

Realistic self-appraisal = 1

Understands racism = 1

Long-range goals = 2

Strong support person = 1

Leadership = 1

Community = 1

Demonstrated legal interests = 2

The distribution of applicants yielded a T score of 32 for the sum of 11. (NC = 32)

Computation of CD score:

1.5 was assigned because the applicant reference group was 10% or less Black.

(CD = 1.5)

Computation of applicant score (AS):

AS = (30 + 32) x 1.5 = 93


[1] In Bakke Justice Poweel stated: “No such facial infirmity (intent or discriminate) exists in an admissions program where race or ethnic background is simply one element—to be weighed fairly against other elements in the selection process….And a court would not assume that a university, professing to employ a facially nondiscriminatory admissions policy, would operate it as a cover for the functional equivalent of a quota system. In short, good faith would be presumed in the absence of a showing to the contrary.” The Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 318-319 (1978).

HEW issued the following interpretation of the efect of the Bakke decision on Title VI of the Civil Rights Act of 1964. In essence, HEW’s interpretation encourages educational institutions “to continue and expand voluntary affirmative action programs to increase their enrollment of minority group members….” Although the interpretation notes that institutions may not set aside a fixed number of places for minority students absent a judicial, legislative or administrative finding of past discrimination, the guidelines expressly permit, “consideration of race, color, or national origin among the factors evaluated in selecting students; increased recruitment in minority institutions and communities; use of alternative admissions criteria when traditional admissions criteria are found to be inadequately predictive of minority student success; [emphasis added] provision of preadmission compensatory and tutorial programs; and the establishment and pursuit of numerical goals to achieve the racial and ethnic constitution of the student body the institution seeks.” [emphasis added]

The guidelines further state that, in addition to these techniques, “institutions may use their authority to broaden admissions criteria generally to evaluate better the qualifications of minority applicants…by giving increased consideration to an applicant’s character, motivation, ability to overcome economic and educational disadvantage, work experience, and other factors.” HWE, “Nondiscrimination Policy Interpretation,” pp. 58509-58511.

[2] Fuller and McNamara sense: “Leona Tyler’s (1978) review of research and philosophy on individual psychological differences provides a broad view of human skills related to, yet distinct from competence in processing information. For example, ‘field-dependent’ individuals, who may be less able to reduce a problem to its component parts, may be more competent to work in small groups rather than alone. Evidence also indicates that those most proficient at solving problems with one ‘right’ answer, for example, questions on standardized tests, may be less able to think divergently, creatively hypothesizing alternative approaches to a problem (Frederiksen, 1978). In the rush to rank applicants by their competence at processing information, other individual differences that could nurture greater diversity among students within professions are ignored.” Bruce Fuller and Patricia P. McNamara, “Defining and Assessing Disadvantagement,” Admitting and Assisting Students After Bakke, ed. Bruce Fuller and Kenneth C. Green (San Francisco: Jossey-Bass, Inc., 1978), p. 68.

See also University of California Task Force Report, September 1977, p. 40.

[3] “Public Policy and Academic Policy,” Selective Admissions in Higher Education,” p. 10 & p. 5.

[4] Astin, “Quantifying Disadvantagement,” p. 75.

[5] Eve Spangler, Marsha A. Gordon, and Ronald M. Pipkin, Token Women: An Empirical Test of Kanter’s Hypothesis, Law Student Activity Patterns Project of the American Bar Foundation (Chicago: University of Chicago, 1978); reprinted from The American Journal of Sociology, vol. 84, no. 1 (July 1978), pp. 160-170.

[6] Ibid.

[7] Green and McNamara observe: “The rarity of minority students and faculty members on a campus, let alone within a department or professional school , contributes to this feeling of alienation and isolation experienced by minority students. Says one black law student, ‘You do feel isolated because there are so few black students. And I guess you become somewhat angered because you know that there are more qualified black students out there who could just as well be here.’” Kenneth C. Green and Patricia P. McNamara, “The Student Experience,” Admitting and Assisting Students After Bakke, ed. Bruce Fuller and Kenneth C. Green (San Francisco: Jossey-Bass, Inc., 1978), p. 32.

[8] See Section IV, supra, for an analysis of MALDEF’s survey instrument distributed to all ABA-approved law schools in California.

[9] In advocating alternative criteria, MALDEF recognizes that some of the suggested alternatives would require a greater financial and time commitment on the part of the law school admissions officers, committees, faculties, and deans than exists presently. Notwithstanding, one can legitimately ponder why the medical schools have been willing to make that additional commitment toward an individualized selection process when many law schools have not.

[10] The Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 313 (1978).

[11] Evans, “Applications and Admissions to ABA Accredited Law Schools,” pp. 566-567. See also Section III, supra.

[12] The above model was developed by Dr. William E. Sedlacek. Dr. Sedlacek is one of the leading researchers in the formulation of noncognitive variables which have been shown to be indicative of minority success in higher education. His work includes: C.M. Pfeifer, Jr. and W.E. Sedlacek, “Nonintellectual Correlates of Black and White Students’ Grades at the University of Maryland,” Cultural Study Center Research Report # 3-70 (College Park, Maryland: University of Maryland, 1970); C.M. Pfeifer, Jr. and W.E. Sedlacek, “Predicting Black Student Grades with Nonintellectua l Measures,” 43 J. Negro Educ. 67-76 (1974); C.M. Pfeifer, Jr. and W.E. Sedlacek , “The Validity of Academic Predictors for Black and White Students at a Predominantly White University,” 8 J. Educ. Measurement 253-261 (1971); and D.O. Prieto, P.G. Bashook, A.G. D’Costa, P.R. Elliot, R.K. Jarecky, B. Kahrahrah, W.F. Leavell, and W.e> Sedlacke, Simulated Minority Admissions Exercise Workbook (Washington, D.C.: Association of American Medical Colleges, 1978), pp. 1-33.

Other researchers whose work supports that of Dr. Sedlacek are: A.R. Baggaley, “Academic Prediction at an Ivy League College, Moderated by Demographic Variables,” Measurement and Evaluation in Guidance, 6 (1974): 232-235; F.H. Borgen, “Able Black Americans in College: Entry and Freshman Experiences,” Merit Scholarship Corporation Research Reports, 6, no. 2 (1970); R.D. Goldman, “Hidden Opportunities in the Prediction of College Grades for Different Subgroups,” 10 J. Educ. Measurement 205-210 (1973); F. Perry, Jr., “Selected Variables Related to Academic Success of Black Freshman Students at the University of Missouri – Columbia,” (Ph.D. diss., University of Missouri, Columbia, Missouri, 1972); and G. Temp, “Validity of the SAT for Blacks and Whites in Thirteen Integrated Institutions,” 8 J. Educ. Measurement 245-251 (1971).

[13] The theoretical framework underlying the use of noncognitive variables in admissions decisions is based on research which indicates that the GPA and LSAT are incomplete instruments on which to make a full evaluation of a candidate, especially a nontraditional applicant. See Section III, Part B, footnotes 7-12, supra.

Moreover, studies support the proposition that if traditional predictors are used, there must be separate equations or cutoffs for each subgroup to achieve optimum validity. Other studies supporting the differential regression equations for race/sex subgroups include: A.S. Farver, W.E. Sedlacek, and G.C. Brooks, Jr., “Longitudinal Predictions of University Grades for Blacks and Whites,” Measurement and Evaluation in Guidance, 7 (1974): 243-250.

[14]As used in this formula, the symbols connote the following: NC represents the noncognitive score. C represents the cognitive score (GPA and LSAT) with CD representing the cultural diversity score. AS denotes the total applicant score on the basis of which offers are made.

[15] “As noted in Justice Powell’s opinion in Bakke vs. University of California [sic], ‘ …we do not compel the University to utilize only “the highest objective academic credentials” as the criterion for admission.’ Institutions do in fact select students at least in part on a variety of other grounds. For example, institutions regularly consider nonacademic characteristics in order to:

--select students likely to exhibit outstanding performance on criteria other than traditional grades (e.g., leadership, scientific creativity, artistic achievement)

--select students who are more likely to persist to a degree

--achieve reasonable representation of important demographic groups (e.g., sex, race)…

--select students who are related to important sources of support to the institution (e.g., relatives of alumni, faculty, or benefactors)”

Hunter M. Breland and Warren W. Willingham, “Presonal Qualities in Admissions,” mimeographed, a cooperative research and development project between the College Board and the Educational Testing Service (Princeton, N.J., December 1978), pp. 5-6.

[16] See Section V, supra, for the Nondiscrimination Policy Interpretation,” governing the interpretation of Title VI in light of Bakke.

[17] William E. Sedlacek, “A Cultural Diversity Selection Model for Law School Admissions,” prepared for the Mexican American Legal Defense and Educational Fund, mimeographed (College Park, Maryland: University of Maryland, 1 October 1979).

[18] Studies corroborate the utility of these factors for student assessment; see footnote 12, supra. Please refer to Addendum A which immediately follows this model for a description of the weighting of the scale values for each of the components making up the noncognitive score.

[19] T score is a standardized score where the mean is set at 50 and the standard deviation is 10. For instance, applicant A had a GPA of 3.6. If the pool of applicants to a particular school had a mean GPA of 3.3 with a standard deviation of 0.2, we would set 3.3 equal to 50 and each unit of 0.2 above or below the mean equal to 10. Thus a GPA of 3.5 would be 1.5 standard deviations, or 0.3 above the mean, so 3.3 + 0.3 = 3.6. T scores allow for scores based on different scales to be compared, added, subtracted, etc. The scoring system employed by many standardized tests such as the SAT and the LSAT is similar to the T score in that the mean in set at 500 and the standard deviation at 100. A more complete discussion of T scores can be found in F.G. Brown, Principles of Educational and Psychological Testing, 2nd ed. (New York: Holt, Rinehart, and Winston, 1976).

 


[X] Close Window

2005 William Sedlacek