The debate about black Minnesota Multiphasic Personality Inventories (MMPIs) and white norms was renewed recently. Pritchard and Rosenblatt, revewing 16 studies, claimed to find no evidence that the MMPI is racially biased. Gynther and Green, reviewing 40 studies, concluded that racial bias in the MMPI varies as a function of sampling, with bias more likely when normal blacks are compared with normal whites but not when abnormal groups are compared. No study yet published has compared black alcoholics with white alcoholics, although problem drinking is classified as a major diagnostic problem. We found that black alcoholics scored significantly lower on a few MMPI scales (e.g., Hysteria), but profile differences were not clinically meaningful, particularly when the covariates of age, education, and socioeconomic status were controlled. Blacks differed appreciably on the Family Environment Scale (FES), however, rating social climates more positively than whites. Findings that blacks obtained negligible differences on a wholly white normed test (the MMPI), contrasted with the significant differences and better adjustment shown by blacks over whites on a test normed with minority-group representation (the FES), interject new considerations into the issue of racial bias. Whereas nonsignificant differences would not appear to support the contention of racial bias on the MMPI, FES results question whether such nonsignificance actually rules out test bias.The debate about using white norms for eluded that degrees of ethnic bias in the interpreting blacks' performances on the MMPI may vary in part as a function of the Minnesota Multiphasic Personality Inven-kinds of samples being compared. Among tory (MMPI) was advanced recently in a normal subjects, blacks tend to score 5-10 lively exchange between Pritchard and Ro-T points higher on the MMPI, particularly senblatt (1980a, 1980b) and Gynther and on the Validity (F), Schizophrenia (Sc), and Green (1980). Pritchard and Rosenblatt Hypomania (Ma) scales. When subjects are concluded that there was no evidence to sup-psychiatric patients, however, differences port the notion of differential performance are less pronounced and more equivocal, peron the MMPI between blacks and whites, haps because profile elevations produced by whereas Gynther and Green (1980) con-maladjustment obscure racial differences.Moreover, when samples consist of heroin This study was conducted with funds furnished by the addicts, the direction of differences is re-General Medical Research Service and the Alcohol versed so that whites score higher than Dependence Treatment Program of the Veterans Adblackg (p enk & R 0 binowitz, 1974).'^S^^SSS^SS^ for his A . J» contradictory nature of black-white processing of MMPI protocols used in this study. Deep differences among normal and abnormal appreciation is expressed to Frank Harris for his un-groups-persisting even after such potenflagging labors at all stages of data reduction and pro-tially confounding variables as age, intelli-^RSiestsforreprintsshould...

The hypothesis that among clinical samples of substance abusers blacks score lower than whites on the Minnesota Multiphasic Personality Inventory (MMPI) was supported when comparing 494 white and 159 black male veterans seeking treatment for polydrug abuse. Blacks scored lower on the Depression, Hysteria, Psychopathic Deviate, and Psychasthenia scales when age, education, socioeconomic status, and intelligence were controlled. The findings do not support the notion of ethnic bias in the MMPI. Rather, the results underscore the need for identifying moderator variables that differentially interact to produce in comparison to whites, lower black profiles among substance abusers but higher black profiles among psychiatric patients.