Behavior Research Methods & Instrumentation 1980, Vol. 12 (5),568 Assessingpredictive test bias with the microcomputer LEON D. LARIMER Community Hospital of the Monterey Peninsula Carmel, California 93921 and MARLEY W. WATKINS Deer Valley School District Phoenix, Arizona 85007 There has been rising concern that many of the standardized educational and psychological tests in current use are biased when employed with culturally or economically disadvantaged populations (Golden, 1978). The significance of this issue has been demonstrated by the appointment of an American Psychological Associa- tion committee to investigate the use of standardized testing with disadvantaged students (Cleary, Humphreys, Kendrick, & Wesman, 1975), by a federal mandate pro- hibiting the use of "biased" methods of assessment (The Education for All Handicapped Children Act of 1975, Note 1), and most recently by litigation relative to the use of intelligence tests to place minority school children in special education classes (Larry P. et al. vs. Wilson Riles et al., Note 2). Peterson and Novick (1976), after reviewing a num- ber of test-bias models, proposed a regression-based approach as the most viable method for assessing selec- tion fairness within the predictive domain. Their sug- gestion was concordant with the recommendation offered by Cleary et al. (1975) that components of predictive test fairness are "directly translatable into regression statistics" (p. 25). While not without some debate (Thorndike, 1971), a regression-based approach for analyzing psychological test bias is being cited with increasing frequency in the testing literature (Bossard, Reynolds, & Gutkin, 1980; Reschley & Sabers, 1979; Reynolds & Hartlage, 1979). This paper presents a BASIC microcomputer pro- gram that calculates a regression model test of pre- dictive bias developed by Potthoff (Note 3). The pro- gram provides both simultaneous and separate tests of regression slopes and intercepts for cases in which a single independent variable is used to predict to a single dependent variable. A significant F for the simultaneous test indicates the presence of predictive bias. Separate tests of slopes and intercepts may then be conducted to determine whether the bias in prediction is constant (intercepts differ) or varies relative to the distance of scores from the mean (slopes differ). Description. The program is written in Applesoft BASIC for the Apple II microcomputer. It occupies 4.3K of RAM memory and requires l.5K of additional RAM for every 100 subjects. Variables are mnemonic in character. The program is fully documented to facili- tate adaption to other popular microcomputers. Input consists of number of groups, as well as predictor (X) and criterion (Y) scores for each group member. Output includes group and total means for X and Y, three F ratios (F1 = simultaneous test of slopes and inter- cepts, F2 =separate test of slopes, F3 =separate test of intercepts), and the probability value associated with each F ratio. Availability. A source listing, complete with sample input and output, may be obtained from Leon D. Larimer, Mental Health Center, Community Hospital of the Monterey Peninsula, P.O. Box HH, Carmel, Cali- fornia 93921, upon receipt of a self-addressed envelope with 30 cents postage affixed. REFERENCE NOTES I. The education for all handicapped children act of 1975. Public Law No. 94-142,89 stat. 773. 2. Larry P. et al. vs. Wilson Riles et al. 343 F Supp. 1306 (D.C.N.D. California, June 20, 1972). 3. Potthoff, R. F. Statistical aspects of the problem of biases in psychological tests (Institute of Statistics Mimeo Series No. 479). Chapel Hill, N.C: University of North Carolina at Chapel Hill, Department of Statistics, 1966. REFERENCES BOSSARD, M. D., REYNOLDS, C. R., & GUTKIN, T. B. A regres- sion analysis of test bias on the Stanford-Binet Intelligence Scale for black and white children referred for psychological services. Journal of Clinical Child Psychology, 1980,9,52-54. CLEARY, T. A., HUMPHREYS, L., KENDRICK, S., & WESMAN, A. Educational uses of tests with disadvantaged students. American Psychologist, 1975,30,15-41. GOLDEN, C. J. Cross cultural second order factor structures of the 16PF. Journal of Personality Assessment, 1978, 42, 167-170. PETERSON, J., & NOVICK, M. An evaluation of some models for culture fair selection. Journal of Educational Measurement, 1976,13,3-29. RESCHLEY, D. J., & SABERS, D. L. Analysis of test bias in four groups with the regression definition. Journal of Educational Measurement, 1979,16,1-9. REYNOLDS, C. R., & HARTLAGE, L. Comparison of WISC and WISC-R regression lines for academic prediction with black and white referred children. Journal of Consulting and Clinical Psy- chology, 1979,47,589-591. THORNDIKE, R. L. Concepts of culture fairness. Journal of Edu- cational Measurement, 1971,8,63-70. (Accepted for publication October 10, 1980.) Copyright 1980 Psychonomic Society, Inc. 568 0005-7878/80/050568-01 $00.35/0