Equivalence of computerized and traditional MMPI administration with substance abusers

摘要

A Latin Squares design was utilized to test the equivalence of computerized versus paper-and-pencil MMPI administration. Sixty-eight alcoholics and seven drug abusers completed both formats. Multivariate analyses of the 14 MMPI scales yielded no significant differences between formats and test administration order. Univariate analyses of scale scores yielded a significant difference between formats for the Cannot Say scale only. This was the result of eliminating the Cannot Say response option with the computer format. Test-retest reliabilities and T-score profiles were also found to be equivalent. Results are discussed in terms of the validity for using computerized MMPI administration in clinical populations.