Psychological Assessment: A Journal of Consulting and Clinical Psychology 1990, VW. 2, No. 2,198-201 Copyright 1990 by the American Psychological Association, Inc. 1040-3590/90/S00.75 WMS-R Validation of the Wonderlic Personnel Test as a Brief Intelligence Measure in a Psychiatric Sample Keith A. Hawkins, Stephen V. Faraone, John R. Pepple, Larry J. Seidman, and Ming T. Tsuang Brockton/West Roxbury Veterans Administration Medical Center Section of Psychiatric Epidemiology and Genetics, Department of Psychiatry, Harvard Medical School and Massachusetts Mental Health Center Although there have been several reports of high correlations between Wonderlic Personnel Test scores and Wechsler Adult Intelligence Scale (WAIS) Full Scale IQ, findings have been inconsistent in psychiatric samples. Sample differences and differences between the WAIS and the revised WAIS (WAIS-R) were considered likely reasons. In this study of relatively nonagitated but chronically ill psychiatric patients (N - 18), Wonderlic IQ estimation accuracy and Wonderlic/WAIS-R corre- lations were consistent with data previously reported for the WAIS, and generally support the value of the Wonderlic as a highly economical measure of general intelligence. However, the inability of 1 subject to manage the Wonderlic format suggests that severe visuospatial impairment can invalidate this test. The Wonderlic Personnel Test (Wonderlic, 1983) isa 12-min, self-administered, pen-and-paper test that has been reported as correlating highly with Wechsler Adult Intelligence Scale Full Scale IQ (WAIS FSIQ) in normal samples (DodriU, 1981). It possesses test-retest stability over a 5-year period equivalent to that of the WAIS (Dodrill, 1983) and allows for IQ predictions no more than 10 points discrepant from actual WAIS IQ scores in 80% to 90% of cases (Dodrill, 1980, 1981, 1983). Moreover, Dodrill (1981) reported that the predictive validity of the Wonderlic is not compromised by sample characteristics such as age, sex, years of education, and emotional adjustment. Con- sidering the disappointing predictive power of many short or abbreviated measures of intellectual ability (Watson & Klett, 1974, 1975), especially insofar as assessment of the individual case is concerned (Dodrill, 1981), the Wonderlic seems to offer an efficient method of assessing global intellectual capacity. As a self-administered and clerically scored instrument, it may be especially useful for mass or initial screening in both research and clinical settings. However, the Wonderlic and Wechsler Adult Intelligence Scale-Revised (WAIS-R; Wechsler, 1981) correlational data have been less impressive than those reported between the Wonderlic and WAIS. Edinger, Shipley, Watkins, and Hammett (1985) obtained a lower Wonderlic/WAIS-R correlation (r = .75) in a psychiatric inpatient sample than Dodrill and Warner (1988) reported between Wonderlic and WAIS (r = .85 to .91) This research was supported in part by the Veterans Administration's Health Services Research and Development Program, National Insti- tute of Mental Health Grants 1 R01MH41879-01 and I R37MH43518- 01 to Ming T. Tsuang and by a National Alliance for Research on Schizophrenia and Depression Award to Keith A. Hawkins. Correspondence concerning this article should be addressed to Keith A. Hawkins, who is now at Yale University, School of Medicine, 34 Park Street, New Haven, Connecticut 06519. for psychiatric and neurologic samples (consisting of psychiat- ric pseudoseizure, psychotic depressive, seizure disorder, and epileptic groups). Edinger et al. also reported a lower "hit rate" (Wonderlic IQ predictions falling within 10 points of WAIS-R) of 71% and reported that IQ scores derived from the Wonderlic fell within the same WAIS-R clinical category (e.g., dull nor- mal, average) in only 40% of cases. Edinger et al. suggested that the different outcomes might reflect population differences and their use of the WAIS-R IQ as the criterion. DodriU (1981) has consistently used the WAIS, on which his Wonderlic IQ conver- sion table is based. Nevertheless, attempts by Edinger et al. to modify Dodrill's conversion table improved prediction only marginally. They concluded that the Wonderlic should not be used as a test of intelligence with psychiatric patients and ques- tioned the usefulness of the Wonderlic with "medical, surgical, or neurological inpatients or with medical and psychiatric out- patients" (Edinger et al., 1985, p. 939). Dodrill and Warner (1988) have argued that Edinger et al.'s use of Wechsler's (1981) intelligence classifications for test vali- dation is inherently unsatisfactory. For example, using this method, a test will "fail" if it provides an IQ estimate of 109 (Average) in cases in which the WAIS-R IQ is 110 (High aver- age); on this basis, the WAIS itself performs poorly when read- ministered 5 years after initial testing (Dodrill, 1983; Dodrill & Warner, 1988). DodriU and Warner concluded that "we are unable to discourage use of the Wonderlic, at least with popula- tions such as those we sampled" (Dodrill & Warner, 1988, p. 146). Quite apart from methodological considerations raised by Dodrill and Warner (1988), it is important to note that Edinger et al.'s (1985) sample included significant but relatively unspeci- fied diversity (i.e., 22 with schizophrenia, 31 with major affec- tive disorders, 7 with organic brain syndromes, and the remain- ing 40%with "various other disorders"), and they did not detail the extent of disturbance in their patients at time of testing. 198 This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.