1960, 16, 421. 4.-WOLFSON, W. and BACHELIS, L. An abbreviated form of the WAIS verbal scale. J . elin. Psychol., T h i s latter finding is significant in that schizophrenics re resented roughly 9270 of this group. The authors are presently investigating this problem. T i e hypothesis of performance change as a function of change in the continuity of item difficulty has been advanced.

PROBLEM I n a recent study, Satz and Mogel@) constructed an abbreviated form of the WAIS in which all scales and subtests were represented. The attempt was to obviate a shortcoming in previous abbreviated forms which used only those combinations of subtests (usually four) that provided maximal correlations with Full Scale IQ(4). Such procedures have limitations, except for research application, by virtue of their emphasis on single I& indices. By representing all scales and subtests in the recent abbreviated form, the clinician is provided with the complete breadth of intellectual functions as measured by the WAIS. This was done by using selected items on nine of the WAIS subtests while leaving the Digit Span and Digit Symbol subtests intact. Correlations with the respective WAIS Scales were: Verbal I& = .99, Performance IQ = .97, and Full Scale I& = .99. I t was further shown that this abbreviated form correlates highly with the standard Full Scale WAIS regardless of intellectual level or diagnostic classification.One methodological problem remains, however, in that a rescoring procedure (as in many of the previous WAIS short forms) was used in validating the present abbreviated form. For example, the assumption is often implied that a subject would perform the same regardless of whether the short form were administered separately or merely rescored from the standard form. In the present abbreviated form, wherein less than 50% of the standard WAIS items are used, factors such as fatigue, set, time, and change in the continuity of item difficulty could have obvious effects on performance. Using a test-retest design, this study attempts to evaluate the influence of these or other factors. This was done by presenting the abbreviated items separately for criterion comparison with the standard WAIS form. METHODSubjects. Files from a large N P Veterans Administration Hospital were drawn for Ss who had previous test-retest scores on the WAIS. This sample (N = 30) represented the control group and consisted of 18 schizophrenics (Mean I& = 83.33)

Satz and Mogel (1962) and Mogel and Satz (1963) devised an abbreviated form of the Wechsler Adult Intelligence Scale (WAIS) which satisfied both clinical usefulness and validity. The procedure involved the selection of items (46%) from 9 of the 11 WAIS subtests, leaving Digit Span and Digit Symbol unchanged. The advantages of this type of short form were twofold: (a) all subtests were represented and (6) considerable time in administration was saved. High correlations between the abbreviated and standard forms were found regardless of intellectual level or diagnostic classification. In a recent study by Yudin (1966), it was suggested that the Satz-Mogel abbreviation was applicable to the Wechsler Intelligence Scale for Children (WISC). Using an emotionally disturbed sample of children, Yudin found extremely high correlations between the standard and abbreviated WISC subtests and scales for different age groups. Intellectual level, however, was shown to lower the magnitude of these correlations, particularly in the upper IQ ranges. If this finding were true, it would limit the usefulness of the abbreviated procedure. The present study was therefore designed to test the replicability of Yudin's findings, with this short form, on a new sample of emotionally disturbed children.The WISC records of ISO emotionally disturbed children were rescored according to the procedure outlined by Yudin. The correction factor, however, was not employed. The total group ranged in age from S years, 11 months, to 15 years, 11 months, with a mean of 12 years, 3 months. The Full Scale IQ of the group ranged from 46 to 137, with a mean of 96.11 and a standard deviation of 17.60. These subject char-1 An extended report of this study may be obtained without charge by Paul Satz, Neuropsychol-