A cross-validation of two short forms of the WAIS-R in a geriatric sample suspected of dementia

Abstract

This study investigated two short forms of the Wechsler Adult Intelligence Scale‐Revised reported by Silverstein (1982). The short form IQs were calculated from test data of 42 geriatric patients with suspected dementia. The short form IQs were compared to the actual Full Scale IQs. Correlations between the two and four subtest short forms and the actual IQs were .93 and .96, respectively. In addition, the two subtest short forms correctly classified 67% of the sample on the basis of Wechsler's intelligence categories, and the four subtest short forms correctly classified 83% of the sample. It was concluded that these short forms should not be used when a precise IQ assessment is required; however, they may be useful for various screening purposes.

Intellectual Decline in Schizophrenia: Evidence from a Prospective Birth Cohort 28 Year Follow-up Study

“…It is of interest that the estimated adult IQ level of this community sample of patients with schizophrenia was 91.2. While the IQ estimate used in this investigation (using the Vocabulary and Block Design subtests, Brooker & Cyr, 1986) is known to overestimate Full Scale IQ by about three points (Margolis, Taylor, & Greenlief, 1986;Roth, Hughes, Monkowski & Crosson, 1984;Ryan, Larsen, & Prifitera, 1983;Thompson, Howard, & Anderson, 1986), it is very close to the average IQ level of approximately 90 reported by Aylward, Walker, and Bettes (1984) in their meta-analysis, and in the average IQ derived from our seven previous studies of neuropsychological functioning in schizophrenia . The results from this combination of clinical, twin (Goldberg et al, 1995), and community samples suggests that the average IQ of approximately 90 is a good estimate of the population mean of persons with chronic schizophrenia.…”

Validity of brief intelligence tests for patients with Alzheimer's disease

“…Misawa‐two and Misawa‐three sometimes estimated the IQ with relatively large differences from the actual IQ. In addition, as with the other studies, 20–24 the IQ based on the WSF were ‘theoretical’ and not derived from repeated administration of the JWAIS‐R. Further investigations are necessary in order to determine whether the WSF truly yield a perfect correlation with the entire JWAIS‐R.…”

“…5,19 Moreover, it should be noted that most investigations performed thus far, both domestic and foreign, evaluated test quality using a mixed sample of patients with diverse diseases. 20–24 Intelligence comprises various aspects, and previous studies have revealed that the vulnerable aspects are mainly determined by the patient's disease. 25–27 Therefore, this type of study should be conducted separately for each disease.…”

Dyadic Short Forms of the Wechsler Adult Intelligence Scale–III

“…The correlation coefficients in the cross-validation sample between actual and derived FSIQs (rs ranging from .90 to .92) were similar to those reported for other dyadic short forms using the WAIS-R (i.e., rs ranging from .86 to .93) in previous studies (Benedict et al, 1992;Boone, 1990;Hoffman & Nelson, 1988;Margolis, Taylor, & Greenlief, 1986;Robiner, Dossa, & O'Dowd, 1988;Roth, Hughes, Monkowski, & Crosson, 1984;Ryan et al, 1983;Satterfield et al, 1994). The dyad I/BD yielded an estimated FSIQ within 5 points of actual FSIQ in 74% of cross-validation cases, whereas the V/MR dyad estimated FSIQ within 5 points of actual FSIQ in only 49% of cases.…”