Utility of half-credit scoring of Russell's revision of the Wechsler Memory Scale.

Abstract

This study examined interrater reliability of Russell's revision of the Wechsler Memory Scale logical memory using an explicit scoring method in a male medical/psychiatric population. Secondly, it examined whether the ratio of full to half-credit scores can be used to differentiate unimpaired patients from patients with documented cerebral dysfunction. Twenty-five unimpaired and 25 brain-impaired patients matched for age and education were used. Pearson correlations indicate excellent interrater reliability. t-tests indicate that the groups were not significantly different from the ratio of full- to half-credit responses on immediate memory, but were significantly different on delayed memory. A posteriori analyses indicated that a small ratio score (less than 3) on delayed recall classified patients with modest success. Results indicated that brain-impaired and unimpaired patients not only differ in regard to level of recall performance, but also differ in regard to pattern of recall performance. Cross-validation of these results is suggested.