WMS-R and MAS correlations in a neuropsychological population

The present study was an attempt to examine the relationship between the Wechsler Memory Scale-Revised (WMS-R; Wechsler, 1987) and the Memory Assessment Scale (MAS; Williams, 1991). The sample consisted of 51 patients referred for neuropsychological examination. Average age was 55.29 (SD = 20.94). Average education was 11.94 years (SD = 3.18). Average Full Scale IQ was 94.75 (SD = 15.44). The results showed that no MAS index correlated higher than.60 with any of the WMS-R indices. There was minimal specific prediction between instruments across each memory domain. Agreement between the tests clinically was low as well. Corrections for such factors as age, education, intelligence or diagnosis lowered the relationships between the tests although not significantly. Corrections to the correlations for the inherent reliability of each test increased correlations between the tests but still resulted in a maximum common variance of 56% (for the visual measures) down to 40% (for General Memory). These results are consistent with the argument that "general memory" is not as useful a construct as that of "general intelligence" and that these tests of general memory measure different underlying constructs. This has important impact on how we conceive and report memory test results.