Faculty Proceedings, Presentations, Speeches and Lectures

Title

Apparently Abnormal IQ-Memory Differences in the Normal Population

Event Location / Date(s)

San Diego, CA / October 16-19, 2013

Document Type

Conference Proceeding

Presentation Date

10-16-2013

Conference Name / Publication Title

33rd Annual Conference of the National Academy of Neuropsychology

Description

Abstract

Objective: Comparison between IQ and memory ability can identify memory impairment caused by neurological disorders. WMS-IV interpretation involves comparison of Memory Indexes to the WAIS-IV General Ability and other Indexes. Published base rates in the standardization sample suggest that 15-point differences between pairs of WAIS-IV/WMS-IV Index scores are normally uncommon, but these data refer to one specific comparison rather than multiple possible comparisons among Indexes. This study reports normative data for the incidence of multiple IQ-Memory Index discrepancies.

Method: The normal incidence of IQ-Memory Index score differences were calculated using Monte Carlo simulations. Correlations among Indexes in the normative sample were used to recreate the distributions of expected scores and subsequently validated against standardization data. The frequency of observed IQ/Memory Index discrepancies was then determined. Results: One or more 15-point IQ > Memory Index difference occurred in 48% of the normal population, and 20-point Index differences occurred in 30% of the standardization sample. 31% had at least two Index pairs that were 15 points discrepant, and 17% had a difference of 20 points between two or more Index pairs. The percentage of the population with differences between specific individual IQ/Memory Indexes did not differ significantly when Monte Carlo results were compared with published standardization data.

Conclusion: WAIS-IV/WMS-IV Index score discrepancies are normally common and therefore lack clinical meaning when multiple such comparisons are made. Empirically supported a priori interpretive hypotheses are necessary to reduce the false-positive diagnoses associated with potentially numerous Index comparisons. Monte Carlo simulation accurately predicts these observed Index score differences.

This document is currently not available here.

COinS