Memory malingering: evaluating WMT criteria.

Abstract

The Word Memory Test (WMT) has been considered one of the best assessment tools for assessing memory malingering in spite of some reported false-positive results. Our goal was to examine the sensitivity and specificity of Green's 2003 criteria for memory malingering using a Portuguese adaptation of the WMT. We applied the WMT to three groups of participants, including 20 healthy subjects asked to simulate memory impairment, 29 healthy adults asked to produce their best performance, and 21 patients with the diagnosis of mild cognitive impairment (MCI). Using Green's 2003 criteria, based on scores below cutoffs on the easy subtests, 67% of the MCI patients were classified as "poor effort." However, the sensitivity and specificity of the diagnosis could be raised to 95% using an alternative set of criteria, relying on comparisons between easy and hard subtest scores. We conclude that Green's original criteria based on easy subtest scores alone seem to be of low specificity for the diagnosis of memory impairment simulation, but the WMT is a good instrument for identifying simulation if the alternative criteria are applied.