Electrophysiological and behavioural indices of simulated recognition memory impairment

This thesis examined the use of event-related potentials as a means of detecting feigned recognition memory impairment. In seven studies, undergraduate students were instructed either to complete a recognition memory test to the best of their abilities, or to simulate accident�related memory loss. These studies extended previous research by investigating electrophysiological differences between the control and malingering tasks (1) for stimuli which differed in linguistic frequency, (2) in tests which varied the format of word presentation, (3) during the initial encoding of the stimuli, and (4) using ERP components not previously considered in studies investigating the detection of malingering. The main results were that simulating individuals appear to use more active or additional cognitive processing during task performance compared to those who respond honestly, with this enhanced effort reflected in an ERP effect indexing earlier recognition of previously-studied words. This earlier recognition, considered to be the result of more elaborative or efficient encoding of the stimuli, was most evident in the easier forms of the recognition test and in malingerers who presented a more believable profile of impairment. The malingerers also demonstrated different electrophysiological responses to items that were incorrectly classified, and a pattern of response latency suggesting that they concealed recognition of previously-studied items. Overall, the results indicate that the simulation of amnesia on a recognition memory task involves qualitatively different processing of the word stimuli, which may be detected in the waveforms and more covert patterns of behaviour of these individuals.