Real -time measurement of dissimulation on self -report measures

Holden (1992) has developed a theory of the underlying process of dissimulation based on response latencies to questionnaire items. Specifically, he stipulates that respondents who are lying or malingering should endorse items that fit their made up schema more quickly than items for which they are aschematic and reject items that fit their made up schema more slowly than items for which they are aschematic. Such respondents will probably reject some items that fit their made-up schema so as not to come across as too good or too bad. He found that it is the rejection of the items that fit the schema, which he also refers to as incongruent responses, that discriminates between someone who is actually lying and someone who is faking lying. Those who did endorse incongruent responses as part of a faked schema took longer to respond than those who were actually lying. This study attempted to replicate and extend Holden's findings by combining response latency, which focuses on process, rather than content with adaptive testing which focuses completely on content. We tested this through the creation of our own personnel questionnaire, which was administered to undergraduate and graduate students. Results revealed a pattern in the types of questions that were able to induce dissimulation. Questions requiring responses that are not easily documented were the most susceptible to dissimulation based on the percentage of change (e.g. Have you ever used a fake illness or injury as a reason for absence from school or work?). However, it was the questions pertaining to specific skills that appeared to be the most likely to discriminate between those participants asked to respond honestly and those participants asked to fake their responses in order to "get the job."*. *Originally published in DAI Vol. 63, No. 8. Republished here with corrected author name.