October 7, 2008

Researchers Create Fake Proof Personality Test

Psychologists from the University of Toronto have developed a personality inventory that can predict who will excel in academic and creative domains, even when respondents are trying hard to fake their answers.

Study authors note that personality questionnaires have a long history of predicting real-world performance, but have been plagued by the problem of biased responding. "It's very common for people to try and make themselves look better than they actually are on these questionnaires, especially if they know they are being evaluated," said Jordan B. Peterson, psychology professor at the University of Toronto and co-author of the paper. "This sort of faking can distort the predictive validity of these tests, with significant negative economic consequences. We wanted to develop a measure that could predict real-world performance even in the absence of completely honest responding."

The research findings demonstrate that traditional personality inventories fail to predict performance outcomes when respondents have strong incentive to fake their scores. The new measure, by contrast, retained its ability to predict success, even when respondents were consciously trying to make themselves look good.

"Personality remains an important factor in predicting performance," said Jacob Hirsh, lead author of the paper and a Ph.D. candidate at the University of Toronto. "Trait conscientiousness has consistently emerged as a major predictor of academic success and workplace performance, while trait openness is a good predictor of creative achievement."

Using formulas derived by Frank Schmidt (Iowa U) and John Hunter of (Michigan State), the studies' authors were able to estimate the potential productivity gain associated with using the new measure in a workplace setting. "Because people differ widely in their individual abilities," notes Hirsh, "even a small degree of accuracy in testing can produce significant economic gains." In the present study, the tests were accurate beyond that small degree. In fact, Schmidt and Hunter's formulas indicate that the use of the bias-resistant test over currently available personality assessment methods could result in a productivity gain of 23 per cent per hired employee, when response faking is an issue ($17,000/year per $75,000 of salary). "Potential gains of this magnitude should not be ignored," said Hirsh. "It is very important that the right people be chosen for any competitive position. This questionnaire is a step in the right direction."

The study findings are published in the October edition of the Journal of Research in Personality.


On the Net: