Interpreting clinical evidence of malingering: A Bayesian perspective

Customary ways of reporting on or testifying about malingering have shortco mings. Stating an opinion "with reasonable medical certainty" tells fact-fi nders little about how much confidence the opinion deserves; stating that a n individual's behavior is similar to that of known malingerers does not co nvey the information that fact-finders really need to know, which is the li kelihood that the evaluee in question is a malingerer, given the evaluator' s findings. Mossman and Hart (Mossman D, Hart KJ: Presenting evidence of ma lingering to courts: insights from decision theory. Behav Sci Law 14:271-91 , 1996) recommend that mental health professionals address this problem by using Bayes' theorem to interpret test data from evaluations. However, thes e authors do not discuss the use of evidence obtained during interviews and from other clinical contexts, nor do they describe a method for quantifyin g imprecision in Bayesian probabilities. This article provides examples of how forensic evaluators might use a Bayesian perspective to interpret clini cal indicia of malingering observed during evaluations of adjudicatory comp etence. The article discusses sources of imprecision in Bayesian posterior probabilities, describes a method for characterizing that imprecision using confidence intervals, and then presents several sample calculations that i llustrate how interview findings change the likelihood of malingering. The article also discusses the implications of the Bayesian approach for forens ic evaluations and for future research on malingered incompetence.