The British Journal of Psychiatry
Sensible hypothesis testing in deluded, depressed and normal subjects.
R P Bentall, H F Young


BACKGROUND Previous research has indicated that deluded patients may experience difficulties when testing hypotheses. In this study, hypothesis-testing strategies were assessed in patients with persecutory delusions, depressed patients and normal controls.

METHOD Subjects were presented problem items describing typical everyday situations with either positive or negative outcomes and were required to choose strategies to prove that one of three variables was responsible for the outcomes.

RESULTS Consistent with previous research into sensible reasoning, subjects chose to manipulate the variable hypothesised to be responsible for the outcome (disconfirmation strategy) more when the outcome was negative than when it was positive, and chose to manipulate the remaining variables (confirmation strategy) more when the outcome was positive. No group differences were observed.

CONCLUSIONS No evidence was found of abnormal hypothesis-testing strategies in deluded patients.