Investigating the effects of problem format and task related experience on evidential reasoning

Abstract

Recent studies of human decision-making have criticized traditional decision-making research for using inexperienced participants and "single-shot" judgment tasks with no feedback. These studies, based on naturalistic and adaptive approaches, have suggested that the poor performance typically demonstrated by traditional research is due not to failures on the part of the human, but failures of the empirical studies to test performance in representative situations. In particular, some researchers have studied how experience in an uncertain environment, or a task format that is more representative of naturally-occurring environments, can improve performance on evidential reasoning tasks. This paper describes research designed to test the effect of both differing task formats and environmental experience on performance on such tasks. Participants performed an evidential reasoning task in one of two formats, before and after receiving related experience in the task environment. As expected, we found that task-related experience did improve performance; however, the effect of task format was not consistent with earlier research.