Tuesday,
January 31, 2012 - 12:30
Schmidt Lecture Hall
Prof. Asher Koriat
Institute of Information Processing and Decision Making, University of Haifa
How do people monitor the correctness of their answers and judgments? A self-consistency model is proposed for the basis of confidence judgments and their accuracy. The model assumes that the process underlying subjective confidence in general-knowledge questions and perceptual judgments has much in common with that underlying statistical inference about the outside world. Participants behave like intuitive statisticians who attempt to reach a conclusion about a population on the basis of a small sample of observations. Subjective confidence is based on the sampling of clues from memory, and represent an assessment of the likelihood that a new sample will yield the same decision. Results consistent with the model were obtained across several two-alternative forced-choice tasks. The model explains some of the basic observations about subjective confidence and generates new predictions.
