A Theory of Diagnostic Inference. I. Imagination and the Psychophysics of Evidence.
Abstract
Diagnostic inference involves the assessment and generation of causal hypotheses to account for observed outcomes/evidence. The importance of diagnosis for prediction, defining 'relevant' variables, and illuminating the nature of conflicting metaphors in inference is first discussed. Since many diagnostic situations involve conflicting evidence, a model is developed for describing how people assess the likelihood that one of two hypotheses is true on the basis of varying amounts of evidence for each. A central notion is that one compares the evidence at hand with the evidence that 'might have been.' This is modeled via an anchoring and adjusting process where the anchor represents 'what is' and the adjustment is based on imaging a contrast case for comparison. Four aspects of this model are then considered. The relation between evidentiary strength and amount of evidence (the evidence function) is shown to mimic a set of power functions. Moreover, the form of the function implies that people will trade-off relative frequency (p) for amount of evidence (n) at small n; that the absolute amount of evidence effects evidentiary strength independent of p; and that 'over' underweighting' of probabilities decreases as amount of evidence increases.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1982
- Accession Number
- ADA115940
Entities
People
- Hillel J. Einhorn
- Robin M. Hogarth
Organizations
- University of Chicago