Applying Signal Detection Theory to Contingency Assessment

Abstract

In most studies of contingency assessment participants judge the magnitude of the relationship between cues and outcomes. This judgment is a conflated measure of the participant’s sensitivity to the cue-outcome relationship, and his or her response bias. A psychophysical model (signal detection theory, SDT) can be used to dissect the independent contributions of sensitivity and bias to contingency judgment. Results of an experiment concerning cue-interaction (blocking) illustrate the utility of applying SDT to understanding contingency assessment. Most accounts of such assessment are associative (derived primarily from Pavlovian conditioning experiments with non-human animals). A psychophysical analysis of contingency assessment is not an alternative to such associative accounts. The SDT analysis supplements (not replaces) learning principles with psychophysical principles.