One way to characterize the shift in the attitude of psychologists toward their work that came with the cognitive revolution is as a decline in interest in “the stimulus. ” In the behaviorist period, understanding the effect of presenting a conditioned or unconditioned stimulus, or a reward, was central, and that effect was usually a more or less overt “response. ” In the cognitive era, the focus has shifted to representations and processing, both nonobservable, and in this respect detection theory is a prototypical cognitive enterprise. In this book, we have repeatedly asked how experimental situations are represented internally, and what sorts of decision processes are applied to them. Details of the stimuli being used have been missing, and in our treatment of data they have not been missed.
This story line is too simple, however, and in the next two chapters we look at two important detection theory scripts that offer the stimulus a lead role. Chapter 11, “Adaptive Methods for Estimating Empirical Thresholds, ” summarizes strategies for determining a stimulus whose detectability or discriminability is at a preset level. Finding the stimulus corresponding to a performance level is the inverse of the one-dimension problems in Part I and assumes the same kinds of representations. The stimulus sets to which adaptive methods have most often been applied are simple perceptual ones, although advancing technology is broadening the scope.
Chapter 12, “Components of Sensitivity, ” is an introduction to the use of detection theory in partitioning discriminability between the stimulus and its processing, and among different types of processing. One of the first applications of SDT was in comparing the performance of human listeners to ideal observers, hypothetical processors who make optimal use of the information in the stimulus in making their decisions. In this early work, sensory applications dominated, but more recently the approach has advanced into cognitive and even social domains.