Abstract
Human Stimulus Generalization of Response Latency
R. Howard
Studies using response rate and choice measures have suggested that human stimulus generalization is categorical: Subjects place new stimuli presented into categories, responding alike to all stimuli within each category. This categorical pattern might, however, have been obtained instead of decremental generalization gradients because response rate and choice are insensitive measures of human stimulus generalization. Perhaps effects of distance of new stimuli from training stimuli along a dimension occur only at the onset of test stimulus presentations. Humans may take progressively longer to decide whether to respond as distance from a training stimulus increases, resulting in incremental gradients of response latency. The present study tested this possibility. Stimuli were presented in discrete trials: University students made one response per trial for occasional monetary reinforcers. After acquiring intradimensional occupational status discriminations, subjects were tested for generalization along the status dimension. Subjects typically showed step-functions with the latency measure. The results support the hypothesis that human stimulus generalization is categorical.