Abstract
Three experiments tested the hypothesis that response selection skill involves associations between individual stimulus features and responses. The Orientation group in Experiments 1 and 2 first practiced responding to the orientation of a line stimulus while ignoring its color, and the Color group practiced responding to the color of the line while disregarding its orientation. When in the ensuing test conditions the Orientation group responded to the color of the line, RTs and errors increased when the then irrelevant line orientation was inconsistent with practice. This confirmed that during practice, Orientation participants had developed orientation feature-response associations they could not fully inhibit. Yet, evidence for color feature-response associations was not observed in the Color group. This was attributed to orientation identification being faster than color identification, even after having practiced responding to colors. Experiment 3 involved practicing to three line stimuli with unique orientation and color combinations. It showed evidence for the independent development of orientation feature-response associations and color feature-response associations. Together, these results indicate that the typical RT reduction with practice in response selection tasks is caused in part by the capacity of participants to learn selecting responses on the basis of the stimulus feature that becomes available first.