The basis of S-R learning: associations between individual stimulus features and responses

Abstract Three experiments are reported testing the hypothesis that response selection skill involves task-dependent associations between a stimulus feature and a response. In the experiments, participants first practiced responding to either the orientation or the color of a line stimulus after which they responded to the other stimulus feature. The question was whether a consistency effect would occur in that response time would be affected by the consistency of the then irrelevant stimulus feature. RTs and errors supported this prediction for stimulus orientation, which confirms development of associations between that feature and the response. There was only limited evidence for color-response associations which could be attributed to the slow identification of the color feature. It appeared that during practice participants could ignore the irrelevant feature but that after practice identification of that feature was mandatory. These results indicate that the typical improvement with practice in selection tasks is caused in part by an association between the most rapidly identified stimulus feature and the following response without the need to wait to identify other stimulus features.