Using a four-choice task, this study was designed to test whether: 1. the order of RT that is predicted by the model could be confirmed, i.e. RT was fastest for congruent mappings, slowest for incongruent mappings, and in between for neutral; and 2. in the case of ensembles with S-R overlap, can the automatic activation of the corresponding response element by the presentation of a stimulus element, as is postulated by the model, be demonstrated.
Result 1. The results confirmed the order of RT’s as predicted by the model.
Result 2. By reducing the time available to process and execute the response, thus increasing the likelihood of errors, the results showed that when the mapping was incongruent most of the erroneous responses were response that would have been correct had the mapping been congruent, thus demonstrating the activation of the response that would have been correct, had the mapping been different. However, by reducing the time available to abort, reprogram, etc. which is called for when the mapping is incongruent (ergo generating errors), automatic activation of the response was verified. Thus, this study confirmed one prediction, and verified an important assumption of the model.