Inter-item manipulations of implicit associative responses in verbal-discrimination learning

Abstract

Two experiments were performed to test predictions from frequency theory regarding the role of implicit associative responses (IARs) in verbal-discrimination learning. Although list content was demonstrated to limit generalization, a list employing associatively related right (R) items and unrelated wrong (W) items was consistently found to be most difficult, while lists composed of associatively related wrong (W) items tended to be learned most readily. These results fail to replicate the findings of Ekstrand, Wallace, & Underwood (1966) and are inconsistent with the frequency theory. Methodological variables having possible implications for the obtained results were suggested for further investigation.