Abstract
Following 10 no knowledge of results (KR) trials Ss were trained for 40 trials in a line-drawing task under conditions of immediate, delayed, or no KR These training trials were followed by 40 trials with no KR (extinction) Analysis of the type of error made during extinction (overshooting or undershooting the correct 3-in. line) revealed that the type of response depended on the training conditions (immediate vs. delayed) and original response bias of S. Delayed-KR resulted in an increased frequency of errors of the same type as the original response bias. Immediate KR resulted in a tendency to make errors in the direction opposite the original response bias. The interaction between delay of KR and S''s response bias argues for the analysis of S''s response bias in future experiments.