Tests were made of the roles of reinforcement and S-R contiguity in paired-associate learning by the use of classical and instrumental conditioning paradigms. Simple comparison in Exp. 1 showed the instrumental procedure to be superior, as predicted. In Exp. 2 the mean delays of reinforcement over successive blocks of 10 correct trias with the classical procedure were imposed on an instrumental group for comparable trials; there was no significant difference in mean speed of response, but the classical group made fewer errors. A second classical group was trained with mean intervals between stimulus and response items matched with those of the instrumental grop; there was no significant difference between the groups in speed of response.
Read full abstract