Search

Factor analyses of recall measures by King have consistently resulted in a two-factor solution with the first dimension being quantitativeand the second, organizationalin nature. To test the generality of the serial position with respect to connected discourse, an eight-sentence passage summarizing the arguments, pro and con, of having a female prime minister was written. It was presented individually to 48 men and 48 women, with the sentences interchanged around conjunctions to control for position effects. While sex differences and the sex × position interactions were not significant, combined data supported the serial-position effect along both the quantitative and organizational dimensions. The organization of the recalls using both dependent accuracy measures was described in terms of positively decelerating functions. Analyses of the recall of sex × pro-con arguments yielded no significant differences on this presumably sex-linked topic with either dimensional measure.

Seventy-eight high school students took part in an opinion change experiment using the topic “use of nuclear weapons.” After reading opposing persuasive communications to Ss “contiguously” prior to the presentation of a reinforcing event, subsequent opinion change was found to be statistically significant and in the direction of the one communication (either pro or con) that was more proximal to the reinforcement. This finding supports an hypothesis derived from Thorndike's earlier Law of Effect. From research on the Spread of Effect, a similar occurrence was hypothesized for the situation where communications followed reinforcement. These latter results were, however, not statistically significant, and the second hypothesis was rejected.