Abstract
Novel communications networks were employed in an attempt to minimize some objections to the Allport and Postman (1946) method of rumor analysis and to test their construct regarding the intensity of rumor, R ~ i × a. The first phase involved the rotation of one of two concentric circles; a sliding parallel Lines technique was used in the second phase. Segments of the information to be exchanged originated from a number of different Ss and were passed on in a multilateral rather than unilateral fashion. Three levels of i and five levels of a were manipulated Scores for both leveling and extraneous details were derived. Extraneous detail scores were found to be unreliable. Obtained leveling score for curves were U-shaped rather than ascending monotonic functions as would be expected from the model. Results were discussed in terms of the possible effects of difficulty and attentional factors.
Get full access to this article
View all access options for this article.