Abstract
Attitude scale construction techniques give, at best, only indirect attention to the prediction capabilities of individual items. Although conventional scaling procedures can be augmented by an item analysis sequence that uses known-affiliation groups, many uncertainties surround this approach. Laboratory analogues of the social processes central to a scale are recommended as a source of data for item analysis. The group autokinetic experiment is suggested as an example of a laboratory analogue of the social processes central to a scale intending to measure attitudes toward dissent.
Get full access to this article
View all access options for this article.
