No Opinion or Not Sure?

Publisher Website
Google Scholar
Abstract
The design of the 1956-58-60 American Panel Study provides a means of demonstrating that “no opinion” and “not sure” are not interchangeable responses. There is, however, no strong general tendency for the “no opinion” response to be carried over from one interview to another two years later to a greater extent than can be explained by a latent-class or latent-trait model, although within a single interview there is a strong tendency for different questions to elicit the use of this response in the same way, quite apart from any general factor, such as political involvement, that could similarly affect knowledgeability about diverse issues. The evidence analyzed here is entirely consistent with the conceptualization of response processes as intrinsically probabilistic.

This publication has 0 references indexed in Scilit: