Abstract
Research using attitude / behavior correlation to predict behavior of subjects in situa tions significant to business communication generally produces low to moderate cor relations that explain relatively small amounts of the variance among variables. Researchers sometimes depend on large sample sizes to produce variation measurable as statistically significant. Magnitude estimation scaling, a method tracing its origin to the field of psychophysics, has had some interesting results in social science re search. This article critically analyzes the technique for its potential use in business communication research. The result finds that the 12-15 percent increase in explained variance by magnitude estimation over categorical scaling methods may be useful in theory building, but may not be sufficient to justify the added expense necessary to en sure data integrity for applied business communication research.
Get full access to this article
View all access options for this article.
