Abstract
Procedures are described for determining the statistical significance of a set of ratings made on bidirectional (or bipolar) scales. A B (bidirectional) coefficient, ranging in value from −1 to +1 and hence providing information on both the magnitude and direction of a set of ratings, is defined. B and its statistical significance can be determined on small or large samples of items or raters. The statistical significance of a mean of B values, computed across items or raters, can also be assessed by applying a normal approximation test. A research illustration involving three types of instructions is given to demonstrate the use of the B coefficient. Tables for determining the two-tailed probabilities associated with selected values of B are also provided.
Get full access to this article
View all access options for this article.