Abstract
The purpose of this study was to determine the adequacy of a simple method for developing parallel or equivalent forms of interest and personality questionnaires which have multiple keys. The method involves the pairing of items for popularity levels, phi coefficients, and item content. The use of this method yielded two 16-subtest forms, each subtest containing 12 items. Using Votaw's test of compound symmetry (7), a stringent statistical criterion which tests simultaneously for equivalence of means, variances, and covariances with an external criterion, it was found that only three of the 16 subtests did not meet this test of equivalence. Further analyses determined that in all three cases there were statistically significantly different means. In one of these cases the variances were also statistically significantly different. It is concluded that the simplified method outlined in this article can be used to develop statistically equivalent subtests for interest and personality questionnaires.
Get full access to this article
View all access options for this article.