Abstract

Equivalency testing, a statistical method often used in biostatistics to determine the equivalence of 2 experimental drugs, is introduced to social scientists. Examples of equivalency testing are offered, and the usefulness of the method to the social scientist is discussed.

PIP:

Equivalency testing. currently used by biostatisticians to determine whether 2 drugs have an equally effective outcome, offers social scientists a method for analysis of various interventions. This is the method of choice when researchers are able to specify a small, non-zero difference between 2 treatments that can serve to define an equivalence interval around a difference of zero. The goal is to reject the null hypothesis and provide evidence for an alternative hypothesis that the difference between 2 groups is smaller than that specified in the null hypothesis. The process involves the definition of equivalency and the performance of 2 simultaneous 1-sided tests of the hypothesis. The Westlake confidence interval approach to equivalency is applied, for illustrative purposes, to data generated by 3 social scientific studies. In the 1st example, the procedure was used to determine whether the mean Minnesota Multiphasic Personality Inventory scores of drug addicted subjects were within 10% of the means of scores of alcoholic subjects. The 2nd study represented an attempt to assess the relative effectiveness of several therapeutic approaches (psychotherapy alone, psychotherapy in combination with antidepressant medication, behavioral approaches, cognitive therapy) to the treatment of depression. Finally, the 3rd study compared baseline characteristic equivalence among nonpregnant black adolescents, pregnant adolescents who carried to term, and those who aborted. It is noted that this method offers a means of evaluating published negative findings to determine whether a reasonable definition of equivalence exists. Moreover, equivalency testing can be used in the social sciences to justify the pooling of study groups.