Many researchers draw the conclusion that one independent variable has more impact than another without testing the null hypothesis that their impact is equal. This paper presents and recommends a technique for testing the relative magnitude of effects, rather than basing conclusions solely on descriptive statistics. (Author/BW)