| The Multitrait-Multimethod Matrix by Campbell and Fiske (1959) provided a foundation for evaluating construct validity. Unfortunately, efforts to apply statistical techniques to analyze the matrix have been unsuccessful. As a result, Sawilowsky (2002) created a quick, distribution-free statistic as a solution to analyzing the matrix. However, a question regarding the robustness of the statistic to departures from independence as well as the power properties of this test required further examination. The purpose of this study was to examine the Type I error rate when independence is violated and the power properties of the Sawilowsky I test. In addition, a modified version of the Sawilowsky I test was developed and examined in terms of its robustness when independence has been violated and its power. Ad hoc critical values were determined to improve the statistical power of this approach to analyzing the Multitrait-Multimethod Matrix. |
