Homogeneity of Variance Assumption: An Empirical Investigation of 14 Approaches in One-Factor Analysis of Variance Models

• In Event: 16.057 - Roundtable Session 5
  In Roundtable Session: 16.057-4 - Missing Data and Investigations of Statistical Assumptions

Abstract

Variance homogeneity is a critical assumption in an analysis of variance or ANOVA, as violations of this assumption lead to distorted Type I error rates, loss of power, and spurious interpretations. Due to previous research on tests proposed for the tenability of variance homogeneity suggesting minimal consensus of the tests’ robustness when variances are heterogeneous and normality does not hold, this simulation study evaluated the performance of fourteen tests for the homogeneity of variance assumption in one-way ANOVA models in terms of Type I error rates and statistical power, under several experimental conditions across normal and non-normal distributions. Results from this study provide researchers with empirical information to guide the selection of a valid test for assessing this critical assumption.

Authors

• Yan Wang, University of South Florida

• Patricia Rodriguez de Gil, University of South Florida

• Aarti P. Bellara, University of South Florida

• Yi-Hsin Chen, University of South Florida

• Tyler Hicks, University of South Florida

• Harold Holmes, University of South Florida

• Eun Sook Kim, University of South Florida

• Jeffrey D. Kromrey, University of South Florida

• Isaac Li, University of South Florida

• Diep Thi Nguyen, University of South Florida

• Thanh Vinh Pham, University of South Florida

• Jeanine Romano, American Board of Pathology