A simulation study was conducted to examine the efficacy of conditional analysis of variance (ANOVA) methods where the initial homogeneity of variance screening leads to the choice between the ANOVA F test and robust ANOVA methods. Type I error control and statistical power were investigated under various conditions.
Wang, Y., Pham, T., Nguyen, D., Kim, E. S., Chen, Y.-H., Kromrey, J., ... Yin, Y. (2018). Evaluating the efficacy of conditional analysis of variance under heterogeneity and non-normality. Journal of Modern Applied Statistical Methods, 17(2), eP2701. doi: 10.22237/jmasm/1555340224