A Monte Carlo simulation study was conducted to examine outliers’ influence on Type I error rates in ANOVA and Welch tests, and the effectiveness of two outlier accommodation methods: nonparametric rank based method and Winsorizing. The study also offered practical recommendations regarding outlier handling with different sample sizes and number of outliers.
Liao, Hongjing; Li, Yanju; and Brooks, Gordon
"Outlier Impact and Accommodation Methods: Multiple Comparisons of Type I Error Rates,"
Journal of Modern Applied Statistical Methods: Vol. 15
, Article 23.
Available at: http://digitalcommons.wayne.edu/jmasm/vol15/iss1/23