Type I Error of Four Pairwise Mean Comparison Procedures Conducted as Protected and Unprotected Tests (original) (raw)

The analysis investigates the Type I error rates of four pairwise mean comparison procedures under both protected and unprotected testing conditions. The study builds on historical concerns regarding the inflation of Type I errors in multiple comparisons, highlighting the inadequacies of traditional methods like the Least Significant Difference Procedure. Various procedures including Dunn-Bonferroni, Dunn-Sidak, Holm, and HSD are evaluated for their error control efficiency, with results indicating a range of performance based on the statistical conditions applied.