Add new comment

Question: "It struck me that when the validity of the SEG is checked, that with a combination of a P-value > 0.05 in the ANOVA (= no significant differences between groups), but where the homoscedasticity condition is not met, BWStat still proceeds to a favorable conclusion: "No significant differences between workers, individual compliance test not required."
Is it correct to assume that BWstat relies exclusively on the result of the ANOVA and not on the combination of both the ANOVA and the Levene's test to make this statement (conclusion validity SEG)?"

Answer: "This is correct indeed. The statement about the differences between workers (or the validity of the SEG) should be based on the differences of the means (ANOVA) as well as the differences of the variances (homoscedasticity). BWStat performs both tests and shows both results. However, the homoscedasticity is not taken into account in the conclusion.

Very well noticed. Thank you for reporting this.
This should be the subject of a bug fix.