This study mentioned the important of the effect size and its confidence interval. The APA made the rule to ask researchers to report effect size and encouraged researchers reporting the confidence interval if possible. For computing the confidence intervals of effect size, the three approaches, such as parametric, percentile bootstrap, and bootstrap bias-corrected and accelerated confidence interval (BCA), were used and evaluated by previous study. This study also evaluated these three approaches for calculating the ω effect size of ANOVA model. The result showed the unexpected outcome is that the BCA approach for ω was generally not the best performer for nonnormal data. Therefore, the researchers should be careful to use BCA approach for estimating confidence interval of effect size of ANOVA when the data doesn’t meet the normal distribution.
Comment:
1) Because the study was similar with the Algina (2006) and Kelley (2005), many details were less to describe. However the label in figure should be explained more clearly. It just hard to image the number of distribution, effect, and N.
2) The expected result should be described before the method, even if this study just explored the phenomenon.