A review of five techniques to derive confidence intervals with a special attention to the Cohen's dp in the between-group design
Keyword(s):
Ad Hoc
◽
Getting confidence interval expressions for some effect size statistics can be a difficult endeavor. There exist multiple approaches to compute a confidence interval, not all being exact. Herein, we review five general classes of techniques that have been used to get confidence interval expression. Some are exact approaches, such as the predicting distribution technique and the inversion technique. Others are quite approximate, being based on proxy distributions and various ad hoc adjustments. Finally, the Bayes technique is exact but can be computationally prohibitive. All five techniques are illustrated with the Cohen's dp effect size statistic in a between-subject design where all five classes of techniques have been employed.