A commentary on reporting effect size and confidence intervals: Response to Palmer and Strelan (2014)

Based on recalculations of effect size (Cohen's d) and confidence intervals around these estimates (ES&CI), Palmer and Strelan (P&S; forthcoming) re-interpret some of the findings from our 2011 Journal of Business Research article (Dutta and Pullig 2011). In response, we critique two aspects of P&S's conclusions, the use of ES&CI in a dichotomous nature and the assumed superiority of the ES&CI approach. In addition, we discuss the implications of the ES&CI results for the substantive domain of our research and findings. Specifically, the ES&CI results do not “overturn” our conclusions, but indicate that for two variables our results should be considered tentative while for two variables the ES&CI approach leads to similar conclusions. We conclude with a discussion of the implications of the larger debate between reporting traditional null hypothesis statistical testing (NHST), ES&CI, or both in behavioral research in marketing and business-related disciplines.