A Web-Based Application for Effect Sizes and Their Confidence Intervals

• In Event: Advances and Accessibility in Statistics

Abstract

There have been increasing calls for applied researchers to see and utilize effect sizes as the primary outcomes of their research. However, this sometimes places a methodological burden on researchers whose primary interests are substantive. Motivated by a desire to help applied researchers better report effect sizes and their confidence intervals, we developed a web application that allows users to calculate effect sizes and confidence intervals for standardized mean differences, binary outcomes, ANOVA, multiple regression, and multilevel models. In this paper, we demonstrate the use of our application in the context of binary outcomes and multilevel models. It is our hope that through our work, applied researchers can better contribute to a cumulative science.

Authors

• James Ohisei Uanhoro, The Ohio State University - Columbus

• Ann A. O'Connell, The Ohio State University