Effect Sizes and Intervention Research

Effect size is an essential component of social work research. The magnitude of the effect size gives critical information on the extent to which an intervention has an impact or groups differ on a measure. Knowing an effect size allows us go beyond statistical significance in interpreting the meaning of observed differences. Furthermore, the American Psychological Association (APA, 2001) suggests that published manuscripts should report effect sizes in addition to inferential statistics and statistical significance. This is particularly important in intervention research. As implementing an intervention requires substantial investment of time and effort, it is essential that scholars communicate the impact an intervention is expected to produce. Many social work researchers, however, neglect effect sizes altogether, fail to interpret their effect size findings, or interpret their effect size findings incorrectly. The purpose of this presentation is to review, discuss, and apply current guidelines for the calculation, interpretation, and reporting of common measures of effect size in intervention research literature. The primary objectives of this workshop are: 1) to assist attendees in deepening their understanding of effect sizes and how they related to intervention research, and 2) to increase attendees' ability to translate and communicate statistical data about effect sizes in practically meaningful terms. The presentation will explain how effect size is distinct from statistical significance and describe the relationship between effect size and statistical power. It will provide a review of commonly reported effect size statistics and the correct interpretation of each. Discussion topics will also include the differences between statistical significance and practical significance, and statistical effect size versus practical effect size. The authors will discuss the use and reporting of effect size through use of examples of selected de-identified articles from peer-reviewed social work journals, and will provide an overview of current trends/guidelines in the reporting of effect sizes. A particular emphasis will be placed on understanding and interpreting effect sizes in everyday language for the benefit of practitioners and clients as well as scholars. Teaching methods incorporated into the workshop include simple effect size computations, software applications, understanding reported effect sizes in published research, question and answer, and discussion.