Guide for calculating and interpreting effect sizes and confidence intervals in intellectual and developmental disability research studies.

This paper includes a nontechnical description of methods for calculating effect sizes in intellectual and developmental disability studies. Different hypothetical studies are used to illustrate how null hypothesis significance testing (NHST) and effect size findings can result in quite different outcomes and therefore conflicting results. Whereas NHST uses probability levels (e.g., p < .05) to evaluate the results of studies, effect size analyses focus on the magnitude of differences between groups or contrasting conditions and the strength of the relationship among variables of interest to report and interpret study results. Two families of effect sizes are described (mean difference, correlation coefficients) that are likely to be applicable to most intellectual and developmental disability studies. Sources of information on effect size calculators are included to provide researchers ready-available data analysis procedures for computing effect sizes and confidence intervals for different types of research designs and studies.