An Introduction to Equivalence Testing in Jamovi for Nonsignificant Results in Speech, Language, and Hearing Research

Evidence-based data analysis methods are crucial in clinical and translational research areas, including speech-language pathology and audiology. Although commonly used, null hypothesis significance testing (NHST) is limited with regards to the conclusions that can be drawn from results, particularly nonsignificant findings. Equivalence testing can be used to complement NHST and imply the presence of an effect large enough to be considered as meaningful. This tutorial provides an introduction to equivalence testing using jamovi, a free graphics-based statistics package that allows researchers to conduct a wide range of statistical analyses, including equivalence testing, in a clear and easy-to-interpret manner.

Simulated examples of equivalence testing of independent-samples t tests, paired-samples t tests, and correlations were conducted in jamovi, with explanations and justifications of choosing the smallest effect size of interest and analysis options provided and statistical output explained and interpreted. These examples also demonstrate what equivalence testing can and cannot infer about a data set.

Analyses of nonsignificant results, through the use of equivalence testing, are underutilized in speech, language, and hearing research. By complementing traditional NHST analyses with equivalence testing, researchers can directly test for the presence (or absence) of an observed effect large enough that may be considered meaningful, and therefore test for both statistical significance and practical/clinical significance, which allows researchers to draw more informative conclusions from their findings and provide clearer information for clinicians and researchers in the field.