Equivalence Testing for Multiple Regression

Psychological research is rife with inappropriately concluding “no effect” between predictors and outcome in regression models following statistically nonsignificant results. This approach is methodologically flawed, however, because failing to reject the null hypothesis using traditional, difference-based tests does not mean the null is true. Using this approach leads to high rates of incorrect conclusions which floods psychological literature. This thesis introduces a novel, methodologically sound alternative; I demonstrate how to apply equivalence testing to evaluate whether predictors have negligible effects on the outcome in multiple regression. I constructed a simulation study to evaluate the performance of two equivalence-based methods and compared it to the traditional test. I further developed two R functions which accompany this thesis to supply researchers with open-access and easy-to-use tools. The use of the proposed equivalence-based methods and R functions is illustrated through examples from the literature, and recommendations for results reporting and interpretations are discussed.