Life After NHST: How to Describe Your Data Without “p-ing” Everywhere

Abstract

In this article we provide concrete guidance to researchers on ways that they can explore and communicate the results of their studies. Although we believe the methods we outline are important for any study, they are particularly useful for researchers who wish to avoid the null hypothesis significance testing paradigm. We articulate three basic principles of data presentation: (a) use graphic displays to facilitate understanding of descriptive statistics, (b) provide measures of variability with measures of central tendency for continuous outcomes, and (c) compute and thoughtfully interpret effect sizes and effect size translations. We then put these principles into action using data drawn from two real social psychological experiments and provide tools (including software code and a new effect size translation) that will help researchers to quickly and efficiently adopt the recommendations that they find sensible.