Designing Studies and Evaluating Research Results: Type M and Type S Errors for Pearson Correlation Coefficient

It is widely appreciated that many studies in psychological science suffer from low statistical power. One of the consequences analyzing underpowered with thresholds significance a high risk finding exaggerated effect size estimates, right or wrong direction. These inferential risks can be directly quantified terms Type M (magnitude) error and S (sign) error, which communicate design choices on estimation. Given study design, factor by statistically significant average exaggerated. probability to find result opposite direction plausible one. Ideally, these errors should considered during prospective analysis phase determine appropriate sample size. However, they also when evaluating studies’ results retrospective analysis. In present contribution, we aim facilitate considerations research practice psychology. For this reason, illustrate how consider using one most common measures psychology: Pearson correlation coefficient. We provide various examples make R functions freely available enable researchers perform for their projects.