Our research in a nutshell:

Inferential statistics play a key role in many sciences. Although the normative workings of these statistical tools are well established, surprisingly little is known about how researchers use them in practice, how often they make mistakes therein, and whether their expectations affect their (reported) statistical results.

We study the prevalence of different types of errors in published research, relate these to practices and journal policies, determine the effects of human factors on meta-analytic outcomes formally and through simulations with artificial and actual data, develop model-based corrections, and conduct interventions to counter their effects.