Abstract

Replicability is currently in the spotlight of the debate around the reliability of research findings in psychology. Well-designed experiments, reliable and valid assessment measures, pre-registration of studies, open access to datasets and to statistical scripts used for running analyses are among the solutions proposed for overcoming the replicability crisis. Close to these methodological recommendations, power-analysis and the use of a Bayesian approach have been also advocated as candidate solutions to the replicability crisis in the psychology field. In this brief report, with a practical exemplification, we illustrate how a more in-depth knowledge of available statistical methods and approaches, and a more thoughtful choice of appropriate data-analysis techniques, depending on the type of data, can contribute to obtain more reliable research findings. To conclude, we propose that working for improving psychology researchers' statistical knowledge represents a pivotal way for promoting more reliable and replicable research findings, thus reducing the replicability crisis.