It has been proposed that preregistration improves the interpretability and credibility of research findings. This talk presents a critical view of this proposal, focusing on the proposed benefits of preregistration vis-à-vis confirmatory and exploratory analyses; HARKing; motivated reasoning; overfitting; deviations from planned analyses; undisclosed multiple testing; p-hacking; the garden of forking paths; optional stopping; test severity; reporting null results; publication bias; and replication rates. In conclusion, preregistration does not improve the interpretability and credibility of research findings when other open science practices are in place, including: (a) rationales for current hypotheses and analytical approaches; (b) publicly available research data, materials, and code; and (c) demonstrations of the robustness of research conclusions to alternative interpretations and analytical approaches

This entry is a one-file data package totaling 2.9 MB, containing a file in .pptx format.If you use this dataset, please cite: Rubin, Mark (2022). Opening Ceremony: Does Preregistration Improve the Interpretability and Credibility of Research Findings?. Erasmus University Rotterdam (EUR). Presentation. https://doi.org/10.25397/eur.19278692