A Practical Guide to Registered Reports for Economists
Date: 2022-06-30, Image: (c) Unsplash
Previous works have shown that about half of the statistically significant results published in econ journals could be false positives, as a result of p-hacking, HARKing, and publication biases. In Registered Reports (RR), journals review the papers before the data are collected and, after review, pre-accept papers. Researchers only then need to collect the data and to run the statistical analyses they committed to do. Some econ journals accept RR, such as the Journal of Development economics, Q-Open, and, more recently, the Journal of Political Economy Micro. There is also a Peer Community In for Registered reports (PCI-RR).
Registered Reports offer the following benefits:
For science: it reduces the risk of data mining (p-hacking, HARKing); it increases reproducibility rates; it improves knowledge accumulation and, thus, policy-making.
For editors: it helps improve the quality of results published in their journals as they become more reliable.
For authors: it helps them improve their design before data collection; it allows them to engage in more risky / resource-intensive studies as studies are conditionally accepted.
Registered Reports might be new to economists and numerous pre-registrations are incomplete (which undermines their very purpose). This practical guide on how to write a Registered Report for economists (working papier) is helpful for the community. It includes most of the good "tips" that researchers use in their own work.
To provide your feedback, especially if you think the authors have missed something important. Send your suggestions at <thibaut.arpinon@univ-rennes1.fr>