Reproducibility in Research

The scientific community has recently expressed concern about the possible existence of a reproducibility crisis in science. Reproducibility, which is an essential criterion of reliability in science, is a complex notion whose different facets need to be defined [1].

This presentation will first define the reproducibility of methods, which indicates that study procedures and data are described with sufficient precision to be exactly repeatable by others. We then turn to the notion of reproducibility of results (or replication). This refers to the actual achievement of identical results when a new study is carried out independently, using procedures identical to those of the initial study. Finally, the notion of inferential reproducibility, which considers the interpretation of results, will be addressed. This last aspect corresponds to the fact that identical conclusions can be drawn from a re-analysis of the original study or from a new experiment.

These concepts will be illustrated with examples from numerous studies assessing reproducibility in different contexts (psychology, cancer biology, clinical research). These works identify several difficulties in reproducing the results of numerous experiments. As a result, the scientific community is proposing solutions to remedy the situation and optimize the reproducibility of our research work [2]. Several examples will be discussed including prospective registration, data sharing and large-scale replication projects.

The seminar will be led by Professor Florian Naudet, University of Rennes, University Institute of France (IUF). The seminar will be on Zoom and held in English.

References

1. Goodman S.N., Fanelli D., Ioannidis J.P. 2016 What does research reproducibility mean? Sci Transl Med 8(341), 341ps312. doi: https://doi.org/10.1126/scitranslmed.aaf5027

2. Munafò M.R., Nosek B.A., Bishop D.V.M., Button K.S., Chambers C.D., du Sert N.P., Simonsohn U., Wagenmakers E.J., Ware J.J., Ioannidis J.P.A. 2017 A manifesto for reproducible science. Nat Hum Behav 1, 0021. doi: https://doi.org/10.1038/s41562-016-0021