Replicability Crisis: What About the Social Sciences?

Anton Perdoncin proposes for the next session of the PROGEDO-Loire seminar, in conjunction with the seminar of the Centre François Viète, a talk that will be of interest to epistemologists, science historians, and human & social sciences quantitativists who are increasingly called upon to practice “open science”.

The crisis of reproducibility refers to the numerous failures of attempts to find, experimentally or empirically, the results published in scientific journals. Do these problems only affect the experimental sciences? Are the SHS concerned, and if so, how? The presentation will first go back to the genesis of this crisis, as well as to the debates opened on the one hand by the growing criticism of the frequentist paradigm and tests of inference and on the other hand by the new injunction of the FAIR principles (Findable, Accessible, Interoperable, Reusable). These elements call into question our practices and invite us to reflect on what open research should be. Secondly, the presentation will propose a practical example of research through the case of the Lubartworld project, questioning the means of conducting reproducible and transparent research in SHS.