Large-scale study on the credibility of scientific findings published

On April 1, 2026, the key results of the international research program Systematizing Confidence in Open Research and Evidence (SCORE) were published. A total of 865 researchers participated in this large-scale initiative. The findings are presented in a series of eight scientific papers, including three articles in Nature as well as five additional preprints.

The SCORE program was funded by the U.S. Defense Advanced Research Projects Agency (DARPA) and coordinated by the Center for Open Science (COS). The aim of the project was to systematically investigate key dimensions of the reliability of scientific research. Particular focus was placed on the reproducibility, robustness, and replicability of research findings, as well as on the question of how well their repeatability can be predicted by human and machine-based approaches.

One of the central publications, involving Dr. Kenneth S.L. Yuen (LIR), “Investigating the analytical robustness of the social and behavioural sciences”, shows that research findings often depend on the analytical pathways chosen. The study therefore recommends giving greater consideration to alternative analyses and making uncertainties more transparent.

The SCORE results provide an important empirical foundation for assessing the reliability of scientific research across disciplines and strengthen the understanding of when research findings can be considered robust.