Landwehr participates in Large-Scale Multi-Analyst Publication in Nature

A new study published in the prestigious journal Nature under the title “Investigating the analytical robustness of the social and behavioral sciences” concludes that scientific conclusions can depend on who performs the statistical analysis and which specific analytical path is chosen from the range of possible analyses.

457 analysts reanalyzed data from 100 studies—with results that were sometimes very different. Only one-third of the teams confirmed the exact results of the original studies. Observational studies, in particular, proved to be less robust than experimental studies.

Jan Landwehr from the Faculty of Economics and Business at Goethe University, who participated in the study as an analyst, comments on the findings as follows: “Just as one should not base important decisions on a single study, it seems significant not to rely solely on the result of a single data analysis. Only when different, well-founded analytical approaches show a consistent pattern does an empirical result appear to be truly robust. Our study is thus also a call for increased scientific collaboration among different research groups and for intensifying scientific exchange.”

The study published in Nature is available here: Link