In silico replication project of ecological and evolutionary studies

Despite the growing concerns about the replicability of ecological and evolutionary studies, no results exist from a field-wide replication project. We conduct the first large-scale in silico replication project, leveraging cutting-edge statistical methodologies. Replicability is 30% to 40% for studies with marginal statistical significance in the absence of selective reporting, whereas the replicability of studies presenting strong evidence is above 70%. The former requires a seven-fold larger sample size to reach the latter’s replicability. We call for a paradigm shift in planning, conducting, and publishing research toward a transparent, credible, and replicable ecology and evolution.