Dr Alfredo Sánchez-Tójar

Recent worldwide many-lab replication attempts in the social sciences and medicine have exposed alarmingly low rates of replicability of scientific findings. This ‘Replication Crisis’, or perhaps better named ‘Credibility revolution’, has been the catalysts of a movement towards more open, more reliable and more transparent science. Although the fields of ecology and evolutionary biology currently lack many-lab replication projects of similar magnitude to those in the social sciences and medicine, it is possible to obtain essential information about replicability by studying publication bias. Publication bias occurs when a subset of research findings, such as statistically non-significant results, are less likely to appear (or appear earlier) in the scientific literature. Publication bias leads to a distorted view of the overall evidence for a hypothesis. In this talk, I will introduce what publication bias is and the different types of publication bias, briefly explain how we can study publication bias, and showcase some of the most outstanding and worrying examples of publication bias (past and present) in ecology and evolutionary biology. I will finish off my talk by introducing and discussing several practices that can help us reduce publication bias and lead to a less distorted understanding of the world.