It is becoming increasingly clear that we as researchers need to analyze large-scale, publicly available open datasets. The replication crisis in psychology vividly illustrated that most studies are strongly underpowered and that results are not replicable (e.g., Munafo et al., 2017). Besides, and maybe less well-known, small samples also increase the probability of finding extreme results, which leads to a literature full of effect sizes that are artificially inflated (e.g., Button et al., 2013).
