Search for a Word

Enter a word above and figures will appear that illustrate trends among the psych/neuro papers containing said word! The figures show several variables, including patterns in p-values. P-values strongly predict replicability, and studies on topics frequently associated with weak p-values (e.g., p ≥ .01) tend to replicate much less.

You can search for "priming" and see how priming papers generally produced weak p-values yet this topic was ubiquitous among top journals and universities before the replication crisis began. However, since then, the field has moved away from priming research. You can alternatively search for "genotype" and find that genetic psychology (and its dubious p-values) peaked in prominence around 2012, although has declined since. By contrast, "perception" research fares better in terms of its p-values and is on the rise (echoed by searching for "perceived"). You can also look at words linked to methods, like "ANCOVA", which has consistently yielded questionable p-values but its usage has gone down over time. Feel free also to toggle the Psych/Neuro switch, and see how "connectivity" and "CRISPR" studies are hot right now; the dataset used for the site includes cognitive, behavioral, and cellular neuroscience.

The statistics computed reflect data on ~250k empirical psychology papers and ~150k empirical neuroscience papers published from 2004-2024. This webpage was prepared alongside the article "One Decade Into the Replication Crisis, How Have Psychological Results Changed?". For a quick summary of this work, take a look at this news piece discussing the study in Science.

You can download the dataset on Psychology paper statistics from the associated OSF repository. You can additionally find the code used to download the papers, organize the dataset, and perform the analysis in this GitHub repository. I'm happy to lend people a hand in working with the data. I'm also happy to help with downloading papers from other fields (e.g., to evaluate p-values in medical research). Feel free to contact me ([email protected])