BEAR
Benchmarks of Empirical Accuracy in Research
BEAR
Benchmarks of Empirical Accuracy in Research
Problem
Quantitative metascience drives important debates about research standards. Most of its crucial contributions have been based on analysing individual datasets, often painstakingly constructed by researchers. Thanks to their work we now have a better understanding of replicability, publication bias, p-hacking, reporting practices, pre-registration, significance rates, and more. But many canonical findings of metascience are based on individual datasets, each constructed in a specific way. How generalisable are these findings?
Solution
BEAR is an open-source database for exploring metascience with over 20 datasets across many scientific fields and millions of data points. BEAR does not generate any new source data; it repackages publicly available datasets into a common structure to make research on replication, exchangeability, meta-analysis, and other questions about empirical research easier and more generalisable.
How to download
BEAR currently works best in R.
To download all data in one table,
BEAR.rds, go to GitHub Releases page.
You can download individual datasets (which can include additional variables and rows), see Datasets page. See Get data for more details.
GitHub ReleasesDatasets
BEAR combines curated registries of studies (e.g. clinicaltrials.gov), random collections of articles in different disciplines, and sets of meta-analyses. These are purposefully heterogeneous sets of varying quality. The database covers many disciplines: economics, political science, biomedicine, clinical trials, ecology, evolution, education, neuroscience, and more.
DatasetsWhat is in each dataset?
At a minimum BEAR contains z-values with clearly labelled identifiers for studies and meta-analyses. Where available, it also includes effect sizes, standard errors, sample sizes, methods (e.g. parallel RCT), measures (e.g. standardised mean difference, odds ratio), and additional covariates (e.g. year, pre-registration status, phase of the clinical trial, journal, discipline).
DocumentationModel
An optional model of signal-to-noise ratios is also included as part of the repository, to quantify basic metascientific parameters like selection, assurance, and replicability of findings. Users are encouraged to fit their own models. We make use of that model in an upcoming paper, "A Statistical Case For Qualified Scientific Optimism" (van Zwet, Gelman, and Więcek 2026).
Read the paperCredits
All credit is due to the researchers who assembled the metascientific datasets and maintain the large-scale repositories from which data are compiled here. Please see individual dataset pages for references and licensing information.