MIT and University of California Berkley have teamed up to provide a critically needed service in these skeptical times: a new journal, Rapid Reviews COVID-19, that provides vetting of not-yet-peer-reviewed COVID-19 research.
This new journal addresses a problem borne of the urgency of the pandemic. Many researchers publish their pandemic-related research on pre-print servers to get quick feedback from their peers. But sometimes this as-yet-unreviewed research turns out to be faulty. The most infamous examples thus far were picked up by two highly respected medical journals, The Lancet and the New England Journal of Medicine. Both had to retract the studies, once it was discovered that the primary data sources were not available for review.
To help counteract this published-before-vetting trend, Rapid Reviews COVID-19 scans the pre-print servers to locate papers on important pandemic topics, then reviews them using machine learning and a network of scholars. This speeds validation of new research, and helps prevent the dissemination of false or misleading scientific news. Congrats to Rapid Reviews COVID-19, our Measurement Maven of the Month!
Now, if only someone could do that for the flawed data that vendors and PR folks present every day. Don’t get me wrong, I don’t think anyone is intentionally skewing their data. It’s just that no one is checking it carefully before it goes into those PowerPoints. (For quick ways to review your data, read “8 Easy Steps to Insure Accurate Data.” For more advanced techniques, you might want to try Benford’s Law or the GRIM test.) ∞