How NIH Director Jay Bhattacharya Plans to Fix the Replication Crisis
What this video covers
In this short interview clip, Bhattacharya lays out, step by step, how he intends to make replication — not publication in top journals — the standard of truth in science, as the centerpiece of what he calls a second scientific revolution.
Key points
- Bhattacharya defines replication broadly: not just rerunning a study with the same data, but independent researchers approaching the same problem from different angles and in different populations, together with the connected science of meta-analysis.
- He contrasts two approaches to the replication crisis — a “policing” approach premised on fraud, which he calls wrongheaded, and the culture-change approach he favors, in which failing to replicate someone’s result is an honor rather than an indictment or a career-ending event.
- Step one: NIH will start funding replication as a viable scientific career, he says, noting that NIH has historically not funded researchers who want to do replication work, making such careers nearly impossible.
- He insists the government must not decide which studies get replicated — that power could become a weapon against controversial ideas — and says the scientific community will crowdsource the “rate-limiting step” ideas, where a result’s truth or falsity sends a whole field in different directions, through NIH’s normal peer-review structure.
- Step two: an NIH journal where scientists can publish replication studies and constructive failed results, with limited gatekeeping and with original authors able to serve as referees whose signed reports are published alongside.
- Step three: a new office of metascience that will measure scientists on broader bases than citations and paper counts — including how often they share data, code, and tissue samples, and how often their work is replicated.