[mchannon]

This leads to a tragedy of the commons. Say a random nation, say, Sweden, devotes 100% of their governmental and university research budgets toward replication.

70% of the studies they attempt are successfully replicated. 20% are inconclusive or equivocal. 10% are clearly debunked.

Now the world is richer, but Sweden? No return on investment for the Swedes, other than perhaps a little advanced notice on what hot new technologies their sovereign funds and investors ought not to invest in.

A bloc of nations, say NAFTA/CAFTA-DR, or the European Union, might be more practical.

That's the carrot. As for the stick, bad lawyers can get disbarred, bad doctors can get "unboarded". Some similar sort of international funding ban/blacklist for bad researchers would be useful.

[mmooss]

Could it be part of the training for young researchers - replicate existing experiments?

[mchannon]

I applaud that approach. The first year of a Ph.D. program could be reformulated to become 75% replicating the research of others, preferably that of unaffiliated research organizations.

A lot of this research is very involved and esoteric, requiring specialized equipment found only in one place, so some would be very hard to replicate. If what Theranos was doing (or claiming to do) was easy to replicate, it would've imploded years prior to when it did. So not all fraud could be detected, but a lot of the low-hanging fraud, especially in the psychological and pharmacological fields, could be quickly identified. Such a system would be a substantial upgrade and I applaud your suggestion. A smaller country could blaze the trail, because "big boys", like the U.S., are too set in their ways.

[shusaku]

I think this suggestion contains the implicit bias that “replication isn’t important or challenging”, hence you leave it to trainees. Actually, replication is incredibly challenging. Put PhD students on it, and they’ll be convinced the original study was fraud for 4 years until they finally have the skill to get it right!

[GeekyBear]

Alternately, look at one recent example of massive waste as a result of accepting fraudulent research as valid.

> Hundreds of millions of dollars and [16] years of research across an entire field may have been wasted due to potentially falsified data that helped lay the foundation for the leading hypothesis of what causes Alzheimer’s disease.

https://globalnews.ca/news/9016221/alzheimers-research-poten...

[echoangle]

Wouldn’t science in total be impossible to fund if this argument were true? What advantage does Sweden have from doing science and publishing if everyone else gets to use it and they could just wait for someone else to do it? If this was how it worked, wouldn’t every scientist work in secret and never publish anything?