Hacker News Reader

Comment by ncmncm | Hacker News Reader

Not to be a poop, but I will want to see confirmation in a US or European lab before I believe this. Incentives for fraud, and likelihood of getting away with it, are very strong in China. A mark in its favor is that Nature has good reasons to try to ensure the work is real, but they have been taken in, in the recent past.

View on HN

Many ethnically Chinese researchers are based in the USA, including Chuan He, one of the senior authors on the study. He is based at the University of Chicago, has a stellar publication record, and is an HHMI investigator.

But you can be rest assured that many labs across the world will be working to replicate this finding.

I think people were talking about where the work was performed, not ethnicity. Chuan He has multiple COI disclosures on this paper.

Incentives for fraud aren't much smaller this side of the pond.

https://www.npr.org/sections/thetwo-way/2016/09/13/493739074...

Such incentives are somewhat more isolated in the west. I'm not sure if it's still the case, but ~10 years ago the Chinese government offered bounties to researchers for publishing in various western journals. The bounties varied in size based on perceived prestige. The result (rampant fraud) was entirely predictable.

Why is predictable that offering a reward to get published in prestigious journals would result in fraud? Why would any scientist who wants to further one's career, not want to publish in prestigious journals? Why would adding an extra incentive on top, result in more fraud than otherwise already would have happened?

The greater reward incentivizes greater risk. Kinda obvious.

That is not obvious at all. That's like saying giving salespeople high bonuses means they're more inclined to break the law (e.g. blackmailing people to sign contracts) as opposed to working harder through legal means.

If the risk of penalty is low enough and breaking the law is easier than working harder through legal means, then I think we could assume they would be more inclined to break the law.

And is the penalty low? China is not known for low penalties.

This article from 2018 said they were upping checks and penalties for academic fraud: https://www.reuters.com/article/us-china-science-idUSKCN1IW0...

This article says that a researcher was punished. The investigation found no plagiarism but did find ethics violations, so even ethic boards are a thing in Chinese academia nowadays. https://www.sixthtone.com/news/1006745/top-chinese-scientist...

I was more responding to your analogy. I couldn't say whether the penalty is too low in this case, and you may be right.

I think it was like $100k US.

Damn I sell a year’s worth of my time for that in the UK… definitely might be worth ruining my reputation for in China if I was immoral.

It is a bit different - this is $100k on top of what is presumably already a fair market rate.

It is a lot more like a massive bonus than a salary. More than enough to have someone unscrupulous start lying, especially if they think their research is marginal quality.

And it's more like ten years salary for a typical Chinese university professor.

Just 3 days ago this was on Hacker News:

"Time to assume that health research is fraudulent until proven otherwise?"

https://news.ycombinator.com/item?id=27884233

Agreed, but consequences are more likely. Not enough, mind. Fraudulently confirming fraudulent work seems less likely than original fraud, though, unless the parties are in cahoots.

Like clockwork. No, showing faults in others does not mitigate yours. Nothing productive came towards verifying this study by showing alleged faults in others.

It does show that it’s not specific to China, contrary to what the previous person alleged.

He alleged only that China produces unreliable scientific results, and he expects peer reviews form non-chinese countries. Some of the most technologically advanced non-chinese peers happen to be the USA and EU. Nobody alleged they are more reliable.

Which is still a baseless accusation, as demonstrated in the comment above.

Which is not disproven by pointing out flaws in others.

Which is proven to be not specific to China by pointing out the same flaws in others.

Nobody claims that this publication doesn't need to be verified; of course it does. The problem is claiming that it needs to be verified because it came from China, which is what's been done above.

Reply on news.ycombinator.com