Harvard Medical affiliate retracts, corrects research after sleuthing by blogger | Hacker News Reader

Harvard Medical affiliate retracts, corrects research after sleuthing by blogger

fortune.com

45 pointsDocFeind22 comments

The title makes it sound like it was just some mistake. The actual details make it sound they they knowingly committed scientific fraud.

Just following glorious path set by University president

Gay committed plagiarism but didn't "steal" her ideas or fabricate quotes to support them. She absolutely should've resigned and is unfit to lead an academic institution but falsifying cancer research data is on a whole other level than lifting quotes.

For another perspective:

https://wset.com/news/nation-world/legal-scholar-livid-over-...

She has a long history of defending and covering up on behalf of people who do forge data though:

https://www.karlstack.com/p/the-curious-case-of-claudine-gay

Many of the academic frauds in question predate Gay's deanship by over a decade.

Prepare for downvotes and rebuttals if you say anything that even remotely disparages Harvard University on HN.

Personally, I think it's quite scandalous that Harvard gave the world's most famous sex-trafficker his own office. I mean, that's the last person on Earth you'd want to give access to a steady supply of young girls and boys, but that's exactly what Harvard did.

And I've been told that we're not supposed to hold Harvard accountable because nobody knew what Epstein was doing... The heads of one of the world's most prestigious universities were all too naive to realize what was going on. Perhaps they thought the girls at his parties were just really friendly? Or perhaps they don't want to admit to anything they'd be legally liable for?

How can one fall so low as to falsify cancer research.

"One white lie will get me that $10m grant, which I will spend curing cancer. Therefore my white lie is clearly worth it." Typical risk/reward ratio.

Not justifying it on scientific purity grounds, but if you genuinely believe that your treatment is effective (and you are heavily incentivized to come up with an effective treatment), but your study result is unconvincing for reasons you don't think are invalidating, it might be easy to convince yourself that "tweaking the presentation" in a way that moves the treatment forward could save lives.

Another win for the internet and citizen journalism.

3 days ago:

https://news.ycombinator.com/item?id=39145728

What happens to the Authors and if this involves grant money, can they be processed for fraud?

That depends, in order to answer that we need to know the Demographics and immutable characteristics of the author(s)

Yes, but they won’t be. Any such trial would make the institution look bad. If there’s anything you need to know about academia in the USA, it’s that they fear losing face more than Communist Chinese party members.

> they fear losing face

Everyone fears losing face. What’s changing is academia is losing its perceived moral footing. That makes them fair (and popular) game for an enterprising prosecutor.

The same people who donate to elect chief prosecutors also donate to these schools. If you ran a campaign on prosecuting scientific fraudsters at Harvard the next election cycle you'd have 20 multi-millionaires trying to scalp you in hopes they can parlay that into their mediocre child's admission.

... used to donate to these schools

Nope. There's way more money in academia than in prosecution

> more money in academia than in prosecution

Doesn’t matter. Different incentives. As another person said, prosecuting academia is right now a winning political position. With donors and the electorate.

I guess it depends on how you perceive "face". Admitting to and taking responsibility for mistakes can be good, but both academia and the PRC are united in their delusion that pretending to never make any mistakes is somehow good, even when it's almost an open secret that those mistakes exist. Doubling down and tripling down, if you will.

The source of this stuff is PubPeer, a website where people can comment on papers and do post-publication peer review. A quick rummage through this site is the quickest way to destroy your confidence in academic science for good. There's no direct equivalent of Photoshopped Western Blots for computer science, but if you want a laugh search for papers generated by SciGen, a perl script that makes joke papers:

https://pubpeer.com/search?q=scigen

Or for a more serious example of problems (medical AI):

https://pubpeer.com/publications/ABB8F161E6691210C44545FF899...

The authors report an AUROC of 1.00. This is extremely high and implies perfect discrimination between cases and controls, which seems unlikely given the highly heterogeneous nature of ASD and suggests some sort of data leakage

...

The highlighted point estimates are not in the middle of their confidence intervals, either by a modest amount (yellow) or by a large amount (pink). In some pink cases, the point estimate is not inside the CI at all.

...

I have rather strong concerns regarding the apparent absence of a hold-out dataset.

Reply on news.ycombinator.com