Dana-Farberications at Harvard University
forbetterscience.com
forbetterscience.com
Mother has stage 4 intrahepatic cholangiocarcinoma.
What's astonishing is that when his work was discovered to have silly photoshop manipulations, he reacted by calling (on Twitter) the people who caught him "steaming turds".
https://forbetterscience.com/2020/01/29/david-sabatini-torme...
I'm a bit suspicious of your attempt to impugn Ackman, who is the leading voice in the fight against academic dishonesty... at Harvard... by women... not married to him... who tolerate Palestinians. His focus is oddly specific, if it's still an important contribution to scientific integrity.
Yes, this is sarcasm.
https://www.science.org/content/article/despite-sexual-haras...
These policies exist and require a zero tolerance application because sexual harassment can be done in a very subtle manner, often purposely, that allows the harasser an escape hatch, but the effects are no less impactful. Further, power dynamics will also very quickly catalyze any personal or working relationship. Thus, as a person in power, it is even more important to not have even the appearance of improper relations.
I can understand how lawyers can be scum and manipulative, but he was the one who let the relationship start the investigation.
Furthermore, the claims that Sabatini had harassed other women are totally unfounded yet outlets continue to insist that Sabatini was a serial harasser.
How are these journals retaining any prestige when this stuff is un-retracted?
The stink should be smeared all over all of them. The publishers should be sued into the ground for fraud given the exorbitant subscription fees. I can’t think of a single reason not to burn it all down.
It is woefully incapable of detecting fraudulent data. How could it? Some fraud has been caught through statistical analysis of datasets, but this takes a lot of time and requires expertise that most academics don't have. Detecting faked images has an even less clear path that people can follow.
Academic fraud is bad. But it isn't like other ecosystems are more effective and capable of detecting fraudulent work.
The journals suck, but mostly for reasons totally unrelated to the limitations of peer review.
Note that there is presently no note on the publications that the papers should not be relied on in any way. The authors are pretending there is no problem which is utterly ridiculous and embarrassingly so. I’m not ok with that as the status quo. I don’t think anyone should be.
> Sidney Farber (September 30, 1903 – March 30, 1973) was an American pediatric pathologist. He is regarded as the father of modern chemotherapy for his work using folic acid antagonists to combat leukemia, which led to the development of other chemotherapeutic agents against other malignancies. Farber was also active in cancer research advocacy and fundraising, most notably through his establishment of the Jimmy Fund, a foundation dedicated to pediatric research in childhood cancers. The Dana–Farber Cancer Institute is named after him.
A basic assumption in science is that the raw facts a scientist presents are true. People question the statistical methods, they question which conclusions are appropriate given the data, but raw data itself is assumed to be accurate. We apply a lot more scrutiny to clinical trials (for obvious reasons), but trying to apply that same standard to all research everywhere would be cost prohibitive and slow research by a huge amount.
A lot of journals use image-duplication detectors now (the post mentions ImageTwin.ai) so I expect this particular form of fraud will stop. I suspect cherry-picking data ("inappropriate data exclusion") is the most common form, and will remain so. It's very difficult (often impossible) to detect.
On the positive side, I suspect the impact of stuff like this is fairly minimal. Everybody in the field knows that experimental biology is difficult and results are untrustworthy. Even excluding outright fraud, it's really easy to just mess something up. Or get a result which is "true" but doesn't generalize, and so it's useless. Good scientists don't trust something just because it's in Nature/Science/Cell, if they really care about the result then they replicate it before building on it. This would remain true even if all deliberate fraud stopped immediately.
[0] https://en.wikipedia.org/wiki/Sch%C3%B6n_scandal
[1] I dunno about the mouse one, it doesn't seem to be a direct copy/paste and a good experimentalist will do a lot to ensure consistent conditions. Everything else is pretty blatant though.
I don't think this is true for everyone. If it is difficult, and easy to mess something up, then all the more reason to have as accurate results as possible. Fudging data isn't acceptable.
"Everybody in the field knows that experimental biology is difficult and results are untrustworthy"
"Good scientists don't trust something just because it's in Nature/Science/Cell, if they really care about the result then they replicate it before building on it"
-> But, I'm Joe Blow, I'm not setting up a lab to replicate every study I read about. This might be fine for a VC, test it before building something. But it forgets how many people might make smaller decisions based on an article in Nature or Cell.
or Extrema: I just read about the Higgs-Boson, do I need to not trust the results until I can build a collider and replicate it.
You mean how lazy and amateurish the fraud that is caught is. I just generally assume most of it is complete bullshit, because you have a group of highly-paid professionals producing write-only documents that nobody gives a damn about. This is a recipe for a constant level of low-grade fraud because people are lazy.
I know Sweden launched a national board to investigate research misconduct [1], I hope more countries do the same.
Rankings influence prestige. Prestige influences perception, even among experts. Is Nature really going to reject that paper? It was written by a Harvard guy, after all.
And next, the number of "high-impact" publications (such as in Nature) is used to rank universities. Good luck...
Not sure what else you can do. Table rankings are actually only applicable for undergrads in practical terms. At research level, You have world leading experts at random universities often, and then people who worked in top universities who advanced the field by none after decades.
I wonder how many universities actually mega climbed ranks in the past decades though
Even in a blind peer review, the subject of study introduces another layer of bias in the process. Let's say two economics papers are submitted that discusses how people spend their last part of their pay check. One studies a pool of people across United States and other studies of people across Nigeria. Which one do you think is going to get positive review from editors and reviewers? The Nigeria paper might not even go to review. The editor is going to say something on the lines of "the focus of the study is too narrow for the journal" but won't say the same when it is across US. This is in addition to the trust issues with any research conducted outside western countries.
For researchers from the "low ranked" universities, the game is rigged against them and there is nothing they can do to swing it to their favour.
But the only value is to get more grants and funding, so they're just complicated beauty competitions.
I know of one, the University of Chicago. And it did so by gaming the ranking metrics. It dropped its long and idiosyncratic application and adopted the Common App. The admissions rate tanked from above 40% less than twenty years ago to about 5% now simply because more students applied. Did the quality of the education change? I doubt it.
P.S. Chicago recently dropped out of the top 10 this past year for the first time since adopting the Common App, due to a change in the ranking formula. Which says more about the inside baseball of rankings than it does about the education on offer.
It's definitely been superficially good for my social life, but I remain very sorry that the Uncommon App is no longer around. It played a nontrivial role in my decision to apply in the first place.
An example of rankings being climbed:
https://www.bostonmagazine.com/news/2014/08/26/how-northeast...
(Not to dis Northeastern. They have some great professors, and do some great teaching and research. And the faculty of a university don't necessarily agree with the businesspeople.)
But if you think that kind of college rankings climbing is a concern, the climbing demand that college has evolved for students is worse, and has implications for tech industry, as I asserted recently:
If you are in a specialist field you generally know who's paper you are looking at - shaping review is a good way of getting ahead.
Yes.
Experiment: randomly pick a “Harvard guy” in biology and check their publications, count what fraction of them are in Nature (or comparable journals).
1. Ray Dalio's "believability" cult:
- https://www.businessinsider.com/bridgewater-rigged-believabi...
- NYT: https://archive.is/ODbAv
2. PageRank. Or, the way that PageRank was a beautiful, effective algorithm only for a few years until people figured out how to game it.
Yes, the bias you describe that makes people respect papers from well known sources more than papers from outsiders is real. But it isn't nearly as extreme as you describe it.
Depressing.
It’s a very profitable industry, and profits corrode science.
It spreads like a cancer. Apologies for the pun but it fits aptly here.
Actually that might be ok if the system allowed just publishing the results and saying "I had this hypothesis and it didn't work, here are the experimental results".
However, the main problem is they watch others around them who bend the rules and get away with it and succeed. The system overall seems to reward cheating, so they get "on board" with it, thus perpetuation the system.
Even more, once it becomes pervasive, the standards are raised for everyone. Now those who don't cut corners, to put it very mildly, are not even able to compete. Honest works becomes impossible. The choice is either leave or start cheating.
https://news.ycombinator.com/item?id=39090959 (34 comments)
I'm usually a science apologist, defender.
Sure, the Harvard president might have misquoted something. And social sciences have a lot of difficulty re-staging the same trials to replicated them, but aren't doing it nefariously.
But here, a hard science, with hard data, and blatantly falsifying it. This is pretty unforgivable.
Science has become distracted by celebrity and money.
What a waste of grant money. I hope the NCI, NSF and NIH fund a program to route out scientific fraud, it is clearly needed.
When I was in grad school some fellow students were admonished for plagiarism but allowed to stay. I always think of who was denied a place by these folks.
Unfortunately after a few failed prosecutions early in its history the ORI now does very little, only a handful of cases per year. Compared to the scale of the problem which is often measured in something like 50%+ or more of papers, it's just irrelevant. There is no "police department" you could staff up that'd be big enough to get a handle on this problem.
That’s hypothetical, but it’s a real trend. Costs go up based on the number of people you have to sift through to find someone eligible.
Things like viagra (10 years ago) or ozempic are still in the honeymoon period where inclusion/exclusion criteria are broad enough that you can enroll lots of people in any given place.
Weird to think I could have ended up working there if I had cut getting my philosophical act together short and set myself to my education immediately. It was a really cool place right in the center of the city near my high school.
Suffix trees are still rad and any DSA-inclined teenager will benefit greatly from learning them :)
Discussion: https://news.ycombinator.com/item?id=39090959
Discussion: https://news.ycombinator.com/item?id=39099565
;)
Generally there is more trust when the theory, the data collection, the data analysis, the model development, the product, and the application, are all completely separate and open. It also advances the science much faster. However, it definitely kills a lot of those monopoly profits.
I propose that there’s really no reason that pharmaceuticals research can’t progress in this manner, with great public benefit. And it would be so simple to enforce: all public grants require everything to be developed in the open, not just data release at the time of publication, or pre-registration, but full release at each stage, no patents, and a preference for funding independent dataset production rather than results. As it is, we pay for the research, then we pay monopoly prices for the drug. It’s like these companies don’t really do anything but collect money.
Also, If anybody knows of a way to donate money to cancer research in a way that doesn’t feed directly into somebody’s profits, please let me know. I have recently become loosely connected to a foundation that has been dumping a lot of money into Dana Farber, and this is highlighting a lot of my concerns as I learn more about the industry.
If reading that makes you upset, do you have a rational reason why? Can you explain how climate science is immune to the social dynamics at work here? Or is it just knee-jerk offense at having your religious beliefs questioned?
This is a genuine question because I may be the one with the religious beliefs, but I am not beyond them being questioned.
https://www.chartr.co/stories/2023-12-08-2-grade-inflation-a...
Apparently teachers' standards have declined for themselves too. Some of that would be a result of being trained at Harvard, a vicious circle.
From a quick google, the acceptance rate to Harvard was 20% higher in the 70's. In 1940 it had an absolute acceptance rate of 85%.
I dont actually know the answer, but from that graph only, the rising rate could be as much from a higher standard of student as lowering standards of marking.
[1] - https://graphics8.nytimes.com/packages/pdf/education/harvard...
That’s an odd way to put “wealthy and privileged”
For instance Harvard's first black graduate graduated in 1870 [1] (meaning he would have passed a test near to exactly like the one shared), and he was a remarkable man. He was left supporting his family as a teenager after his father ran away for the Gold Rush, and was unable to attend public school due to his skin color. Yet, he persevered and lived one amazing life far above and beyond becoming Harvard's first black grad. But people like him are far and few between - and those were the sort of people applying.
[1] - https://en.wikipedia.org/wiki/Richard_Theodore_Greener
Those were a tiny minority of the people applying.
Nothing in that entrance exam requires an academic outlier. All it requires is a classical liberal arts education (Latin, Greek, classic literature, history, geography, geometry)
Most people who attended higher education could do so because wealthy children received such an education above and beyond what the poor received (and because the could also afford college thanks to that wealth).
And on that note of excellence, that test is going to completely destroy most of anybody, let alone high school grads. The breadth of knowledge demanded paired with the extensive depth of such is something very few are practically capable of obtaining without an absolutely obscene amount of effort. Keep in mind this wasn't some sort of a take home exam or whatever, these exams were done on the spot - in an 8 hour window.
The math section has some hard wanky arithmetic and by-hand computations using logarithms. Modern students would struggle because that hasn't been a focus of math education since the invention of the calculator. Your average physics professor would struggle with that because no one needs to do that stuff by hand any more. A modern Harvard applicant would ace that algebra section. The geometry section is 50/50 - there's still plenty of geometry in modern education but it no longer involves memorising Euclid.
On the other hand there's no calculus/analysis which is probably the biggest focus of higher-level high school math - something an applicant of that era would be flummoxed because it wasn't part of classical math education. There's also a complete dearth of any of the natural sciences which was again neglected in a classical education.
I'm telling you that that's not an intrinsically difficult exam, it just covers areas that were the focus of an upper class classical education of the time. You're confusing a different educational focus with intrinsic difficulty. If your math education up to the age of 18 was 90% long-form manual computations and memorising Euclid's Elements you'd ace that section too.
I tried to find a list of Harvard graduates of the era to demonstrate this, but the best I could come up with is this [1] list of Harvard presidents who all tended to be distinguished graduates themselves. It's quite unfortunate, because these are ostensibly the cream of the crop - which kind of plays into your class stuff, but I think even this sample also works to largely reject it. In any case it's quite interesting to read their bios!
For instance check out guys like Thomas Hill [2]. He had very minimal formal education, yet was a sponge for knowledge, and learned what was necessary to achieve - and achieve he did. Another is Jared Sparks [3]. Educated in common schools (which most certainly were not teaching Latin/Greek and other advanced topics) before he went to work at a variety of trades and then (presumably after saving up some money) attended a private school at the age 20+, before finally making his way to Harvard.
Basically it's not like you just went to school, graduated, and then applied with the knowledge you'd gained - at least not in general. In a world where education was optional, those who pursued it tended to do so out of passion and from that passion exceptional competence also tended to emerge.
[1] - https://en.wikipedia.org/wiki/President_of_Harvard_Universit...
Fully support the observation that university meant something different in those days though, the standards were much higher. Although I'm not sure if that would apply at Harvard since their class sizes probably haven't changed by all that much.
We don't routinely teach Classics in high school, so it's not surprising that those sections would be beyond most students today.
The sections on that exam dealing with Mathematics seem appropriate or possibly slightly on the easy side; I would expect an equivalent exam today to include at least a bit of calculus, but over the past 150 years there has been a shift in pedagogy towards calculus and away from geometry so it's probably just a sign of the times.
Initially I thought it was all a rich kids game, but the later maths and geometry questions require you to know what you are doing. Of course thats easier with expensive tutors, but it’s entirely possible you’ll never grasp it.
There are apparently advanced classes like Math 55 that had huge dropout rates that had to be cleaned up though. So those courses have been completely changed instead of just awarding higher marks I'd imagine.
It could be caused by lowered standards or more competition or better student performance.
My buddy said he had to take a botany class as a filler for some academic requirement for graduation. He said the entire class was to pick a tree out on campus and then describe how it changed over the course of the semester.
He maintained he had more rigorous classes in HS then he had at Harvard and said the progressive politics were so pervasive, that as long as you towed the ideological line, then you got favorable grades.
It would appear not much as changed since he was there in the early aughts.
I'm not a big fan of standardized tests, but they can identify serious issues to address especially as it pertains to geographical, ethnic, etc. groups and allocate funding accordingly. If we keep lowering bars instead of addressing the issues, this is going to get very bad.
https://www.oregonlive.com/education/2023/10/oregon-again-sa...
https://www.opb.org/article/2021/09/20/examining-oregon-deci...
It’s ironic that this stance is actually racist, since it implies certain races perform worse.
Lots of blame levied at a strawman white culture. If only there’d be a concomitant critique of the problematic groups at hand. Alas, that’s verboten under the current regime.
Lowering expectations for those under their patronage only makes advantage easier for me and mine. Good luck to them and their token positions.
It is useful to define terms. Merriam-Webster has what I believe is a quite useful definition of racism: "a belief that race is a fundamental determinant of human traits and capacities and that racial differences produce an inherent superiority of a particular race".
That is a useful definition because it helps to separate a racist interpretation of test scores from a non-racist interpretation. The racist interpretation is to jump to a conclusion of fundamental inferiority in academics. The non-racist interpretation is to acknowledge and emphasize the environmental factors that can lead an historically oppressed/displaced/enslaved minority to underperform academically, to acknowledge that African-born immigrants do quite well on standardized testing, and also to understand that standardized tests are normed on academic success, which creates a causality issue when measuring groups that have historically not been allowed in academic institutions. Perhaps all of those arguments don't hold water: or at least the pitcher holding them has holes of its own. But it is quite erroneous to jump to a conclusion of racial biological determinism, and that is what the racist conclusion would be.
There is a particular kind of fallacy where a person can, seemingly unconsciously, take what are perfectly valid observations, like the test-taking disparities, and spin them into non-evidential overarching beliefs. This happens all too commonly when it comes to scientific (or at least metrical) observations and racism. People were all too eager to embrace phrenology, IQ testing, evolutionary biology, etc. when it confirmed their cultural preconception that their race was superior. All of those areas, even phrenology, had some scientific merit to them (and still do in the case of IQ testing and evolutionary biology), but that merit is lost when the conclusions get spun into non-evidential belief systems like racist ideology. Seemingly (I can't peer into their heads, really) skepticism is lost in a chase for the personal gratification of a belief of fundamental superiority.
Discussions around academic standards are a case in point. Academic standards change quite frequently. The Oregon requirement we're discussing was only instituted in the early 2010's. When race is not part of the discussion, it is entirely normal for educators to say that a new standardized test doesn't really have evidence it serves the population and to remove it. There is a very lively debate amongst educators about how excessive measurement, i.e. standardized test taking, can interfere with the normal give-and-take of teaching. But, if the evidence is that non-native-English speakers are underserved, or that African-Americans are underserved, suddenly it is 'racist' to remove a requirement or hurdle. What I saw in the two articles is that those in favor of removing the requirement saw a lack of evidence it was beneficial, and saw evidence that it was not beneficial; while those opposed had no evidence other than it would graduate 'dumber' students; ignoring the fact that presumably dumber students were being graduated before the 2010's when the program was instituted.
In short, it is often a smokescreen to say that observations about race are racist, because that makes it impossible to act against inequality, or sometimes even to act at all. Once again, racism is the belief of one race's inherent superiority over another.
There are many reasons for this to be the case [1] not just the classic "this race is inferior at test taking".
[1] Pre WWI, WWII Germany famously tested students for acedemic performance. During WWII and times of food scarcity from Allied blockades test performance dropped by greater than 10%. Same students, same race, different conditions.
Typically well fed, well exercised, unstressed, groups with sound economic security, home ownership, etc perform very well on tests.
Let's just point out that testing in question took place across all areas of Germany, not just those famously occupied by the elite intelligensia that were moved on, also the rural areas occupied by the beer swilling fraus and blue eyed leden hosen knee slappers, and that test results improved with food quality.
It doesn't take an Einstein to ANOVA.
In the beginning the picture of copypasted images and doctored results is very clear.
But when it goes to the passage about genius, and it is so strange. Out of place, weird rant about chosen word…
Whatever bastard Sabatiny is, the author’s approach also leaves bad taste in mouth
Not my field, but my understanding of how these things work in big medical research factories: first few authors tend to be young researchers (maybe med students or even undergrads) trying to match residency or get into grad school. They do much of the work actually assembling the submission. The later names on the author list (who this article is taking to task) run labs or oversee research groups. Should they correct the record when it's pointed out? Yes. (But the snark and tenor of the post doesn't exactly convince someone they can admit an oversight in good faith.)
Should they be vigilant enough to check and notice these things? Of course. Some of the fakes are not subtle. Others, like the copy-paste of empty space in the lane to cover some undesirable result? Way harder to spot with the naked eye. I don't think there was great automated tech to detect image duplication in the 00's when these were published.
So your med student fudges data on a paper. The ethical answer is to expel them—"the world needs plenty of bartenders." But it appears big institutions these days are pretty invested in the sunk costs of prestige, dislike admitting error in admission or hiring, and prioritize go-along-get-along environments. It could be career limiting if students don't get any/enough pubs working in your lab. It'd a lot of hearings and paperwork to report him, plus I heard his uncle's a donor. If she got kicked out she'd lose her visa. And if I reported them, I'd be obliged to report everyone, and I'd be sunk in discipline hearings three times a year. So much easier to just... not look very hard.
It's bad science and bad ethics, but if you want better, reform the incentives. "Public" shaming by a niche newsletter... might be better than nothing, but doesn't qualify as an incentive.
I would think they can't admit an oversight due to the institutional incentives you mention; the snark is irrelevant. If anything, it encourages publicity for the oversight, which is the only thing that might change the incentive.
To the real brass tacks incentives: yeah, it's "someone is angry on the internet" vs, "I will have to deal with a discipline process with documentation and meetings and maybe depositions and adversarial lawyers. That's not my bag, I'm a scientist. There will be volatile young people and bad feelings communicated in person, plus gossip among my close coworkers. Also undesirable. If this becomes a repeated pattern, learners might start avoiding my lab, and deans/my superiors might start asking very awkward questions." Yeah, stacked against that, angry person on the internet is a weak incentive. Even if they're right.
And the snark does matter. Because this guy writes like a YouTube comments section, and that's not how you talk to adults or solve problems in elite institutions. So the contrast in styles draws lines of "us" vs "them." And it's natural to care more about the opinions and esteem of your in-group (who talk like you) than the out-group (who deride you).
Let's all remember what we are talking about here: every single one of us will know someone who will die several years early because of scientific fraud. It is reasonable to be angry as hell.
Indeed, we've seen how the polite approach is so hard to ignore/avoid that the adults in elite institutions solve problems before snarky folks wake up!
The first author 'undergrads' or 'med students' are not assembling the paper no way. For sure the last authors know what is going on.
This is no excuse, and allowing it to be an excuse is at the root of why things are going so wrong in Science. If your name is on the paper you 100% must own every word.
There should be some agreed sanctions for this kind of thing. I would recommend that if any paper is withdrawn all the authors should be prohibited from receiving government funding for 3 years, and they should be banned from publication & citation for 5 years.
I am thinking this would change the culture quite radically.
In software, this would be the equivalent of not importing a package unless you first checked the code line by line.
The system works okay now, but there's a lot of room for improvement. I generally think the best way to improve it is to require more open data so that consumers of research can validate the findings of papers.
In the old days a single dodgy paper meant ostracism. Literally - no meeting invites, no more grants, no promotion, no moves to another department. Nothing. You sat where you were in shame until you retired. You had to do a lot of teaching and admin as well.
I do not think that this is the case now, and folks know it and are acting that way.