Allegations of Research Fraud at the Highest Levels in China

Get your science fix here: research, quackery, activism and all the rest
Post Reply
User avatar
Pucksoppet
Snowbonk
Posts: 599
Joined: Mon Nov 11, 2019 8:13 pm
Location: Girdling the Earth

Allegations of Research Fraud at the Highest Levels in China

Post by Pucksoppet » Sun Dec 01, 2019 6:22 pm

Picked up an interesting article at Soylent News: Research Fraud at the Highest Levels in China

It links to an article on For Better Science: The Teachings of Chairman Cao, as well as giving further details.

I'm beginning to wonder if deliberate fraud, and mismanagement of statistics are going to overwhelm good science. Are there any grounds for optimism?

User avatar
Pucksoppet
Snowbonk
Posts: 599
Joined: Mon Nov 11, 2019 8:13 pm
Location: Girdling the Earth

Re: Allegations of Research Fraud at the Highest Levels in China

Post by Pucksoppet » Sun Dec 01, 2019 6:40 pm

Following on from a few links in comments in the above, I stumbled across this:

Not even trying: the corruption of real science

It's why I think places such as this are important, as are the efforts of Ben Goldacre and Nick Brown. We need more people like them in positions of influence.

Allo V Psycho
Catbabel
Posts: 734
Joined: Sat Nov 16, 2019 8:18 am

Re: Allegations of Research Fraud at the Highest Levels in China

Post by Allo V Psycho » Sun Dec 01, 2019 7:45 pm

Pucksoppet wrote:
Sun Dec 01, 2019 6:40 pm
Following on from a few links in comments in the above, I stumbled across this:

Not even trying: the corruption of real science

It's why I think places such as this are important, as are the efforts of Ben Goldacre and Nick Brown. We need more people like them in positions of influence.
Interesting, but yes, bitter.
I've not worked in lab science for a while, which is what I think he is addressing. I did work in fairly similar fields back then, and his time scale stretches back to my lab science days, so am not completely unqualified to comment.
Where I agree:
I think a major problem has been the rise of 'managerialism': the view that what people do is inherently sloppy/lazy, and they need to be 'monitored' or 'measured' all the time to bring them up to scratch. The Research Excellence Framework is a good example of this.
We had a Faculty meeting about REF early in the cycle, where it was laid down that we had to produce one 3 or 4 star paper every year. And we had to work on approved themes, to match the University narratives. Since I'm sufficiently successful in my current field that I meet REF needs (and don't give a sh.t about senior managers), I felt I had to intervene on behalf of junior staff, and point out that this is absolutely not how research works. Sometimes nothing works, even when you have the right ideas. Sometimes you have the wrong ideas. I spent about three years chasing something completely worthless in retrospect. Yet because research is frustrating and difficult, you have to work on something you love, wherever it takes you, not on a departmental theme. The pressure to 'get grants' (as opposed to 'doing research' ) doesn't help either. A colleague of mine who had just published a brilliant paper in Nature was publicly described at a Departmental meeting as a 'research zero' because she worked in a very inexpensive field that didn't need grants. The only way to guarantee you would meet your departmental REF target - would be to fake the data. And in the long run, you would probably still be found out.
Where I don't:
As far as I can see, the people entering science (and it is not easy: it takes real dedication and effort, unlike the jobsworths Charlton seems to be describing) are the same curious, interested and honest people they ever were. Most of them could make more money in other occupations, with less work. But they want to know what reality is, and that keeps them curious, and, as a by-product, honest.
And I disagree strongly with his comments about narrowing specialisation. On the contrary, the fields I worked in have undergone a massive convergence. Where there was once molecular biology, cell biology, cancer biology, developmental biology etc., there is now just one massive overlapping field, where everybody's work informs everybody else's.

It's true I think that we are in a period of 'normal science' in lab biology, rather than a Kuhnian 'paradigm shift' phase, so much of the work seems a little boring. (This is why I changed fields). But that is what science is mostly like. A paradigm shift only comes along at an interval of decades.

If Shpalman is reading this thread, I'd be very interested to know what he thinks, from a very different field.

mikeh
Fuzzable
Posts: 277
Joined: Mon Nov 11, 2019 5:48 pm

Re: Allegations of Research Fraud at the Highest Levels in China

Post by mikeh » Mon Dec 02, 2019 12:06 am

Been many a story about research fraud in China over recent years, see for example this article
https://www.sciencemag.org/news/2018/10 ... th-penalty

In sort-of-related anecdotal stuff, was speaking with a couple of (English nationality) teachers about their time teaching in China. In separate incidents in separate towns at separate times, they had identical stories of why their time there ended. The conversations with their parents, headmaster and the local town Mayor went thus -
Parent - "you've written here my son has failed his exams!"
Teacher - "yes, afraid so"
Parent - "but you must change the mark"
Teacher - * laughs politely * .... "oh you're being serious? No I am not doing that."
Headmaster - "can you change that boys marks please? He is not failing"
Teacher - * laughs politely * .... "oh you're being serious? No I am not doing that."
Enter Town Mayor
Mayor - "can you change that boys marks please? He is not failing"
Teacher - * laughs politely * .... "oh you're being serious? No I am not doing that."
Mayor - "you're sacked"
Headmaster - "yes, you're sacked"
Parent - "Ha, you're sacked"

The point of relating that story (where the child in question was part of an influential family in that town) is that blatant fiddling of data in ingrained into Chinese culture , one suspects to a far greater extent than in the education system of other countries. If you are 'someone', by association or by genetics, then you get what you need, it's simple, just change the grade, that's all you need to do. I've heard many a tale of the 'Publish or Perish' mantra and how it exhibits in Chinese universities with the pressures on staff and students to get their stuff into good journals (and often specifically, English-speaking journals; for years, the Chinese published loads and loads of research, but because so little of it was in English, no one had the faintest idea what they were doing).

User avatar
shpalman
Princess POW
Posts: 8241
Joined: Mon Nov 11, 2019 12:53 pm
Location: One step beyond
Contact:

Re: Allegations of Research Fraud at the Highest Levels in China

Post by shpalman » Mon Dec 02, 2019 1:15 pm

Allo V Psycho wrote:
Sun Dec 01, 2019 7:45 pm
... If Shpalman is reading this thread, I'd be very interested to know what he thinks, from a very different field.
Well there are a couple of different issues in these posts. One is what we might call overoptimistic statistical manipulation rather than actual fabrication of data. This is less likely to happen in my actual specific field because in the end what matters is if the devices actually work or not, and they have to demonstrably work rather than only turn out to work just a bit better than noise following statistical analysis. Not to say that noise isn't sometimes published as if it were data: just look at Fig. 6 of A micromachining-based technology for enhancing germanium light emission via tensile strain if you have access. But that got quickly pulled up (high-profile articles which claim to achieve long-sought results attract rivals ready to shoot them down hard given the slightest possibility...).

In more fundamental physics, well, I refer the reader to backreaction, the blog of Sabine Hossenfelder. While this is interesting, I'm mainly referring to her posts about experimental and theoretical particle physics: http://backreaction.blogspot.com/2019/1 ... ticle.html

The tl;dr is that theorists invent new particles to make their theories look nicer, and keep themselves busy writing lots of papers about these new particles, while the experimentalists fail to see anything, so the theorists decide they must have been just beyond the reach of the current particle accelerator (again) so it's worth it to build a new one (again). There's no reason that these new particles have to actually exists beyond making theorist's theories look nicer, though, so there's no reason to expect that the particles exist at all let alone that they'd be accessible if we only upgraded CERN a bit.

The other issue is the one of research assessment exercises and the like; the approach taken by the department here is a bit more magnanimous, especially as (well the rules are slightly different every time) everyone submits a certain number of works for consideration but the department can't submit the same work twice, so collaborative efforts are shared amongst us so that everyone gets their share of credit. Additionally, it's been known to happen that people who haven't really done much for a paper get added to the list of authors for political reasons (maybe they're more focused on teaching rather than research but still need to demonstrate some research output) which is a bit dodgy; however, I would rather have more people on the author list rather than less, I don't believe that my contribution is "diluted" by this.
having that swing is a necessary but not sufficient condition for it meaning a thing
@shpalman@mastodon.me.uk

User avatar
shpalman
Princess POW
Posts: 8241
Joined: Mon Nov 11, 2019 12:53 pm
Location: One step beyond
Contact:

Re: Allegations of Research Fraud at the Highest Levels in China

Post by shpalman » Mon Dec 02, 2019 1:16 pm

Don't, however, get me started on technology spin-offs...
having that swing is a necessary but not sufficient condition for it meaning a thing
@shpalman@mastodon.me.uk

Bewildered
Fuzzable
Posts: 257
Joined: Tue Nov 12, 2019 2:51 pm

Re: Allegations of Research Fraud at the Highest Levels in China

Post by Bewildered » Sat Dec 07, 2019 10:21 pm

I can’t click through to the article because I am currently in China where it is blocked by the firewall and my VPN is being effectively blocked too for the most part. I am here to discuss a job in academia, I will get a medical, see what the place is like and discuss whether I sign a contract for it...

Edit: however this place is currently not blocked by the great firewall, which is good.

Bewildered
Fuzzable
Posts: 257
Joined: Tue Nov 12, 2019 2:51 pm

Re: Allegations of Research Fraud at the Highest Levels in China

Post by Bewildered » Sat Dec 07, 2019 11:02 pm

Allo V Psycho wrote:
Sun Dec 01, 2019 7:45 pm
Pucksoppet wrote:
Sun Dec 01, 2019 6:40 pm
Following on from a few links in comments in the above, I stumbled across this:

Not even trying: the corruption of real science

It's why I think places such as this are important, as are the efforts of Ben Goldacre and Nick Brown. We need more people like them in positions of influence.
Interesting, but yes, bitter.
I've not worked in lab science for a while, which is what I think he is addressing. I did work in fairly similar fields back then, and his time scale stretches back to my lab science days, so am not completely unqualified to comment.
Where I agree:
I think a major problem has been the rise of 'managerialism': the view that what people do is inherently sloppy/lazy, and they need to be 'monitored' or 'measured' all the time to bring them up to scratch. The Research Excellence Framework is a good example of this.
We had a Faculty meeting about REF early in the cycle, where it was laid down that we had to produce one 3 or 4 star paper every year. And we had to work on approved themes, to match the University narratives. Since I'm sufficiently successful in my current field that I meet REF needs (and don't give a sh.t about senior managers), I felt I had to intervene on behalf of junior staff, and point out that this is absolutely not how research works. Sometimes nothing works, even when you have the right ideas. Sometimes you have the wrong ideas. I spent about three years chasing something completely worthless in retrospect. Yet because research is frustrating and difficult, you have to work on something you love, wherever it takes you, not on a departmental theme. The pressure to 'get grants' (as opposed to 'doing research' ) doesn't help either. A colleague of mine who had just published a brilliant paper in Nature was publicly described at a Departmental meeting as a 'research zero' because she worked in a very inexpensive field that didn't need grants. The only way to guarantee you would meet your departmental REF target - would be to fake the data. And in the long run, you would probably still be found out.
Where I don't:
As far as I can see, the people entering science (and it is not easy: it takes real dedication and effort, unlike the jobsworths Charlton seems to be describing) are the same curious, interested and honest people they ever were. Most of them could make more money in other occupations, with less work. But they want to know what reality is, and that keeps them curious, and, as a by-product, honest.
And I disagree strongly with his comments about narrowing specialisation. On the contrary, the fields I worked in have undergone a massive convergence. Where there was once molecular biology, cell biology, cancer biology, developmental biology etc., there is now just one massive overlapping field, where everybody's work informs everybody else's.

It's true I think that we are in a period of 'normal science' in lab biology, rather than a Kuhnian 'paradigm shift' phase, so much of the work seems a little boring. (This is why I changed fields). But that is what science is mostly like. A paradigm shift only comes along at an interval of decades.

If Shpalman is reading this thread, I'd be very interested to know what he thinks, from a very different field.
I don’t know if it is specifically managerialism that cause fraud, and as per above I can’t read the article and I feel
like this is more a general complaint about this problem*. However I strongly agree with a lot of the problems you write here. Also these thing are everywhere, and you seem to be speaking from uk experience, though it varies a bit and my impression is that China is on the worse end of this, but Australia and US are also very bad, with Germany being the place I see as least affected. I know that in China there was previously a lot of pressure to publish before, so there was an insentive for high volume low quality papers, but this was recognised by administrators so now they rank more by journal impact factor and citation count, which is also problematic for many reasons. I was reading a Facebook post by a senior person in my field saying he refused to referee for nature because in our field we should not publish there — the reason being its not possible to explain what we do in a way that is at all useful to anyone actually doing reseearch directly related and the peer review would be impossible because the technical details would all be missing. However nature has a much higher impact factor than in any journal we can publish in and I have even received referee reports on grants (clearly the referee is not someone actually in our field) saying we don’t have enough nature papers, as well as getting that same unrealistic criticism from my head of department when discussing possibilities to be made permanent.

FWIW the pressure to publish on fashionable topics and ones that will deliver early results, rather than what you think is good long term is almost unavoidable in my field, unless you just happen to have a very good record on paper from doing what you wanted. It is very sad and I feel there is no choice but to find a balance but the joy gets sucked out more and more.

I also agree with you about the people enemtering science. However I am already seeing that the ones who survive are ones who most quickly realise they need to “play the game” to some extent, rather than necessarily being those with best ideas, ability or most rigorous calculations etc.


* I say that not as a criticism, but because I am about to do exactly the same thing.

User avatar
jimbob
Light of Blast
Posts: 5276
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: Allegations of Research Fraud at the Highest Levels in China

Post by jimbob » Sat Dec 07, 2019 11:30 pm

shpalman wrote:
Mon Dec 02, 2019 1:15 pm
Allo V Psycho wrote:
Sun Dec 01, 2019 7:45 pm
... If Shpalman is reading this thread, I'd be very interested to know what he thinks, from a very different field.
Well there are a couple of different issues in these posts. One is what we might call overoptimistic statistical manipulation rather than actual fabrication of data. This is less likely to happen in my actual specific field because in the end what matters is if the devices actually work or not, and they have to demonstrably work rather than only turn out to work just a bit better than noise following statistical analysis. Not to say that noise isn't sometimes published as if it were data: just look at Fig. 6 of A micromachining-based technology for enhancing germanium light emission via tensile strain if you have access. But that got quickly pulled up (high-profile articles which claim to achieve long-sought results attract rivals ready to shoot them down hard given the slightest possibility...).

In more fundamental physics, well, I refer the reader to backreaction, the blog of Sabine Hossenfelder. While this is interesting, I'm mainly referring to her posts about experimental and theoretical particle physics: http://backreaction.blogspot.com/2019/1 ... ticle.html

The tl;dr is that theorists invent new particles to make their theories look nicer, and keep themselves busy writing lots of papers about these new particles, while the experimentalists fail to see anything, so the theorists decide they must have been just beyond the reach of the current particle accelerator (again) so it's worth it to build a new one (again). There's no reason that these new particles have to actually exists beyond making theorist's theories look nicer, though, so there's no reason to expect that the particles exist at all let alone that they'd be accessible if we only upgraded CERN a bit.

The other issue is the one of research assessment exercises and the like; the approach taken by the department here is a bit more magnanimous, especially as (well the rules are slightly different every time) everyone submits a certain number of works for consideration but the department can't submit the same work twice, so collaborative efforts are shared amongst us so that everyone gets their share of credit. Additionally, it's been known to happen that people who haven't really done much for a paper get added to the list of authors for political reasons (maybe they're more focused on teaching rather than research but still need to demonstrate some research output) which is a bit dodgy; however, I would rather have more people on the author list rather than less, I don't believe that my contribution is "diluted" by this.
The classic example I can think of is the Schön case
Have you considered stupidity as an explanation

Bewildered
Fuzzable
Posts: 257
Joined: Tue Nov 12, 2019 2:51 pm

Re: Allegations of Research Fraud at the Highest Levels in China

Post by Bewildered » Sun Dec 08, 2019 12:45 am

shpalman wrote:
Mon Dec 02, 2019 1:15 pm

In more fundamental physics, well, I refer the reader to backreaction, the blog of Sabine Hossenfelder. While this is interesting, I'm mainly referring to her posts about experimental and theoretical particle physics: http://backreaction.blogspot.com/2019/1 ... ticle.html

The tl;dr is that theorists invent new particles to make their theories look nicer, and keep themselves busy writing lots of papers about these new particles, while the experimentalists fail to see anything, so the theorists decide they must have been just beyond the reach of the current particle accelerator (again) so it's worth it to build a new one (again). There's no reason that these new particles have to actually exists beyond making theorist's theories look nicer, though, so there's no reason to expect that the particles exist at all let alone that they'd be accessible if we only upgraded CERN a bit.
I also can't click through to see the blog because of the great firewall, but I have read a bunch of her blogs saying stuff which sounds like your summary and I do actually work in High Energy Physics. I think you should be careful when reading blogs like that, they come from a very particular perspective and are criticising areas of research they are not directly involved in. This is loosely connected to a lot of bitterness that got created due to the feeling that only string theory was funded and not other ideas for quantum gravity with Sabine on one side of that. she is taking much wider aim generally, but what you wrote is a very skewed picture coming from a fairly shallow analysis of what people in HEP do.

First it is not just theoretical arguments motivating the new physics at higher energies, there is also the observational problem of dark matter, for example and wanting to know if the physics changes at higher energies is a very obvious line of enquiry, regardless of what theoretical ideas motivate it, just because we have seen new things after increasing the energy scale many times, and it is an unknown frontier yet to be explored.

Regarding the theory though, just to make sure we don't start from a perspective that unless there is an experimental problem we should ignore theorists. the Higgs mechanism was proposed because there was a theoretical inconsistency rather than an experimental problem and it led to a nobel prize discovery. I guess you didn't have that in mind though and Sabine's criticism is certainly more nuanced, but there is absolutely nothing wrong with looking for better theories that unify things or explain them in a simpler way. An example closer to what I think you mean, is grand unified theories. The standard model of particle physics correctly describes all know gauge interactions, however it has quite a strange structure, with three different forces coming from three different gauge symmetries that look quite different to each other, while there are many possible gauge symmetries one could write down, so why these specific three? It is natural to ask why are there these three forces and why specifically do they look like this, or ask if they could be unified into a single force. You seem to be criticiizing any science motivated this way, which seems bizarre to me.

Sabine has a problem with at least one particular argument, the hierarchy problem, which can be phrased as the problem that standard model of particle physics cannot explain why the weak scale (~ set by the Higgs mass) is so much lighter than the Planck scale where gravitational effects become important. She is not alone in this, a lot of people don't like it though the reasons are quite varied, but most HEP theorists working on new physics, including me, do find the problem compelling and arguments about this go back to the 1970s . The criticism you apply seems here seems to be that we just shouldn't try to answer such questions, which again I think is bizarre. Or that maybe its more that you don't like the fact that they have not found new physics where it was initially searched for, but haven't given up trying to answer these problems. ie we should ignore the fact that experimental results don't answer the question and pretend this makes it meaningless, even though its clear there are many ways round the arguments that people put forward to say they expected this to lead to discovery.

Some of Sabine's argument are also plain wrong and I have the impression she is just unaware of relevant papers in the literature that make the hierarchy problem more rigorous than the original intuitive formulation.

Post Reply