science fraud is common place

You again have some weird concept of my 'entire position' which has nothing to do with anything I've claimed.
Hi MCH. If any poster has misinterpreted your position then the fix is easy. Just state your position fully and coherently.

My interpretation of your position (if I am wrong then please state your position fully and coherently) is that you think that detection of any level of scientific fraud means that scientific fraud is "common place".
 
None of those things you give are quality controls. Neither peer review, replication, or community review. What qualities exactly are they supposed to assure?

How often are results replicated? Several people have commented above that replication is sometimes impossible, or too expensive, or cannot get funding, or is not interesting enough to do, etc, etc.

Peer review is for interest, not for quality.

You cannot start quality controls at publication - it has to start at data collection.

I think I can half agree. Peer review is quality control for sure, but has a blind spot for original data in some circumstances.

Quality is not just about the original data. The peer reviewers are examining the whole picture... basically, does the conclusion really follow from the data and method?

Regarding original data... peer reviewers have been able to detect fraudulent original data when it is 'unlikely' for statistical reasons (eg: we know some of the Targ&Puthoff telepathy data was adjusted prior to publication because it is 'too random')
 
heh yes I know; I find it quite fun that even 'in science' we know how to do rigour and scrutiny, but then only do it in very specific disciplines,
That is where I think you go a bit wrong, MCH. As far as I have seen "rigour and scrutiny" are common throughout science. But the definition of that "rigour and scrutiny" differs to be appropriate for the discipline, e.g. I certainly did not ask permission of any atom before enlisting it in an experiment :D.
 
wow, wrs, that sounds so ... absurdly sarcastic and completely missing the point :eek:!

Darn, I forgot to include the /sarcasm................


You do know that a requirement for empirical data to be generally accepted in science is that it is independently replicated?

I know that is a claim made by science. I don't know that it's an absolute practice. I believe that in cases where the data is proprietary that it's most certainly not the practice to let an objective third party replicate it.
 
Hi MCH. If any poster has misinterpreted your position then the fix is easy. Just state your position fully and coherently.

My interpretation of your position (if I am wrong then please state your position fully and coherently) is that you think that detection of any level of scientific fraud means that scientific fraud is "common place".
How does that follow?!

My position - such as it is - is that we don't know what the level of fraud in research is. That's not a reason to assume, or claim, that it's low, or is anything like the detection rate. And I would be interested in looking at ways of (1) detecting it and (2) if it's significant (for dodgy meanings of 'signficant') reducing it.
 
I think I can half agree. Peer review is quality control for sure, but has a blind spot for original data in some circumstances.

Quality is not just about the original data. The peer reviewers are examining the whole picture... basically, does the conclusion really follow from the data and method?

Regarding original data... peer reviewers have been able to detect fraudulent original data when it is 'unlikely' for statistical reasons (eg: we know some of the Targ&Puthoff telepathy data was adjusted prior to publication because it is 'too random')
So I must disagree that peer review is a quality control for the work; yes it sometimes controls certain aspects of the paper.
 
My initial premise is that fraud is commonplace in humanity. I base this on the fact that fraud is illegal. Therefore it must arise enough to be noticeable and thus a problem. We were discussing noticeable earlier in the thread. I said no higher than 6% maybe as low as 1-2%. Scientists are human and therefore fraud should be just as commonplace.

This is nonsense. If I had data showing that 5% of retail-store cashiers steal from the till, does that imply that 5% of bank tellers also steal from the cash drawer? (No, because it's hard to get away with the latter.) If I had data showing that 25% of accountants take naps at their desks, does that imply that 25% of receptionists take naps at their desks? (No, because receptionists' desks are much more public than accountants' desks.)
 
That is where I think you go a bit wrong, MCH. As far as I have seen "rigour and scrutiny" are common throughout science. But the definition of that "rigour and scrutiny" differs to be appropriate for the discipline, e.g. I certainly did not ask permission of any atom before enlisting it in an experiment :D.
heh, well that's an ethical issue that is between you and your conscience. And any atoms that might take you to tribunal.

I just don't think that the 'rigour and scrutiny' that is common to research is sufficient to label most research as 'scientific'.
 
How many papers are replicated? If only a few are replicated, then the detection rate will not be high.
The results of observations, experimental results, clinical results, etc. in papers have to be replicated before the papers are generally accepted. So every generally accepted paper has been replicated. I cannot give you any number - look then up if you want, e.g. counting the non-cited experimental papers.
It is not a "few", it is the "overwhelming majority".

N.B. We are not talking about experiment A is done and published, experiment B is done, is published and confirms A, and than everything stops!
The confirmation of empirical data never stops. Scientists often repeat experiments just to get more accurate results, e.g. measuring electron mass. There are undergraduate physics students in labs today that are testing Galileo's experiments and Newton's Laws!
 
We have some data in this thread that we can use to discuss estimates of rates of fraud.
Rather short discussion though, MCH :D, e.g.
0.0096% of biomedical papers in 2007 were retracted for fraud. Biomedical fraud rate in 2007 is thus between 0.0096% and 100%. Fraud rate in all disciplines between 0% and 100%. Repeat for any other year.
 
Alright, it's the economist, but:

http://www.economist.com/news/brief...elf-correcting-alarming-degree-it-not-trouble

" A few years ago scientists at Amgen, an American drug company, tried to replicate 53 studies that they considered landmarks in the basic science of cancer, often co-operating closely with the original researchers to ensure that their experimental technique matched the one used first time round.

According to a piece they wrote last year in Nature, a leading scientific journal, they were able to reproduce the original results in just six.

Months earlier Florian Prinz and his colleagues at Bayer HealthCare, a German pharmaceutical giant, reported in Nature Reviews Drug Discovery, a sister journal, that they had successfully reproduced the published results in just a quarter of 67 seminal studies.

Oh my gosh, do we now need to consider that scientists are just as prone to carelessness and not seeing their mistakes and seeing what they are looking for as the rest of humanity? Oh, nooooooooooooooooooooo...............

article said:
Fraud is very likely second to incompetence in generating erroneous results, though it is hard to tell for certain. Dr Fanelli has looked at 21 different surveys of academics (mostly in the biomedical sciences but also in civil engineering, chemistry and economics) carried out between 1987 and 2008. Only 2% of respondents admitted falsifying or fabricating data, but 28% of respondents claimed to know of colleagues who engaged in questionable research practices.
 
Last edited:
The results of observations, experimental results, clinical results, etc. in papers have to be replicated before the papers are generally accepted. So every generally accepted paper has been replicated. I cannot give you any number - look then up if you want, e.g. counting the non-cited experimental papers.
It is not a "few", it is the "overwhelming majority".

N.B. We are not talking about experiment A is done and published, experiment B is done, is published and confirms A, and than everything stops!
The confirmation of empirical data never stops. Scientists often repeat experiments just to get more accurate results, e.g. measuring electron mass. There are undergraduate physics students in labs today that are testing Galileo's experiments and Newton's Laws!
Hmmm but I think that's an assumption that doesn't actually happen in practice. See eg the link to the economist above - yes, sometimes people go to the effort of replicating but they often don't.

Worse, we don't have a way of telling when results are 'generally accepted'. Citations are not an indication of replication - let's face it, citation counts can indicate refutation rather than confirmation.

It depends heavily on the discipline. Some have taken on full quality controls throughout - medicine is one, I personally call that a science - and some are struggling. My rule of thumb is the data collection test: are there independent verifications that the data given in the work is the data collected? If not, then it's not scientific. It might be a discipline that produces science in a sporadic, unreliable way - ie, like gossip.
 
Last edited:
Rather short discussion though, MCH :D, e.g.
0.0096% of biomedical papers in 2007 were retracted for fraud. Biomedical fraud rate in 2007 is thus between 0.0096% and 100%. Fraud rate in all disciplines between 0% and 100%. Repeat for any other year.
Heh yes, but put it in context to what I was answering.

I also suggesting using scientific studies of college students, or other studies, for estimated fraud rates.
 
We are not forgetting it. We (well, I) just don't think that researchers are significantly different from the general public.
You missed the point I made, MCH:
What you and B]wrs[/b] are forgetting is that scientists are not the general public. They are under an extreme amount of inspection and they know this.
(my emphasis added)
The Bob example I gave before lists some of the inspection:
Bob notices some interesting data and creates a hypothesis to explain it. Then
* Bob looks for problems in the hypothesis.
* Bob's colleague Alice looks for problems in the hypothesis.
* Bob presents the hypothesis at a symposium to his colleagues.
* Bob puts a poster up at a conference.
* Bob presents the hypothesis at another conference.
* After dozens of scientists have had a look for problems with the hypothesis, Bob writes it up and submits a paper to a journal.
* Now the peer review process starts!
* The paper is published.
* Hundreds of scientists look for problems in the paper.
* Every time that the paper is cited, it is implicitly reviewed again by the authors of the new paper.

It is not unreasonable to start with the assumption that a researcher is bog-standard human. It is unreasonable to stop there :eek:! You have to also understand the researcher's situation.

And remember - we are talking about scientific fraud. The inspection is not limited to the execution of an experiment as you imply.

Of course I am also ignoring another side of "scientists are not the general public". Scientists go through years of training to be honest in an environment where fraud is held to be the deadliest sin. This does not happen to the general public.
 
Oh my gosh, do we now need to consider that scientists are just as prone to carelessness and not seeing their mistakes and seeing what they are looking for as the rest of humanity? Oh, nooooooooooooooooooooo...............
And, ironically, it's scientists that did the research.

I have a half-written paper for peer review showing how peer review is worthless. I can see there being a problem with this.
 
You missed the point I made, MCH:

(my emphasis added)
The Bob example I gave before lists some of the inspection:


It is not unreasonable to start with the assumption that a researcher is bog-standard human. It is unreasonable to stop there :eek:! You have to also understand the researcher's situation.

And remember - we are talking about scientific fraud. The inspection is not limited to the execution of an experiment as you imply.

Of course I am also ignoring another side of "scientists are not the general public". Scientists go through years of training to be honest in an environment where fraud is held to be the deadliest sin. This does not happen to the general public.

No I think we're at cross purposes. Where the inspections actually do really apply, then we can count the results as being (more) reliable. My point is that the inspections often don't. Especially the small inspections: did this data actually come from the instrument?

It's the inspections that make the difference, not some label ('scientist') applied to the researcher.

Where the inspections don't happen, we revert to ordinary human behaviour.
 
Of course I am also ignoring another side of "scientists are not the general public". Scientists go through years of training to be honest in an environment where fraud is held to be the deadliest sin. This does not happen to the general public.

Would that be like priests going to seminary for 8 years and knowing that homosexuality, self-abuse and boy molestation are all mortal sins?
 
None of those things you give are quality controls. Neither peer review, replication, or community review.
That is wrong, MCH. You seem to be under the impression that quality control only happens in experiments.
If you cannot replicate something then its quality is abysmal.
Peer review is the review of the quality of the paper. Here is a good explanation: The 4 Jobs of a Referee in Peer Review
Community review is the review of the quality of the paper.
That is quality control.
 
Last edited:
Of course I am also ignoring another side of "scientists are not the general public". Scientists go through years of training to be honest in an environment where fraud is held to be the deadliest sin. This does not happen to the general public.
Sorry, missed this. I just don't accept it without extraordinary supporting evidence; there are other incentives to scientists to publish, to be 'right', to raise funds, etc that we can reasonably expect to override honesty in small ways, at least.
 
That is wrong, MCH. You seem to be under the impression that quality control only happens in experiments.
If you cannot replicate something then its quality is abysmal.
Peer review is the review of the quality of the paper.
Community review is the review of the quality of the paper.
That is quality control.
yes I should have been more specific - I did say this later above - but quality controls of a paper are not quality controls of the work. It's too late. My concern is that quality controls don't happen *even* in experiments.

Often things aren't replicated. What then?
 

ISF - Join now!

Every member here is approved by hand. No bots, no spam, just people who care about evidence and honest debate.

Membership is free!

Create your free account

Back
Top Bottom