3 ms·

The reluctance of the American academic bureaucracy to admit to the prevalence of fraud in academic research is understandable, but there are many examples. So

The reluctance of the American academic bureaucracy to admit to the prevalence of fraud in academic research is understandable, but there are many examples. Some of the recent notables are the serotonin hypothesis of depression, and aspects of the amyloid theory of Alzheimer's. Data falsification is not uncommon, and it's often done for obvious reasons - a graduate student may want that positive result that finishes off their PhD work and gets them a needed publication, a professor may want to support their past work with new data and results, an academic startup may want to prop up their latest drug development results to draw in investors and big corporations, etc. The reason academic institutions don't want to talk about it is the same reason that religious institutions don't want to talk about child abuse - it makes a mockery of their whole enterprise. With the former, it's the search for truth via the scientific process, and with the later it's upholding moral standards and guiding society. Both institutions cover up such things in order to preserve their public image - which, of course, allows such behaviors to persist. Note also that the corporatization of research, i.e. the focus on generating lucrative patents and the resulting emphasis on keeping raw data secret, tends to amplify opportunities for fraud.

Misdicorl 4y ago
Citing a few highly damaging cases of fraud isn't enough to make your point. You have to show that it's more prevalent and damaging in the US system than in other systems. That seems an extraordinarily unlikely bar to meet based on my cursory knowledge of where large scale fraud is happening. But perhaps you know more and would like to enlighten me.
krastanov 4y ago
In the last 15 years academic institutions have definitely started talking about it, especially in new mandatory classes on ethics for their students.
jll29 4y ago
> The reason academic institutions don't want to talk about it is the same reason that religious institutions don't want to talk about child abuse - it makes a mockery of their whole enterprise. This is why in science, replication of past work is essential. Every new Ph.D. student should replicate a paper's results in their area of interest as a warm-up exercise. In Information Retrieval (the part of Computer Science concerned with search for information - search engines etc.), conferences like ECIR have regular reproduceability tracks, where such works of scrutiny get published, and are highly regarded. There needs to be a cultural change to the effect that genuine errors and fraud can happen, but we can weed it out to an extent by independently repeating past work.
psyklic 4y ago
I wasn't aware that the serotonin hypothesis was "fraudulent," nor can I find credible evidence of this on Google. Instead, it seems that the scientific process worked as intended. A reasonable hypothesis was initially proposed, then opposing evidence accumulated over time. This seems moreso an issue with the pharmaceutical industry. For example: "In 2005, two academics published a piece of research in which they compared the information on pharmaceutical websites with the pronouncements of certain researchers, and found a ‘disconnect’ between the marketing and the experts’ views." - https://joannamoncrieff.com/2022/07/24/how-to-take-the-news-that-depression-has-not-been-shown-to-be-caused-by-a-chemical-imbalance/ https://joannamoncrieff.com/2022/07/24/how-to-take-the-news-...

Misdicorl 4y ago

Citing a few highly damaging cases of fraud isn't enough to make your point. You have to show that it's more prevalent and damaging in the US system than in other systems. That seems an extraordinarily unlikely bar to meet based on my cursory knowledge of where large scale fraud is happening. But perhaps you know more and would like to enlighten me.

krastanov 4y ago

In the last 15 years academic institutions have definitely started talking about it, especially in new mandatory classes on ethics for their students.

jll29 4y ago

> The reason academic institutions don't want to talk about it is the same reason that religious institutions don't want to talk about child abuse - it makes a mockery of their whole enterprise. This is why in science, replication of past work is essential. Every new Ph.D. student should replicate a paper's results in their area of interest as a warm-up exercise. In Information Retrieval (the part of Computer Science concerned with search for information - search engines etc.), conferences like ECIR have regular reproduceability tracks, where such works of scrutiny get published, and are highly regarded. There needs to be a cultural change to the effect that genuine errors and fraud can happen, but we can weed it out to an extent by independently repeating past work.

psyklic 4y ago

I wasn't aware that the serotonin hypothesis was "fraudulent," nor can I find credible evidence of this on Google. Instead, it seems that the scientific process worked as intended. A reasonable hypothesis was initially proposed, then opposing evidence accumulated over time. This seems moreso an issue with the pharmaceutical industry. For example: "In 2005, two academics published a piece of research in which they compared the information on pharmaceutical websites with the pronouncements of certain researchers, and found a ‘disconnect’ between the marketing and the experts’ views." - https://joannamoncrieff.com/2022/07/24/how-to-take-the-news-that-depression-has-not-been-shown-to-be-caused-by-a-chemical-imbalance/ https://joannamoncrieff.com/2022/07/24/how-to-take-the-news-...