3 ms·

Are you referring to the still ignored replication crisis? https://en.wikipedia.org/wiki/Replication_crisis https://en.wikipedia.org/wiki/Replication_crisis E

Are you referring to the still ignored replication crisis? https://en.wikipedia.org/wiki/Replication_crisis https://en.wikipedia.org/wiki/Replication_crisis Experts say, scientists say x may, yet noone talks about the underlying shakiness of all their data.

rualca 5y ago
"shakiness" is not the problem at all because papers are expected to serve as a perpetual source of observations intended to serve as food for thought. Therefore, as knowledge at the bleeding edge is always incomplete and being constructed as we speak, it's very nature is shaky. The problem being discussed is far greater and with far reaching repercussions. The problem is somewhere between bad science pushing bogus ideas and outright scientific fraud. The Royal Society's motto is quite literally "take no one's word for it". That's the Hallmark of science. Yet, today's academia is completely ignoring their most basic tenets and just accepting whatever bullshit idea is peddled by anyone without the faintest requirement of effort to verify it's validity?
verdverm 5y ago
There was a study of psych papers that had worse results than the CS one (which might be cited here https://cacm.acm.org/magazines/2020/8/246369-threats-of-a-replication-crisis-in-empirical-computer-science/fulltext https://cacm.acm.org/magazines/2020/8/246369-threats-of-a-re...) Having trouble finding them again... I experienced the effects of this as a PhD student. While trying to write a paper and compare results, I was unable to reproduce the competing ideas results. I came across one author who published better results without code or data any time anyone compared to and beat his results. I suspect he is highly fraudulent. He only seems to publish in Springer rather than they typical conferences for the field

3 ms·

Are you referring to the still ignored replication crisis? https://en.wikipedia.org/wiki/Replication_crisis https://en.wikipedia.org/wiki/Replication_crisis E

by yeuxardents 5y ago

rualca 5y ago
"shakiness" is not the problem at all because papers are expected to serve as a perpetual source of observations intended to serve as food for thought. Therefore, as knowledge at the bleeding edge is always incomplete and being constructed as we speak, it's very nature is shaky. The problem being discussed is far greater and with far reaching repercussions. The problem is somewhere between bad science pushing bogus ideas and outright scientific fraud. The Royal Society's motto is quite literally "take no one's word for it". That's the Hallmark of science. Yet, today's academia is completely ignoring their most basic tenets and just accepting whatever bullshit idea is peddled by anyone without the faintest requirement of effort to verify it's validity?
verdverm 5y ago
There was a study of psych papers that had worse results than the CS one (which might be cited here https://cacm.acm.org/magazines/2020/8/246369-threats-of-a-replication-crisis-in-empirical-computer-science/fulltext https://cacm.acm.org/magazines/2020/8/246369-threats-of-a-re...) Having trouble finding them again... I experienced the effects of this as a PhD student. While trying to write a paper and compare results, I was unable to reproduce the competing ideas results. I came across one author who published better results without code or data any time anyone compared to and beat his results. I suspect he is highly fraudulent. He only seems to publish in Springer rather than they typical conferences for the field