4 ms·
> Most doesn't mean there isn't a lot of local fraud. Elizabeth Holmes comes to mind. Let's not forget that fraudulent Alzheimer's paper that misdirected Alzh
by WalterBright 2y ago
> Most
doesn't mean there isn't a lot of local fraud. Elizabeth Holmes comes to mind.
Let's not forget that fraudulent Alzheimer's paper that misdirected Alzheimer's research for decades.
The Pons and Fleischmann cold fusion fraud?
Does it matter if most come from foreign scientists, when prestigious journals publish them? What are we supposed to do, reject papers with foreign sounding names? Why aren't the journals doing proper vetting?
Attempts to suppress any counterclaims to the official scientific opinions during covid would make anyone skeptical.
Then there's this:
https://www.oaklandreport.org/p/the-misrepresentation-of-oakland https://www.oaklandreport.org/p/the-misrepresentation-of-oak...
https://www.wsj.com/articles/the-media-say-crime-is-going-down-dont-believe-it-e5b07784 https://www.wsj.com/articles/the-media-say-crime-is-going-do...
One should be careful about statistics reported by an entity that has a strong self-interest in the outcomes.
- matthewdgreen 2y agoImagine I told you the US was overrun by violent crime and murder, and when asked for examples I listed: * A single horrible murder took place decades ago * A second (possibly less terrible) manslaughter case that took place in the 1980s * A variety of violent acts in dangerous areas in developing countries, today I’m not even going to dispute that there is academic fraud. Just that your examples tend to illustrate how relatively rare it is when one filters for reputable researchers, journals and institutions, and also takes into account the denominator (which is enormous) rather than highlighting some spectacular examples and making broad claims about the world falling apart.
- WalterBright 2y agoI provided two cites of manipulation of aggregate statistics. I've also seen articles (didn't keep cites) about the "replication crisis" of large numbers of research papers with results that nobody can replicate. > making broad claims about the world falling apart More like claims of losing trust in "trust us, because we're scientists!"
- bumby 2y agoDo we have good data to the contrary, though? Is there evidence that most science is good? The replicability crisis has certainly given some evidence that much of published research could be classified as 'not good' in some domains. (And maybe there is an important distinction to make between 'bad' science and 'fraudulent' science). If there isn't good evidence to show that most science is good, the best we can probably claim is that we don't know.
- matthewdgreen 2y agoI don’t think “there is an important distinction between ‘bad’ science and ‘fraudulent’ science” should have the word “maybe” attached to it. Of course there is a difference, just as there is a huge distinction between premeditated murder and accidental deaths. The correct level of both should be zero, but one is vastly more damaging and also more rare than the other. As an aside: the problem I have with the OP is that it isn’t arguing about scientific quality, it’s arguing that science is rife with actual fraud. Typically in the midst of such conversations the conversation switches from actual accusations of fraud to the replication crisis in small-effect-size social science experimental work, which seems to be more an issue of people being careless than malicious —- and importantly, is an issue that the scientific community is addressing through improved standards. With respect to your question: “is there evidence that most science is good”, I’m not even sure how you would answer that question. All I know is that if you compare the world of my grandmother to the world of today, the two are so different as to be unrecognizable, and scientific advances are responsible for most of the difference. That progress continues every single day, and we’re both alive because of science was at least good enough to make our continued existence possible.
- bumby 2y agoIt reads like most of your issue is with my phrasing, which was deliberate to soften the comment so it doesn’t come across as accusatory. >I’m not even sure how you would answer the question. It should probably be answered in the same way we’d answer the converse. Something like “here is the data and the tests that show the data quality is high and the results are replicable.” My point was, if you can’t show that, then the claims that the work is good may not be scientifically valid. To put a finer point on it, we know there is replicability crisis in many corners of science. But we don’t likely know how much of that is due to bad science or outright fraud unless we do the science-y work to check. And unfortunately many of the incentives are aligned against that very thing.