Hacker News Reader

Comment by tptacek | Hacker News Reader

When people talk about stories like this they're invariably discussing the numerator without the denominator. Count up every story about research malpractice that has ever been on HN, and you're not even in the vicinity of 1% of all the research done every year; a truly huge amount of research is published every year, in every field.

View on HN

The question is how much of that stuff is fraudulent or at least suspect: you'll find that most scientists are skeptical of a large fraction of papers published in their field, there is a lot of crap that gets through peer review, and most will not get openly questioned because it's a lot of effort and quite high-risk, especially if it's a result that's of low importance.

I am probably even crazier

I see that the mathematics ultimately provide the basis for a lot of that research, which can be summed up as measuring various things, recording the measurements, and publishing the whole thing. But it's all built on the same essential math at the foundation.

and the mathematics is duplicated a lot to, but this is hard to see, you'd need a 'real' expert, of the kind that can be hand counted

Yes, the label "peer review" is meaningless. There are good peer review processes at top journals and conferences and getting through those is a meaningful accomplishment. Then there are also worthless peer review processes at lower quality avenues. Getting through those is like getting a participation trophy. Yes, you need to show up on time, but not much more than that.

The real peer review happens afterwards, when other researchers actually engage with your work.

This is true, but as in many areas, power laws apply. The vast majority of research has very low impact and mostly serves fulfilling publication quotas. There is likely plenty of scientific fraud happening in lower tier publications as well, but no one notices because no one cares.

Reply on news.ycombinator.com