3 ms·
Disclosure: also have PhD and leading a lab for of PhD students I think not the PhDs are the problem, but the amount of papers. We just need all the PhD studen
by riedel 3y ago
Disclosure: also have PhD and leading a lab for of PhD students
I think not the PhDs are the problem, but the amount of papers. We just need all the PhD students to produce those papers to keep the industry going (much like a snow ball system). There is no incentive to scale down, so we keep scaling up.
The problem IMHO ist that we have scaled our methods to cope with enormous amount of mostly empirical results. P-value hacking is implicit in this system. IMHO we can throw away two thirds of all published research in many fields because it is not reliably reproducible/generalizable. If we could/would use all those PhD students to put science on better grounds that would be great. However, we expect all those PhDs to produce new results, so we get what we wished for.
- logifail 3y ago> We just need all the PhD students to produce those papers to keep the industry going (much like a snow ball system). I'm not sure we agree on the definition of "we just need" in that! > There is no incentive to scale down, so we keep scaling up. Scaling up waffle just produces more waffle, and no-one actually wants that. We seem unable to agree on how to judge progress in academia, so we end up counting papers. Which means everyone's trying to publish waffle.
- jltsiren 3y agoJudging progress is easy enough. Trusting the judgment is hard, because the world is small and conflicts of interest are inevitable. People who are established experts on a topic generally know each other. Because taxpayers don't want to fund someone's research just because the applicant's friends say so, they need ways of judging the merits without the actual expertise to do so.