Crosspost from @SC_Griffith@twitter.com:
this is a perfect science shitpost https://twitter.com/ScienceAdvances/status/1621252177057439745
🐦🔗: https://twitter.com/SC_Griffith/status/1623380240255819808
@tiago is it my impression, or is stuff like this becoming the norm in computational social science in prestigious journals? I see it all the time in nature human behavior, PNAS just this week as well....
@lmrocha I also think it's becoming more common.
You can get by with surprisingly shoddy methodology in CSS. All that matters is the size and originality of your dataset, it seems. Given that, you can claim whatever you want and no one cares.
There are entire fields of science where nonsense like this is normal. This is coming over to CSS, and editors either do not care or do not seem to be strong enough to catch this (this is not even a peer review problem, it's editorial).
@tiago yeah, and then publishing in venues like these gets you invited for keynotes, etc. It seems like these days one needs to just speak of polarization, disinformation or discrimination and have results that reinforce accepted wisdom :)
@lmrocha Indeed. It's often things like “Bears really do shit in the woods!” after tracking 6 million bears for a period of 20 years. And they won't release the data publicly.
