Apparently there is evidence that interdisciplinary work is viewed as more valuable after the fact, but less valuable before. (cc @davidmanheim, evidence supporting your tweets earlier)pic.twitter.com/BT7Cg8Pw35
Show this thread
-
Something I do wonder about in connection with the equal odds rule across a scientists' career: does it take into account the fact that in many fields citation rates are (I believe) rising? If so, the numbers should be rescaled.
Show this thread
The probability density for a paper's number of citations, normalized by the average number of citations for its discipline. The is essentially identical across disciplines(!) This is honestly quite remarkable.pic.twitter.com/SQvHcmabee
-
-
It's striking how much the paper focuses on citations and similar measures, but not on things like understanding, quality of explanations, deep new ideas, and so on; nor on economic impacts. Lots of proxy measures, rather than the reasons we want to do science in the first place
Show this thread -
Put another way, it's not so much about the science of science, as the science of the extremely limited set of things we can measure about science. Citations are not uninteresting, but there's a reason Boltzmann's grave has S = k log W on it, not his h-indexpic.twitter.com/RVhOGtTaEA
Show this thread -
Anyways, it's a useful survey paper, with much that was new to me in it. And it was fun to make a first pass over it "in public" like this
Show this thread
End of conversation
New conversation -
-
-
The fact that such scaling laws exist that are stable against different scientific cultures in various disciplines is remarkable.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Agree. I thought that figure was the most interesting (and surprising) thing in the paper.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Hmmn, yeah, but it is lognormal, and it would have to be sthing like that. Not normal in this case, because non-negative and heavy tail. And they have folded out one DOF by normalizing, is there a DOF left that magically agrees across fields?
-
Here's a model (not serious). Value of a paper is Gaussian, while payoff to being better than the next guy is exponential.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.