3 ms·
You don't really need 10 years. Each researcher at UC might be confounded by their local network, but you could argue that is just a casual variable anyway. As
by ImaCake 6y ago
You don't really need 10 years. Each researcher at UC might be confounded by their local network, but you could argue that is just a casual variable anyway. As for statistical power, UC has >30 researchers and each one of those researchers would access several papers a week. Take the ratio of sci-hub vs non-sci-hub papers accessed per week or month and compare to the month before UC's contract with elsevier ran out. Take the mean and do a t-test. Don't even need timeseries!
I suspect plotting sci-hub/non-sci-hub papers accessed per week for several weeks before and after the contract expiry would dispense with the need for any statistical analysis.
- smichel17 6y agoOr maybe there's been little noise about the change because everyone was using sci-hub already anyway! I recently introduced a part-time professor to sci-hub, when he was having trouble getting his vpn connection to work, so he could access a paper in a journal. Open access is just easier.
- sitkack 6y agoDoes sci-hub release geographic group-by ? She should, I have a corp all access pass to basically everything, and I still sci-hub. The racist, classist control of information has to stop. It should basically be a reverse paywall.