I just realized there's a better way to control for some of these effects using multivariate models. Another update will come later.
-V -- tor-talk mailing list - [email protected] To unsubscribe or change other settings go to https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-talk
