I named a variable wrong, which has no substantive effect on the 'paper ballot' coefficient but flips colplus/black coef. in a sensible waypic.twitter.com/k6va7wF9ui
Effect of paper ballots in Wisconsin goes from 7 pts, like NY article, to 0 if you control for race education, density (true w&w/o weights,)pic.twitter.com/3ZVfDa44Zn
-
-
-
hmm almost like tweeting out a hot-take analysis is not a great idea
-
i blame the wine, not the heat of the take
-
a very fair counterpoint
End of conversation
New conversation -
-
-
You are using d-shift as dependent variable? Why? That doesn't seem to be their dependent variable. Try with dem margin?
-
it's the same, though i interpreted their dv to be dshift
-
I think at this point, they should spill their data. Also, only WI is this close, no, with enough no paper machines?
-
It looks to me like all electronic voting machines in WI include printer for VVAT https://www.verifiedvoting.org/verifier/#year/2016/state/55 …
End of conversation
New conversation -
-
-
why don't you use R studio?
-
tools > global options > appearance > editor theme. Don't know which one
@Nate_Cohn uses but it's not the default white bg. -
thanks! That's why I didn't recognize it
End of conversation
New conversation -
-
-
Note that Ron Rivest is advocating audits in several states as well. He's not exactly a tin-foil hat guy either.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
are you running this on all counties, or just w/ population <50k, and/or used electronic voting? R output doesn't show n-size.
-
looks like Nate Silver is restricting by population, so good to know that effect disappears when all counties included
-
Oh wait. Does output change if you run multiple linear regression for each county? Are there counties where effect isn't null?
End of conversation
New conversation -
-
-
Lol you're about 15 minutes behind the other Nate
-
dun dun dun. Tuesday night, Nate v Nate, stats nerds- go!
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.