Nate Cohn

@Nate_Cohn ↗

Deleted No

Hibernated No

Last Checked May 24, 2022

Created

Thu Mar 03 22:11:36 +0000 2022

Likes

Retweets

Source

Twitter Web App

View Raw Data

JSON Data

View on Twitter

Likely Available

Nate Cohn @Nate_Cohn

@davidshor @lxeagle17 @SpecialPuppy1 @hodgesmr @mehdirhasan @ashe_cs it's kind of the canonical issue, albeit a novel and extreme enough example of it that you probably wouldn't anticipate it unless you tried it yourself — PolitiTweet.org

Posted March 3, 2022

Preceded By

Nate Cohn @Nate_Cohn

@davidshor @lxeagle17 @SpecialPuppy1 @hodgesmr @mehdirhasan @ashe_cs one last point, with relevant to the logit: it's quite clear to me that census tract col+ had a lot of predictive value on vote flipping, beyond individual level educational attainment. helps explain GA v MS to some extent — PolitiTweet.org

Posted March 3, 2022

Followed By

Nate Cohn @Nate_Cohn

@arnonmishkin @davidshor @lxeagle17 @SpecialPuppy1 @hodgesmr @mehdirhasan @ashe_cs it would directly account for urbanicity, since these models are fit at the individual level, have geographic predictors and are adjusted by precinct. it would only indirectly account for racial resentment to the extent it effects precinct vote, etc. — PolitiTweet.org

Posted March 3, 2022