Nate Cohn

@Nate_Cohn ↗

Deleted No

Hibernated No

Last Checked April 2, 2021

Created

Tue Mar 02 19:47:49 +0000 2021

Likes

Retweets

Source

Twitter Web App

View Raw Data

JSON Data

View on Twitter

Likely Available

Nate Cohn @Nate_Cohn

@pollcat well, the conclusion that error on issues < error on pres. ballot only holds if you assume that the source of nonresponse is presidential vote choice. if it's something else, it's possible that it's more highly correlated with another issue question than presidential vote — PolitiTweet.org

Posted March 2, 2021

Preceded By

Nate Cohn @Nate_Cohn

@pollcat i'm not sure i'd go so far as to say that this exercise is misleading, but i do think this is making a huge and undefended assumption: that presidential vote choice is the source of nonresponse bias, and not another variable correlated with presidential vote choice — PolitiTweet.org

Posted March 2, 2021

Followed By

Nate Cohn @Nate_Cohn

@pollcat and i think it's very easy to imagine an 'x' factor on such a large scale, given what we already see in various benchmarking exercises/authoritative government data. we wouldn't believe polls could be off by 15 pts on the BA+ share of adults if we didn't have proof! — PolitiTweet.org

Posted March 2, 2021