Nate Silver

@NateSilver538 ↗

Deleted No

Hibernated No

Last Checked Dec. 22, 2020

Created

Thu Dec 03 22:26:14 +0000 2020

Likes

558

Retweets

Source

Twitter Web App

View Raw Data

JSON Data

View on Twitter

Likely Available

Nate Silver @NateSilver538

Yeah, it's the pollsters who do election polls who put their necks on the line. So I have little patience for polling firms that *don't* do election polling saying that somehow the problems that affect election polling don't affect their polls. — PolitiTweet.org

And in fact some issue pollsters (*cough* Gallup) got some elections wrong and then stopped because getting that ri… https://t.co/XOJWX6yN8G

Posted Dec. 3, 2020

Preceded By

Nate Silver @NateSilver538

@DKThomp There's also an argument (I'm not 100% sure I buy it) that shorter polls that focus on horse-race questions are going to have fewer problems with low-trust voters than longer ones that ask dozens of questions. — PolitiTweet.org

Posted Dec. 3, 2020

Followed By

Nate Silver @NateSilver538

It's true that for issue polls, you don't have to model turnout, which is one fewer complication. But : 1—It's not clear turnout modeling is a big source of recent poll error 2—Issue polls have challenges election polls *don't* have, e.g. question wording is much more important — PolitiTweet.org

Posted Dec. 3, 2020