3/ State polls were a bigger problem, but that’s more a function of missing data. Not enough polls lead to bad forecasts.
-
-
actually the case in population, you won't get a result that represents the population. Right?
1 reply 0 retweets 0 likes -
Replying to @SharylAttkisson @RayGuy3
1/ But how do you know what the “correct” breakdown of party is to weight for? You could do that for a registered voter poll, but...
1 reply 0 retweets 0 likes -
2/ for likely voters, there are much more predictive screens than party ID.
1 reply 0 retweets 0 likes -
3/ and then you’re expecting voters to accurately report party ID, unless you’re calling from voter rolls instead of random digit dialing.
2 replies 0 retweets 0 likes -
Right. But why bother to ask (which they all do) if they assume there's no reliability or legitimacy?
1 reply 0 retweets 0 likes -
Replying to @SharylAttkisson @RayGuy3
Because the shifts in party ID are a very interesting *result* of the polling, not a valid way to weight the polling.
3 replies 0 retweets 0 likes -
Perhaps. But I would argue the party sample, even if self-ID'd can be *very* relevant to results on political topics.
1 reply 0 retweets 0 likes -
Replying to @SharylAttkisson @RayGuy3
Oh, it’s very revealing about the overall mood of the electorate, along with who’s likely to actually vote any given year...
2 replies 0 retweets 0 likes -
…it’s just not a statistically valid way to weight your results on horse race questions.
5 replies 0 retweets 0 likes
Serious question: I looked at quite a few political polls in 2016 and since. I never found one that interviewed more Repubs or even
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.