You should be polling “likely voters” especially in a midterm election where turnout is historically low compared to presidential election years. Polling “Registered voters” is nothing but for ratings and unneeded drama. Most people do not understand the difference.
-
-
- 1 more reply
New conversation -
-
-
Tbh, I don’t think those being polled necessarily tell the truth. I don’t think polls are relevant anymore. The mid-terms will be a good test.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Poll the dead ones first! Then Poll the Living. Subtract the Dead Votes and there is your tally!
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Fox does pretty well in
@FiveThirtyEight's pollster ratings: https://projects.fivethirtyeight.com/pollster-ratings/ …pic.twitter.com/QY0hjTKtTU
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
100,000 minimum.
- 1 more reply
New conversation -
-
-
1,000 people. 500 from rural locations, 500 from suburban and urban areas. 20 per state, split as above. That's a poll I'd believe. Until then I don't trust polls from anyone in the media.
-
That's biased as well. 20 per state weights Montana too much and Texas too little. 500 in rural weights rural too much because less than half of the population lives in rural areas. A random sampling as done here is the correct mathematically sound statistical method.
-
The problem is (while my numbers were indeed tossed out almost at random) a "random sampling" is invariably taken in one city. One town. Etc.
-
You have a citation for that?
@foxnewspoll seems pretty transparent in their methodology. I'm no Fox News apologist, but I am a science proponent. And this is sound statistical methodology.http://www.foxnews.com/politics/2012/08/07/fox-news-poll-methodology-statement.html … -
Hmm. Never actually read that for Fox News, for some reason. There again, at work, I mostly monitor Leftist outlets. Not a bad system. You'd get, statistically, the occasional poll weighted due to RNG, but otherwise it's not bad. My apologies
@FoxNews
End of conversation
New conversation -
-
-
Polls are unreliable for several reasons. Most political polls are weighted by party affiliation & ask "loaded" questions. Registered voters vs. likely voters = different results. Calling only landlines isn't "random." Can't tabulate voter ignorance & people who lie.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I don't agree with the poll numbers. I have a sneaky suspicion there are more people out there like me. After the 2016 election I have no trust in polls
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Trump is your President.... and her
career is Dead.
Flush your fake polls.pic.twitter.com/cVC2yjpeZfThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
How about a 110 thousand. I'm amazed people believe all these polls of basically 1500 people out of millions in this country. Ridiculous
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
She has a point. I've been in the failing survey polling industry for 30+ yrs. They catch some trends, but not many close predictions.
@GillianHTurner@MikeEmanuelFox@MariaBartiromoThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Bret is right. You only need 400 for a statistically significant poll. Now, if these are landlines, you're not getting random results, but its close.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Borrowed quote, can't recall from where: Note, a statistically significant result may be of "No Practical Significance".
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
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.