General Discussion
Related: Editorials & Other Articles, Issue Forums, Alliance Forums, Region ForumsNot even close to panicking about the new NBC poll but
can anyone tell me why, this late in the election, they are polling registered voters instead of likely voters? It is my understanding that we are past the time in the cycle where likely voters are more indicative of final results.
In case you haven't heard about this morning's NBC poll, Harris and the other guy are tied 48-48 among registered voters.
Abnredleg
(743 posts)LV tends to favor Harris in the polls, although I dont know why NBC didnt release LV numbers.
PoindexterOglethorpe
(26,345 posts)Recreation Vehicle is what I find, and I'm fairly sure that's not what you mean.
hauckeye
(713 posts)PoindexterOglethorpe
(26,345 posts)getagrip_already
(17,012 posts)A high percentage of registered voters dont vote, so its an attempt to get a more probable group that will actually vote.
But it will miss young voters who are recently registered, as well as issue voters motivated to show up for an election they find important.
It does nothing to address nonresponse bias, or other difficulties contacting a true random population of people who will show up to vote.
agingdem
(8,497 posts)they are irrelevant...it's the vote
senseandsensibility
(20,014 posts)using my critical thinking skills, thank you. You're free to ignore them of course.
berksdem
(653 posts)It did not turn out so well
agingdem
(8,497 posts)I'll be honest with you..she knew she was going to run years before she announced so she had plenty of time to "clean up" any foreseeable problems...the Clinton foundation and "her emails"...it wasn't just Comey's death knell announcement..it was also Anthony Weiner's scandal and the fact that he was married to Hillary's campaign staffer
and there was the media and by media I mean Joe and Mika...they hated the Clintons, Hillary in particular, so they, with the help of Andrea Mitchell and Chuck Todd, reinvented their party buddy Trump, a tabloid TV buffoon, and sold him as a viable candidate who would grow in office ..and their constant her emails drumbeat reinforced their Trump is a serious man bullshit.. once he was in office Trump turned on them and they were forced to rewrite history..
a shitty perfect storm....
iemanja
(54,368 posts)But like you, I'm not panicking.
senseandsensibility
(20,014 posts)iemanja
(54,368 posts)Registration has closed in many states. Whether people say they are 'likely" to vote is irrelevant. If they aren't registered, they can't vote.
Abnredleg
(743 posts)So you still need to discern LV.
iemanja
(54,368 posts)NoRethugFriends
(2,847 posts)So likely voters is a better measure of course
iemanja
(54,368 posts)hlthe2b
(105,663 posts)they only want "evidence" for the continuing "horserace" narrative.
senseandsensibility
(20,014 posts)servermsh
(1,313 posts)All 50 states are not relevant.
-misanthroptimist
(1,108 posts)Who is going to remember this poll the day after Election Day? In the meantime, everyone is glued to their screen to watch the horse race. If one horse is way out front, who would watch?
PoindexterOglethorpe
(26,345 posts)anciano
(1,497 posts)to stay breathlessly tuned in for what the next poll will say. It's always about the money.
former9thward
(33,303 posts)Data has to be produced to show who is likely and then applied to the voters contacted. That is why likely voters in polls are rarer -- at least for public consumption. The campaigns use likely voters in their polling.
Fiendish Thingy
(17,780 posts)Instead of verifying someones past history of voting as a predictor of voting in the current election, many pollsters are resorting to self assessment to determine LVs.
That is, they simply ask the respondents how likely are you to vote?
Fiendish Thingy
(17,780 posts)Instead of verifying someones past history of voting as a predictor of voting in the current election, many pollsters are resorting to self assessment to determine LVs.
That is, they simply ask the respondents how likely are you to vote?
kerry-is-my-prez
(8,897 posts)Im concerned because Harris has been going down in the battleground polls.
Response to senseandsensibility (Original post)
Name removed Message auto-removed
Fiendish Thingy
(17,780 posts)karynnj
(59,856 posts)The actual sample frame from which voters are sampled is usually registered voters. Obviously, there are no lists of likely voters. This list will have errors - people who have moved, or died or recently registered after the list was created. I think all polls use stratified samples where they use the best population estimates of each strata weight. Realize that there are three kinds of error in this process. The simplest to understand is sampling error which has well defined and computed formulae. There is also the potential error in the weight to give the strata. An addition source of error is that there is an implicit assumption that the people in any given strata who are willing to be polled (only about 2 to 3 percent) are similar to the rest of the people in the strata. NOTE - only the first error is included in the MOE.
The Likely Voter model uses a model that tries to give more weight to observations with high likelihood of voting and less weight to observations seen as less likely to vote. These models are heuristic and are rarely defined by the polling companies.
I want to see the RV polls so I can compare it to past RV polls. Here, I assume that a given poll will not change the population weights. So if there is a difference or a trend, it is because there is a shift in preference or maybe a change in who is willing to be polled. However, the Likely voter model - especially when looked at with the registered voters polls - will pull in information on the choice of the people who actually will show up to vote.
rso
(2,406 posts)The problem with using likely voters is that it would not include for example newly registered young people, especially young women, who decided to register after Roe was overturned.
Klarkashton
(1,369 posts)W_HAMILTON
(8,258 posts)Here is a link to the hard data on early voting, which will continue to be updated throughout the early voting period leading right up to Election Day: https://election.lab.ufl.edu/early-vote/2024-early-voting/