General Discussion
Related: Editorials & Other Articles, Issue Forums, Alliance Forums, Region ForumsAxios-Ipsos Latino poll: Warning signs for Dems
AxiosWhy it matters: The findings underscore Latinos' collective hold on the American swing vote and Democrats' relatively modest advantage is an ominous sign for those who fear a further erosion of their party's support among this large, diverse ethnic group.
Inflation and crime remained by far the biggest worries for survey respondents. Abortion ranked significantly lower, and has not climbed since the Supreme Court overturned Roe v. Wade in June.
But, but, but: Democrats have some targeted advantages. Respondents support President Biden's student loan forgiveness and the Inflation Reduction Act by about 2-to-1 margins.
Survey of 1,005 U.S. Hispanic/Latino adults conducted Sept. 30 to Oct. 8, 2022
Democratic candidate 33%
Republican candidate 18
Dont know 23
Would not vote 17
Independent / other candidate 7
Data: Ipsos; Note: Data rounded to nearest whole
vercetti2021
(10,156 posts)Is this all it's going to be for the next fucking 3 weeks on DU? we're fucked, we're fucked, we're fucked?
Jesus fucking christ. Polls means absolutely jack and shit.
Go vote and don't let apathy win
brooklynite
(94,489 posts)....which are relevant to how we finish the three weeks before the election. Whether that news is "happy" or "sad".
Laura PourMeADrink
(42,770 posts)Butterflylady
(3,541 posts)karynnj
(59,501 posts)This is most definitely not a simple random sampling problem. It is not even a simple stratified sample problem, where strata, where responses should be more homogeneous within the strata than responses in the entire population are.
The response rate for polls is very low - with some pollsters speaking of reaching only about 4 percent of those sampled - even with repeated attempts. Decades ago when that percent was much higher, pollsters defended the necessary assumption that those reached were similar to those not reached within the same demographic group. This pollsters then use the estimates from those reached as the values for their demographic strata.
However, this is not the end of the process. The frame from which people are sampled is registered voters or all adults. The former will miss any recently registered voters if they are not on the lists obtained. The latter will include the question of whether a person is registered. (Consider that people tend to give answers they think make themselves look good.) In addition to all the problems just with the sampling, they need to have a heuristic model of who will actually vote.
I (obsessively) look at 538 and other polls - even though I realize that of races that are not shoe ins, the majority are essentially too close for polls to accurately call. Note that the confidence intervals is from the sampling process -- and unintentional biases and other factors are at play as well. I know that it is not lack of sophistication, intellect, honesty, or effort.
However, the one thing I do think is likely valid and that is the change over time from the same pollster polling races. The thing that is most disheartening at this point is that the polls for so many races seem at this point to be getting worse for us. That is very likely real and something that indicates the simplistic Republican messages are working better than our more complex messages of how Democratic leadership has done things that are and will help people.
onenote
(42,685 posts)The DUers that criticize polling as crap when they don't like the results but are happy to treat as gospel other polls, such as those showing Democrats closing the gap unexpectedly in some races.
People should remember this: Forewarned is forearmed. Don't like what the polls are suggesting? Work harder to ensure the polls are wrong.
DemocratSinceBirth
(99,710 posts)Happy and unhappy ones.
SlimJimmy
(3,180 posts)Using "registered" voters in polling tends to lean Democratic, but it's a false sense of hope. Polls are excellent indicators of what's happening at that moment in time. Don't like the polling? GOTV and make a difference.
sarisataka
(18,570 posts)If I don't like the results the poll is wrong. Either the pollsters don't know what they are doing and/or they are lying. I can safely ignore such polls.
After the election if we win I can gloat about how the polls were wrong and the media just wanted it to look like a close race. If we lose I can complain the polls were wrong and they lied to suppress Democratic voters.
OTH If I like the results of polls they are clearly accurate and as unquestionable as if I watched with my own eyes the deity carve them into stone tablets.
FBaggins
(26,727 posts)The question is - do you hunt for posts on those polls and insist that anyone who fails to hide the polls in the first group and only post the polls in the second group... is some undercover RW nutjob trying to demoralize Democrats?
sarisataka
(18,570 posts)I SMH at those who deny every poor poll as we have been bitten more than once ignoring such.
IMO the pie-in-the-sky polls do more to reduce voter turnout. I believe always assume you are 10 points behind and need to convince everyone you know to vote.
Polybius
(15,373 posts)Wow.