The poll was released June 1 after being conducted between April 18-25. It surveyed 1,500 American adults via with a margin of error of 2.53%. wow 1500 american adults out of 200 million? great representative sample
The sample size is not unusually small for a poll like this. The results for the approval for "assassinating a politician who is harming the country or our democracy" show more of a difference between "younger" and "older" people more than a difference by party preference. https://www.splcenter.org/news/2022...reat-replacement-hard-right-ideas#methodology
That does not make them accurate. All you said is that it is a typical sample size. And you proved the sample size for each class is significantly lower because out of 1500 people they were broken into Dem and Rep and then by men and women and then by age. We dont even know sample of older or younger. So the sample size for each defined quality was way less than 1500. 1500 out of 100,000,000 registered voters.....maybe even 400 male REP and 400 Male Dem within that sample..... yeah that is what we call a bullshit poll sample size. it is impossible to get an accurate R-squared or Z score from such a subjective question and sample size.