Donald Trump is leading Kamala Harris by just one point among men, according to a new poll.
The survey, conducted by Outward Intelligence between August 11 and 15, shows that 46 percent of men back Trump compared to Harris’ 45 percent.
Among women, Harris was 13 points ahead, on 53 percent to Trump’s 40 percent.
The poll surveyed 1,858 likely voters and had a margin of error of +/- 2.3 percent.
2000 is really solid to achieve a reasonable margin of error. It’s a pretty simple formula taught in any basic stats class.
It’s also far too small to be a reasonable example of voter opinion in a region as large as the US. If this was quantified with a region that the sample is from if might be better. But minimal sample size for making statements as broad as this just makes the stat garbage and exactly the sort of thing media should leave out of articles due to how biased a sample this small could be.
Critically, that is assuming that the sampling is perfectly random sample of the thing being measured. Political polls in particular are going to struggle since participation is voluntary, motivations for poll participation differs than the both, different levels of anonymity, etc.