Did anyone bother to read the results? This is one of the worst and laziest polls I've seen. It has Gideon doing better in the 2nd district than the 1st, and a 45 point age gap... a 45 point age gap between the Collins leading Gideon by 26 points with <30's and Gideon leading by 19 with >65's. The worst offender - no education breakdown, AT ALL. No presidential poll to compare/contrast. etc. It may as well be a fake poll.
You’re reading way too much into crosstabs.
That's your response? I'm trying to display how bad it's overall quality is to people who clearly believe any poll without reading it. This is not a bias problem but a credibility problem. It should be disregarded, completely. I can't believe I'm getting called out for being too detail-driven on a no life forum like this.
I am not calling you out for being detail-driven. I am merely pointing out that a lot of your problems with this poll (and several other polls you have trashed over the past several months) are related to crosstabs, which are inherently unreliable. Let’s take a look at the sample sizes of the crosstabs you pointed out:
CDs: not mentioned but if there’s a 53% CD-01/47% CD-02 split like 2016, we can say we have roughly n = 240 in ME-02, a margin of error of 6.3%. The CD-01 sample at 272 is at 5.9%. Given that the poll has Gideon only marginally doing better in CD-02 anyway, this is nowhere near significant.
Age: The under 29 sample is literally 66 people! That’s an MoE of over a dozen points! Even the 65+ crosstab has a margin of error of close to 8! You can virtually wipe out the age gap in this poll just factoring that in.
The lack of education breakdown is a valid concern especially in a state like Maine, I will agree with that. And I would have liked a Presidential poll although not having it there doesn’t make it a junk poll.
