This is excellent work, and I'm so glad someone tested this out.
"Telling people that Mr. Biden and Mr. Trump have been vaccinated...has an impact on some but not others...our findings suggest that presidents can nudge some of them further along that path."
"Republicans who outright say 'no' do not budge when told of [Trump's] endorsement of vaccines; however, among Republicans more on the fence, we see movement."
Also, recontact data: "Among those who previously said they were planning to get vaccinated, 16% now report having received at least one dose in our survey's control condition. Moreover, one in five of those who previously said maybe to getting a vaccine now say yes outright."
The thing I have no idea what to make of is the Biden cue moving some Dems toward "maybe."
Adventures in question order:
"[W]hen refusals of gay and lesbian people are equated with refusing service to other minority groups, respondents might be more likely to view religiously based service refusals of gay and lesbian people as discriminatory in nature."
This is, additionally, yet another entry in "why you should always, always, always, always ask to see the toplines first."
Important takeaway here! It's way too easy to frame varying results on issue polling as "these opinions aren't real" when the better answer is "the way people think about this issue is complex, and that's reflected in the data." In the end, it tells us more, not less.
Partisan split is an easy takeaway here, but the way that *everyone* is overestimating (half of GOP/60% of inds are in top two categories!) I think lends itself to a bigger point about the difficulty of estimating probabilities, and translating small %s into so many lives.
Also, people (myself often included) are phenomenally bad at estimating percentages in general. I actually wonder whether people would have done better generally if instead of the numbers they got a more subjective/intuitive scale (e.g. "most," "about half," etc.)
Quite possibly not! But would be an interesting experimental design...
This is one thing that'll be really interesting to keep an eye on. I've thought for a while that if there's a social-trust driven non-response issue with election polls, that same bloc of non-responders would be likely to inflate vaccine numbers by a few points.
Will obviously be useful to have more comparison points - FWIW, when I looked at this a few weeks back when I was doing coronavirus polls for HuffPost/YouGov, our national numbers were tracking pretty well with national vaccination rates, although not sure how it's evolved since.