barmar, on 2021-March-17, 23:54, said:
This is of course a danger in any poll about future actions -- we undoubtedly have a similar problem when polling about how people will vote. But pollsters know about this effect and they know how to adjust for it. Also, making the poll large enough should reduce the significance of these answers.
Furthermore, the precise numbers aren't always that important, it's the relative numbers that matter. The fact that the Republican hesitancy rate is 3x the Democratic rate is very significant, even if a large number of the Republican responders go back on their answers. It's unlikely that 2/3 of the GOP answers were wrong.
I agree with your second paragraph much more than I agree with the first. Pollsters, unless they are idiots, know that answers are sometimes swayed by social expectations, true enough. In the 1950s, when I was growing into adulthood, all unmarried young women were virgins. Of course. But with the vaccine, I have not seen any polls (ok, I have not studied the issue) that describe any attempt to take account of social influence on the accuracy of the answers. And how would they? For me, it seems best to just accept that there might be a pretty substantial difference between the answers given and the actual behavior.
I think that your second point, that we should not get bogged down with the fact that not everyone tells the truth to a pollster, is very important. We can then focus on the question: What do we do about the fact that it appears that many people, never mind just how many, but it seems safe to say many, are reluctant to get the vaccine? I suggested one answer above. I read that Hank Aaron got the vaccine and did so publicly. A good start, but he is even older than I am. Prominent respected African-Americans could make a big difference in what African-Americans do Same with church leaders and with Trump supporters. I saw that Trump made some halfway supportive comment about taking the vaccine but we need more and better from leaders on the right.
A message from me to the leadership of cultural religious and political groups: There is no way around it guys. What you say and do, and make a point of saying and doing publicly, about the vaccine is going to make one hell of a difference. You can pitch in and help, or you can sit on the sidelines and watch the results. Your choice will have a substantial effect.
Yes, I already said the above but I think it is important.
Among the many reasons people lie to pollsters is just for the fun of it. I can recall when I was (muc0 younger pollsters would stop me in the mall and ask various questions and I would give answers. Sure I think the Earth is flat, doesn't everyone?
Not proper behavior I know and I would not do it now.
There are also times when I am not sure what the right answer is. I have agreed to answer a daily poll for a covid study. One of the questions allows for three answers about my interaction with others on the previous day. The choices are:
a. did not wear a mask.
b. wore a mask when I interacted at a distance of less than 6 feet.
c. did not interact.
Now suppose, as happens, that I am out for a walk, I encounter someone, and we converse for a bit standing 20 feet apart. Well, c doesn't seem right because I did interact. But looking at a and b, I am thinking they want to know if I wore a mask when the interaction was less than 6 feet apart. If I were to check a, it appears to me I am saying something like "Mask, no of course I don't bother with a mask". So I check c, trusting it to mean that I had no interactions at a distance of less than 6 feet. I am not sure that this is the right interpretation, but clicking a, that I did not wear a mask, seems to seriously misdescribe my approach to interactions. We need a d, always wore a mask in any interaction that is anywhere near 6 feet or less but sometimes interacted without a mask at greater distances. And usually with my head turned slightly to the side.
Polls are tough, with best of intentions all around.