Not accounting for the fact that some people may know the correct answer does not amount to “throwing out” their opinions. Quite the opposite—such people may well be present in the sample, and since they will answer “one” regardless of what stereotypes they hold, the result in the paper may be an underestimate of the extent of distrust-based stereotyping of atheists. In theory, it might get more complicated if different subgroups (e.g. religious vs nonreligious) differ in the fraction of individuals who know the right answer, but this fraction is probably very low across the board.
[EDIT: I think I may have misunderstood what you said in the second paragraph, causing me to overestimate the degree of disagreement here. If this is right, the above might not make much sense to you; in that case just ignore everything except my point that the fraction of people who know the right answer is probably too low to matter.]
A subsequent reading led to a more charitable interpretation, which prompted me to amend my post to avoid possible confusion (about me sounding like I disagree while making points that HonoreDB would agree with). The paragraph in question (second paragraph in the original post) has since been edited, which might add to your confusion.
I thought their method was pretty neat.
Not accounting for the fact that some people may know the correct answer does not amount to “throwing out” their opinions. Quite the opposite—such people may well be present in the sample, and since they will answer “one” regardless of what stereotypes they hold, the result in the paper may be an underestimate of the extent of distrust-based stereotyping of atheists. In theory, it might get more complicated if different subgroups (e.g. religious vs nonreligious) differ in the fraction of individuals who know the right answer, but this fraction is probably very low across the board.
[EDIT: I think I may have misunderstood what you said in the second paragraph, causing me to overestimate the degree of disagreement here. If this is right, the above might not make much sense to you; in that case just ignore everything except my point that the fraction of people who know the right answer is probably too low to matter.]
It may be helpful to write explicitly in what sense you have misunderstood. I am now confused about how to understand your comment.
A subsequent reading led to a more charitable interpretation, which prompted me to amend my post to avoid possible confusion (about me sounding like I disagree while making points that HonoreDB would agree with). The paragraph in question (second paragraph in the original post) has since been edited, which might add to your confusion.