Get your irony meters ready. So, the media articles misrepresent this study quite a bit and oversell its findings (thus actually contributing to why people might not defer to scientific reports over claims of laypersons).
What they did was take 2 independent variables and cross them, resulting in two Public-Service-Announcements, one that supported and one that was against vaccines, and then each announcement was paired sometimes with a supportive lay comment and sometime with a anti lay comment.
(They had other conditions with high credentialed commenters, but that isn't what the headlines or thread are about).
Attitudes toward getting oneself vaccinated were scored on a 1-7 scale with anything above 4 being a pro attitude and below it an anti-attitude.
What the headlines (and the thread title) imply is that people responded mostly consistent with whatever position the comments were, regardless of what the PSA said. Here is the general pattern of means for the 4 conditions, as implied by the headline (not the actual data)
Pro-PSA /Pro-comment: 5.00
Pro-PSA /Anti-comment: 2.00
Anti-PSA /Pro-comment: 4.00
Anti-PSA /Anti-comment: 1.00
Such data would show that the position of the comments creates a bigger difference between means, than a difference in the position of the PSA, and that when the PSA and comments oppose each other, people take the side of the comments. This is not what happened.
Here are the actual means:
Pro-PSA /Pro-comment: 4.18
Pro-PSA /Anti-comment: 5.04
Anti-PSA /Pro-comment: 5..33
Anti-PSA /Anti-comment: 2.33
Notice that mean #1 should be lower than mean 2, yet it is higher. Also, since group #2 and #4 both have anti-comments, they should look somewhat similar, and yet #2 is much higher than #4, and #2 is notably on the pro side of vaccines (above 4) even though the comments are anti-vaccine.
The only group that really supports the headlines and thread title is group #3, who are decidedly pro vaccine which matches the comments they saw but not the PSA they saw. This driving their effects in the various regression analyses they do.
It is important to remember that they only studies this one topic and that most people coming into the study would have leaned toward the pro-vaccine side (something they did not measure of account for). Thus, one can interpret the findings as showing that so long as either the PSA or the comments were pro vaccine, people favored the pro side that they already favored coming into the study. The only time they really switched to a priori view was when they were exposed to both a PSA and comments that were all anti-vaccine. The preliminary study they ran showed that people thought that the fake
"National Vaccine Information Center" was a highly credible, reliable, honest, truthful, and trustworthy organization. IOW, people were likely fooled into thinking this fake organization was a mainstream scientific and likely government backed organization, because that is what is sounds like. This is who the anti vaccine PSA was authored by, so when people read that plus all comments supporting it and no one opposing it, they were persuaded.
IT is interesting and troubling that people can be so easily conned by a fake scientific organization name, but it is less troubling than the very incorrect headline implying that people discounted CDC pro-vaccine PSAs just because of anti commenters.
Of course with lousy science reporting like this, it is understandable that people might trust laypersons as much as scientists, thus my warning about irony meters.
