Well, this is really a problem with English and pronouns isn’t it?
The result of the study is that scientists guess that if we constructed a rationality-and-objectivity scale (RAO) then we’d find the average RAO among scientists is higher than the average RAO among non-scientists.
What we are inclined to read in the headline is that individual scientists tend to think of themselves as having higher RAO than non-scientists.
I think there are very good reasons to think the former is itself an objective assessment. I think there is little reason to think that the latter is even true (that is, that individual scientists, on average, have a higher opinion of their own RAO than individual non-scientists do).
