This is really interesting. Nice find. Here's a very brief summary: Participants were shown a few words and then four different identifiers, one of which matched the words and three of which did not. The example in the paper is:

phrase: "full pathname"

identifiers: fill_pathname, full_mathname, full_pathnum, full_pathname

From the conclusions section: "Although, no difference was found between identifier styles with respect to accuracy, results indicate a significant improvement in time and lower visual effort with the underscore style. The interaction of Experience with Style indicates that novices benefit twice as much with respect to time, with the underscore style. This implies that with experience or training, the performance difference between styles is reduced. These results add to the findings of Binkley et al.’s study"

So, finding the correct identifier was just as likely regardless of style but the correct identifier was found faster if it was using underscores as opposed to camelCase.

I think the underscores acted as "whitespace" in this study which in my experience is the most important factor in how quickly you can visually parse code. However, from a subjective/aesthetic perspective, I still prefer using camelCase over underscores.