Not sure which particular study to point you toward (there are probably plenty out there on variously related topics), but to me, it seems pretty unremarkable that people seek to avoid pejorative characterizations. Then again, it's pretty remarkable how far we'll go to do this! We're social animals with a lot riding on our reputations, hence we're quite sensitive to others evaluations of us, and we're generally eager to please, or at least to be respected. Individual differences in these concerns exist in big ways though, as should surprise no one who's heard anything good about personality psychology lately.
Consider checking out the literature on self-monitoring. It's a well-studied individual difference in the tendency to alter one's self-presentation to please others, or at least control others impressions of oneself, basically (it's actually a somewhat complex construct, or two or three, depending on how you measure it). Low self-monitors tend to act how they act without worrying too much about it, about who might be watching, or about what others might think about them. (Come to think of it, one might even call low self-monitoring a characteristic resistance to social psychological phenomena in general! There's a hypothesis worth testing...but don't quote me on it until you do!) Conversely, high self-monitors make good actors / actresses: they know who's watching, how they look in front of "the camera," and they know how to please their audiences (or, at least, they think they do; it's often measured by self-report). [Edit]: Just in case it's unclear, I've described the meanings of "high" and "low" scores on a continuous dimension (or maybe two or three) of individual differences here; real people can be anywhere in-between too, or maybe even more extreme.
Self-monitoring might not be exactly what you're talking about—it's a characteristic adaptation to a pretty broad range of social situations—but I bet you could predict exactly what you're talking about with a nontrivial degree of accuracy using self-report questionnaire scores on self-monitoring.
P.S. Whenever I mention self-monitoring, I can't resist adding that I'm a former research assistant of the guy who came up with it. It entitles me to add that Mark Snyder is an awesome psychologist in general!