In self-stereotyping, I explained why this phenomenon is difficult to demonstrate. Although some empirical demonstrations exist (e.g., by Cadinu; Knowles; van Veelen), the scope of the phenomenon is limited. When people perceive similarities between themselves and a group to which they belong, these perceptions tend to be a matter of projection from the self to the group rather than a matter of introjection (self-stereotyping) from the group to the self.
Here, I elaborate on why this is so. Social projection is a heuristic that allows you to infer, predict, or estimate the properties of a group or a particular person on the basis of your own properties. If you prefer watching House instead of Grey’s Anatomy, you may guess that fellow group members feel the same way. Projective inferences tend to be correct (i.e., better than random guesses) because your personal traits (properties, preferences, behaviors) tend to be correlated with the traits of your group. Note that this is not a matter of stereotyping, but of accuracy. Inasmuch as there is variation in the prevalence of traits in a group, many individuals have majority traits and few individuals have minority traits. This is not just true; it is truistic. As a consequence, a person’s profile of traits (having or not-having them) is correlated with the group rates and with the trait profiles of other individuals even if all the sampling from the trait pool is random (i.e., your chances of having a trait held by 80% [70%, 60% . . .] are .8 [.7, .6 . . .]. Hence, if you know your traits, you can infer, predict, estimate the traits of others.
Projective estimates of others can be checked for accuracy. You can do this in real life. After predicting if Joey loves House just like you do, you can ask him. If he prefers Anatomy, you have learned something about Joey. Evaluating the statistics of many respondents making many estimates about many traits, researchers can compute projection correlations between self-judgments and group judgments, accuracy correlations between group judgments and actual trait prevalence in the group, and validity (or typicality) correlations between self-judgments and trait prevalence. Inasmuch as the validity correlations are positive, accuracy correlations increase with projection correlations. Researchers can examine projective bias by computing the projection correlation while statistically controlling for a person’s validity correlation. If this correlation is positive, the person projects from the self to the group beyond what can be accounted for by his or her own actual similarity with the group (Krueger, 2008, has more detail).
Try to map this arrangement onto the self-stereotyping hypothesis and see how hard it is. Claims about self-stereotyping make no distinction between actual and perceived trait prevalence. Indeed, self-categorization theory, the main theoretical base for self-stereotyping, dismisses the notion of actual group characteristics on a priori grounds. There are only perceptions. What about the self-concept? Here too, all we have is perceptions. Hence, it is theoretically (and practically) impossible to infer your own traits from the stereotypical traits of the group and then look up whether you actually possess these traits. It is impossible to make any distinction between self-stereotyping and self-stereotyping bias, and it is impossible to change self-perception to make it more accurate. Claims about self-stereotyping thus reduce to declarations that correlations between self- and group perception are inferences from the group to the self rather than inferences from the self to the group.
Or do they? I mentioned that there have been demonstrations of self-stereotyping. What are these? Participants are vigorously reminded of one of their group memberships (woman, firefighter, philatelist, or whatever) and of group-descriptive traits (social orientation, fearlessness, attention to detail). Then, it is shown that respondents rate these traits as more self-descriptive than do controls (who have no salient group membership or undergone trait priming). In other words, it is possible to demonstrate the assimilation of the self-concept to a group stereotype – and that is important to know. But it is rather like showing that we will ascribe a love of Mariachi music to Pablo when learning that Pablo hails from Guadalajara. Showing that X can cause Y when pressed does not mean that it does so a lot in the wild. To assume that it does is an instance of the reverse-inference fallacy (Krueger, 2014).
Herr August Landmesser refused the Teutonic salute. He set one of the most poignant examples of resistance to group pressure. Had Landmesser not resisted, would he have self-stereotyped? Arguably, he would have conformed. Even if his conformity had been private, that is, even if he had come to endorse an act, and the beliefs associated with it, this change of his self-concept falls short of the definition of self-stereotyping. By definition, self-stereotyping is an entirely intrapsychic event, and hence cannot reduce to explicit group pressure. The actual presence of a pressuring crowd smacks of an objective reality, which self-categorization theory diligently dismisses.
Note. I found the epigraphic picture on a site discussing Italian and Italian-American life. The author notes the dangers of self-stereotyping (accepting a negative view of one's own group as held by a dominant outgroup).
Krueger, J. I. (2008f). The robust beauty of simple associations. In J. I. Krueger (Ed.), Rationality and social responsibility: Essays in honor of Robyn M. Dawes (pp. 111-140). New York, NY: Psychology Press.
Krueger, J. I. (2014). Reverse inference. In S. O. Lilienfeld & I. D. Waldman (eds.), Psychological science under scrutiny: Recent challenges and proposed solutions. New York, NY: Wiley.