Matt Keller and I have a new paper out in which we present two methods to measure biases in individuals’ self-perception. These methods have the advantage that they neither involve introspection nor any external standards of comparison. One of these methods (Study 1) allows systematic modeling of specific personality dimensions in participants’ own faces in a theory-driven way and measurement of self-enhancement regarding these dimensions. The other method (Study 2) allows measurement of self-enhancement in a purely data-driven way by extracting the dimensions of self-enhancement from random noise patterns applied to participants’ own faces. Results from two studies reveal that individuals self-enhance regarding both Big Two personality dimensions (i.e., agency and communion).
The Figure above visualizes the female (A) and male (B) color (row 1), shape (row 2), and full self-enhancement vector applied to the female (A) and the male (B) average face from the Basel Face Model (row 3) and to individual participants’ faces (row 4) extracted in Study 2.
These novel methods might advance theory regarding self-enhancement in the long run, because they allow investigation of self-enhancement (and self-protection) regarding various dimensions (e.g., facial, personality, typicality of a certain group membership), various groups of individuals (e.g., from different cultural backgrounds, age groups), and individuals in different situations (e.g., by temporarily manipulating self-esteem or
group membership), thus providing information about inter-group, inter-individual, and intra-individual differences in self-enhancement.
With regard to the benefits of self-enhancement, such as being happy and caring about the self and others (Taylor & Brown, 1988), detecting the groups or individuals who are successful, and the situations that facilitate doing so, seems to be a critical endeavor for future research.
One Comment
congratulations on the publication and the novel method(s) – i can imagine a lot of people could be interested in using these. do you also supply an easy way to construct stimuli/analyse data using these principles that the academic community can use?