Changing the personality of a face: Perceived Big Two and Big Five personality factors modeled in real photographs.

Abstract

General, spontaneous evaluations of strangers based on their faces have been shown to reflect judgments of these persons’ intention and ability to harm. These evaluations can be mapped onto a 2D space defined by the dimensions trustworthiness (intention) and dominance (ability). Here we go beyond general evaluations and focus on more specific personality judgments derived from the Big Two and Big Five personality concepts. In particular, we investigate whether Big Two/Big Five personality judgments can be mapped onto the 2D space defined by the dimensions trustworthiness and dominance. Results indicate that judgments of the Big Two personality dimensions almost perfectly map onto the 2D space. In contrast, at least 3 of the Big Five dimensions (i.e., neuroticism, extraversion, and conscientiousness) go beyond the 2D space, indicating that additional dimensions are necessary to describe more specific face-based personality judgments accurately. Building on this evidence, we model the Big Two/Big Five personality dimensions in real facial photographs. Results from 2 validation studies show that the Big Two/Big Five are perceived reliably across different samples of faces and participants. Moreover, results reveal that participants differentiate reliably between the different Big Two/Big Five dimensions. Importantly, this high level of agreement and differentiation in personality judgments from faces likely creates a subjective reality which may have serious consequences for those being perceived—notably, these consequences ensue because the subjective reality is socially shared, irrespective of the judgments’ validity. The methodological approach introduced here might prove useful in various psychological disciplines.

Modelling perceptions of criminality and remorse from faces using a data-driven computational approach

“…Todorov, Dotsch and colleagues (2013) have further validated models of attractiveness, competence, extraversion, and likeability. Using pictures of real people, Walker and Vetter (2009) Walker & Vetter, 2016).…”

“…Correlations between criminal appearance as well as remorseful appearance with threat, dominance, and trustworthiness for both the Princeton and the Basel approach are reported in Table 1. For the Basel face model, it was additionally possible to correlate criminal and remorseful appearance with models of sadness and masculinity/femininity (with data collected by Walker & Vetter, 2016). Within the Basel model, perceptions of criminality strongly correlate with perceived masculinity, whereas perceptions of remorse strongly correlate with the perception of sadness (more precisely "feeling down"/ translation of the German term used: "niedergeschlagen").…”

“…The data to identify criminal appearance in the Basel faces were taken from a study on personality judgments of faces conducted by Walker and Vetter (2016).…”

“…For models to be obtained with the Princeton approach, we conducted studies in which we let participants rate each face twice to calculate a participant's test-retest reliability in order to ultimately reduce error variance in participants' ratings. For the Basel face models, we used existing ratings on criminality (obtained by Walker & Vetter, 2016) and remorse (obtained by Funk et al, under review), as described below.…”

Modification of First Impression Formation and “Personality” by Manipulating Outer Appearance

“…Similarly, the perception of attractiveness, competence, creativity, cunning, extraversion, meanness, trustworthiness, or intelligence can be influenced by setting up different facial expressions [11,12] . Even personality judgements based on the Big Five can be changed by different postures and facial expressions in a photo [13] .…”