Project

person perception; faces; statistical models; personality

Lead
Lay summary
People draw personality trait inferences about unfamiliar individuals on the basis of their facial appearance. These inferences are often inaccurate. However, they affect face memory and predict social behavior in different domains such as hiring decisions, election outcomes, eyewitness testimonies, and criminal sentencing. Recently, first steps have been undertaken to investigate the role of specific facial characteristics underlying personality trait judgments with the help of statistical face models (Oosterhof & Todorov, 2008; Walker & Vetter, 2009). Both approaches have been successful in modeling personality traits in faces in that a high degree of interpersonal consensus in forming these judgments was achieved. However, there are also inter-individual differences in forming personality trait judgments on the basis of facial appearance. Sources of dissent among judges are, for example, self-resemblance, resemblance of stimulus faces to familiar persons, or judge’s gender, and motives (Zebrowitz, Voinescu, & Collins, 1996). This project aims at investigating how personality trait judgments are formed on the basis of faces from two perspectives: First, we will focus on stimulus characteristics, that is, we will use a newly developed statistical face model to identify the physiognomic correlates of a wide range of personality trait judgments. Second, we will focus on judges’ characteristics and investigate different sources of inter-individual consensus and dissent in direct comparison. By bringing together approaches and methods from social psychology and computer graphics and vision, we are able to investigate the processes of impression formation on the basis of facial features in a comprehensive way. This project will not only result in new findings regarding physiognomic correlates of personality trait judgments and inter-individual sources of consensus and dissent on social judgments, but also in more precise personality trait vectors that can be used in future research in different psychological fields, such as social cognition, social neuroscience, or the media psychology.

Direct link to Lay Summary

Last update: 21.02.2013

Active participation

Self-organised

Communication	Title	Media	Place	Year

Title	Year

People draw personality trait inferences about unfamiliar individuals on the basis of their facial appearance. These inferences are usually inaccurate. Therefore, they are labeled as perceptual illusions, having their origin in face overgeneralization effects (Montepare & Dobisch, 2003; Zebrowitz, Fellous, Mignault, & Andreoletti, 2003) and stereotyping (Hoffman & Hurst, 1990; Montepare & Zebrowitz McArthur, 1998; Sczesny & Kühnen, 2004). How-ever, these inferences affect face memory and predict social behavior in different domains such as hiring decisions, election outcomes, eye-witness testimonies, and criminal sentencing. Surprisingly little is known about the specific physical characteristics underlying these social judgments. Some research has been done on how facial information is related to perceptions of attractiveness and to categorization by age, gender, ethnicity, and emotional states. Recently, first steps have been undertaken to investigate the role of specific facial characteristics underlying per-sonality trait judgments with the help of statistical face models. In one approach different personality trait judg-ments were reduced to the two dimensions explaining most variance in these judgments, namely valence and domi-nance (Oosterhof & Todorov, 2008). Another approach differentiated between different personality trait judgments such as aggressiveness or trustworthiness, and manipulated them separately in photographs of faces (Walker & Vetter, 2009). Both approaches have been successful in modeling personality traits in faces in that a high degree of inter-personal consensus in forming these judgments was achieved. However, there are also inter-individual differ-ences in forming personality trait judgments on the basis of facial appearance. Sources of dissent among judges are, for example, self-resemblance, resemblance of stimulus faces to familiar persons, or judge’s gender, and motives (Zebrowitz, Voinescu, & Collins, 1996). The proposed project aims at investigating how personality trait judgments are formed on the basis of faces from two perspectives: First, we will focus on stimulus characteristics, that is, we will use a newly developed statistical face model to identify the physiognomic correlates of a wide range of personality trait judgments. Second, we will focus on judges’ characteristics and investigate different sources of inter-individual consensus and dissent in the framework of a comprehensive conceptual model. The identification of specific facial characteristics underlying different personality trait judgments will result in an enriched statistical face model. This model is highly flexible and is able to generate fully controlled and natural-looking variations of any individual face, differing only with respect to certain personality traits that are ascribed to the depicted individual. The possibility of manipulating the personality traits ascribed to individuals on the basis of their faces, independent of identity information, allows for addressing many research questions in different fields of psychology for the first time. Moreover, judges’ character-istics, such as their motives or their own physical appearance, will be investigated as sources of inter-individual consensus and dissent in forming personality trait judgments in a direct comparison. Finally, having identified judges’ gender role attitudes as a potential additional source of consensus and dissent for social judgments based on faces, we will therefore investigate their impact on likeability and other trait inferences based on faces signaling gender norm violations.