Learning to judge a book by its cover: Rapid acquisition of facial stereotypes

https://doi.org/10.1016/j.jesp.2021.104225Get rights and content

Highlights

  • A brief training can result in the formation of novel facial stereotypes related to trustworthiness.

  • The novel facial stereotypes are highly automatic and activate early on in processing.

  • Trait evaluations from faces are malleable and can be formed by learning and experience.

Abstract

People are able to quickly and automatically evaluate faces on different traits, such as trustworthiness. There is a growing literature demonstrating that factors such as learning and experience play a role in shaping these judgments. In the current work, we assess the malleability of our trait evaluations by associating arbitrary facial features with trustworthy or untrustworthy behaviors. Across five studies, we demonstrate that this learning can impact trait evaluation and effectively form novel facial stereotypes, which exert effects on evaluations as strong as intrinsic facial trustworthiness. With only a brief training, participants' rapidly acquired novel facial stereotypes, which were activated automatically and early on in processing, and which biased participants' trust behavior and hiring decisions. These results suggest that our trait evaluations of faces are shaped by an implicit learning mechanism that abstracts the co-occurrence between facial features and trait-related behaviors, resulting in the creation of novel facial stereotypes.

Section snippets

Study 1A

In the present study, we test whether arbitrarily learned feature–behavior associations affect the trait evaluation of faces, resulting in novel facial stereotypes. First, participants engage in a short learning phase wherein faces with thin sellions were associated with trustworthy behaviors 80% of the time and faces with wide sellions were associated with untrustworthy behaviors 80% of the time. We then assess the amount of money participants pay in an economic trust game to a new set of

Study 1B

It is possible that the effects in payment in Study 1A are somehow limited to the feature–behavior associations used for that study, namely that narrow sellions were learned to be trustworthy and wide sellions learned to be untrustworthy. In Study 1B, we aim to replicate the effects with the counterbalanced associations (narrow sellion = untrustworthy, wide sellion = trustworthy).

Study 2

In Study 1, we showed that an arbitrary, newly learned facial stereotype can impact to what extent participants trusted targets in the context of an economic game. In Study 2, we extended these results to another consequential domain in which facial stereotyping can readily bias outcomes, namely hiring and candidate selection processes. We implement the same training regimen as in Study 1, but we follow it with a line-up choice regarding which target individual participants would like to select

Study 3

In Studies 1A/1B and 2, we demonstrate that a brief learning period resulted in a novel facial stereotype that affected evaluations based on payment in an economic trust game and in choosing a personal financial advisor. However, while these newly learned associations clearly biased the outcomes of such explicit evaluations, it is unclear to what extent they activated automatically or biased evaluations early on in processing. Facial trustworthiness has long been known to activate automatic

Study 4

The previous studies show that a brief training resulted in a novel facial stereotype that can impact explicit judgments of trustworthiness and has an early impact on evaluative processing in parallel with “intrinsic” facial trustworthiness features. One open question is to what extent this novel facial stereotype is automatized and has an impact on more implicit measures of evaluation. This would cast doubt on the possibility that the previous effects are mere artifacts of demand

Study 5

One distinct possibility throughout the previous studies is that participants were explicitly aware of the sellion-width association with trustworthiness and that the shifts in evaluations observed in each study were driven by heuristics or rules guided by this awareness or by demand characteristics. Another potential issue to resolve is the extent to which the effects of the training phase are truly generalizing to the novel exemplars with the learned feature during the evaluation phase. In

Meta-analysis

To characterize the overall strength of the newly learned facial stereotype effect, we meta-analyzed Studies 1A, 1B, 4, and 5 using fixed effects, with effect size weighted by sample size (Goh, Hall, & Rosenthal, 2016). We did not include the mouse-tracking study (Study 3) as the predicted empirical pattern in that study (Sellion Width × Facial Trustworthiness interaction on mouse-trajectory attraction) was qualitatively different than the other studies (effect of Sellion Width on evaluations).

General discussion

Overall, we demonstrate that newly learned associations between an arbitrary facial feature and valenced behaviors resulted in the creation of a novel facial stereotype associated with sellion width, which exerted an equal – if not stronger – effect to the long-established trustworthiness of the face on evaluations. These results arose in payments allocated in an economic trust game and in the choice of hiring a financial advisor, demonstrating how these newly learned facial stereotype

Acknowledgements

This work was supported by the NSF BCS-1654731. We thank Michael Berkebile and Maryam Beshar for their assistance with the studies. We thank Dr. Chu Chang Chua for her continued guidance.

References (60)

  • A. Todorov et al.

    The efficiency of binding spontaneous trait inferences to actors’ faces

    Journal of Experimental Social Psychology

    (2003)
  • S.C. Verosky et al.

    When physical similarity matters: Mechanisms underlying affective learning generalization to the evaluation of novel faces

    Journal of Experimental Social Psychology

    (2013)
  • M. Bar et al.

    Very first impressions

    Emotion

    (2006)
  • D.S. Berry et al.

    Perceiving character in faces: The impact of age-related craniofacial changes on social perception

    Psychological Bulletin

    (1986)
  • V. Blanz et al.

    A morphable model for the synthesis of 3D faces

  • E. Bliss-Moreau et al.

    Individual differences in learning the affective value of others under minimal conditions

    Emotion

    (2008)
  • J.F. Bonnefon et al.

    The modular nature of trustworthiness detection

    Journal of Experimental Psychology. General

    (2013)
  • K.-W. Chua et al.

    Facial stereotype bias is mitigated by training

    Social Psychological and Personality Science

    (2020)
  • E.J. Cogsdill et al.

    Inferring character from faces: A developmental study

    Psychological Science

    (2014)
  • W. De Neys et al.

    Split-second trustworthiness detection from faces in an economic game

    Experimental Psychology

    (2017)
  • R. Dotsch et al.

    Statistical learning shapes face evaluation

    Nature Human Behaviour

    (2016)
  • J.L. Eberhardt et al.

    Looking deathworthy: Perceived stereotypicality of black defendants predicts capital-sentencing outcomes

    Psychological Science

    (2006)
  • J.L. Eberhardt et al.

    Seeing black: Race, crime, and visual processing

    Journal of Personality and Social Psychology

    (2004)
  • A.D. Engell et al.

    Implicit trustworthiness decisions: Automatic coding of face properties in the human amygdala

    Journal of Cognitive Neuroscience

    (2007)
  • V. Falvello et al.

    The robustness of learning about the trustworthiness of other people

    Social Cognition

    (2015)
  • F. Faul et al.

    G* power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences

    Behavior Research Methods

    (2007)
  • R.H. Fazio

    On the automatic activation of associated evaluations: An overview

    Cognition and Emotion

    (2001)
  • R.H. Fazio et al.

    On the automatic activation of attitudes

    Journal of Personality and Social Psychology

    (1986)
  • S.T. Fiske et al.

    A model of (often mixed) stereotype content: Competence and warmth respectively follow from perceived status and competition

    Journal of Personality and Social Psychology

    (2002)
  • J.B. Freeman

    Abrupt category shifts during real-time person perception

    Psychonomic Bulletin & Review

    (2014)
  • Cited by (7)

    • Contextual modulation of appearance-trait learning

      2023, Cognition
      Citation Excerpt :

      It is noteworthy that the AB-A and AB-B effects seen in Experiment 4, while highly significant, are also a little weaker than seen in Experiments 1 and 2. There is increasing consensus that learning plays a key role in the emergence of first impressions (Chua & Freeman, 2022; Cook et al., 2022; Falvello et al., 2015; FeldmanHall et al., 2018; Lee et al., 2021; Sutherland et al., 2020; Sutherland, Burton, et al., 2020; Verosky & Todorov, 2010). At present, however, little is known about the nature of the learning processes responsible.

    • The cultural learning account of first impressions

      2022, Trends in Cognitive Sciences
      Citation Excerpt :

      Rather than ‘judging a book by its cover’, the resulting judgements are akin to ‘judging a book by its first page’. Crucially, however, newly acquired face–trait associations generalise to novel faces of broadly similar appearance [50–53]. This suggests that our first impressions of strangers are influenced by our knowledge of familiar others and their traits.

    • The effects of visual attention on social behavior

      2024, Social and Personality Psychology Compass
    View all citing articles on Scopus

    This paper has been recommended for acceptance by Dr. Jack Rachael.

    View full text