Measuring Our Changing First Impressions

People make rapid judgments about the characteristics of others based on their facial expressions. Although these first impressions may seem superficial, they have been found to predict legal, political, and financial outcomes.

Research suggests that many of the first impressions we form are based on perceptions of the trustworthiness, dominance, and attractiveness of the person we are viewing.

Although people assume that the social judgments we make based on faces reflect enduring characteristics of the person, research shows that they vary, and that this variance can be as large as the variance in impressions formed from viewing pictures of different people.

Research examining variance in face-based impressions often uses ambient images (i.e., natural photographs of faces). These images, however, can vary in many ways, including emotional expression, image viewpoint, gaze direction, hairstyle, lighting, color, and background — to name a few.

In a 2017 article in the British Journal of Psychology, researchers Clare A. Sutherland (University of Western Australia), Andrew W. Young (University of York), and APS Fellow Gillian Rhodes (University of Western Australia) focused on two specific sources of within-person variability in facial impressions: emotional expression and viewpoint. They sought to understand how these two factors influence the impressions of trustworthiness, dominance, and attractiveness that we form from images of faces.

The researchers utilized faces from the Karolinska Directed Emotional Face database — one of the most widely used databases of facial emotion. The database contains images of people displaying different emotional expressions taken from multiple viewpoints. Participants rated the trustworthiness, dominance, or attractiveness of faces with emotional expressions (e.g., happiness, anger, fear, disgust, sadness, and neutral) taken from various viewpoints (forward-facing, three-quarter profile, and full profile).

The researchers found more variability in social impressions between pictures of the same person than between pictures of different people. Emotional expression was found to be the main source of this variability. Although the individual influence of viewpoint on social impressions was minor, it was found to interact with emotional expression to modify social impressions. For example, happy faces were thought to be more trustworthy, and angry and disgusted faces less trustworthy, when a person was viewing a forward-looking face compared with a three-quarter-profile or a full-profile face. Angry faces were perceived as more dominant, and sad faces as less dominant, when a person was viewing a forward-looking face compared with a three-quarter-profile or a full-profile face. Happy faces were perceived as being more attractive, and disgusted faces as being less attractive, when a person was viewing a forward-looking face compared with a three-quarter-profile or a full-profile face.

The authors suggest that the viewpoint of the pictured face may convey directionality of the emotion, such that forward-facing photos suggest that the emotion is being directed at the viewer. This might be an adaptive trait (e.g., a person who is angry at you may pose an immediate threat, while a person who is happy with you may not).

An additional benefit of this work is its contribution to the emotional-face database it utilized. While ratings of trustworthiness, dominance, and attractiveness existed for neutral forward-facing images in the database, no such ratings existed for faces of other emotions or for neutral faces pictured from other viewpoints. The authors have made the social-impression rating for the images used in their study (i.e., images displaying anger, disgust, happiness, sadness, and fearful expressions from front-facing, three-quarter, and full-profile viewpoints) accessible online. This byproduct of their study can help other researchers understand how social impressions are formed from faces or assist those wishing to control for these factors in their own studies.

Reference
Sutherland, C. A., Young, A. W., & Rhodes, G. (2017). Facial first impressions from another angle: How social judgements are influenced by changeable and invariant facial properties. British Journal of Psychology, 108, 397–415.