Cheerleader effect

In 2014, the first research was reported by Drew Walker and Edward Vul.[3] Across five studies, participants rated the attractiveness of male and female faces when shown in a group photo, and an individual photo, with the order of the photographs randomised. The cheerleader effect was quantified as the difference between the attractiveness ratings assigned in the experimental condition (in a group photo) and the control (in an isolated image) condition. It was found that participants consistently rated the person as more appealing in the group photograph compared to the individual picture.

This effect occurs with male-only, female-only and mixed gender groups, and both small and large groups. In addition, the effect occurs to the same extent with various group sizes of four and 16 people. Participants in studies looked more at the attractive people than the unattractive people in the group.

Drew Walker and Edward Vul proposed that this effect arises due to the interplay of three cognitive phenomena:[3]

1. The human visual system takes "ensemble representations" of faces in a group. This explanation was backed up by Timothy F. Brady and George A. Alvarez's findings in 2011.[12] In the study, participants were displayed with 30 sets of circles, and circles of various sizes surrounded a tested circle. When asked to determine the tested circle's size, observers' memory of its size is biased by the mean size of all circles shown to them to estimate, showing that people do not encode visual images in memory independently.
2. Perception of individuals is biased towards this average. People's visual systems subconsciously and automatically calculate the average facial impression so that any extreme is ruled out.[13]
3. "Attractive faces are only average."[14] Results showed that composited faces were rated as more attractive and typical without extreme features. Humans develop this preference for "prototype" face from early life since they are easily identified, and individuals could extract social information from these most facelike stimuli to aid social interaction.[15]

When all three of these phenomena are taken together, researchers proposed that the cheerleader effect results from a "hierarchical encoding" and that the hierarchical structure of visual working memory makes observers summarize the group into an ensemble average. Specifically, the individual faces will seem more attractive in a group, as they appear more similar to the average group face, which is more attractive than members' faces.[3]

However, this causation proposal of "hierarchical encoding" was doubted by Carragher et al. in 2019, who found that this effect also occurred in contexts that "were incompatible with hierarchical encoding."[16] They then proposed another explanation: "Social inference mechanism." It implies that the social context of being surrounded by friends may elicit observers' positive inferences on the target's trait, like "friendly or likable, which causes an increase to their perceived attractiveness.[17] However, a recent study in 2021 tested this hypothesis in trustworthiness judgments. It is found that the trait inferences on one's facial trustworthiness did not experience the cheerleader effect.[18]

A 2015 study conducted by van Osch et al. confirmed the existence of the cheerleader effect obtained by Walker and Vul.[19] Based on the effect, the research team offered two other potential explanations for it:[19]

• Selective attention to attractive group members. "People selectively attending to" and have longer fixation time on the most appealing members within a group so that they tend to make group rating based on "an average of the ratings of the most attractive group members"instead of taking every member's attractiveness within the group into account.[20]
• The Gestalt principle of similarity. It suggests that an initial perception of people's attractiveness is towards a group with a similar attractiveness degree as a whole, followed by the perception of its individual member's.

They claim that selective attention fits with the gathered data better.[19] The explanation based on the Gestalt psychology was objected to in this study since researchers found that the effect only occur in the group with large variation in attractiveness. This finding thus was inconsistent with this principle of perceiving similar attractive people as a group to evaluate.

A 2015 replication of Walker and Vul's study failed to show any significant results for the group attractiveness effect. The research team hypothesized potential reasons for this. Firstly, this may be due to cultural differences, since the replication study was performed in Japan.[21] Secondly, the effect size was affected by the variation in the composition of members in a group. Researchers suggested that the cheerleader effect was less likely to occur for people with the similar attractiveness level in a group since the selection attention would not happen to bias participants' memory towards a higher attractiveness average.[19]