Cross-dimensional assimilation effect in facial trustworthiness evaluation: The asymmetry of in-group favoritism under the moderation of socioeconomic status

Abstract

<p indent="0mm">It is common for within-dimensional assimilation effect in the social evaluation of faces under social context. Recent studies have also found a cross-dimensional assimilation effect in the facial trustworthiness evaluation with the socioeconomic status (SES) background information as social context. However, whether the SES of evaluators moderates this cross-dimensional assimilation effect remains unclear, since the relationship between evaluators and evaluatees has largely been ignored in studies of the assimilation effect. Therefore, the present study aimed to investigate whether the SES of evaluators moderates the impact of the SES background information for evaluated faces on facial trustworthiness evaluation, and thereby to explore the broader theoretical issue on the interaction between evaluator and evaluatee on assimilation effect. There were two phases in the experiment. In the first phase, participants completed a status-color association training task, to learn how to associate colored backgrounds (purple/blue) with different levels of SES (high-SES/low-SES). This was to enable the participants to accurately perceive the SES information of the evaluated faces based on the background color in the subsequent evaluative task. In the second phase, they completed a facial trustworthiness rating task that used a 100-point scale to evaluate the trustworthiness of the faces displayed on colored backgrounds indicating high- or low-SES (<italic>T</italic>_high-SES/<italic>T</italic>_low-SES). After that, the facial trustworthiness baseline (<italic>T</italic>₀) was also measured by presenting participants the isolated faces without the colored backgrounds. The cross-dimensional assimilation effect of SES information of evaluated faces on facial trustworthiness (<italic>T</italic>_change) was calculated as follows: <italic>T</italic>_change=<italic>T</italic>_high-SES–<italic>T</italic>₀ or <italic>T</italic>₀–<italic>T</italic>_low-SES. One-sample <italic>t</italic>-test analysis revealed that the presence of high/low SES background information for the evaluated faces led to the evaluators’ trustworthiness evaluation being closer to the faces’ SES background information. That is, both the high and low SES information for the evaluated faces led to a significant assimilation effect on facial trustworthiness evaluation. But the assimilation of low SES information among low-SES participants is non-significant. Furthermore, a 2 (SES background information for evaluated faces: High vs. low) × 2 (SES of evaluators: High vs. low) repeated-measures ANOVA of the <italic>T</italic>_change showed a significant interaction of evaluators and evaluatees’ SES. Simple effect analysis revealed that the assimilation of low-SES information on facial trustworthiness evaluation among low-SES participants is significantly smaller than that among high-SES participants, which was also supported by the item-based ANOVA analysis. In summary, the results of the experiment suggested: (1) Both high-SES and low-SES information for evaluated faces exerted an assimilation effect on facial trustworthiness evaluation; (2) the assimilation effect is moderated by the SES of evaluators with an asymmetry of the in-group favoritism among high-SES evaluators and low-SES evaluators. The present study also contributed to the literature on the assimilation effect, by showing that: (1) Assimilation effect can occur not only between context and evaluative targets in the within dimension but also in cross-dimensions; (2) the social relationship between evaluator and evaluatee should be taken into account when evaluating the assimilation effect of social context.