Partner similarity and social cognitive traits predict social interaction success among strangers

Abstract

Social interactions are a ubiquitous part of engaging in the world around us, and determining what makes an interaction successful is necessary for social well-being. This study examined the separate contributions of individual social cognitive ability and partner similarity to social interaction success among strangers, measured by a cooperative communication task and self-reported interaction quality. Sixty participants engaged in a 1-h virtual social interaction with an unfamiliar partner (a laboratory confederate) including a 30-min cooperative 'mind-reading' game and then completed several individual tasks and surveys. They then underwent a separate functional MRI session in which they passively viewed video clips that varied in content. The neural responses to these videos were correlated with those of their confederate interaction partners to yield a measure of pairwise neural similarity. We found that trait empathy (assessed by the interpersonal reactivity index) and neural similarity to partner both predicted communication success in the mind-reading game. In contrast, perceived similarity to partner and (to a much lesser extent) trait mind-reading motivation predicted self-reported interaction quality. These results highlight the importance of sharing perspectives in successful communication as well as differences between neurobiological similarity and perceived similarity in supporting different types of interaction success.

Keywords: empathy; neural similarity; perceived similarity; social interaction; social skills.

Fig. 1.

Main effects and interaction models predicting communicative success. The joint main effects model (IRI + neural similarity) and interaction term model (IRI × neural similarity) compared to single main effects and covariate-only models. Confederate, number of runs and motion similarity were included in all models. Note: The x-axis is a log scale, not linear.

Fig. 2.

Associations between predictive measures and communicative success. Average distance scores were inverted prior to plotting, so higher scores indicate greater success. (A) IRI scores plotted against communicative success. (B) Neural similarity to partner plotted against communicative success. (C) Regression lines for the association between IRI and communicative success with the interactive effect of neural similarity. Hotter colors indicate higher neural similarity, and plot lines are ±1 standard deviation (SD) from mean neural similarity. All models include confederate as a covariate, and the models with neural similarity include the number of runs and motion similarity as a covariate.

Fig. 3.

Brain regions predicting communicative success in the joint main effects model. Average distance scores were inverted prior to plotting, so higher scores indicate greater success. (A) Regression lines for models which yielded a BF ≥ 10 when neural similarity within the region was included. Hotter colors indicate higher neural similarity, and plot lines are ±1SD from mean neural similarity. All models include confederate, number of runs and motion similarity as covariates. (B) Map of regions where the joint main effects model yielded a BF > 3 when neural similarity within the region was included.

Fig. 5.

Associations between predictive measures and perceived interaction quality. (A) MRMS scores plotted against perceived interaction quality. (B) Perceived similarity to partner plotted against perceived interaction quality. Points are jittered randomly by 0.2 units on the x-axis. (C) Regression lines for the association between MRMS and perceived interaction quality with the effect of perceived similarity. Hotter colors indicate higher perceived similarity, and plot lines are ±1SD from mean perceived similarity. All models include confederate as a covariate.

Fig. 4.

Main effects and interaction models predicting perceived interaction quality. The joint main effects model (MRMS + perceived similarity), interaction term model (MRMS × perceived similarity) and single main effects models compared to the confederate only model. Note: The x-axis is a log scale, not linear.