Date Published: May 7, 2019
Publisher: Public Library of Science
Author(s): Eiluned Pearce, Rafael Wlodarski, Anna Machin, Robin I. M. Dunbar, Ethan Morgan.
Most studies of social cognition have focused on dyadic relationships, and rather few have looked at how we engage with individuals in the wider social world into which we are embedded. Here we use principle component analysis (PCA) and path analysis to explore how different aspects of human sociality interact. We demonstrate two distinct clusters in both sexes relating to (i) romantic relationships and (ii) wider social engagement, such as that with the local community. These two domains of relationship were associated with different dispositional traits: individual variation in impulsivity in the former, and in empathy and avoidant attachment in the latter. Although these clusters were broadly similar across both sexes, clearer differentiation is evident in males. In females only, support network size was positively related to the anxious dimension of attachment and, unlike in males, was not related to feelings of inclusion in the local community. This suggests that support networks may play different roles in the two sexes, indicating a productive line of future research. These findings have important practical applications: loneliness interventions that target the specific type of relationship that is felt to be lacking and the associated dispositional traits are likely to be more effective than more generic approaches.
The social world is by far the most complex aspect of our environment. In part, its challenge arises from the fact that it is dynamic and subject to unpredictable and continuous change over time, as the individuals that make it up fall in and out of favour with each other. Handling this complexity requires cognitive skills, such as empathy and mentalising, that are not required for other non-social tasks. These skills are cognitively demanding in terms of both information processing and neural recruitment [1–7]. Our ability to navigate successfully through this social world also depends crucially on the ability to inhibit prepotent responses: an inadvertent comment or an injudicious social interaction can easily destabilise not only our own dyadic relationships with others, but also relationships between third parties in one’s social network 
Overall, the results indicate a distinction between romantic relationships and non-sexual relationships. The first two factors extracted through PCA are broadly similar in males and females, aligning with engagement with wider social networks (Factor 1) as opposed to romantic relationships (Factor 2). Each of these factors includes associated dispositional traits: degree of empathy and avoidant attachment in the case of wider social engagement (Factor 1), and impulsivity with respect to sociosexual orientation (Factor 2). Anxious attachment loaded onto a third factor in both sexes. If only the strongest loadings are considered, this suggests three factors that can be broadly delineated as (i) wider social engagement, (ii) romantic relationships and (iii) the anxious dimensions of attachment.