The research addressed a gap in how depression has been studied socially. Previous work on social information processing in depression has focused primarily on self-focused or individual contexts, leaving unclear how adolescents with major depressive disorder process actions in social group settings. To probe this, the team used computer animations that simulated social group interactions in which participants observed positive actions, depicted as helping, and negative actions, depicted as hindering. After viewing the scenarios, participants judged whether words describing helping, hindering, or neutral interactions were real words. The key measure was the facilitation effect, defined as faster responses to valenced words compared with neutral words. Faster reactions to words matching the actions just observed were used to infer that the observed group behavior had been spontaneously generalized to other group members, without deliberate instruction.

For hindering actions, the 30 adolescents with major depressive disorder showed facilitation effects regardless of group membership, indicating that they broadly generalized negative actions across the observed social scene. Healthy controls, by contrast, exhibited no facilitation effect for hindering actions at all. The study also found that scores on the Children’s Depression Inventory, a self-report measure of depressive symptoms, were positively correlated with the facilitation effects in the adolescents with major depressive disorder, an association suggesting that the breadth of generalization tracked with reported depressive symptom levels.

Helping actions revealed a different picture for the two groups. Adolescents with major depressive disorder again generalized broadly, exhibiting facilitation effects for helpful behaviors across group boundaries. Healthy controls also showed a facilitation effect for helping, but only for members of the same group. This contrast indicates that in healthy adolescents, action expectations are valence-dependent, with helpful actions extended within groups while negative actions are not spread in the same way, whereas the depressed group extended both types of actions without regard to group membership.

A control condition underscored that these effects depended on actually observing the group actions. In a separate group of 31 participants tested without prior presentation of information on group actions, the facilitation effects in adolescents with major depressive disorder disappeared for both helping and hindering actions. This indicates that the generalized expectations measured in the main experiments arose from the observed interactions rather than from a response tendency independent of the animated scenarios.

Taken together, the authors conclude that adolescents with major depressive disorder spontaneously generalize both harmful and helpful actions to others regardless of group membership, while healthy controls demonstrate valence-dependent generalization, extending only helpful actions within groups. Within the scope of this study, the pattern points to a negativity bias and broader generalization in action expectations among depressed adolescents, extending the study of social information processing in depression beyond self-focused contexts and into group-level social inference.

Subject of Research: Negativity bias in how depressed adolescents generalize observed group actions to other group members

Article Title: Negativity bias in the generalization of actions to group members by depressed adolescents

Article References: Zhou, Y., Liu, X., Xu, Z., Chen, J., Zhang, X., Huang, H., Zhou, D., & Yin, J. (2026). Negativity bias in the generalization of actions to group members by depressed adolescents. BMC Psychology. https://doi.org/10.1186/s40359-026-05502-5

Image Credits: AI Generated

DOI: 10.1186/s40359-026-05502-5

Keywords: depression, adolescents, major depressive disorder, social cognition, negativity bias, generalization, group dynamics, lexical decision task, Children's Depression Inventory, social information processing, Negativity, bias