About this srcd poster session

---

Panel information
Panel 13. Moral Development

Abstract

---

Theoretical models stipulate that adults evaluate others based on two fundamental dimensions: morality and competence (Abele & Wojciszke, 2007; Fiske et al., 2007). Developmental research shows that children prefer prosocial over antisocial (Van de Vondervoort & Hamlin, 2018; Hamlin et al., 2018) and competent over incompetent agents (Myslinska-Szarek & Warneken, 2024). While this prior work has examined how children make inferences about an agent’s future behavior within a single domain, little is known about whether children generalize behaviors across domains. This project aims to fill this gap by testing whether children who see an agent acting either morally or competently in one situation generalize this information to future moral and competent behaviors. Specifically, we focus on two key questions: (1) How do children generalize behavior within and across the domains of morality and competence? (2) Are there age-related differences in how children apply information about individual behavior to predict future actions?
To address these questions, we conducted two experiments with children 5 to 9 years old. We used between-subjects design manipulating where, during an introductory phase, children either learned about two characters’ level of completing a task (competence as the source domain) or their level of fairness (prosociality as the source domain) (Figure 1). In the subsequent test phase, the same characters were shown in other scenarios that required competent or prosocial behaviors. We asked children how competent and prosocial each character would behave in these novel scenarios, how much they liked or disliked each character, and who they would choose as friends.
In a pilot study with N = 36 children (Mage = 7.2 years, SD = 1.7), we validated our experimental procedure and the effectiveness of manipulations of prosociality and competence. Results showed that children generalized behaviors when the source and target domains were the same (within-domain generalization), but not when the source and target domains differed (across-domain generalization).
For our main study, we pre-registered a target number of N = 75 children at 5 to 9 years of age. Preliminary results from N = 64 children (85% of the planned sample) suggest that while younger children tend to be less differentiating by generalizing information within and across domains, with age, children generalize behaviors within but not across domains (b = 3.00, SE = 0.97, p = .012). Specifically, older children expect that a character presented as competent in one task will be competent in another task, and a prosocial character was expected to be more prosocial in another situation (b = 2.13, SE = 0.13, p < .001). However, the same children did not generalize behaviors across domains – did not assume that competent characters would be more prosocial than less competent ones, and vice versa, prosocial characters would not be more competent than less prosocial ones (b = 0.18, SE = 0.13, p = .173).
These findings shed light on how children’s understanding of morality and competence shapes their social judgments. Younger children tend to generalize more holistically, while older children develop a more domain-specific approach, which may play a role in selecting appropriate social partners.

Author information

Author	Role
Dr. Katarzyna Myslinska Szarek, University of Michigan,	Presenting author
Felix Warneken, University of Michigan	Non-presenting author

⇦ Back to session