Children’s Integration of Prior Evidence in Generalizations About Antisocial Behaviour

• In Event: Poster Session 07
  In Poster Session: PS 07 Topic - Moral Development

Abstract

A core challenge that we must contend with as we develop is using the information we gain to make predictions about the future. Given the central importance of social interaction to the human experience, accurately inferring the intentions and future behaviour of others is particularly valuable, and our capacities for social inference develop alongside those of understanding our physical environment (Buchsbaum et al., 2012). Nevertheless, social stimuli are often more unpredictable due to their complexity and dynamic nature; thus, we may use information through automatic perception of traits and cues, background knowledge of normative social behaviour, and active perspective-taking to predict how others are likely to behave (FeldmanHall & Shenhav, 2019).
As a result, generalizing from experience may become difficult when dealing with unpredictable actors, such as those who commit moral or normative violations. This is true especially for children, whose capacity for social inference is still evolving. Resolving the challenge of making predictions about moral or norm violators may thus be aided both by building a prior belief about an actor (“what kind of person are they?”) and collecting information from a person’s disposition or behaviour at a given moment (“what do they look like they’re going to do?”)
Pilot data (N = 13) suggests that children draw strong generalizations about those who previously committed moral violations when presented with dispositional cues such as laughing and pointing at a potential victim, but are much less likely to do so when these cues are not present, t = –3.22, p < 0.01. This may reflect a kind of base-rate neglect, with children relying more heavily on normative beliefs of how one should behave in the absence of cues that could allow for stronger prediction. Alternately, children may be integrating new dispositional information in a Bayesian fashion and simply have a strong prior belief that actors who do not show signs of behaving immorally will follow social norms.
To distinguish these two predictions, we are extending this pilot data with a new study that tests children’s generalizations from dispositional cues about characters who previously committed a moral violation and those who did not. In this study, children ages 6 to 8 will be presented with 3 different vignettes that include an actor committing a moral violation (e.g. hitting, stealing), a kind act (e.g. giving cookies), or neutral act (e.g. going jogging). Children will then be asked how normative the act was before presenting a new scenario where the outcome must be inferred. The actor’s dispositional cues will be either neutral, hidden, or will be consistent with an antisocial behaviour such as teasing.
Power analyses suggest that a sample size of 96 will be necessary, and data collection will occur from September to March. This data will provide new evidence for how children make inferences about the unpredictable behaviour of moral and norm violators, giving new insight into how children’s emerging capacity for social reasoning influences their understanding not only of prosocial agents, but antisocial ones as well.

Authors

• Rebekah Gelpí, University of Toronto
  Non-Presenting Author

• Denise Arefhaghi, University of Toronto
  Presenting Author

• Jessica A Sommerville, University of Toronto
  Non-Presenting Author