Poster #199 - Three- to Four-Year-Olds use Probability to Infer Other People’s Surprise

• In Event: 3-106 - Poster Session 11
  In Poster Session: PS 11 Section - Social Cognition

Integrative Statement

The emotion of surprise is closely linked with probability, as improbable events are more surprising than probable events. This fact is useful for anticipating how other people will feel – we can infer that people will be surprised when unlikely events occur. However, recent research suggests that children may not make these probability-based inferences of surprise until around the age of 6 or 7. For example, 7-year-olds understand that a girl who has a low chance of receiving a red gumball will be more surprised to receive it than a girl who has a high chance. Further, 6-year-olds connect probability and surprise when prompted to consider the chances of receiving a red gumball (Doan, Friedman, & Denison, in press). These probability-based inferences of surprise emerge quite late, and years after children are first able to recognize when outcomes are improbable (Denison, Reed, & Xu, 2013; Téglás, Girotto, Gonzalez, & Bonatti, 2007). This suggests that it may depend on the development of other conceptual knowledge, such as more mature reasoning about beliefs. Earlier research shows that children sometimes infer surprise from false belief (MacLaren & Olson, 1993; Ruffman & Keenan, 1996), so perhaps probability-based inferences of surprise actually depend on inferences about beliefs. Here we report findings that cast doubt on this possibility, and instead suggest that children’s probability-based inferences of surprise may emerge before their belief-based inferences of surprise.
We tested 120 children ages 3 to 4 to see if they can use probability to infer surprise when probability is made especially salient. If children correctly infer surprise, this would suggest that their probability-based inferences of surprise do not depend on belief attributions because 3-year-olds do not typically consider beliefs (e.g., Wellman & Banerjee, 1991; Wellman & Bartsch, 1988), and children do not make belief-based inferences of surprise until the age of 5 (e.g., MacLaren & Olson, 1993).
Children saw a scenario where two characters each received a red gumball from different gumball machines. In the Probabilistic condition, one machine contained mostly purple gumballs (36 purple, 4 red), and the other contained mostly red gumballs (opposite distribution). In the Deterministic condition, one machine contained mostly purple (36 purple, 4 red), and the other machine contained all red. Children in both conditions were told that one character is surprised they got a red gumball and were asked to choose who was surprised. A GEE revealed only a main effect of condition, Wald X2(df=1, N=120) = 5.70, p=.017. Children in the Deterministic condition correctly chose the character that was surprised (character with a lower chance of receiving a red gumball) more so than children in the Probabilistic condition (see Fig.1). Only children in the Deterministic condition performed above chance level, z=-2.04, p=.041.
These results show that even very young children can use probability to infer surprise, but only when the comparison is to a completely unsurprising (i.e., deterministic) event. Importantly, these inferences are not dependent on belief understanding, suggesting that probability and theory of mind may provide independent routes for inferring surprise in others.

Authors

• Tiffany Doan, University of Waterloo
  Presenting Author

• Ori Friedman, University of Waterloo
  Non-Presenting Author

• Stephanie Denison, University of Waterloo
  Non-Presenting Author