Preschool children selectively explore following a surprising claim

• In Event: Poster Session 01
  In Poster Session: PS 01 Topic - Social Cognition

Abstract

Testimony from other people allows children to learn about non-observable entities such as germs or God (e.g., Harris & Koenig, 2006). It also allows children to learn about observable entities. When children receive testimony from others about observable entities that contradicts their prior beliefs, children may seek first-hand information to decide whether to revise those beliefs or not, just as they do upon encountering surprising phenomena first-hand (e.g., Bonawitz, et al., 2012). Indeed, 6- to 8-year-old but not 4- to 5-year-old children are significantly more likely to explore objects when their intuitions about them have been contradicted rather than confirmed (Ronfard et al., 2018). One reason for this age-related pattern may be that the claim used by Ronfard et al. (2018) was difficult to check as it required children to compare two objects. This study was designed to clarify when, and under what conditions, children begin to incorporate empirical testing into their evaluation of what they are told.

To test whether preschool children would explore following a surprising claim that was easier to check, we presented children (N = 109, M = 5.46 years, R = 4.02-6.94) with three objects: a rock, a piece of sponge-like material and a hacky sack (Figure 1). We then asked children whether the rock was hard or soft. All children stated that the rock was hard. Half of the children then heard confirming testimony: “That’s right, this rock is hard”. The remaining children heard counterintuitive testimony: “Actually, this rock is soft, not hard”. Next, all children were given an opportunity to explore the objects (e.g., by touching or picking them up) when the experimenter left the room.

To investigate children’s exploration, we conducted three pre-registered analyses. All three analyses revealed an effect of testimony (counterintuitive vs. confirming) but no effect of age (continuous). As compared to the confirming testimony condition, children in the counterintuitive testimony condition were significantly more likely to touch the rock first, p < .001, Figure 2a; significantly quicker to touch the rock, p = .003, Figure 2b; and spent more time exploring the rock, Condition X Object interaction, p = 0.017, Figure 2c.

In light of prior work showing an age-related increase in children’s exploration of surprising claims requiring the comparison of two objects (Ronfard et al., 2019), two interpretation of these data are possible. First, children may be getting better at exploration that involves the comparison of two objects rather than one object. Alternatively, children may be getting better at realizing that an unexpected claim can be checked empirically. In other words, the behavior of younger and older children in the current study may look similar but its ultimate purpose is different. Older children explore to determine whether what they have been told is really true. Younger children explore the surprising properties of the objects they have just been told about – and which they take to be true. Whichever hypothesis proves correct, it is evident that that are notable age changes in children’s empirical testing of unexpected testimony.

Authors

• Samantha Cottrell, University of Toronto
  Presenting Author

• Eric Torres, Harvard Graduate School of Education
  Non-Presenting Author

• Paul Harris, Harvard University
  Non-Presenting Author

• Samuel Ronfard, University of Toronto
  Non-Presenting Author