Poster #10 - Black gumball or pink gumball: Preschoolers’ understanding of multiple possibilities

• In Event: Poster Session 10

Abstract

An important aspect of human abstract reasoning is our ability to consider multiple possibilities (e.g., the cat might be outside or in the house), and this ability underpins our representation of modal concepts and modal language (e.g., possibly, might). How do young children represent possibilities in their minds? Recent studies in cognitive development suggest that preschoolers may not be able to represent multiple possibilities, and therefore lack modal concepts such as possible, impossible, and necessary (Leahy & Carey, 2020). Under this hypothesis, dubbed a “minimal representation of possibility”, children under four cannot consider multiple possibilities; instead, they conflate what might be true (e.g., the cat might be outside) with what is true (e.g., the cat is outside). This behavior has been demonstrated across various modal reasoning tasks (Leahy et al., under review; Leahy, under review; Leahy & Carey 2020; Mody & Carey, 2016).

The current study aimed to see if three-year-old children could employ modal concepts in a task adapted from previous probability tasks. We designed a task that has the same logical structure used to test modal reasoning (Leahy & Carey, 2020; Mody & Carey, 2016). In Experiment 1, children (N=16) were told that their goal was to get a pink gumball and they were asked to choose between pressing the lever for one of two gumball machines. Two types of test trials were included: (1) a gumball machine with a single pink gumball, and a gumball machine with two gumballs: one pink and one black, and (2) a gumball machine with two pink gumballs and a gumball machine with one pink and one black gumball (Figure 1a.). Children chose the correct machine 92% of the time, above chance (50%; p < .001), and above the performance predicted by the minimal representation of possibility hypothesis (75%; p = 0.0014). This supports the hypothesis that three-year-olds can consider multiple possibilities.

Experiment 2 tested if performing the gumball task improved performance on a modal reasoning task. Three-year-old children performed the gumball task, and then a 3-exit slides task (Leahy, under review) (Figure 1b). We hypothesized that perhaps computing possibilities over objects in the gumball task may help children compute over possible locations in the slides task. Children (N=36) chose the correct machine 86% of the time, above chance (50%; p < .001), and above the performance predicted by the minimal representation of the possibility hypothesis (75%, p < .001). On the slides task, they chose the correct exit 70% of the time, above chance (50%; p < .001), but not significantly better than the 60% performance in Leahy (under review). Thus, we replicated the findings of Experiment 1, but the children’s performance did not significantly improve on the slides task.

Results support the hypothesis that three-year-old children represent multiple possibilities, and they employ modal concepts under some conditions. We suggest that quantifying possibilities over objects (as in the gumball task) may be easier than quantifying over trajectories or locations (as in the slides task). Future research will explore this hypothesis further.

Authors

• Stephanie Nicole Alderete, University of California - Berkeley
  Presenting Author

• Fei Xu, University of California - Berkeley