The Development of Children’s Representations of Coincidences

• In Event: Poster Session 14
  In Poster Session: PS 14 Topic - Cognitive Processes

Abstract

Coincidences feel special—we have all experienced delight in discovering that someone shares our birthday, or in running into a friend somewhere unexpected. Coincidences grab our attention, and make us puzzle about whether they hold some deeper meaning. What makes coincidences feel so noteworthy? The small amount of existing research on adults’ coincidence representations suggests that adults’ coincidence judgments are based on how likely the target events are to co-occur under certain hypotheses, and how plausible the observer feels the hypotheses are (Griffiths & Tenenbaum, 2007). Yet, little is known about how coincidence representations develop; specifically, whether young children represent coincidences, and if so, what factors influence their representations.
Here we explored whether and how children represent coincidences by examining adults’ coincidence judgments, then comparing them to those of children aged 4-10 years. In Experiment 1, 117 adults read stories describing co-occurrences. For instance, in one story a boy went to a park with his pet bulldog and saw that other people in the park also had bulldogs. We manipulated the number of co-occurrences (either two, three, a few, or all of the people in the park had bulldogs), and the presence/absence of a potential explanation for the co-occurrences (e.g., half of the stories mentioned that there was a bulldog judging competition). We found that adults were more likely to rate events as a coincidence when no plausible explanation was offered (H(1) = 5.91, p < .001). Further, the strength of their ratings increased as the number of co-occurrences increased, but only up to a point—when there were too many co-occurrences (e.g., all of the people in the park had bulldogs), ratings declined (inverted U in Figure 1). These results suggest that adults generated their own explanations when the number of co-occurrences grew too large, and that when an explanation is present for co-occurring events (even surprising ones), the events no longer count as coincidences.
In Experiment 2, we examined the development of coincidence representations. We tested 94 4-10 year old children using the methods of Experiment 1. We found that, like adults, both older children (N = 47, M = 8.30 years) and younger children (N = 47, M = 5.44 years) were more likely to rate co-occurrences as coincidences if no explanation was given (older children: H(1) = 30.27, p < .001; younger children: H(1) = 5.91, p = .015). Furthermore, older children’s coincidence judgments showed an inverted U: like adults, they were more likely to judge co-occurrences as coincidences as the number of co-occurrences increased; but when there were too many co-occurrences, their coincidence judgments declined (Figure 1). In contrast, younger children’s coincidence ratings did not exhibit sensitivity to the number of events.
In sum, our results suggest that even young children represent coincidences—as early as we tested, children recognized that coincidences involve multiple events that lack an explanation for the co-occurrences. However, more fine-grained adult-like representation of coincidence, including sensitivity to the number of co-occurrences, may only emerge in older children.

Authors

• Qiong Cao, Johns Hopkins University
  Presenting Author

• Lisa Feigenson, Johns Hopkins University
  Non-Presenting Author