Curiosity has emerged as an essential skill for children to thrive in the twenty-first century, with organizations including the World Economic Forum and the OECD listing it among essential skills for success. Little is known, however, about how to measure children’s curiosity, limiting researchers’ ability to evaluate whether early childhood programs build this essential trait.
In this paper, the authors offer a new tool for measuring curiosity in preschool-aged children. They create the Animal Game, an online game consisting of trials in which children choose between learning about either a mystery animal or a familiar animal. Their choices result in two measures of curiosity: a Novelty Score, which counts the trials in which the child chose the mystery animal, and reflects one’s pull toward novelty and the unknown; and a Novelty + Knowledge-Seeking (NKS) Score, which counts the trials in which the child both chose the mystery animal and, once it was revealed and named, asked to learn more about it, and reflects the goal-directed drive to fill a specific gap in one’s knowledge.
The authors administer the Animal Game to a racially and socioeconomically diverse sample of 594 children recruited nationally through housing authorities, libraries, health clinics, and social media. They assess the game along a range of dimensions, including whether children’s choices are consistent across trials and over time, whether the Novelty and NKS scores appear to capture distinct measures of curiosity, whether the game yields different results across genders or parental education levels, and finally whether children’s scores are associated with their parents’ predictions for those scores or with their own vocabulary level.
The authors find the following:
- The Animal Game appears to capture curiosity consistently. Children differed from one another in how often they chose the mystery animal, suggesting the game effectively distinguishes between different levels of curiosity, while each child’s choices tended to hold across trials, suggesting the differences reflect a stable tendency rather than random variation.
- Reaching for the unknown (interest-type curiosity, represented by the Novelty Score) came more readily to preschoolers than wanting to understand it (deprivation-type curiosity, represented by the NKS score). Children chose the mystery animal in about half of the ten trials on average, but asked to learn more on fewer than three, and a quarter never once asked about an animal they had chosen to reveal. The structure of children’s responses indicates the game captures two distinct dimensions of curiosity.
- A score meant the same thing for every group of children. One concern with the assessment is that a particular item might land differently for different children — for example, one animal pair pulled in children of college-educated parents while another pulled in everyone else, the two groups’ scores wouldn’t be comparable even if they averaged the same. The authors find no such pattern: the same animal pairs drew children in, to the same degree, for boys and girls and for children of more- and less-educated parents.
- Parents’ assessments of their children’s curiosity were not predictive of their children’s actual scores. Parents rated how much their child prefers new activities and topics, and how curious their child is relative to other preschoolers, with neither rating predicting how children played the game.
- Children who sought out knowledge also understood more words. Both curiosity scores were positively associated with children’s vocabulary, measured by asking children to point to the picture matching a spoken word. The association was somewhat larger for the NKS Score, consistent with the idea that actively closing knowledge gaps is more tightly coupled to the adult–child conversations through which vocabulary grows.
- Curiosity matters for early language regardless of familial resources. The link between curiosity and vocabulary barely moved when the authors accounted for household income, parental education, preschool enrollment, parenting stress, and parents’ beliefs about learning. It was also comparable in size for children across the income and education range.
The upshot is that curiosity can now be treated as a viable target of early childhood intervention, with early evidence suggesting it’s possible to inspire curiosity in children. The same children assessed here were participants in a randomized trial of a text message–based parenting program, which found the program increased children’s knowledge-seeking on the Animal Game while leaving their preference for novelty unchanged. The drive to close a specific knowledge gap may therefore be the more responsive of the two dimensions, though the mechanisms remain to be identified.