The Exploration Advantage: Children’s instinct to explore allows them to find information that adults miss.

Abstract

Humans have a long childhood in comparison to all other species. Across disciplines, researchers agree that humans’ prolonged immaturity is integral to our unique intelligence. The studies presented here support the hypothesis that human beings’ extended childhood pays off in the form of an ability to learn more about changing environments. Across two studies (n = 213), children and adults played a game where they chose among four different cartoon monsters yielding different numbers of star rewards. Adults focused on maximizing reward, while children chose to explore longer, even at the cost of earning fewer stars. As a result, adults won significantly more stars than children did. However, in the ‘dynamic’ version of the task, the rewards given out by the monsters changed halfway through: the monster that had been giving out the fewest stars began giving out the most. Because children continued to explore whereas adults ignored the low-reward monster, children were much more likely than adults to detect the change. This illustrates that while exploration may be costly in the short term, it leads to a more flexible understanding of the world in the long term, particularly when that world is changing.

Children are more exploratory and learn more than adults in an approach-avoid task

“…In a classic "multiarmed bandit task," participants choose over a number of trials between a fixed number of options, each of which produces a fixed reward (e.g., one option provides 3 candies and another option provides 8 candies). In this task, adults repeatedly choose the most rewarding option after a brief initial period of exploration, while children as young as 3 years old are more likely to persist in choosing both good and bad options throughout the task (Blanco & Sloutsky, 2019, 2020Sumner, Li, et al, 2019;Sumner, Steyvers, et al, 2019). Using a modified multiarmed bandit task, another study found that 7-to 11-year-olds showed more directed exploration (actively choosing options that resolve uncertainty) than adults.…”

“…In addition, very little research has investigated the connection between explore-exploit decision making and subsequent learning (but see Sumner, Li, et al, 2019). If children do explore more than adults across tasks, does this help them learn?…”

“…Some research suggests that 7-to 11-year-olds are less likely than adults to generalize from explored actions to new actions in multiarmed bandit tasks (E. Schulz et al, 2019)meaning children may be less likely than adults to make false generalizations in the first place. In addition, there is some evidence that 5-to 12-year-olds are more likely than adults to notice changes in the reward structure of a multiarmed bandit task (Sumner, Li, et al, 2019)-meaning even if they do make false generalizations, children might notice the information that adults miss and correct their initial false beliefs. And finally, if children are able to capitalize on this information through later exploitation, they may even eventually earn more rewards than adults do.…”

Embracing curiosity eliminates the exploration-exploitation dilemma

“…The answer to this question is empirical. If our account does well in describing and predicting animal behavior, that would be some evidence for it ( Sumner et al, 2019 ; Wang and Hayden, 2019 ; Jaegle et al, 2019 ; Gottlieb and Oudeyer, 2018 ; Kidd and Hayden, 2015 ; Berlyne, 1950 ; Colas et al, 2020b ; Rahnev and Denison, 2018 ; Wilson et al, 2020 ; Cogliati Dezza et al, 2017 ; Berger-Tal et al, 2014 ). If it predicts neural structures ( Cisek, 2019 ; Kobayashi and Hsu, 2019 ), that would be some evidence for it (we discuss this more below).…”

“…The choice between exploring and exploiting is a kind of decision that is faced routinely by learners of all kinds, including foraging bees, business organizations, humans, worms, monkeys, rodents, birds, children, and computer algorithms ( Gupta et al, 2006 ; Sutton and Barto, 2018 ; Woodgate et al, 2017 ; Lee et al, 2011 ; Schulz et al, 2018 ; Calhoun et al, 2014 ; Wang and Hayden, 2019 ; Sumner et al, 2019 ; Auersperg, 2015 ). But is reward really fundamental to it?…”