Individual Submission Summary
Share...

Direct link:

Disentangling Mechanisms Influencing Children’s Choices Across Childhood

Fri, April 9, 4:20 to 5:50pm EDT (4:20 to 5:50pm EDT), Virtual

Abstract

Recent findings suggest that young children (4-year-olds) have extremely elevated levels of exploration, and that their patterns of choices show a surprising level of systematicity. But the mechanisms enabling and supporting this systematic behavior are not well understood. Work in adults suggests that uncertainty-directed exploration relies on executive functioning processes that are immature in young children. Young children instead appear to rely on a simpler strategy that may often approximate directed exploration in which they are more likely to choose options that they have least recently sampled.

The current study seeks to understand what drives this behavior as well as how it develops. More specifically this study has two goals: 1) to disentangle the influences of several possible contributors to young children’s choices, 2) to understand development in decision-making by examining how the strength of these factors vary with age across childhood.

Children (N= 300) from ages 4 to 9-years-old and adults (N = 51) took part in a reward-based decision-making task. On each trial of this task, participants were presented with two options and chose one, earning points for their choice. The goal was to collect as many points as possible, and children earned more stickers for collecting more points. The options presented on each trial were pulled from a set of four possible options that varied, critically, both in terms of reward outcome (i.e. number of points) and in how frequently they appeared on a trial. For example, the most frequent option was present on 80% of trials while the least frequent was only present on 20% of trials. Across four different blocks of the task, the mapping of reward values to frequencies was manipulated between options. Location (left vs. right) in which each option appeared was randomized from trial to trial. This design allows us to disentangle the relative influences of reward value, frequency (as an indicator of novelty or uncertainty), and simple location-based switching (e.g. left-right-left-right, regardless of option identity) on children’s choices.

We use a regression model to compare the relative influences of these factors across age groups. The youngest children (4-year-olds) show about equal influence of all three factors. With age simple (location-based) switching drops down, having little to no influence in older children and adults, while the influence of reward value rises dramatically with age in children (see Figure 1). The effect of option frequency remains constant across age groups.

These findings suggest a shift in strategy over childhood, with different mechanisms primarily driving choices at different ages: from an early tendency toward behavioral variability, to a strategy focused on uncertainty and novelty, to one largely motivated by rewards. Put another way, children transition from strategies characterized by undirected exploration, to directed exploration, to exploitation over the course of development.

Authors