Poster #25 - Children make increasingly optimal decisions about risk throughout development

• In Event: 3-159 - Poster Session 12
  In Poster Session: PS 12 Section - Cognitive Process

Integrative Statement

How do risk preferences develop throughout childhood? Previous work suggests that younger children are more risk-seeking than older children (Paulsen et al., 2012), with risk-aversion increasing slowly between childhood and adulthood (Rakow and Rahim, 2010). Another way to consider the development of risk preferences, however, is to consider the optimal amount of risk one should incur in a given task, and how the ability to take optimal risks correlates with age and inhibitory control.

Our study examined the development of risk preferences in 5–10-year-old children (N = 66) using a simplified, non-computerized version of the Columbia Card Task (Figner et al., 2009). In this task, we introduced children to a deck of 10 cards. Nine cards had a red Skittle icon on them, and one card had a red X. Participants were told that each card with a Skittle on it would earn them one Skittle, but the red X would cause them to lose all their winnings from the current round. Children played either 4 or 8 rounds of this game. In each round, they would hit a “card” button to be dealt a card without replacement, always face down, and would hit a “stop” button when they were done. Cards were then flipped over and Skittles paid out accordingly (Fig 1). Taking into account the expected value of each flip, the optimal number of card flips is five.

Following the card task, children participated in the Happy-Sad task (Lagattuta et al. 2011), a Stroop-like measure of inhibitory control in which they had to say “sad” when dealt a “happy face” card, and “happy” when dealt a “sad face” card.

In this preliminary analysis, we examined the average number of card flips per trial to see how much children deviated from the optimal five flips. We find that as children get older, they are more likely to approach the optimal five flips (B = 0.631, SE = 0.0773, p < 0.01), suggesting increasingly optimal decisions about risk throughout development (see Fig 2). Additionally, younger children take fewer cards than older children (B = 0.194, SE = 0.102, p = 0.06), suggesting more risk aversion, perhaps to a suboptimal degree. Importantly, while we document a strong main effect of age on Happy-Sad task performance (B = 0.052, SE = 0.008, p < 0.001), age is still a significant predictor of deviations even after controlling for Happy-Sad task performance (B = -0.273, SE = 0.093, p < 0.01), suggesting that performance on the card task is not simply a reflection of inhibitory control ability.

These results suggest that as children age, they make increasingly optimal decisions about risk. Further, their risk preferences are not simply downstream consequences of inhibitory control development and can be better represented as increasingly sensitive and informed decisions about the optimal amount of risk to incur. A framework that considers risk only in terms of “risk-seeking” or “risk-aversion,” without accounting for the optimality of risk decisions, may miss important changes across development.

Authors

• Dorsa Amir, Boston College
  Presenting Author

• Richard E Ahl, Boston College
  Non-Presenting Author

• Yarrow Dunham, Yale
  Non-Presenting Author