Children's risk preferences vary across sexes, social contexts, and cultures

Abstract

People exhibit more risk-prone behaviors when together with peers than when in private. The interplay of social context effects and other variables that alter human risk preferences (i.e., age, sex, or culture) remains poorly understood. Here, we explored risk preferences among Namibian Hai||om and Ovambo children (N = 144; Age_Range = 6-10 years). Participants chose between risky and safe options in private or during peer presence. In a third condition, children collaborated with peers before their risk preferences were assessed in those peers' presence. Children from both societies were risk-averse, but Hai||om children showed greater risk aversion than their Ovambo counterparts. Across cultures and ages, boys were less averse to risks than girls. This effect was most pronounced during peer presence, whereas collaboration did not additionally affect risk preferences. These results suggest a dynamic interplay of individual, social, and cultural factors shaping children's risk preferences.

Fig. 1. Illustration of study protocol across conditions.

a Individual condition; b observation condition with peer present; c collaboration condition with peer present; Illustration by Leonore Blume, used with permission.

Fig. 2. Retrieval of risky or safe option.

a Children’s choice of either safe reward (transparent ball on the left-hand side) or risky reward (blue opaque ball on the right-hand side); b Risky choice; c Safe choice; Illustration by Leonore Blume, used with permission.

Fig. 3. Risk Preferences depending on sex, condition, and culture.

Posterior probability distributions illustrating the estimated probability that a child will choose the risky option depending on their sex, condition, and culture. Densities are based on the full model to illustrate all potential interactions between predictors. Children’s age is set at the mean. Dotted vertical lines represent chance level at probability_Chance = 0.50; n = 125 participants.

Fig. 4. Risk preferences across ages, conditions, and culture.

Posterior probabilities illustrating the estimated probability that a child will choose the risky option depending on their age, condition, and culture. Densities are based on the full model to illustrate all potential interactions between predictors. Children’s sex is centered at zero. Solid lines present posterior means, surrounding areas present 95%-HPDs. Dotted horizontal lines represent chance level at probability_Chance = 0.50; n = 125 participants.