Disentangling within- and between-subject correlations in cognitive models: The essential role of hierarchical estimation.

Prosociality and morality are closely linked to emotions. Immoral actions such as hindering or distributing resources unfairly can induce negative feelings in the victims or recipients, who may even conceive a desire for subsequent retaliation or punishment (Fehr and Gächter ; Sanfey et al. ). Emotions can also motivate moral actions or decisions. For example, people who anticipate that a future moral violation will bring about shame or guilt may try to attenuate the transgression (Tangney et al. ). Moreover, the expectation of positive emotions from others can promote prosocial behaviours, while the prediction of others' negative emotions can discourage immoral actions (Aknin et al. ). Emotions are key factors in children's moral development as well (Dahl and Brownell ; Killen and Smetana ; Malti and Dys ). Young children construct their understanding of morality and prosociality based on the emotional consequences of acts. Emotional understanding also guides the children's behaviours that are based on concerns for others' welfare and fairness, such as helping, sharing, and comforting. In this study, we focused on the connection between emotions and children's sharing in resource allocation. Specifically, this study aims to investigate whether consideration of others' emotions can affect children's decision‐making in resource distribution. Developmental research has focused on young children's sharing behaviour in resource allocation situations. The dictator game has been frequently used to measure children's voluntary sharing behaviour and sense of fairness (Castelli et al. ; Gummerum, Hanoch, and Keller ). In this one‐shot game, the divider unilaterally decides how given resources will be allocated, allowing the divider to theoretically maximize their own profits by keeping all the given resources. Results from previous studies, however, have suggested that this rarely happens. Adults typically donate on average 20%–30% of the given resources in dictator games (Forsythe et al. ; Henrich et al. ; Hoffman et al. ). Similar results have been found in experiments using dictator games on children. Specifically, previous research showed that children and adolescents donated as much as adults did (e.g., Benenson et al. ) or slightly more generously than adults (e.g., Gummerum, Keller, et al. ). These studies imply that human beings, even from young ages, not only pursue the maximization of their own profits but also consider others' interests or prosocial norms when dividing resources. Previous research also suggests the developmental patterns of children's sense of fairness in resource allocation. Specifically, research has shown that as children get older, they allocate more resources to others on average, choose selfish allocation behaviour less, and make equal splits more (e.g., Benenson et al. ; Fehr et al. ; Gummerum et al. ). Understanding others' mental states and emotions is closely linked to children's resource allocation. Previous studies suggested that consideration of others' mental states may influence children's sharing behaviours. For example, Takagishi et al. () found that preschoolers with a more mature theory of mind showed more altruistic sharing than other children. Ongley and Malti () demonstrated that children's sympathy (i.e., other‐oriented concern) was a predictor of their sharing behaviour. These studies suggest that children consider the consequences of their sharing behaviour, particularly for others' mental states. Previous research and theories have demonstrated that affective responses from others are the key sources of children's moral understanding and prosocial behaviour (Killen and Smetana ; Malti and Dys ). Therefore, it is possible that children in the previous studies may have considered the effects of their actions on others' emotions (e.g., ‘If I made unfair distribution, the other person would feel unhappy/sad/angry’). However, the previous studies discussed above merely showed the connection between understanding others' mental states and children's altruistic sharing. They did not directly examine whether children's distributive behaviours vary depending on whether children consider others' emotions.