توسعه اعتماد و نوع دوستی در دوران کودکی

Knowing when to trust is an essential skill, but little is known about its cognitive development. No previous studies have examined the development of trust while controlling for age differences in altruism. We hypothesized that older children are more likely to trust, and that this age-related increase is not due to an increase in altruism. In two experiments, we compared the choices of kindergarten (4–5 years) and elementary school (9–10 years) children in economic games. Age was positively related to both trust and altruism, but more strongly to the former. The age difference in trust was robust when we controlled for partner age and the ability to delay gratification. We further hypothesized that older children are more attuned to the probability of reciprocity. Indeed, older children were more sensitive to changes in the game’s structure and the trustee’s characteristics, suggesting that they are not only more trusting, but more discerning in their decisions of when to trust.

One of the most important problems in social decision-making is knowing when to trust. Dilemmas of trust come in many forms, and they affect consumers, organizations, and close relationships (Rotter, 1970). Trusting someone requires the abilities to forego an immediate reward or payoff and to accurately predict how a stranger, the trustee, will react. Effectively managing trust requires complex cognitive skills, including perspective-taking (Davis, 1983) and the ability to delay gratification (Metcalfe & Mischel, 1999). Researchers across the social sciences have identified trust as an important topic (Luhmann, 2000 and Mayer et al., 1995), but relatively little is known about its cognitive development. The present research addresses this issue by examining the social decision-making of children. 1.1. Trust Most research on interpersonal trust has focused on adults (Evans & Krueger, 2009), but it is also important in the lives of children, who must decide how to spend their pocket money, whether to trade or lend toys, and choose which peers to collaborate with in games and school projects (Webley, 2005). Across these contexts, trust is defined as the willingness to accept vulnerability based on a positive expectation of another person’s behavior (Rousseau, Sitkin, Burt, & Camerer, 1998). An act of trust involves the consideration of potential outcomes, what may be gained or lost from trusting, and the estimation of the probability with which those outcomes will occur. That is, the trustor must predict whether the other person is trustworthy. The economic trust game is a widely accepted tool for the measurement of trust (see Fig. 1a for an example). This game involves two individuals with distinctive roles. First, the trustor chooses between the status quo and trust. The choice of the status quo amounts to an act of distrust, and the interaction ends with the trustor receiving a modest, but guaranteed outcome. The alternative response, trust, can yield a better or a worse outcome than the status quo, its value depends on how the second party reacts. When trust occurs, the trustee has a choice between an equitable, mutually beneficial outcome (reciprocity) and an inequitable, personally advantageous outcome (betrayal). For the trustor, reciprocity is the best possible outcome, whereas betrayal is the worst. The dilemma is that a decision must be made without knowing the trustee’s future response. There are legitimate reasons for the trustor to be skeptical of the trustee, whose best possible outcome is betrayal. Full-size image (12 K) Fig. 1. The economic trust game (a) and the dictator game (b). Figure options Not surprisingly, acts of trust are more common within positive interpersonal relationships (Glaeser, Laibson, Scheinkman, & Soutter, 2000) or when third-party information suggests that the other party is trustworthy (Bolton, Katok, & Ockenfels, 2004). However, even when these factors are absent—as they are when strangers interact—adults show moderate levels of trust (Berg et al., 1995 and Johnson and Mislin, 2011). It remains controversial whether this finding is evidence that people trust too much or too little. Normative game theory asserts that rational behavior is strictly self-interested, and that therefore any expectation of reciprocity is groundless (Binmore, 2007). From this perspective, any act of trust is irrational. By contrast, sociological norm theory (Cialdini, Reno, & Kallgren, 1990) asserts that many, if not most, people respect the norm of reciprocity and that trustors can anticipate this respect (Bicchieri, Xiao, & Muldoon, 2011). Comparing empirical data with maximum possible outcomes, some psychologists have argued that the irrationality lies in people not trusting enough (Evans and Krueger, 2011 and Fetchenhauer and Dunning, 2010). In the present paper, we investigate how children make decisions in anonymous dilemmas of trust. Our first hypothesis is that children are predominantly distrusting until they develop the cognitive skills necessary to find value in trust (Eisenberg, Fabes, & Spinrad, 2006). Specifically, we expect that trust depends on perspective-taking (Epley, Morewedge, & Keysar, 2004), abstract reasoning (Piaget, 1977/2001), and the ability to delay gratification (Metcalfe & Mischel, 1999). We hypothesize that cognitive development generally improves the value of trust relative to the status quo, as the benefits of trust are uncertain, delayed, and contingent on the behavior of an unknown partner. Younger children, who prefer immediate rewards and lack the ability to predict the trustee’s future response, prefer the status quo. Two previous studies investigated the development of trust in children. Harbaugh, Krause, Liday, and Vesterlund (2003) compared economic trust across four age groups (ranging from 3rd to 12th graders). Overall, older age groups did not show higher levels of trust. As the experimenters noted, however, the complexity of the experimental design could have been responsible for these null results. Rather than playing a binary choice trust game, such as the one depicted in Fig. 1a, the children played an investment game that required them to decide how many tokens to invest. In addition to being easier for children to understand, the binary choice game also enables experimenters to manipulate structural aspects of trust, such as the trustee’s temptation not to reciprocate. Sutter and Kocher (2007) conducted a study comparing 8-year-olds, adolescents, and adults, and found that the likelihood of trust increased from childhood to adulthood. In this study, the participants always interacted with a same-age partner, which makes it difficult to be sure whether the age difference signaled a developmental change in trust, or whether the trustors responded to an assumed increase in the trustworthiness of their partners. The latter possibility cannot yet be excluded because Harbaugh and colleagues reported that trust was greater when the trustee was older. To pursue the developmental hypothesis further, we begin by testing the basic finding that older children are more trusting, and then investigate whether the effects of age can be explained by the age of the trustee. It is important to distinguish between trust as a mental state and trust as an economic behavior: Although we hypothesize that young children are generally distrusting in economic situations, they may have strong feelings of trust. The emotional state of trust is the willingness to accept information, making it instrumental in language and skill acquisition ( Koenig & Harris, 2005). This capacity manifests itself in early infancy, and Erikson (1950) famously called trust “the first task of the ego” (p. 29). For infants and young children, trust may be a biologically prepared orientation, which only gradually yields to the ability to apply trust selectively. To capture this maturational process, Vanderbuilt, Lui, and Heyman (2005) observed how children of different ages interacted with adults who had previously proven to be either helpful or intentionally misleading. Three-year-olds trusted indiscriminately. At age five, however, children began to distrust the misleading adults. Yet, even many 5-year-olds erroneously trusted the unreliable adult, suggesting that their skills were not fully developed. In adults, the mental state of distrust not only depends on sophisticated cognitive processes, it also has surprising adaptive consequences. For example, priming distrust leads to enhanced creativity and improved problem solving ( Mayer and Mussweiler, 2011 and Schul et al., 2008). Although there is a systematic relationship between feelings of trust and economic behavior (Huang & Murnighan, 2010), there is no one-to-one correspondence. Many factors beyond feelings of trust affect economic decisions, including incidental affect (Lount, 2010), group processes (Brewer, 2008), and social norms (Dunning, Fetchenhauer, & Schlösser, 2012). In fact, models assuming egocentrism suggest that feelings of trust are secondary to the calculation of costs and benefits. According to this view, trustors overemphasize information regarding their own potential outcomes, while neglecting the trustee’s outcomes and hence the probability of seeing their own trust reciprocated (Evans & Krueger, 2011). Therefore, it is reasonable to expect that older children are more trusting in an economic game, even though they may be more inclined to have distrustful feelings. Evidence of a positive relationship between age and trust would further clarify the difference between trust as an economic behavior vs. trust as a mental state. 1.2. Altruism To understand the development of trust, it is necessary to distinguish it from the parallel development of prosocial behavior, or actions taken with the primary intention to help or benefit others (Batson & Powell, 2003). If there are meaningful age differences in altruistic motivations, those differences are likely to have consequences for trust, as there is a structural relationship between acts of trust and altruism. In our example of the trust game (Fig. 1a), the combined payoffs for betrayal and reciprocity are greater than the combined payoffs for the status quo. The trustor may not personally benefit from trust, but there are guaranteed increases in social efficiency (the total wealth of both players) and the other player’s wealth. Hence, an act of trust could be related to either self-interest, prosocial concern for the trustee, or a combination of the two. One way of accounting for the role of altruism in trust is to measure behavior in a truncated trust game, where the second player has no opportunity to reciprocate (Fig. 1b). Such a situation amounts to a dictator game, where the trustor chooses between the status quo and the worst possible outcome of trust, namely “betrayal.” When trust games are yoked with comparable truncated games, rates of trust are higher than rates of selfless giving ( Cox, 2004). That is not to say that prosocial motives are irrelevant. Among adults, concern for others’ well-being predicts trust ( Kanagaretnam, Mestelman, Nainar, & Shehata, 2009). While there is mixed support for the hypothesis that trust increases with age, there are numerous findings linking cognitive development with altruism. Gummerum, Hanoch, Keller, Parsons, and Hummel (2010) found that older children (in a comparison of 3, 4, and 5-year-olds) were more willing to share stickers in a dictator game. Yet, altruism does not plateau at such an early age, it continues develop throughout childhood: Kogut (2012) found that the willingness to share increased between the ages of 3 and 8, and Martinnson, Nordblom, Rützler, and Sutter (2011) found that sharing increased further between the ages of 10 and 15. Similarly, Fehr, Rützler, and Sutter (2011) found that from ages 8 to 17, children become less spiteful and more altruistic. Consistent with prior results, we hypothesize that older children behave more altruistically in a dictator game. Our hypotheses thus far suggest that age is associated with greater altruism and trust. However, consistent with the view that altruism plays a secondary role in trust decisions (Cox, 2004), we also hypothesize that age has differential effects on the two types of decisions. We expect that the effects of age in trust go beyond changes in prosociality. The increasing tendency to trust reflects, at least in part, self-interested decision-making. 1.3. Trust and perspective-taking We propose that the developmental trajectories of trust and altruism diverge because trust decisions are also motivated by self-interested, strategic perspective-taking (Galinsky, Maddux, Gilin, & White, 2008; Trötschel, Huffmeier, Loschelder, Schwartz, & Gollwitzer, 2011). We posit that older children are more likely to trust because they are more focused on the trustee and the expected probability of reciprocity. In addition to anticipating an overall increase in trust, this hypothesis leads to the prediction that older children are more sensitive to variables that affect the probability of reciprocity. In other words, older children will be more trusting, and more discerning in when to trust. As in the case of trust, reciprocity is most likely to occur when there is a preexisting relationship (Glaeser et al., 2000) or the possibility of future interactions (Dal Bó, 2005), but there is still a moderate level of reciprocity among strangers (Johnson & Mislin, 2011). In anonymous interactions, the best predictor of reciprocity is the trustee’s temptation, that is, the extra money obtained from betrayal (Malhotra, 2004 and Snijders and Keren, 1999). Demographic variables, such as gender and age, are also correlated with the trustee’s decision (Croson and Gneezy, 2009 and Sutter and Kocher, 2007). Women and older trustees are more likely to reciprocate. The ability to infer the probability of reciprocity from such situational and demographic cues requires a sophisticated theory of mind (Wellman, 2012). Our hypothesis is that older children are more sensitive to the factors that affect the probability of reciprocity, such as the trustee’s temptation and age. 1.4. Gender differences in economic behavior An additional goal of our research was to test for gender differences in the economic behavior of children. In studies of adult decision-making, women tend to be less trusting but more prosocial than men (Croson & Gneezy, 2009). It is unknown, however, when these differences emerge. Identifying the point of origin in economic gender differences has important implications for education. For example, collaborative projects that require trust among peers could potentially favor male students. Gender differences in trust demonstrate how trust as a feeling can be dissociated from trust as an economic behavior. Research on individual differences in personality has shown that women are more agreeable than men (Costa, Terracciano, & McCrae, 2001), which suggests that women are also more trusting, as trust is categorized as a facet of agreeableness. However, when economic behavior is studied, men tend to show higher levels of trust (for a review see Croson & Gneezy, 2009). Arguably, men are more trusting in economic situations because they are more willing to accept the risk of betrayal (Bohnet & Zeckhauser, 2004). The pattern is reversed for altruism and trustworthiness (Croson & Gneezy, 2009). With respect to altruism, research shows that 3–5-year old girls are more willing to share stickers in a dictator game (Gummerum, Hanoch, Keller, Parsons, & Hummel, 2010). A motivational explanation is that women are generally less competitive than men (Gneezy & Rustichini, 2004). However, Dreber, von Essen, and Ranehill (2011) compared 7–10 years old children and found no gender differences in competitiveness across several tasks. Alternatively, girls may be socialized in accordance with gender roles that encourage relationship oriented behavior (Ruble, Martin, & Berenbaum, 2006). 1.5. Overview We report two cross-sectional experiments that examine the developmental trajectories of trust and prosocial behavior. Our primary hypotheses are that older children are more trusting and altruistic, but that age differences in altruism do not fully account for the effects of age on trust. In addition to studying whether children are more trusting overall, we also examine whether older children are more discerning when placing trust. We manipulated two cues related to the probability of reciprocity, with the hypothesis that older children are more sensitive to these cues. Finally, we examine the development of gender differences in decision-making.