The Pizzagame: A virtual public goods game to assess cooperative behavior in children and adolescents

Abstract

Social dilemmas are characterized by conflicts between immediate self-interest and long-term collective goals. Although such conflicts lie at the heart of various challenging social interactions, we know little about how cooperation in these situations develops. To extend work on social dilemmas to child and adolescent samples, we developed an ageappropriate computer task (the Pizzagame) with the structural features of a public goods game (PGG). We administered the Pizzagame to a sample of 191 children 9 to 16 years of age. Subjects were led to believe they were playing the game over the Internet with three sets of two same-aged, same-sex coplayers. In fact, the co-players were computer-generated and programmed to expose children to three consecutive conditions: (1) a cooperative strategy, (2) a selfish strategy, and (3) divergent cooperative-selfish strategies. Supporting the validity of the Pizzagame, our results revealed that children and adolescents displayed conditional cooperation, such that their contributions rose with the increasing cooperativeness of their co-players. Age and gender did not influence children and adolescents' cooperative behavior within each condition.However, older children adapted their behavior more flexibly between conditions to parallel the strategies of their coplayers. These results support the utility of the Pizzagame as a feasible, reliable, and valid instrument for assessing and quantifying child and adolescent cooperative behavior. Moreover, these findings extend previous work showing that age influences cooperative behavior in the PGG.

Prosocial Behavior and Depression: a Case for Developmental Gender Differences

“…Notably, in this study, more girls identify as “prosocials” than “proselfs,” and prosocial individuals are more likely to be trusting and trustworthy [63•], indicating that adolescent boys may be trusting because they are actually less risk averse. Other studies in healthy adolescents have reported no gender difference in prosocial behavior using a Public Goods Game [64•] and a Trust Game in the role of trustee [65]. Similar to adulthood, this discrepancy in findings might be explained by differential sensitivity to experimental cues by girls compared to boys; however, this remains to be empirically tested in adolescents.…”

Getting less than their fair share: Maltreated youth are hyper‐cooperative yet vulnerable to exploitation in a public goods game

“…Informed consent and assent were obtained from caregivers and youth prior to participation. Following our established procedure to safeguard against limited task comprehension (Keil et al., ), we excluded 79 children and adolescents (46.1% maltreated) because they erred on more than one of nine comprehension questions following instructions on the strategic makeup of the game. Excluded participants did not differ from the final sample population with regard to gender ( p = 0.103, d = 0.16), but regarding the proportion of maltreatment ( p = 0.009, d = 0.26) age ( p ≤ 0.001, d = −0.66), monthly household income ( p = 0.012, d = 0.26), and caregiver's school education ( p ≤ 0.001, d = 0.51).…”

“…Given that our a priori inclusion threshold based on Keil et al. () resulted in exclusion of a sizable number of participants, we conducted sensitivity analyses while relaxing the threshold (up to three errors in the comprehension questions), which only resulted in exclusion of n = 20 participants (40% maltreated). This sensitivity analysis involved a larger sample of N = 388 children and adolescents who were comparable to the excluded children with regard to the proportion of maltreatment ( p = 0.406, d = 0.01), gender ( p = 0.169, d = 0.14), as well as caregiver's monthly household income ( p = 0.106, d = 0.16), and merely differed from excluded children in terms of age ( p ≤ 0.001, d = −1.22) and caregiver's school education ( p = 0.001, d = 0.81).…”

“…Despite ample research in adults (for overviews see Chaudhuri, ; Ledyard, ; Zelmer, ) much less is known about child and adolescent behavior in these situations. Recent work has begun to show that children from community samples initially contribute about the same amount of resources as adults and also display the tendency to cooperate conditionally (i.e., cooperate only when others cooperate; Harbaugh & Krause, ; Keil et al., ; Vogelsang, Jensen, Kirschner, Tennie, & Tomasello, ). Furthermore, older children as well as adolescents have been shown to more flexibly adjust their strategy as a function of their peer's behavior compared to younger children (Keil et al., ; van Hoorn, van Dijk, Meuwese, Rieffe, & Crone, ).…”

“…In this study we analyze cooperative behavior of maltreated and nonmaltreated children and adolescents within a computerized PGG called the Pizzagame (Keil et al., ). In the Pizzagame , participants are led to believe they are playing an online computer game with three sets of two same‐aged, same‐sex coplayers who are in fact computer generated and following preprogrammed scripts.…”