New Research in Psychological Science

Children Delay Gratification for Cooperative Ends
Rebecca Koomen, Sebastian Grueneisen, and Esther Herrmann

Children seem better at delaying gratification while cooperating with others than while pursuing an individual goal, this research suggests. Five- to six-year-olds from Germany and Kenya were paired and, while in separate rooms, instructed that if they resisted eating one cookie while the experimenter was out, they would receive another cookie as a reward. In the interdependence condition, both dyad members were told that they would have to resist the temptation in order for each of them to get a second cookie. In the dependence condition, they were told that their partners did not have a cookie to resist eating and that the outcome of both of them getting a second cookie solely depended on them. In the solo condition, each child was told that a second cookie would be given as a reward as long as he or she resisted eating their first cookie. Results indicated that both German and Kenyan children were better able to resist eating the first cookie in the interdependence condition than in the solo condition. The ability to resist eating the cookie in the dependence condition was between the levels of the other two conditions, suggesting that in the interdependence condition, children delayed gratification because they were depending on their partners and because of the feelings of commitment this tended to generate. These findings indicate that human children might be equipped to respond to interdependencies in ways that improve the success of cooperation.

Same Data Set, Different Conclusions: Preschool Delay of Gratification Predicts Later Behavioral Outcomes in a Preregistered Study
Laura E. Michaelson and Yuko Munakata

In a recent study, Watts, Duncan, and Quan (2018) replicated the classic finding that preschoolers’ ability to delay gratification in the marshmallow test was related to academic achievement, but the relation between delayed gratification and academic achievement decreased when the authors controlled for early cognitive ability or home environment. These findings were seen as a failure to replicate previous results and as suggesting that children’s ability to delay gratification might not be closely related to academic achievement. Michaelson and Munakata independently analyzed the same data set and found distinct results from those found by Watts and colleagues—preschoolers who delayed gratification in the marshmallow test had better academic achievement, fewer problem behaviors, and better social skills in adolescence than those who did not. The authors also showed that social support seemed to explain these relationships better than self-control. They discuss how the analytic decisions they made seem to justify the discrepancy of results between their study and Watts et al.’s and offer a novel interpretation for the classic findings that associate early ability to delay gratification with higher academic achievement, suggesting that the marshmallow test predicts future achievement because it reflects aspects of a child’s early environment (social support) that are important over the long term.

Declines in Religiosity Predict Increases in Violent Crime—but Not Among Countries With Relatively High Average IQ
Cory J. Clark, Bo M. Winegard, Jordan Beardslee, Roy F. Baumeister, and Azim F. Shariff

In two studies, Clark and colleagues found that religion seems to regulate violent behavior in societies in which the average IQ is lower but not among societies in which the average IQ is higher. In the first study, they analyzed the associations between religiosity (i.e., the percentage of the population that practices a religion) and homicide rates between 1945 and 2010 in 176 countries that varied in average estimated IQ. Results indicated that declines in religiosity were associated with simultaneous increases in homicide rates only for countries with lower average IQ. In the second study, Clark and colleagues analyzed modern data from 195 countries and reached similar conclusions. These results indicate that intelligence might moderate and change the relationship between religiosity and moral behavior. Religion seems to have a greater effect in diminishing violence in societies with relatively lower average intelligence. The authors warn that these findings might not reflect every type of cultural system and would likely have been different for regions experiencing religious conflicts. They also suggest that these findings might inspire a reflection on the prescriptive values of Western, educated, industrialized, rich, and democratic (WEIRD) societies and the secularization that accompanies such values.

Two Preregistered Direct Replications of “Objects Don’t Object: Evidence That Self-Objectification Disrupts Women’s Social Activism”
Matthias De Wilde, Annalisa Casini, Philippe Bernard, Robin Wollast, Olivier Klein, and Stéphanie Demoulin

In 2013, Calogero found that women who reported more self-objectification tended to adopt a gender-specific system-justification belief, which led to diminished participation in collective action. Calogero asked women to rank the importance of observable attributes (e.g., weight) and unobservable attributes (e.g., health) to their self-concept and used the difference between the sum of ranks of unobservable attributes and the sum of ranks of observable attributes as a measure of self-objectification. Calogero also scored women’s responses to a gender-specific system-justification scale (e.g., agreement with statements such as “In general, relations between men and women are fair”) and a collective-action-intention scale (e.g., frequency of actions such as “Sign a petition in the street for women’s rights and gender equality”). Here, De Wilde and colleagues report a mini meta-analysis including Calogero’s study, four unpublished studies, and two preregistered direct replications of Calogero’s study that used larger samples of women. Their meta-analysis suggested that the association between self-objectification and gender-specific system-justification belief is much smaller than that reported by Calogero, and their replications revealed no association between these two variables. These findings challenge the association between self-objectification and passivity and the role attributed to self-objectified women in the maintenance of patriarchy by adopting a gender-specific system-justification belief and participating less in collective action.

Design Drives Discovery in Causal Learning
Caren M. Walker, Alexandra Rett, and Elizabeth Bonawitz

How does the design of an object foster the learning of causal rules? Walker and colleagues introduced 3-year-olds to a novel toy that played music when certain pairs of blocks were used. Children were supposed to learn whether a pair of identical blocks or a pair of different blocks produced the music. The critical manipulation was whether the blocks went on top of the toy or were inserted into side openings. Children had difficulty learning the right pairs that produced music when the blocks were placed on top of the toy but fared much better when the blocks went into side openings. In a second experiment, Walker and colleagues tested whether design could also facilitate adults’ causal inferences. Participants were presented with machines similar to the toys in the first experiment and were supposed to learn a distinct conjunctive cause (i.e., that two blocks together activated the machine to play a sound). Adults usually dismiss conjunctive causes in favor of disjunctive relations (i.e., that only one object in a set is needed). Just like the children, adults who were presented with an object with the side openings were more likely to learn the conjunctive cause. These findings suggest that when an object’s design helps learners generate hypotheses they might not have considered otherwise (e.g., two blocks rather than only one are needed for the machine to operate), it facilitates learning. These results support the importance of “setting the stage for learning” when researchers and practitioners design learning environments, Walker and colleagues propose.