Daniel Balliet’s research while affiliated with Vrije Universiteit Amsterdam and other places

Research in the social and behavioral sciences is accumulating at an exponential rate. One of the challenges facing scientists is how to quantitatively determine, in an unbiased manner, which predictors contribute the most to explaining variation in a specific phenomenon. To address these issues and improve the predictor importance estimation, we propose an enhanced grouped permutation feature importance (GPFI) method using a state-of-the-art ensemble machine learning regression model. A simulation study utilizing four artificial datasets demonstrated that the mean absolute percentage error of importance estimation was reduced to 5.0% with the proposed method compared to 30.0% with the conventional GPFI method. As an applied example, we utilized the Cooperation Databank to assess the relative importance of 106 predictors of cooperation, including parameters of the study’s experimental paradigm (e.g., group size, incentive structure, and repeated interaction) and sample characteristics (e.g., gender, age, and ethnicity), and found that the proposed technique was able to identify the top predictors of cooperation. These results clarify the implications of such information for understanding and promoting cooperation. The analytical methods developed in this study can be applied across the social and behavioral sciences, especially in well-developed topics that involve accumulated empirical studies.

People frequently participate in interdependent tasks (i.e., tasks in which the outcome of one person is reliant on the other person's actions), in which people can behave in ways that benefit others (i.e., cooperate) to achieve mutually beneficial outcomes in daily life. The ability to select appropriate cooperative partners for these tasks is essential to achieve successful outcomes. Yet, little is known about individual partner preferences for interdependent tasks and whether these preferences change in response to situational affordances of the task (i.e., which traits can affect task outcomes). Here, we report four studies (N = 1021) that investigate the relationship between partner preference, person perceptions, and partner selection in interdependent tasks that afford the expression of warmth- or competence-related traits to affect outcomes. Over four studies, participants were randomly assigned to an interdependent task affording for warmth- or competence-related traits, then rated the most important traits in a partner (Study 1–4), evaluated potential partners' warmth and competence (Study 3–4), and selected partners (Study 3–4). Overall, participants strongly prefer warmth-related traits in a partner, but partner preferences also vary depending on task affordance. Specifically, people demonstrated a stronger preference for partner trustworthiness in tasks affording warmth-related traits and preferred highly competent partners in tasks affording competence-related traits. Additionally, preferences for partner traits strengthened the relationship between the perceived partner trait afforded by the situation and partner selection. We discuss these findings in relation to theories of partner selection and cooperation, as well as the implications of these results to develop tools and interventions to help people optimize their partner selections.

Evolutionary theory suggests that partner selection – the ability to identify and preferentially interact with individuals willing (warmth) and able (competence) to work towards mutual benefits – is a key driver of cooperative behavior. However, partner selection is complex, requiring the integration of various information, such as impression formation and task affordances. Despite its importance, there is limited research on the effect of these factors on partner selection for cooperative tasks. Thus, this paper investigates how person perceptions (warmth and competence), task affordances, and facial and acoustic nonverbal behavior inform partner selection for cooperative tasks. For this purpose, we asked participants to select partners for a task that either expressed warmth- or competence-related traits. Participants had a 3-minute (online) conversation with up to five individuals, reported their evaluations, selected partners for the task, and then engaged in the task. Results indicate that person perceptions guide partner selection, with each trait being more predictive in relevant tasks. Additionally, we found that the perceptions of warmth, but not competence, can be predicted by facial and acoustic cues during conversations. Lastly, we find that in the context of online social interactions, individuals were more cooperative towards selected participants than unselected. We discuss these implications in the context of the theory of partner selection and offer insights on how these results can be used in future efforts for designing socially intelligent artificial systems that support partner selection decisions.

Prisoner’s dilemma (PD) games exist in a variety of configurations. Although the normative solution for any one-shot PD game is defection, empirical research shows that games differ in the amount of cooperation they elicit from human players. Different researchers have developed a variety of indexes, which are functions of the specific payoffs in a given PD, aiming to measure the level of anticipated cooperation elicited by that game. The purpose of the first study in this article was to collect, organize, and compare the mathematical properties of these various indexes. Our results show that many superficially different indexes are the same and that in general six fundamental PD indexes exist. We propose that the six families of indexes correspond to psychological processes and preferences involving prosociality, greed, fear, and risk. In the second study, we provide a meta-analysis of the relationship between selected indexes and cooperation in PD games (k = 280). The amount of variance in cooperation explained by each index ranged from 0.00% to 7.57%. The first family of indexes (described as indexes of the incentive for cooperation in the first study) had the most consistently strong relationships with cooperation. In general, the findings of the meta-analysis aligned with predictions based on our own review and taxonomy of six index families proposed in Study 1.

Two theoretical perspectives (i.e., the risk management perspective and the resource perspective) offer competing predictions that higher class individuals—relative to lower class individuals—tend to be less versus more prosocial, respectively. Different predictions can also be drawn from each perspective about how the class–prosociality association varies across sociocultural contexts. To date, each perspective has received mixed empirical support. To test these competing perspectives, we synthesized 1,106 effect sizes from 471 independent studies on social class and prosociality (total N = 2,340,806, covering the years 1968–2024) conducted within 60 societies. Supporting the resource perspective, we found higher class individuals to be slightly more prosocial (r = .065, 95% confidence interval [.055, .075]); this association held for children, adolescents, and adults and did not significantly vary by any sociocultural variable. In testing the methodological moderators, we found no significant difference in the class–prosociality association in studies measuring objective social class (r = .066) and those measuring or manipulating subjective social class (r = .063). Nevertheless, the observed class–prosociality association was stronger when assessing prosocial behavior involving actual commitment of material or nonmaterial resources (r = .079) compared to prosocial intention (r = .039), and stronger under public (r = .065) than private (r = .016) circumstances. These findings generally support the resource perspective on class-based differences in prosociality—that the relatively higher cost of prosocial behavior, combined with heightened experience of deprivation, results in lower levels of prosociality among individuals with a lower social class background.

Humans evolved to solve adaptive problems with kin and nonkin across fitness-relevant domains, including childcare and resource sharing, among others. Therefore, there is a great diversity in the types of interdependences humans experience across activities, relationships, and ecologies. To identify human psychological adaptations for cooperation, we argue that researchers must accurately characterize human fitness interdependence (FI). We propose a theoretical framework for assessing variation in FI that applies to the social interactions humans would have experienced across situations, relationships, and ecologies in the ancestral past and continue to experience today. According to this model, FI is characterized along four dimensions: (a) corresponding versus conflicting interests (b) mutual dependence versus independence, (c) asymmetrical versus symmetrical dependence (i.e., power), and (d) coordination. Because humans evolved to be highly mutually dependent on others to solve myriad adaptive problems, even compared to our closest living relatives, there is immense variability in the types of interdependences humans experience in daily life. Here, we describe the kinds of variation in interdependence humans experience, paying particular attention to social life in small-scale societies. In demonstrating the diversity of conflicts and coordination problems humans manage, we contend that humans evolved psychological adaptations to infer from signals, cues, and properties of the environment the four dimensions of FI under degrees of uncertainty to reduce the costs of cooperation. We conclude by discussing the theoretical implications of FI theory and emphasize that when individuals understand that others depend on them, it gives way to a new means of leverage to influence how others behave toward them.