Research

Humans

Wisdom of crowds

A powerful method for boosting decision accuracy is to combine the independent decisions of different people. This “wisdom of the crowd” approach is often more accurate than the judgements of single individuals. In our lab, we examine when and why different ways of aggregating information lead to improved collective performance, combining empirical research methods (e.g., large online studies, lab studies, and virtual reality) with advanced theoretical modeling. We also apply the insights gained from this research to real-world domains with high stakes, including medical diagnostics, fingerprint recognition, and geopolitical forecasting, with the aim of enabling decision makers to make better decisions by joining forces.

Social influence

People rarely make decisions in isolation. Instead, they tap into the information and knowledge provided by their social environment to make better decisions. Under certain conditions, though, the social environment can also pose a threat—for example, when crowds stampede or political opinions polarize. The ability to successfully navigate the social environment is a key competence for adaptive decision making in an uncertain world. In our lab, we try to understand how individuals integrate information provided by others. This requires an understanding of how and when people search for social information, how they integrate conflicting personal and social information, how information flows through populations, and how these processes are shaped by the interplay between social structure and individual characteristics. To answer these questions, we combine empirical studies observing real-world behavior with theoretical simulations that scale up the observed behaviors to larger populations.

Next to highly controlled lab experiments, we also work with more naturalistic decision making contexts, to test how theories of social influence translate to more realistic decision making environments. This includes immersive reality group experiments (e.g., in Minecraft and Unity), and, as of late, tracking large groups of icefishers in Finland. Here we combine tracking technology with organism-borne videos and heart rate monitors to understand the individual, collective and environmental drivers of human foraging dynamics in the wild.

Cognitive modeling of collectives

To better understand the cognitive processes driving collective decision making, we use cognitive modeling. Recently, we developed a social version of the drift diffusion model (DDM). The DDM assumes that individuals accumulate evidence over time until they research a decision threshold and make a decision. This evidence accumulation process has proven very successful in accounting for a wide range of individual-level decisions in domains including perception, memory, categorization, preference, and inference. We developed a dynamic social version of the DDM whereby multiple individuals simultaneously accumulate evidence and where the choices of one individual enter the evidence accumulation of other, undecided, individuals. We tested the model in an experiment showing that the model was able to recover the dynamics of the complete social decision-making process, accurately capturing how individuals integrate personal and social information dynamically over time. Using process models is a powerful tool to understand social and collective decision making, by linking (cognitive) processes at the individual and collective level. We are currently making these models more realistic, by incorporating individual heterogeneity and social structure, and applying them to real-world social contexts. Other (cognitive) approaches include signal detection theory (in the context of false news discrimination) and reinforcement learning.

Animals

The role of individuality in collectives

When observing a group of insects, fish, birds or humans, it quickly becomes apparent that individuals within these groups are not identical. Or, as poetically put by E.O. Wilson: “When you have seen one ant, one bird, one tree, you have not seen them all.”. Whereas especially behavioral differences were for a long time seen as random noise, it is becoming increasingly clear that behavioral differences between individuals of the same species can have adaptive reasons. This is especially interesting for social contexts, since the behavior of collectives is ultimately determined by the individuals, and the interactions between individuals. In our group, we study the importance of individuality in social and collective decision making. For example, we have shown that boldness predicts who takes the lead in collective movements, who uses social information, and who plays producer or scrounger during foraging in geese flocks. Together with colleagues, we recently developed a brief test to measure social information use in humans. This test can be used to quickly measure the tendency of individuals to rely on the information of others, and help to uncover the sources of variation explaining individual differences in this important trait. We recently deployed this test in several museums gathering data across cultures and development.

In animal communities, like in human communities, individuals generally do not interact randomly but have preferred interaction partners. Such non-random structure can have major implications for a range of ecological and evolutionary processes, including dispersal, invasion, and the spread of diseases and information. In our lab, we study which factors drive social structure, and how social structure develops from early-life into adulthood. We also study the functionality of social networks, specifically looking at information transfer. Together with an international research group, we are studying social information transfer in the wild in the Trinidadian guppy, one of the key vertebrate model systems in ecology and evolution. This system uniquely allows to manipulate group composition in the wild, gather high replication at the group level, and translocate entire groups, making this a powerful system for studying socio-ecological processes. Translocating entire groups across different environments, we have shown that individuals consistently vary in their foraging success, that sociality positively predicts foraging success, and that this is crucially mediated by the group’s sex-composition. Currently, we are looking at how differences in evolutionary predation pressure have shaped network structure and information transfer.

Collective hunting

Together with an international group of researchers, we study cooperative hunting in billfish. By studying these fascinating predator-prey dynamics, we hope to gain a better understanding of the evolutionary origins of (i) collective hunting, (ii) prey’s collective defense mechanisms, and (iii) the bill structure. Chasing these beautiful creatures over several years of the coast of Mexico has been challenging, fun, rewarding and occasionally frustrating 😉 Using high-speed video footage, we have been able to show that sailfish groups hunt by alternating attacks on sardine schools. While only 1 in 4 attacks results in prey capture, multiple prey are injured in almost all attacks, leading to an increase of injured fish in the prey school over time. How quickly prey are captured correlates with the school’s injury level, suggesting that sailfish benefit from other conspecifics’ attacks on the prey. When attacking, sailfish use their bill as a stealth weapon to either tap on individual prey targets or slash through prey groups with powerful lateral motions, being among the highest accelerations ever recorded. This rapid bill movement is either left- or rightward and it turns out that individual have a dominant side when slashing. More strongly lateralized individuals enjoy a higher capture success. This is puzzling as predictability in hunting behavior may give the prey an opportunity to anticipate the attack. Whereas single sailfish are indeed highly predictable, this predictability declines with increasing group size because there is no population-level lateralization. Current work aims at studying more species of billfish, using a comparative analyses of morphological and behavioral differences to understand the evolutionary origin of the bill.

Funders