The wisdom of the crowd refers to the phenomenon in which a group of individuals, each making independent decisions, can collectively arrive at highly accurate solutions—often more accurate than any individual within the group. This principle relies heavily on independence: if individual opinions are unbiased and uncorrelated, their errors tend to cancel out when averaged, reducing overall bias. However, in real-world social networks, individuals are often influenced by their neighbors, introducing correlations between decisions. Such social influence can amplify biases, disrupting the benefits of independent voting. This trade-off between independence and interdependence has striking parallels to ensemble learning methods in machine learning. Bagging (bootstrap aggregating) improves classification performance by combining independently trained weak learners, reducing bias. Boosting, on the other hand, explicitly introduces sequential dependence among learners, where each learner focuses on correcting the errors of its predecessors. This process can reinforce biases present in the data even if it reduces variance. Here, we introduce a new meta-algorithm, casting, which captures this biological and computational trade-off. Casting forms partially connected groups (“castes”) of weak learners that are internally linked through boosting, while the castes themselves remain independent and are aggregated using bagging. This creates a continuum between full independence (i.e., bagging) and full dependence (i.e., boosting). This method allows for the testing of model capabilities across values of the hyperparameter which controls connectedness. We specifically investigate classification tasks, but the method can be used for regression tasks as well. Ultimately, casting can provide insights for how real systems contend with classification problems.

This model examines language dynamics within a social network using simulation techniques to represent the interplay of language adoption, social influence, economic incentives, and language policies. The agent-based model (ABM) focuses on interactions between agents endowed with specific linguistic attributes, who engage in communication based on predefined rules. A key feature of our model is the incorporation of network analysis, structuring agent relationships as a dynamic network and leveraging network metrics to capture the evolving inter-agent connections over time. This integrative approach provides nuanced insights into emergent behaviors and system dynamics, offering an analytical framework that extends beyond traditional modeling approaches. By combining agent-based modeling with network analysis, the model sheds light on the underlying mechanisms governing complex language systems and can be effectively paired with sociolinguistic observational data.

The model represents urban commuters’ transport mode choices among cars, public transit, and motorcycles—a mode highly prevalent in developing countries. Using an agent-based modeling approach, it simulates transport dynamics and serves as a testbed for evaluating policies aimed at improving mobility.

The model simulates an ecosystem of human agents who decide, at each time step, which mode of transportation to use for commuting to work. Their decision is based on a combination of personal satisfaction with their most recent journey—evaluated across a vector of individual needs—the information they crowdsource from their social network, and their personal uncertainty regarding trying new transport options.

Agents are assigned demographic attributes such as sex, age, and income level, and are distributed across city neighborhoods according to their socioeconomic status. To represent social influence in decision-making, agents are connected via a scale-free social network topology, where connections are more likely among agents within the same socioeconomic group, reflecting the tendency of individuals to form social ties with similar others.

…

This model demonstrates how different psychological mechanisms and network structures generate various patterns of cultural dynamics including cultural diversity, polarization, and majority dominance, as explored by Jung, Bramson, Crano, Page, and Miller (2021). It focuses particularly on the psychological mechanisms of indirect minority influence, a concept introduced by Serge Moscovici (1976, 1980)’s genetic model of social influence, and validates how such influence can lead to social change.

The “Descriptive Norm and Fraud Dynamics” model demonstrates how fraudulent behavior can either proliferate or be contained within non-hierarchical organizations, such as peer networks, through social influence taking the form of a descriptive norm. This model expands on the fraud triangle theory, which posits that an individual must concurrently possess a financial motive, perceive an opportunity, and hold a pro-fraud attitude to engage in fraudulent activities (red agent). In the absence of any of these elements, the individual will act honestly (green agent).

The model explores variations in a descriptive norm mechanism, ranging from local distorted knowledge to global perfect knowledge. In the case of local distorted knowledge, agents primarily rely on information from their first-degree colleagues. This knowledge is often distorted because agents are slow to update their empirical expectations, which are only partially revised after one-to-one interactions. On the other end of the spectrum, local perfect knowledge is achieved by incorporating a secondary source of information into the agents’ decision-making process. Here, accurate information provided by an observer is used to update empirical expectations.

The model shows that the same variation of the descriptive norm mechanism could lead to varying aggregate fraud levels across different fraud categories. Two empirically measured norm sensitivity distributions associated with different fraud categories can be selected into the model to see the different aggregate outcomes.

The Communication-Based Model of Perceived Descriptive Norm Dynamics in Digital Networks (COMM-PDND) is an agent-based model specifically created to examine the dynamics of perceived descriptive norms in the context of digital network structures. The model, developed as part of a master’s thesis titled “The Dynamics of Perceived Descriptive Norms in Digital Network Publics: An Agent-Based Simulation,” emphasizes the critical role of communication processes in norm formation. It focuses on the role of communicative interactions in shaping perceived descriptive norms.

The COMM-PDND is tuned to explore the effects of normative deviance in digital social networks. It provides functionalities for manipulating agents according to their network position, and has a versatile set of customizable parameters, making it adaptable to a wide range of research contexts.

The Social Identity Model of Protest Emergence (SIMPE), an agent-based model of national identity and protest mobilisations.

I developed this model for my PhD project, “Polarisation and Protest Mobilisation Around Secessionist Movements: an Agent-Based Model of Online and Offline Social Networks”, at the University of Glasgow (2019-2023).

The purpose of this model is to simulate protest emergence in a given country where there is an independence movement, fostering the self-categorisation process of national identification. In order to contextualised SIMPE, I have used Catalonia, where an ongoing secessionist movement since 2011 has been present, national identity has shown signs of polarisation, and where numerous mobilisations have taken place over the last decade. Data from the Catalan Centre of Opinion Studies (CEO) has been used to inform some of the model parameters.

…

Using data from the British Social Attitude Survey, we develop an agent-based model to study the effect of social influence on the spread of meat-eating behaviour in the British population.

This model aims to explore how gambling-like behavior can emerge in loot box spending within gaming communities. A loot box is a purchasable mystery box that randomly awards the player a series of in-game items. Since the contents of the box are largely up to chance, many players can fall into a compulsion loop of purchasing, as the fear of missing out and belief in the gambler’s fallacy allow one to rationalize repeated purchases, especially when one compares their own luck to others. To simulate this behavior, this model generates players in different network structures to observe how factors such as network connectivity, a player’s internal decision making strategy, or even common manipulations games use these days may influence a player’s transactions.

Substitution of food products will be key to realising widespread adoption of sustainable diets. We present an agent-based model of decision-making and influences on food choice, and apply it to historically observed trends of British whole and skimmed (including semi) milk consumption from 1974 to 2005. We aim to give a plausible representation of milk choice substitution, and test different mechanisms of choice consideration. Agents are consumers that perceive information regarding the two milk choices, and hold values that inform their position on the health and environmental impact of those choices. Habit, social influence and post-decision evaluation are modelled. Representative survey data on human values and long-running public concerns empirically inform the model. An experiment was run to compare two model variants by how they perform in reproducing these trends. This was measured by recording mean weekly milk consumption per person. The variants differed in how agents became disposed to consider alternative milk choices. One followed a threshold approach, the other was probability based. All other model aspects remained unchanged. An optimisation exercise via an evolutionary algorithm was used to calibrate the model variants independently to observed data. Following calibration, uncertainty and global variance-based temporal sensitivity analysis were conducted. Both model variants were able to reproduce the general pattern of historical milk consumption, however, the probability-based approach gave a closer fit to the observed data, but over a wider range of uncertainty. This responds to, and further highlights, the need for research that looks at, and compares, different models of human decision-making in agent-based and simulation models. This study is the first to present an agent-based modelling of food choice substitution in the context of British milk consumption. It can serve as a valuable pre-curser to the modelling of dietary shift and sustainable product substitution to plant-based alternatives in Britain.