Mathematical modelling of behaviour to inform policy for societal challenges: Talk abstracts
Invited speakers
Kavita Vedhara (Cardiff University)
Title: Behavioural Science: Warts and all
Abstract: An overview of some of the current challenges with behavioural data to provoke discussion on how behavioural science and mathematical modelling communities may work together more effectively.
Anne Kandler (Max Planck Institute for Evolutionary Anthropology)
Title: Inferring social learning processes from empirical data: Which details matter?
Abstract: Applying theoretical cultural evolution frameworks to real-world data to understand the underlying mechanism that - potentially - produced them is an interesting inferential challenge. Problems may arise due to characteristics of the available data such as its level of aggregation (e.g., population-level or individual-level data), its sparsity or its spatial/temporal resolution. However, problems may also arise due to misspecifications of the mathematical framework used to analyse the data, i.e., due to the omittance of crucial properties of the cultural system that can affect the observed data. Here, we mainly discuss the model misspecification problem. Focusing on the coarse distinction between unbiased and biased social learning, we use models of unbiased social learning coupled with realistic assumptions about demographic properties or the structure of the learning process to show that population-level frequency data generated by these models may not conform to theoretical expectations under unbiased social learning. In more detail, we show that using expectations blind to (i) demographic characteristics of the population (age structure, social structure or changes in population size) or (ii) details of the learning process (learning of packages of cultural variants vs. the learning of single variants) may generate wrong conclusion about the underlying learning process.
Suzy Moat and Tobias Preis (University of Warwick)
Title: Quantifying human behaviour with online data
Abstract: Our everyday usage of the internet leaves volumes of data in its wake. Can we use this data to help us reduce delays and costs in measuring human behaviour, or even to measure behaviour we couldn’t measure before? Here we will outline a number of studies carried out at the Data Science Lab at Warwick Business School, investigating whether online data can help us monitor disease levels or better understand the relationship between the beauty of outdoor environments and our wellbeing. We will discuss some of the challenges in generating estimates of human behaviour from online data when our relationship with the internet continues to evolve at such rapid pace.
Alice Milne (Rothamsted Research)
Title: Modelling land management
Abstract: European ash (Fraxinus excelsior), one of the most common trees in the UK has been seriously impacted by the fungal disease Ash Dieback (ADB). Additionally, ash trees are now facing a new threat from the potential invasion of the Emerald Ash Borer (Agrilus planipennis, EAB), a pest beetle that is ranked in the top three most dangerous invasive pests by EFSA. Preventing the introduction of such threats is often difficult, therefore, early detection and successful management are key areas where science can deliver. Scientific research tends to focus on single pest species, ignoring possible interactions of multiple pests/diseases. In addition, surveillance and management strategies also often neglect fail to account for the ability and willingness of managers to adopt advice on the control or management of pests and diseases. To address these challenges, we developed a model of the distribution and spread of ADB and the potential introduction and dispersal of EAB in a spatially explicit landscape. We integrated this with a novel model of stakeholder behaviour based on social science research. Stakeholder perspectives and decision-making was were modelled dynamically allowing them to realistically change over time in response to peer views and external influences. The linked models allowed us to identify the impacts of deploying surveillance and management resources under various scenarios to deliver successful strategies for the detection and management of ash treescapes. These results can be applied to other ecologically important systems where social and epidemiological interactions impact the delivery of appropriate management and control of pests and diseases.
David Haw (University of Liverpool)
Title: Human behaviour and transmission modelling: how to calibrate an new generation of integrated models
Abstract: We build on an integrated economic/epidemiological model by introducing a parametric description of behavioural change in response to quantities measurable in real time during an outbreak. A proof of concept will be given based on data from the first 18 months of COVID-19 mitigation in the UK.
Ruth McCabe (Imperial College London)
Title: Communicating uncertainty in modelling to non-technical audiences
Abstract: While mathematical models of epidemics are widely used to inform public health decision-making, the uncertainty in results are often poorly communicated, in particular to non-technical audiences. In this talk, I will share work from my DPhil thesis surrounding the communication of uncertainty to both decision-makers and the public and discuss how our community might improve the way in which we share our results.
Contributed talks
Elaine Ferguson (University of Glasgow)
Title: How does human behaviour impact dog rabies outbreaks?
Abstract: Canine rabies typically circulates at low levels in the dog populations of endemic low- and middle-income countries; a characteristic difficult to reconcile with the presence of large free-roaming dog populations, an absence of acquired immunity, and limited dog vaccination. We hypothesised that large outbreaks may be prevented by human populations becoming more sensitised to rabies when there have been recent cases in the local area; so that exposed or rabid dogs become more likely to be restrained before they can transmit.Using 18 years of contact tracing data from Serengeti District in Tanzania, we fitted a General Linear Mixed Model to determine the impact of recent local rabies cases on the probability that a rabid dog is restricted prior to biting. We then ran simulations of rabies transmission in Serengeti District using versions of a spatially-explicit individual-based model both with and without the fitted incidence-dependent dog restriction mechanism. The time series of rabies cases and the outbreak sizes generated with incidence-dependent restriction were much closer to those observed in the data, indicating a key role of human behaviour in maintaining low rabies incidence.
Mark Lynch (University of Warwick)
Title: Separating compartmental behaviour for non-selfish individuals in an epidemic
Abstract: How can I care about others in an epidemic? Population behaviour in an epidemic can be viewed as a “differential game”: Rational, well informed individuals seek to maximise their own utility function by modifying their behaviour. This behaviour affects the course of the epidemic in an SIR compartmental model. Typically, this analysis involves a single behavioural class, or multiple behavioural classes depending on fixed attributes such as occupation or age. Less studied are models in which individuals change their behaviour depending on which compartment they are in. Given a utility, one can solve the related game theoretic problem to derive Nash equilibrium system dynamics. When the costs in the utility only pertain to the individual, infected individuals never socially distance because modifying their behaviour cannot improve their situation. We study the case where individuals care about other members of the population. Their behaviour will then change in order to protect others from incurring the cost of being infected and/or of having to socially distance. We quantify the degree to which individuals must care about the population in order to rationally target disease eradication through social distancing.
Luisa Fernanda Estrada Plata (University of Warwick)
Title: Learning a Social Network by Influencing Opinions
Abstract: We study a campaigner who wants to learn the structure of a social network by observing the underlying diffusion process and intervening on it. Using synchronous majoritarian updates on binary opinions as the underlying dynamics, we offer upper bounds on the campaigner's budget for learning any network with certainty, considering both observation and intervention resources, and further improving them for the case of clique networks. Additionally, we investigate the learning progress of the campaigner when her budget falls below these upper bounds. For such cases, we design a greedy campaigning strategy aimed at optimising the campaigner's information gain at each opinion diffusion step.
Matt Ryan (Commonwealth Scientific and Industrial Research Organisation - CSIRO)
Title: BaD transmission modelling: Incorporating human behaviour into a simple transmission model
Abstract: The interactions between human behaviour and the spread of infectious diseases creates complex feedback loops between behaviour and infection transmission. That is, where human behaviour affects human behaviour, human behaviour affects infection spread, and infection spread affects human behaviour. Despite this, many transmission models either ignore the intricacies of human behaviour or consider it as a static constant. Here, I introduce our behaviour and disease (BaD) modelling approach for incorporating human behaviour models from behavioural science into standard transmission models. Specifically, I will discuss how to incorporate the socio-psychological Health Belief Model into the standard susceptible-infectious-recovered-susceptible (SIRS) transmission model for a visible protective behaviour. I will also demonstrate how BaD modelling can be used to investigate the effects of non-pharmaceutical intervention strategies that target specific components of the Health Belief Model.
Fanqi Zeng (University of Oxford)
Title: Is the mafia the new Sheriff of Nottingham? Social order and crime in an English city
Abstract: Can a mafia ensure social order and help reduce ordinary crime? This work investigates urban crime patterns over time by conducting in-depth interviews with local officials in the city of Nottingham. It also explores a novel dataset of the public’s phone calls to the police (2012-2019), encompassing spatial-temporal information and police-labelled crime types. We find that Nottingham’s ward of Bestwood is the site of an entrenched, mafia-type organised crime group, while the most similar ward of Bulwell is free from such type of organised crime. A matched comparison between the two wards and an evaluation of crime distributions across the two sites, focusing on eight major crimes over this period, reveal that Bestwood’s mafia exercises neighbourhood governance over volume crime, which lowers certain crimes compared with Bulwell. Ultimately, the study shows that governance-type organised crime groups also exist in the Global North in addition to the well-known area like Italy.