Building a Knowledge Base for the Model

In this chapter, after summarising the key conceptual challenges related to the measurement of asylum migration, we briefly outline the history of recent migration flows from Syria to Europe. This case study is intended to guide the development of a model of migration route formation, used throughout this book as an illustration of the proposed model-based research process. Subsequently, for the case study, we offer an overview of the available data types, making a distinction between the sources related to the migration processes, as well as to the context within which migration occurs. We then propose a framework for assessing different aspects of data, based on a review of similar approaches suggested in the literature, and this framework is subsequently applied to a selection of available data sources. The chapter concludes with specific recommendations for using the different forms of data in formal modelling, including in the uncertainty assessment.

Introduction

Population processes, including migration, are complex and uncertain. We begin this book by providing a rationale for building Bayesian agent-based models for population phenomena, specifically in the context of migration, which is one of the most uncertain and complex demographic processes. The main objectives of the book are to pursue methodological advancement in demography and migration studies through combining agent-based modelling with empirical data, Bayesian statistical inference, appropriate computational techniques, and psychological experiments in a streamlined modelling process, with the overarching aim to contribute to furthering the model-based research agenda in demography and broader social sciences. In this introductory chapter, we also offer an overview of the structure of this book, and present various ways in which different audiences can approach the contents, depending on their background and needs.

Uncertainty and Complexity: Towards Model-Based Demography

This chapter focuses on the broad methodological and philosophical underpinnings of the Bayesian model-based approach to studying migration. Starting from reflections on the uncertainty and complexity in demography and, in particular, migration studies, the focus moves to the shifting role of formal modelling, from merely describing, to predicting and explaining population processes. Of particular importance are the gaps in understanding asylum migration flows, which are some of the least predictable while at the same time most consequential forms of human mobility. The well-recognised theoretical void of demography as a discipline does not help, especially given the lack of empirical micro-foundations in formal modelling. Here, we analyse possible solutions to theoretical shortcomings of demography and migration studies from the point of view of the philosophy of science, looking at the inductive, deductive and abductive approaches to scientific reasoning. In that spirit, the final section introduces and extends a research programme of model-based demography.

Towards More Realistic Models

This chapter is devoted to the presentation of a more realistic version of the model, Risk and Rumours, which extends the previous, theoretical version (Routes and Rumours) by including additional empirical and experimental information following the process described in Part II of this book. We begin by offering a reflection on the integration of the five elements of the modelling process, followed by a more detailed description of the Risk and Rumours model, and how it differs from the previous version. Subsequently, we present selected results of the uncertainty and sensitivity analysis, enabling us to make further inference on the information gaps and areas for potential data collection. We also present model calibration for an empirically grounded version of the model, Risk and Rumours with Reality. In that way, we can evaluate to what extent the iterative modelling process has enabled a reduction in the uncertainty of the migrant route formation. In the final part of the chapter, we reflect on the model-building process and its implementation.

Agent-Based Modelling and Simulation with Domain-Specific Languages

Conducting simulation studies within a model-based framework is a complex process, in which many different concerns must be considered. Central tasks include the specification of the simulation model, the execution of simulation runs, the conduction of systematic simulation experiments, and the management and documentation of the model’s context. In this chapter, we look into how these concerns can be separated and handled by applying domain-specific languages (DSLs), that is, languages that are tailored to specific tasks in a specific application domain. We demonstrate and discuss the features of the approach by using the modelling language ML3, the experiment specification language SESSL, and PROV, a graph-based standard to describe the provenance information underlying the multi-stage process of model development.

Uncertainty Quantification, Model Calibration and Sensitivity

Better understanding of the behaviour of agent-based models, aimed at embedding them in the broader, model-based line of scientific enquiry, requires a comprehensive framework for analysing their results. Seeing models as tools for experimenting in silico, this chapter discusses the basic tenets and techniques of uncertainty quantification and experimental design, both of which can help shed light on the workings of complex systems embedded in computational models. In particular, we look at: relationships between model inputs and outputs, various types of experimental design, methods of analysis of simulation results, assessment of model uncertainty and sensitivity, which helps identify the parts of the model that matter in the experiments, as well as statistical tools for calibrating models to the available data. We focus on the role of emulators, or meta-models – high-level statistical models approximating the behaviour of the agent-based models under study – and in particular, on Gaussian processes (GPs). The theoretical discussion is illustrated by applications to the Routes and Rumours model of migrant route formation introduced before.

The Boundaries of Cognition and Decision Making

This chapter outlines the role that individual-level empirical evidence gathered from psychological experiments and surveys can play in informing agent-based models, and the model-based approach more broadly. To begin with, we provide an overview of the way that this empirical evidence can be used to inform agent-based models. Additionally, we provide three detailed exemplars that outline the development and implementation of experiments conducted to inform an agent-based model of asylum migration, as well as how such data can be used. There is also an extended discussion of important considerations and potential limitations when conducting laboratory or online experiments and surveys, followed by a brief introduction to exciting new developments in experimental methodology, such as gamification and virtual reality, that have the potential to address some of these limitations and open the door to promising and potentially very fruitful new avenues of research.

Principles and State of the Art of Agent-Based Migration Modelling

Migration as an individual behaviour as well as a macro-level phenomenon happens as part of hugely complex social systems. Understanding migration and its consequences therefore necessitates adopting a careful analytical approach using appropriate tools, such as agent-based models. Still, any model can only be specific to the question it attempts to answer. This chapter provides a general discussion of the key tenets related to modelling complex systems, followed by a review of the current state of the art in the simulation modelling of migration. The subsequent focus of the discussion on the key principles for modelling migration processes, and the context in which they occur, allows for identifying the main knowledge gaps in the existing approaches and for providing practical advice for modellers. In this chapter, we also introduce a model of migration route formation, which is subsequently used as a running example throughout this book.

Open Science, Replicability, and Transparency in Modelling

Recent years have seen large changes to research practices within psychology and a variety of other empirical fields in response to the discovery (or rediscovery) of the pervasiveness and potential impact of questionable research practices, coupled with well-publicised failures to replicate published findings. In response to this, and as part of a broader open science movement, a variety of changes to research practice have started to be implemented, such as publicly sharing data, analysis code, and study materials, as well as the preregistration of research questions, study designs, and analysis plans. This chapter outlines the relevance and applicability of these issues to computational modelling, highlighting the importance of good research practices for modelling endeavours, as well as the potential of provenance modelling standards, such as PROV, to help discover and minimise the extent to which modelling is impacted by unreliable research findings from other disciplines.

Bayesian Model-Based Approach: Impact on Science and Policy

In this chapter, we summarise the scientific and policy implications of the Bayesian model-based approach, starting from an evaluation of its possible advantages, limitations, and potential to influence further scientific developments, policy and practice. We focus here specifically on the role of limits of knowledge and reducible (epistemic), as well as irreducible (aleatory) uncertainty. To that end, we also reflect on the scientific risk-benefit trade-offs of applying the proposed approaches. We discuss the usefulness of proposed methods for policy, exploring a variety of uses, from scenario analysis, to foresight studies, stress testing and early warnings, as well as contingency planning, illustrated with examples generated by the Risk and Rumours models presented earlier in this book. We conclude the chapter by providing several practical recommendations for the potential users of our approach, including a blueprint for producing and assessing the impact of policy interventions in various parts of the social system being modelled.