A-Z of Digital Research Methods

Catherine Dawson

• Publisher: Routledge
• Year: 2019
• ISBN:
  9781351044653
• Edition: 1
• Topic: Biowissenschaften
• Subtopic: Wissenschaftliche Forschung & Methodik

This accessible, alphabetical guide provides concise insights into a variety of digital research methods, incorporating introductory knowledge with practical application and further research implications. A-Z of Digital Research Methods provides a pathway through the often-confusing digital research landscape, while also addressing theoretical, ethical and legal issues that may accompany each methodology.

Dawson outlines 60 chapters on a wide range of qualitative and quantitative digital research methods, including textual, numerical, geographical and audio-visual methods. This book includes reflection questions, useful resources and key texts to encourage readers to fully engage with the methods and build a competent understanding of the benefits, disadvantages and appropriate usages of each method.

A-Z of Digital Research Methods is the perfect introduction for any student or researcher interested in digital research methods for social and computer sciences.

CHAPTER 1

Agent-based modelling and simulation

Overview

Agent-based modelling and simulation (ABMS) is a method that enables researchers to create a computer model and simulation of active entities and their behaviour and interaction with each other and their environment. These interacting, autonomous and adaptive agents can be individuals, households, groups, organisations, vehicles, equipment, products or cells, in social and evolutionary settings, for example. ABMS is a specific type of computer model and simulation: other types and a general discussion of modelling and simulation can be found in Chapter 7. There are different terms that are used to describe the same, or similar techniques, and these include agent-based computational modelling, agent-based modelling (ABM), agent-based simulation modelling, agent-based social simulation (ABSS) and agent-based simulation. It is important to note that models provide representations whereas simulations use models (or simulate the outcomes of models) for study and analysis: ABMS is a term that covers both.

ABMS enables researchers to build models of systems from the bottom up (micro to macro), with the aim of producing simulations in which patterns, structures and behaviours emerge from agent interaction. Models and simulations can be exploratory, descriptive and predictive: they can be used to provide insight into behaviour and decision-making, make predictions about future behaviour and trends or help to analyse, validate or explain data collected from other sources. Models and simulations can be relational, dynamic, responsive and adaptive. For example, agents:

• respond to the actions of others;
• respond to environmental stimuli;
• influence each other;
• learn from each other;
• learn from their experiences;
• adapt their behaviour as a result of other agents’ behaviour;
• adapt their behaviour to suit their environment.

Researchers from a variety of disciplines and areas of study use ABMS including sociology and social psychology (Chattoe-Brown, 2014; Conte and Paolucci, 2014; Eberlen et al., 2017), geography (Millington and Wainwright, 2017), health and medicine (Auchincloss and Garcia, 2015), economics (Caiani et al., 2016), politics (Fieldhouse et al., 2016), the sports sciences (Lauren et al., 2013); the environmental sciences (Kerridge et al., 2001; Sun and Taplin, 2018) and the computer sciences (Abar et al., 2017). Examples of research projects that have used ABMS include a study that adapts principles of developmental biology and agent-based modelling for automated urban residential layout design (Sun and Taplin, 2018); research into pedestrian flow and movement (Kerridge et al., 2001); research into the interaction between the development of creative industries and urban spatial structure (Liu and Silva, 2018); research that helps to predict rates of burglary (Malleson et al., 2009); a study to model, simulate and test tactics in the sport of rugby union (Lauren et al., 2013); and research into voter turnout (Fieldhouse et al., 2016).

If you are interested in finding out more about ABMS, and using it for your research, a good reference to begin with is Silverman et al. (2018), which is an open access book that provides in-depth coverage of methodological issues and complexities associated with ABM and the social sciences. A useful reference for those working within, or studying, economics is Hamill and Gilbert (2016), which provides a practical introduction and history to ABM methods and techniques. Another is Caiani et al. (2016), which provides a practical guide and basic toolkit that highlights practical steps in model building and is aimed at undergraduates, postgraduates and lecturers in economics. A useful reference for those working within geography (and who are interested in mixed methods approaches) is Millington and Wainwright (2017: 68) who discuss ‘mixed qualitative-simulation methods that iterate back-and-forth between “thick” (qualitative) and “thin” (simulation) approaches and between the theory and data they produce’. Auchincloss and Garcia (2015) provide a brief introduction to carrying out a simple agent-based model in the field of urban health research. Chapter 7 provides a useful overview of computer modelling and simulation and contains additional references and relevant questions for reflection. If you are interested in predictive modelling, more information can be found in Chapter 45.

Questions for reflection

Epistemology, theoretical perspective and methodology

• Miller (2015: 175) proposes critical realism as a philosophical perspective to understand, orient and clarify the nature and purpose of agent-based modelling research. Does this perspective have resonance with your research and, if so, in what way? How might this perspective help you to evaluate, validate and assess models?
• Do you intend to use agent-based modelling as a standalone research method, or do you intend to adopt a mixed methods approach? Is it possible to integrate diverse forms of data (and interdisciplinary data) with agent-based modelling? Chattoe-Brown (2014) believes so, illustrating why and how from a sociological perspective, and Millington and Wainwright (2017) discuss mixed method approaches from a geographical perspective.
• How might ABMS be used to complement and improve traditional research practices? Eberlen et al. (2017) will help you to reflect on this question in relation to social psychology.
• Can phenomena emerging from agent-based models be explained entirely by individual behaviour? Silverman et al. (2018) provide a comprehensive discussion on this and other methodological considerations.
• Do models represent the real world, or are they a researcher’s interpretation of the real world?
• What are the strengths and weaknesses of ABMS? Conte and Paolucci (2014) will help you to address this question in relation to computational social science and Eberlen et al. (2017) discuss these issues in relation to social psychology.

Ethics, morals and legal issues

• Is it possible that modelling can be to the detriment of individuals? Can model outcomes lead to unethical or inappropriate action that can cause harm to individuals? Can individuals be singled out for action, based on models? What happens when predictions are based on past behaviour that may have changed? Can individuals correct model inputs?
• Have data been volunteered specifically for modelling purposes?
• Is it possible that individuals could be identifiable from models?
• Millington and Wainwright (2017: 83) ask a pertinent question that needs to be considered if you intend to use ABMS: ‘how might new-found understandings by individuals about their agency be turned back to geographers to understand the role of agent-based simulation modelling itself as an agent of social change?’

Practicalities

• How will you go about building your model? Jackson et al. (2017: 391–93) provide a seven-step guide to creating your own model:
  • Step 1: what are your world’s dimensions?
  • Step 2: how do agents meet?
  • Step 3: how do agents behave?
  • Step 4: what is the payoff?
  • Step 5: how do agents change?
  • Step 6: how long does your world last?
  • Step 7: what do you want to learn from your world?

• Do you know which is the most appropriate agent-based modelling and simulation toolkit for your research? How do you intend to choose software and tools? A concise characterisation of 85 agent-based toolkits is provided by Abar et al. (2017).
• How accurate is your model? How important is accuracy (when action is to be taken, or decisions made, based on your model outcomes, for example?)
• How do you intend to verify and validate your model (ensuring the model works correctly and ensuring the right model has been built, for example)?

Useful resources

There are a wide variety of agent-based modelling and simulation software and digital tools available. A few examples available at time of writing are given below (in alphabetical order).

• Adaptive Modeler (www.altreva.com);
• AnyLogic (www.anylogic.com);
• Ascape (http://ascape.sourceforge.net);
• Behaviour Composer (http://m.modelling4all.org);
• Cougaar (www.cougaarsoftware.com);
• GAMA (https://gama-platform.github.io);
• JADE (http://jade.tilab.com);
• NetLogo (https://ccl.northwestern.edu/netlogo);
• OpenStarLogo (http://web.mit.edu/mitstep/openstarlogo/index.html);
• Repast Suite (https://repast.github.io);
• StarLogo TNG (https://education.mit.edu/portfolio_page/starlogo-tng);
• Swarm (www.swarm.org/wiki/Swarm_main_page).

The Society for Modeling and Simulation International (http://scs.org) and the European Council for Modelling and Simulation (www.scs-europe.net) provide details of conferences, workshops, publications and resources for those interested in computer modelling and simulation (see Chapter 7 for more information about these organisations and for additional tools and software that can be used for computer modelling and simulation).

Key texts

Abar, S., Theodoropoulos, G., Lemarinier, P. and O’Hare, G. (2017) ‘Agent Based Modelling and Simulation Tools: A Review of the State-of-Art Software’, Computer Science Review, 24, 13–33, May 2017, 10.1016/j.cosrev.2017.03.001.

Auchincloss, A. and Garcia, L. (2015) ‘Brief Introductory Guide to Agent-Based Modeling and an Illustration from Urban Health Research’, Cadernos De Saude Publica, 31(1), 65–78, 10.1590/0102-311X00051615.

Caiani, A., Russo, A., Palestrini, A. and Gallegati, M. (eds.) (2016) Economics with Heterogeneous Interacting Agents: A Practical Guide to Agent-Based Modeling. Cham: Springer.

Chattoe-Brown, E. (2013) ‘Why Sociology Should Use Agent Based Modelling’, Sociological Research Online, 18(3), 1–11, first published August 31, 2013, 10.5153/sro.3055.

Chattoe-Brown, E. (2014) ‘Using Agent Based Modelling to Integrate Data on Attitude Change’, Sociological Research Online, 19(1), 1–16, first published March 5, 2014, 10.5153/sro.3315.

Conte, R. and Paolucci, M. (2014) ‘On Agent-Based Modeling and Computational Social Science’, Frontiers in Psychology, 5(668), first published July 14, 2014, 10.3389/fpsyg.2014.00668.

Eberlen, J., Scholz, G. and Gagliolo, M. (2017) ‘Simulate This! An Introduction to Agent-Based Models and Their Power to Improve Your Research Practice’, International Review of Social Psychology, 30(1), 149–160. 10.5334/irsp.115.

Fieldhouse, E., Lessard-Phillips, L. and Edmonds, B. (2016) ‘Cascade or Echo Chamber? A Complex Agent-Based Simulation of Voter Turnout’, Party Politics, 22(2), 241–256, first published October 4, 2015, .

Hamill, L. and Gilbert, N. (2016) Agent-Based Modelling in Economics. Chichester: John Wiley & Sons Ltd.

Jackson, J., Rand, D., Lewis, K., Norton, M. and Gray, K. (2017) ‘Agent-Based Modeling: A Guide for Social Psychologists’, Social Psychological and Personality Science, 8(4), 387–395, first published March 13, 2017, 10.1177/1948550617691100.

Kerridge, J., Hine, J. and Wigan, M. (2001) ‘Agent-Based Modelling of Pedestrian Movements: The Questions that Need to Be Asked and Answered’, Environment and Planning B: Urban Analytics and City Science, 28(3), 327–341, first published June 1, 2001, 10.1068/b2696.

Lauren, M., Quarrie, K. and Galligan, D. (2013) ‘Insights from the Application of an Agent-Based Computer Simulation as a Coaching Tool for Top-Level Rugby Union’, International Journal of Sports Science & Coaching...