A Common Protocol for Agent-Based Social Simulation

Journal of Artificial Societies and Social Simulation, The (Impact Factor: 1.16). 01/2006; 9.
Source: RePEc

ABSTRACT Traditional (i.e. analytical) modelling practices in the social sciences rely on a very well established, although implicit, methodological protocol, both with respect to the way models are presented and to the kinds of analysis that are performed. Unfortunately, computer-simulated models often lack such a reference to an accepted methodological standard. This is one of the main reasons for the scepticism among mainstream social scientists that results in low acceptance of papers with agent-based methodology in the top journals. We identify some methodological pitfalls that, according to us, are common in papers employing agent-based simulations, and propose appropriate solutions. We discuss each issue with reference to a general characterization of dynamic micro models, which encompasses both analytical and simulation models. In the way, we also clarify some confusing terminology. We then propose a three-stage process that could lead to the establishment of methodological standards in social and economic simulations.

1 Bookmark
  • [Show abstract] [Hide abstract]
    ABSTRACT: Artificial markets (AMs) are an emerging form of agent-based simulation (ABS), in which agents represent individual consumers, firms, or industries interacting under simulated market conditions. The validity of the method depends on the ability of researchers to construct simulated agents that faithfully capture the key behavior of targeted entities. Without such a correspondence the simulation cannot be considered to be a valid representation of market dynamics. To date, no such correspondence has been established. Yet, for artificial markets to achieve their potential as a tool for marketing practice it is crucial that closer ties be forged with mainstream methods for consumer behavioral research, especially qualitative methods. The primary contribution of this article is a novel method combining qualitative marketing research (inductive case studies, grounded theory, and sequence analysis) and software engineering techniques to synthesize simulation-ready theories of consumer behavior. We provide a step-by-step explanation and a demonstrative example of theory-building from the consumer technology adoption domain. The outcome is a theory of consumer adoption behavior that is sufficiently precise and formal to be expressed in Unified Modeling Language (UML). The article concludes with a discussion of the limitations of the method, recommendations for its implementation in the study of diffusion of innovation (DOI) and suggestions for further research. The arguments and findings in this article that pertain to artificial markets can be generalized with respect to most agent-based simulations, including those applied to the study of diffusion of innovation. The results of an ABS of innovation diffusion cannot be relied upon unless the agents are based on a theory of adoption that grounded in empirical observations of the targeted entities - regardless of whether those entities are consumers, firms or industries. Qualitative research of adoption behavior is thus a us- - eful precursor to successful agent-based approaches to studying the diffusion of innovations.
    Technology Management for Global Economic Growth (PICMET), 2010 Proceedings of PICMET '10:; 01/2010
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Recently it has been suggested that low fertility countries may be caught in a trap that is hard to get out of. One important mechanism in such a trap would be social interaction and its effect on the ideal family size. Such social interaction mechanisms are hard to capture in formal models, therefore we use an agent-based simulation model to investigate the issue. In our experimental setup a stable growth and population path is calibrated to Swedish data using the Swedish social policy setup. The model is provoked into a fertility trap by increasing relative child costs linked to positive growth. Even rather large increases in child benefits are then insufficient to get out of the trap. However, the small number of children temporarily enables the economy to grow faster for several decades. Removing the adaptation of social norms turns out to disarm the trap. II a été suggéré récemment que les pays à basse fécondité pourraient être victimes d'un piège dont ils auraient du mal à se dégager. Un mécanisme essentiel dans ce piège serait l'interaction sociale et son effet sur la taille idéale de famille. Des mécanismes de ce type sont difficiles à représenter dans un modèle formel, et c'est pourquoi nous avons eu recours à un modèle de simulation multi-agents pour explorer le processus. Dans notre dispositif expérimental, un modèle de croissance et de population stable est calibré aux données suédoises, en utilisant la configuration suédoise de politique sociale. Le modèle est entraîné dans un piège de fécondité en élevant les coûts relatifs de l'enfant en lien avec la croissance positive. Dans ce cas, même des augmentations importantes des prestations familiales sont insuffisantes pour sortir du piège. Toutefois, le petit nombre d'enfants permet temporairement à l'économie de croître plus rapidement pendant quelques décennies. L'arrêt de l'adaptation aux normes sociales conduit à une neutralisation du piège.
    European Journal of Population 01/2010; 26(2):183-205. · 1.75 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Effective management of natural resources requires understanding both the dynamics of the natural systems being subjected to management and the decision-making behaviour of stakeholders who are involved in the management process. We suggest that simulation modelling techniques can provide a powerful method platform for the transdisciplinary integration of ecological, economic and sociological aspects that is needed for exploring the likely outcomes of different management approaches and options. A concise review of existing literature on ecological and socio-economic modelling and approaches at the interface of these fields is presented followed by a framework coupling an individual-based ecological model with an agent-based socio-economic model. In this framework, each individual of the species of interest is represented on a spatially-explicit landscape, allowing the incorporation of individual variability. The socio-economic model also simulates inter-agent variability through the assignment of different attitudes and decision-making options for different agents; these may represent farmers, estate managers, policy-makers, the general public and/or other stakeholders. This structure enables variation in attitudes and circumstances of individual stakeholders, together with interactions between stakeholders, to be simulated. We discuss strengths and limitations of such an approach, and the information requirements for building a robust model to inform a real management situation.
    iEMSs 2012 - Managing Resources of a Limited Planet; 07/2012

Full-text (3 Sources)

Available from
May 29, 2014