A Common Protocol for Agent-Based Social Simulation

Journal of Artificial Societies and Social Simulation, The (Impact Factor: 1.16). 01/2006; 9(1).
Source: RePEc


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.

Download full-text


Available from: Roberto Leombruni
    • "A way to break up this black box of detailed models is to test for the effects of single model components on model performance. This includes the consideration of the match between simulation output and data, i.e., the validation, which ideally should be performed with independent data on patterns at different hierarchical levels (Grimm et al., 2005; Richiardi et al., 2006; Raux et al., 2014; Putra et al., 2015). In other domains of environmental modelling, testing the model performance is state of the art, and many methods have been developed to analyse different aspects of model performance (see the review by Bennett et al., 2013). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Agent-based models (ABMs) have tended to become more complicated in recent years, going along with challenges of fully understanding model behaviour. While partly answered using simple and abstract models, the question of the role of different aspects of model detail for controlling model outcomes still has not been explored with empirical ABMs. We therefore use a detailed model built upon an empirical survey of residential mobility to explore the effects of three domains of model detail - agent heterogeneity, structural model detail and detail in the data used - on model performance. Evaluation is done by measuring the ability to predict empirical patterns for different topics, population, relocations and vacancies at different scales. Our diverse results indicate that a good data basis is crucial, heterogeneity strongly controls small-scale patterns and different design aspects must be tested thoroughly. This outcome calls for caution when interpreting the results of such models.
    No preview · Article · Jan 2016 · Environmental Modelling and Software
    • "This paper shows the potential of systematic DOE by Lorscheid et al. (2012) as a proposed standard that focuses specifically on the analysis of simulation models and results. Other approaches, such as that of Richiardi et al. (2006) and Manuj et al. (2009), present standards that encompass the complete simulation research process. The overview, design and details (ODD) 2 protocol by Grimm et al. (2006) focuses on 2 Grimm et al. (2010) and Müller et al. (2013) further developed the ODD protocol. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Simulation offers management accounting research many benefits, such as the ability to model and to experiment with complex and large systems. At the same time, the acceptance of this method is hampered by a feeling of complexity often associated with simulation models and their behavior, as well as with challenges in communicating the models’ results. This study shows how these challenges can be addressed via the systematic use of design of experiment (DOE) principles. The DOE process framework is applied to a simulation model of a cost accounting system that is used to quantitatively evaluate two different methods for the allocation of service costs. As a result, we not only demonstrate the potential and benefits of simulation in the field of management accounting, but also show how DOE principles can help to improve understandings of simulation model behavior and the communication of simulation results.
    No preview · Article · Jul 2015
  • Source
    • "Sensitivity analysis is considered as good modeling practice when performed as part of model verification and has been widely used to assess quantitative models in many studies [19]. This paper is organized as follows. "
    [Show abstract] [Hide abstract]
    ABSTRACT: As one of the most important urban commuter transportation modes, urban rail transit (URT) has been acting as a key solution for supporting mobility needs in high-density urban areas. However, in recent years, high frequency of unexpected events has caused serious service disruptions in URT system, greatly harming passenger safety and resulting in severe traffic delays. Therefore, there is an urgent need to study emergency evacuation problem in URT. In this paper, a method of bus dynamic coscheduling is proposed and two models are built based on different evacuation destinations including URT stations and surrounding bus parking spots. A dynamic coscheduling scheme for buses can be obtained by the models. In the model solution process, a new concept-the equivalent parking spot-is proposed to transform the nonlinear model into an integer linear programming (ILP) problem. A case study is conducted to verify the feasibility of models. Also, sensitivity analysis of two vital factors is carried out to analyze their effects on the total evacuation time. The results reveal that the designed capacity of buses has a negative influence on the total evacuation time, while an increase in the number of passengers has a positive effect. Finally, some significant optimizing strategies are proposed.
    Full-text · Article · Nov 2014 · Computational Intelligence and Neuroscience
Show more