Start Date
10-12-2017 12:00 AM
Description
The Information Systems discipline builds upon a well-established set of theories and models. A frequently used methodological procedure is to build a model based on a theory and to apply a multivariate method such as Structural Equation Modeling to test the fit of the model and the significance of its parameters. This approach does not take into account the important role of temporal dynamics. Taken by itself, the Technology Acceptance Model cannot explain the Diffusion of Innovations model, and vice versa, in spite of the fact that both deal with adoption patterns. In this paper we call for a methodological paradigm shift and show how Agent Based Modeling in combination with Structural Equation Modeling can help to create models which combine the strengths of dynamic modeling approaches with rigorous multivariate statistics. This will lead to more meaningful models that incorporate temporal dynamics and enable new ways to test validity.
Recommended Citation
Treiblmaier, Horst, "The World Isn’t Static, So Why Are We? How Agent Based Modeling Helps to Create and Test Dynamic IS Models and Theories" (2017). ICIS 2017 Proceedings. 2.
https://aisel.aisnet.org/icis2017/ResearchMethods/Presentations/2
The World Isn’t Static, So Why Are We? How Agent Based Modeling Helps to Create and Test Dynamic IS Models and Theories
The Information Systems discipline builds upon a well-established set of theories and models. A frequently used methodological procedure is to build a model based on a theory and to apply a multivariate method such as Structural Equation Modeling to test the fit of the model and the significance of its parameters. This approach does not take into account the important role of temporal dynamics. Taken by itself, the Technology Acceptance Model cannot explain the Diffusion of Innovations model, and vice versa, in spite of the fact that both deal with adoption patterns. In this paper we call for a methodological paradigm shift and show how Agent Based Modeling in combination with Structural Equation Modeling can help to create models which combine the strengths of dynamic modeling approaches with rigorous multivariate statistics. This will lead to more meaningful models that incorporate temporal dynamics and enable new ways to test validity.