Contributions to model validation: hierarchy, process, and cessation

Stefan N. Groesser, Markus Schwaninger

System Dynamics Review · 2012 · 104 citations · 47 references

Concepts

TL;DR

Dynamic modeling and simulation face a persistent challenge of ensuring model validity, and while many studies have explored validation tests, terminology, and epistemology, they leave novice modelers uncertain about which tests to use, when to apply them, and when to stop. This paper aims to address these gaps by presenting a hierarchy of validation tests, an integrated validation process, and a heuristic for determining when to cease formal validation. The authors develop a complexity-based hierarchy of validation tests, combine them into a stepwise validation process, and formulate a decision heuristic that guides modelers on when to terminate validation activities. The study concludes by outlining future research directions to refine and extend the proposed validation framework. © 2012 System Dynamics Society.

Abstract

Abstract In the domain of dynamic modeling and simulation, the assurance of model validity is a prominent challenge. An extensive number of contributions concerning model tests, terminology, and the epistemological foundations of validation have been elaborated. These contributions, however, do not fully answer the questions for novice modelers, namely, which validation tests to choose, when and how to apply them, and at what point to cease their formal validation efforts. Our intention here is to help close this gap by introducing a complexity hierarchy of validation tests, an integrative validation process , and a decision heuristic about when to stop formal validation efforts. The paper concludes by providing directions for future research. Copyright © 2012 System Dynamics Society

References

47