Lorig F. Hypothesis-Driven Simulation Studies: Assistance For The Systematic Design And Conducting Of Computer Simulation Experiments

New York: Springer, 2019. — 407 p.Fabian Lorig develops a procedure model for hypothesis-driven simulation studies which supports the design, conducting, and analysis of simulation experiments. It is aimed at facilitating the execution of simulation studies with regard to the replicability and reproducibility of the results. In comparison to existing models, this approach is based on a formally specified hypothesis. Each step of the simulation study can be adapted to the central hypothesis and performed in such a way that it can optimally contribute to the verification and thus to the confirmation or rejection of the hypothesis.Preface and AcknowledgementsList of Figures
List of Tables
Abstract
Introduction and BackgroundProblem Statement
Research Questions and Objectives
Contribution
Outline
Foundations and Methods of Simulation
History of Simulation
Fundamentals of Simulation
Simulation Model
Simulation Experiment
Application of Simulation
Areas of Application
Appropriateness and Advantages of Simulation
Procedure Models for Simulation Studies
Assistance and Automation of Simulation
Description Languages and Theoretical Frameworks
Interchange Formats and Specification Languages
Guidelines
Systematizations
Toolkits and Software Frameworks
Domain-Specific Toolkits and Frameworks
Multi-Purpose Toolkits and Frameworks
Methodological Shortcomings and Research Gap
Hypothesis-Driven Simulation Studies
Requirements Analysis on Hypotheses in Simulation
Scientific Hypotheses
Formalization of Scientific Hypotheses
Epistemological Demands of Simulation
Requirements on Scientific Hypotheses in Simulation
Hypotheses in Simulation Studies
Methodological Integration of Research Hypotheses
Structural Components of Simulation Studies
Aggregation and Interpretation of Results
Methodological Shortcomings
Conclusions
Hypothesis-Driven Simulation Studies
Integrated Procedure Model
Introduction of the Study’s Scenario
Specification of Hypotheses
Design and Structural Decomposition of Experiments
Aggregation and Analysis of Outputs
Implications for the Assistance of Simulation Studies
Logical Connection of Services
Abstract Architecture of an Assistance System
Conclusions
Services for the Assistance of Simulation Studies
Simulation Model
Model Metadata
Input Variables and Parameters
Output Variables and Performance Measures
Specification of the Simulation Model
Formal Specification of Hypotheses
Documentation of the Solution Process
Design of Experiments
Factor Screening
Experimental Design
Replication Estimation
Execution of Experiments
Simulation Framework
Scaling and Parallel Execution of Simulation Runs
Random Numbers
Outliers and Missing Values
Aggregation of Results
Hypothesis Testing
Conclusions
Application and Evaluation
Introduction of Simulation Model and Definition of Scenario
NetLogo Simulation Framework
Supply Chain Simulation Model
Scenario and Research Hypothesis
Scenario : Maximum Required Storage Capacity
Conventional Investigation of the Model’s Behavior
Design of Experiments
Conducting of Experiments
Analysis of Experiments
Scenario : Customer Demand and Storage Capacity
Conventional Investigation of the Model’s Behavior
Design of Experiments
Conducting of Experiments
Analysis of Experiments
Conclusions
Summary and Contribution
Outlook and Future Work