Common mistakes in System Dynamics: Manual to create simulation models for business dynamics, environment and social sciences.

Creating a simulation model with System Dynamics is not easy, there is the risk of making serious mistakes that force the model to remain unfinished after having dedicated days of work.

There are books and courses which show the steps to be taken in the process of creating a simulation model, but it is observed that some errors are repeated frequently. This book offers a different approach, instead of explaining how to create a simulation model, it shows the mistakes that are usually made.

The book is designed for users who are looking for a quick manual to identify the most common mistakes made when creating simulation models by applying System Dynamics, to correct them before presenting their research or work.

Content

1. Common Mistakes in System Dynamics
1.1. System Dynamics Modeling Process
1.2. Identifying and Avoiding Common Mistakes
1.3. Improve Reliability and Usability
2. Building Causal Loop Diagrams
2.1. Not following style rules
2.2. Mistaking Correlation for Causality
2.3. Overcomplicating the CLD
2.4. Ambiguous or Inconsistent Variable Names
2.5. Do not complete the loops
2.6. Inappropriate Polarities (+ or - signs)
2.7. Forgetting Delays
2.8. Variable that depends on many other
2.9. Variables in a negative sense
2.10. Variables that do not influence anything
2.11. Variables with signs
3. Stock and Flow Diagrams
3.1. Not following style rules
3.2. Incorrect variable type
3.3. Incorrect Flow Direction
3.4. Units Inconsistency
3.5. Leaving Outflows/Inflows Unconnected
3.6. Incorrectly connected flows
3.7. Levels that depend on auxiliaries
3.8. Duplicate variables
3.9. Neglecting Boundary Conditions
3.10. Arrows with signs
3.11. Use capital letters for everything
3.12. Clouds that depend on variables
3.13. Variables that depend on two tables
3.14. Physically impossible results
3.15. Variable that depends on itself
3.16. Simultaneous equations
3.17. Very obvious errors
4. Equations
4.1. Undefined or Unclear Variables
4.2. Lack of Dimensional Consistency
4.3. Confusing Constants with Variables
4.4. Over-simplifying Equations
4.5. Failure to Include Delays and Nonlinearities
4.6. Incorrect Handling of Initial Conditions
5. Simulating Models
5.1. Using Incorrect Time Step
5.2. Failing to Test for Sensitivity
5.3. Poorly Documented Assumptions
5.4. Inadequate Validation and Verification
5.5. Ignoring Extreme Conditions
5.6. Simulation with Uncalibrated Parameters
6. Simulation Results
6.1. Overconfidence in Model Accuracy
6.2. Misinterpreting Feedback Loops
6.3. Ignoring Time Horizons
6.4. Focusing on Isolated Variables
6.5. Misunderstanding Limits to Growth
6.6. Soft Variables and Intangible Factors
7. Communicating Model Results
7.1. Overusing Jargon
7.2. Poor Visualization of Results
7.3. Underexplaining Policy Implications
7.4. Ignoring the Audience's Perspective
7.5. Overstating the Importance of Simulation
8. Documentation
8.1. Lack of Transparency
8.2. Inconsistent Model Versioning
8.3. Unclear Assumptions or Data Sources
9. A Real-World Model
10. Best Practices for Avoiding Mistakes
11. FAQs
12. Checklist to Avoid Mistakes

Author

Juan Martín García is teacher, consultant, and a worldwide recognized expert in System Dynamics, with more than twenty years of experience in this field. Ph.D. Industrial Engineer (Spain) and Postgraduated Diploma in Business Dynamics at Massachusetts Institute of Technology MIT (USA). He teaches Vensim online courses in http://vensim.com/online-courses/ based on System Dynamics.