| |
Text with CD-ROM |
From the Preface
Accelerating economic, technological, social, and environmental change challenge managers and policy makers to learn at increasing rates, while at the same time the complexity of the systems in which we live is growing. Many of the problems we now face arise as unanticipated side effects of our own past actions. All too often the policies we implement to solve important problems fail, make the problem worse, or create new problems.
Effective decision making and learning in a world of growing dynamic complexity requires us to become systems thinkers–to expand the boundaries of our mental models and develop tools to understand how the structure of complex systems creates their behavior.
This book introduces you to system dynamics modeling for the analysis of policy and strategy, with a focus on business and public policy applications. System dynamics is a perspective and set of conceptual tools that enable us to understand the structure and dynamics of complex systems. System dynamics is also a rigorous modeling method that enables us to build formal computer simulations of complex systems and use them to design more effective policies and organizations. Together, these tools allow us to create management flight simulators–microworlds where space and time can be compressed and slowed so we can experience the long-term side effects of decisions, speed learning, develop our understanding of complex systems, and design structures and strategies for greater success.
Features and Content
University and graduate-level texts, particularly those focused on business and public policy applications, have not kept pace with the growth of the field. This book is designed to provide thorough coverage of the field of system dynamics today, by examining
You will learn about the dynamics of complex systems, including the structures that create growth, goal-seeking behavior, oscillation and instability, S-shaped growth, overshoot and collapse, path dependence, and other nonlinear dynamics. Examples and applications include
The goal of systems thinking and system dynamics modeling is to improve our understanding of the ways in which an organization’s performance is related to its internal structure and operating policies, including those of customers, competitors, and suppliers and then to use that understanding to design high leverage policies for success. To do so this book utilizes
System dynamics is not a spectator sport. Developing systems thinking and modeling skills requires the active participation of you, the reader, via
The book can be used as a text in courses on systems thinking, simulation modeling, complexity, strategic thinking, operations, and industrial engineering, among others. It can be used in full or half-semester courses, executive education, and self-study. The book also serves as a reference for managers, engineers, consultants, and others interested in developing their systems thinking skills or using system dynamics in their organizations.
System dynamics is grounded in control theory and the modern theory of nonlinear dynamics. There is an elegant and rigorous mathematical foundation for the theory and models we develop. System dynamics is also designed to be a practical tool that policy makers can use to help them solve the pressing problems they confront in their organizations. Most managers have not studied nonlinear differential equations or even calculus, or have forgotten it if they did. To be useful, system dynamics modeling must be accessible to the widest range of students and practicing managers without becoming a vague set of qualitative tools and unreliable generalizations. That tension is compounded by the diversity of backgrounds within the community of managers, students, and scholars interested in system dynamics, backgrounds ranging from people with no mathematics education beyond high school to those with doctorates in physics.
If You Don’t Have a Strong Mathematics Background, Fear Not
This book presents system dynamics with a minimum of mathematical formalism. The goal is to develop your intuition and conceptual understanding, without sacrificing the rigor of the scientific method. You do not need calculus or differential equations to understand the material. Indeed, the concepts are presented using only text, graphs, and basic algebra. Mathematical details and references to more advanced material are set aside in separate sections and footnotes. Higher mathematics, though useful, is not as important as the critical thinking skills developed here.
If You Have a Strong Mathematics Background, Fear Not
Realistic and useful models are almost always of such complexity and nonlinearity that there are no known analytic solutions, and many of the mathematical tools you have studied have limited applicability. This book will help you use your strong technical background to develop your intuition and conceptual understanding of complexity and dynamics. Modeling human behavior differs from modeling physical systems in engineering and the sciences. We cannot put managers up on the lab bench and run experiments to determine their transfer function or frequency response. We believe all electrons follow the same laws of physics, but we cannot assume all people behave in the same way. Besides a solid grounding in the mathematics of dynamic systems, modeling human systems requires us to develop our knowledge of psychology, decision making, and organizational behavior. Finally, mathematical analysis, while necessary, is far from sufficient for successful systems thinking and modeling. For your work to have impact in the real world you must learn how to develop and implement models of human behavior in organizations, with all their ambiguity, time pressure, personalities, and politics. Throughout the book I have sought to illustrate how the technical tools and mathematical concepts you may have studied in the sciences or engineering can be applied to the messy world of the policy maker.
I welcome your comments, criticisms, and suggestions. Suggestions for additional examples, cases, theory, models, flight simulators, and so on, to make the book more relevant and useful to you are especially invited. Email comments to .
Preface 1. Learning in and about Complex Systems
1.1 Introduction
1.1.1 Policy Resistance, the Law of Unintended Consequences, and the Counterintuitive Behavior of Social Systems
1.1.2 Causes of Policy Resistance
1.1.3 Feedback
1.1.4 Process Point: The Meaning of Feedback
Challenge: Dynamics of Multiple-Loop Systems
1.2 Learning is a Feedback Process
1.3 Barriers to Learning
1.3.1 Dynamic Complexity
1.3.2 Limited Information
1.3.3 Confounding Variables and Ambiguity
1.3.4 Bounded Rationality and the Misperceptions of Feedback
1.3.5 Flawed Cognitive Maps
1.3.6 Erroneous Inferences about Dynamics
1.3.7 Unscientific Reasoning: Judgmental Errors and Biases
Challenge: Hypothesis Testing
1.3.8 Defensive Routines and Interpersonal Impediments to Learning
1.3.9 Implementation Failure
1.4 Requirements for Successful Learning in Complex Systems
1.4.1 Improving the Learning Process: Virtues of Virtual Worlds
1.4.2 Pitfalls of Virtual Worlds
1.4.3 Why Simulation is Essential
1.5 Summary
2. System Dynamics in Action
2.1 Applications of System Dynamics
2.2 Automobile Leasing Strategy: Gone Today, Here Tomorrow
2.2.1 Dynamic Hypothesis
2.2.2 Elaborating the Model
2.2.3 Policy Analysis
2.2.4 Impact and Follow-up
2.3 On Time and Under Budget: The Dynamics of Project Management
2.3.1 The Claim
2.3.2 Initial Model Development
2.3.3 Dynamic Hypothesis
2.3.4 The Modeling Process
2.3.5 Continuing Impact
2.4 Playing the Maintenance Game
2.4.1 Dynamic Hypothesis
2.4.2 The Implementation Challenge
2.4.3 Results
2.4.4 Transferring the Learning: The Lima Experience
2.5 Summary: Principles for Successful Use of System Dynamics
3. The Modeling Process
3.1 The Purpose of Modeling: Managers as Organization Designers
3.2 The Client and the Modeler
3.3 Steps of the Modeling Process
3.4 Modeling is Iterative
3.5 Overview of the Modeling Process
3.5.1 Problem Articulation: The Importance of Purpose
3.5.2 Formulating a Dynamic Hypothesis
3.5.3 Formulating a Simulation Model
3.5.4 Testing
3.5.5 Policy Design and Evaluation
3.6 Summary
4. Structure and Behavior of Dynamic Systems
4.1 Fundamental Modes of Dynamic Behavior
4.1.1 Exponential Growth
4.1.2 Goal Seeking
4.1.3 Oscillation
4.1.4 Process Point
Challenge: Identifying Feedback Structure from System Behavior
4.2 Interactions of the Fundamental Modes
4.2.1 S-shaped Growth
4.2.2 S-Shaped Growth with Overshoot
Challenge: Identifying the Limits to Growth
4.2.3 Overshoot and Collapse
4.3 Other Modes of Behavior
4.3.1 Stasis, or Equilibrium
4.3.2 Randomness
4.3.3 Chaos
4.4 Summary
5. Causal Loop Diagrams
5.1 Causal Diagram Notation
5.2 Guidelines for Causal Loop Diagrams
5.2.1 Causation versus Correlation
5.2.2 Labeling Link Polarity
Challenge: Assigning Link Polarities
5.2.3 Determining Loop Polarity
Challenge: Employee Motivation
5.2.4 Name Your Loops
5.2.5 Indicate Important Delays in Causal Links
5.2.6 Variable Names
5.2.7 Tips for Causal Loop Diagram Layout
5.2.8 Choose the Right Level of Aggregation
5.2.9 Don’t Put All the Loops into One Large Diagram
5.2.10 Make the Goals of Negative Loops Explicit
5.2.11 Distinguish between Actual and Perceived Conditions
5.3 Process Point: Developing Causal Diagrams from Interview Data
Challenge: Process Improvement
5.4 Conceptualization Case Study: Managing Your Workload
5.4.1 Problem Definition
5.4.2 Identifying Key Variables
5.4.3 Developing the Reference Mode
5.4.4 Developing the Causal Diagrams
5.4.5 Limitations of the Causal Diagram
Challenge: Policy Analysis with Causal Diagrams
5.5 Adam Smith’s Invisible Hand and the Feedback Structure of Markets
Challenge: The Oil Crises of the 1970s
Challenge: Speculative Bubbles
Challenge: The Thoroughbred Horse Market
5.4.1 Market Failure, Adverse Selection, and the Death Spiral
5.6 Explaining Policy Resistance: Traffic Congestion
5.6.1 Mental Models of the Traffic Problem
5.6.2 Compensating Feedback: The Response to Decreased Congestion
5.6.3 The Mass Transit Death Spiral
5.6.4 Policy Analysis: The Impact of Technology
5.6.5 Compensating Feedback: The Source of Policy Resistance
Challenge: Identifying the Feedback Structure of Policy Resistance
5.7 Summary
6. Stocks and Flows
6.1 Stocks, Flows, and Accumulation
6.1.1 Diagramming Notation for Stocks and Flows
6.1.2 Mathematical Representation of Stocks and Flows
6.1.3 The Contribution of Stocks to Dynamics
6.2 Identifying Stocks and Flows
6.2.1 Units of Measure in Stock and Flow Networks
6.2.2 The Snapshot Test
Challenge: Identifying Stocks and Flows
6.2.3 Conservation of Material in Stock and Flow Networks
6.2.4 State-Determined Systems
6.2.5 Auxiliary Variables
6.2.6 Stocks Change only Through their Rates
6.2.7 Continuous Time and Instantaneous Flows
6.2.8 Continuously Divisible versus Quantized Flows
6.2.9 Which Modeling Approach Should You Use?
6.2.10 Process Point: Portraying Stocks and Flows in Practice
6.3 Mapping Stocks and Flows
6.3.1 When Should Causal Loop Diagrams Show Stock and Flow Structure?
Challenge: Adding Stock and Flow Structure to Causal Diagrams
Challenge: Linking Stock and Flow Structure with Feedback
6.3.2 Aggregation in Stock and Flow Mapping
Challenge: Modifying Stock and Flow Maps
Challenge: Disaggregation
6.3.3 Guidelines for Aggregation
6.3.4 System Dynamics in Action: Modeling Large-Scale Construction Projects
6.3.5 Setting the Model Boundary: "Challenging the Clouds"
6.3.6 System Dynamics in Action: Automobile Recycling
6.4 Summary
7. Dynamics of Stocks and Flows
7.1 Relationship between Stocks and Flows
7.1.1 Static and Dynamic Equilibrium
7.1.2 Calculus without Mathematics
7.1.3 Graphical Integration
7.1.4 Graphical Differentiation
Challenge: Graphical Differentiation
7.2 System Dynamics in Action: Global Warming
7.3 System Dynamics in Action: The War on Drugs
7.3.1 The Cocaine Epidemic after 1990
7.4 Summary
8. Closing the Loop: Dynamics of Simple Structures
8.1 First-order Systems
8.2 Positive Feedback and Exponential Growth
8.2.1 Analytic Solution for the Linear First-Order System
8.2.2 Graphical Solution of the Linear First-Order Positive Feedback System
8.2.3 The Power of Positive Feedback: Doubling Times
Challenge: Paper Folding
8.2.4 Misperceptions of Exponential Growth
8.2.5 Process Point: Overcoming Overconfidence
8.3 Negative Feedback and Exponential Decay
8.3.1 Time constants and half lives
Challenge: Goal-seeking behavior
8.4 Multiple-Loop Systems
8.5 Nonlinear First-Order Systems: S-Shaped Growth
Challenge: Nonlinear Birth and Death Rates
8.5.1 Formal Definition of Loop Dominance
8.5.2 First-Order Systems Cannot Oscillate
8.6 Summary
9. S-Shaped Growth: Epidemics, Innovation Diffusion, and the Growth of New Products
9.1 Modeling S-Shaped Growth
9.1.1 Logistic Growth
9.1.2 Analytic Solution of the Logistic Equation
9.1.3 Other Common Growth Models
9.1.4 Testing the Logistic Model
9.2 Dynamics of Disease: Modeling Epidemics
9.2.1 A Simple Model of Infectious Disease
9.2.2 Modeling Acute Infection: The SIR Model
9.2.3 Model Behavior: The Tipping Point
Challenge: Exploring the SIR Model
9.2.4 Immunization and the Eradication of Smallpox
Challenge: The Efficacy Of Immunization Programs
9.2.5 Herd Immunity
9.2.6 Moving Past The Tipping Point: Mad Cow Disease
Challenge: Extending the SIR Model
9.2.7 Modeling the HIV/AIDS Epidemic
Challenge: Modeling HIV/AIDS
9.3 Innovation Diffusion as Infection: Modeling New Ideas and New Products
9.3.1 The Logistic Model of Innovation Diffusion: Examples
9.3.2 Process Point: Historical Fit and Model Validity
9.3.3 The Bass Diffusion Model
Challenge: Phase Space of the Bass Diffusion Model
9.3.4 Behavior of the Bass Model
Challenge: Critiquing the Bass Diffusion Model
Challenge: Extending the Bass Model
9.3.5 Fad and Fashion: Modeling the Abandonment of an Innovation
Challenge: Modeling Fads
9.3.6 Replacement Purchases
Challenge: Modeling the Life Cycle of Durable Products
9.4 Summary
10. Path Dependence and Positive Feedback
10.1 Path Dependence
Challenge: Identifying Path Dependence
10.2 A Simple Model of Path Dependence: The Polya Process
10.2.1 Generalizing the Model: Nonlinear Polya Processes
10.3 Path Dependence in the Economy: VHS vs. Betamax
Challenge: Formulating a Dynamic Hypothesis for the VCR Industry
10.4 Positive Feedback: The Engine of Corporate Growth
10.4.1 Product Awareness
10.4.2 Unit Development Costs
10.4.3 Price and Production Cost
10.4.4 Network Effects and Complementary Goods
10.4.5 Product Differentiation
10.4.6 New Product Development
10.4.7 Market Power
10.4.8 Mergers and Acquisitions
10.4.9 Workforce Quality and Loyalty
10.4.10 The Cost of Capital
10.4.11 The Rules of the Game
10.4.12 Ambition and Aspirations
10.4.13 Creating Synergy for Corporate Growth
10.5 Positive Feedback, Increasing Returns, and Economic Growth
10.6 Does the Economy Lock in to Inferior Technologies?
10.7 Limits to Lock In
10.8 Modeling Path Dependence and Standards Formation
10.8.1 Model Structure
10.8.2 Model Behavior
10.8.3 Policy Implications
Challenge: Policy Analysis
Challenge: Extending the Model
10.9 Summary
11. Delays
11.1 Delays: An Introduction
Challenge: Duration and Dynamics of Delays
11.1.1 Defining Delays
11.2 Material Delays: Structure and Behavior
11.2.1 What is the Average Length of the Delay?
11.2.2 What is the Distribution of the Output around the Average Delay Time?
11.2.3 Pipeline Delay
11.2.4 First-Order Material Delay
11.2.5 Higher-Order Material Delays
11.2.6 How Much is in the Delay? Little’s Law
11.3 Information Delays: Structure and Behavior
11.3.1 Modeling Perceptions: Adaptive Expectations and Exponential Smoothing
11.3.2 Higher-Order Information Delays
11.4 Response to Variable Delay Times
Challenge: Response of Delays to Changing Delay Times
11.4.1 Nonlinear Adjustment Times: Modeling Ratchet Effects
11.5 Estimating the Duration and Distribution of Delays
11.5.1 Estimating Delays when Numerical Data are Available
11.5.2 Estimating Delays when Numerical Data are not Available
11.5.3 Process Point: Walk the Line
11.6 System Dynamics in Action: Forecasting Semiconductor Demand
11.7 Mathematics of Delays: Koyck Lags and Erlang Distributions
11.7.1 General Formulation for Delays
11.7.2 First-Order Delay
11.7.3 Higher-Order Delays
11.7.4 Relation of Material and Information Delays
11.8 Summary
12. Coflows and Aging Chains
12.1 Aging Chains
12.1.1 General Structure of Aging Chains
12.1.2 Example: Population and Infrastructure in Urban Dynamics
12.1.3 Example: The Population Pyramid and the Demographic Transition
12.1.4 Aging Chains and Population Inertia
12.1.5 System Dynamics in Action: World Population and Economic Development
12.1.6 Case Study: Growth and the Age Structure of Organizations
12.1.7 Promotion Chains and the Learning Curve
12.1.8 Mentoring and On-The-Job Training
Challenge: The Interactions of Training Delays and Growth
12.2 Coflows: Modeling the Attributes of a Stock
Challenge: Coflows
12.2.1 Coflows with Nonconserved Flows
Challenge: The Dynamics of Experience and Learning
12.2.2 Integrating Coflows and Aging Chains
Challenge: Modeling Design Wins in the Semiconductor Industry
12.3 Summary
13. Modeling Decision Making
13.1 Principles for Modeling Decision Making
13.1.1 Decisions and Decision Rules
13.1.2 Five Formulation Fundamentals
Challenge: Finding Formulation Flaws
13.2 Formulating Rate Equations
13.2.1 Fractional Increase Rate
13.2.2 Fractional Decrease Rate
13.2.3 Adjustment to a Goal
13.2.4 The Stock Management Structure: Rate = Normal Rate + Adjustments
13.2.5 Flow = Resource * Productivity
13.2.6 Y = Y^{*} * Effect of X_{1} on Y * Effect of X_{2} on Y * … * Effect of X_{n} on Y
13.2.7 Y = Y^{*} + Effect of X_{1} on Y + Effect of X_{2} on Y + … + Effect of X_{n} on Y
13.2.8 Fuzzy MIN Function
13.2.9 Fuzzy MAX Function
13.2.10 Floating Goals
Challenge: Floating Goals
Challenge: Goal Formation with Internal and External Inputs
13.2.11 Nonlinear Weighted Average
13.2.12 Modeling Search: Hill-Climbing Optimization
Challenge: Finding the Optimal Mix of Capital and Labor
13.2.13 Resource Allocation
13.3 Common Pitfalls
13.3.1 All Outflows Require First-Order Control
Challenge: Preventing Negative Stocks
13.3.2 Avoid IF…THEN…ELSE Formulations
13.3.3 Disaggregate Net Flows
13.4 Summary
14. Formulating Nonlinear Relationships
14.1 Table Functions
14.1.1 Specifying Table Functions
14.1.2 Example: Building a Nonlinear Function
14.1.3 Process Point: Table Functions Versus Analytic Functions
14.2 Case Study: Cutting Corners Versus Overtime
Challenge: Formulating Nonlinear Functions
14.2.1 Working Overtime: The Effect of Schedule Pressure on Workweek
14.2.2 Cutting Corners: The Effect of Schedule Pressure on Time per Task
14.3 Case Study: Estimating Nonlinear Functions With Qualitative and Numerical Data
Challenge: Refining Table Functions with Qualitative Data
14.4 Common Pitfalls
14.4.1 Using the Wrong Input
Challenge: Critiquing Nonlinear Functions
14.4.2 Improper Normalization
14.4.3 Avoid Hump-shaped Functions
Challenge: Formulating the Error Rate
Challenge: Testing the Full Model
14.5 Eliciting Model Relationships Interactively
14.5.1 Case Study: Estimating Precedence Relationships in Product Development
14.6 Summary
15. Modeling Human Behavior: Bounded Rationality or Rational Expectations?
15.1 Human Decision Making: Bounded Rationality or Rational Expectations?
15.2 Cognitive Limitations
15.3 Individual and Organizational Responses to Bounded Rationality
15.3.1 Habit, Routines, and Rules of Thumb
15.3.2 Managing Attention
15.3.3 Goal Formation and Satisficing
15.3.4 Problem Decomposition and Decentralized Decision Making
15.4 Intended Rationality
15.4.1 Testing for Intended Rationality: Partial Model Tests
15.5 Case Study: Modeling High-Tech Growth Firms
15.5.1 Model Structure: Overview
15.5.2 Order Fulfillment
15.5.3 Capacity Acquisition
Challenge: Hill Climbing
15.5.4 The Sales Force
15.5.5 The Market
15.5.6 Behavior of the Full System
Challenge: Policy Design in the Market Growth Model
15.6 Summary
16. Forecasts and Fudge Factors: Modeling Expectation Formation
16.1 Modeling Expectation Formation
16.1.1 Modeling Growth Expectations: The TREND Function
16.1.2 Behavior of the TREND Function
16.2 Case Study: Energy Consumption
16.3 Case Study: Commodity Prices
16.4 Case Study: Inflation
16.5 Implications for Forecast Consumers
Challenge: Extrapolation and Stability
16.6 Initialization and Steady State Response of the TREND Function
16.7 Summary
17. Supply Chains and the Origin of Oscillations
17.1 Supply Chains in Business and Beyond
17.1.1 Oscillation, Amplification, and Phase Lag
17.2 The Stock Management Problem
17.2.1 Managing a Stock: Structure
17.2.2 Steady State Error
17.2.3 Managing a Stock: Behavior
17.3 The Stock Management Structure
17.3.1 Behavior of the Stock Management Structure
17.4 The Origin of Oscillations
17.4.1 Mismanaging the Supply Line: The Beer Distribution Game
17.4.2 Why Do We Ignore the Supply Line?
17.4.3 Case Study: Boom and Bust in Real Estate Markets
17.5 Summary
18. The Manufacturing Supply Chain
18.1 The Policy Structure of Inventory and Production
18.1.1 Order Fulfillment
18.1.2 Production
18.1.3 Production Starts
18.1.4 Demand Forecasting
18.1.5 Process Point: Initializing a Model in Equilibrium
Challenge: Simultaneous Initial Conditions
18.1.6 Behavior of the Production Model
18.1.7 Enriching the Model: Adding Order Backlogs
18.1.8 Behavior of the Firm with Order Backlogs
18.1.9 Adding Raw Materials Inventory
18.2 Interactions among Supply Chain Partners
18.2.1 Instability and Trust in Supply Chains
18.2.2 From Functional Silos to Integrated Supply Chain Management
Challenge: Reengineering the Supply Chain
18.3 System Dynamics in Action: Reengineering the Supply Chain in a High-Velocity Industry
18.3.1 Initial Problem Definition
18.3.2 Reference Mode and Dynamic Hypothesis
18.3.3 Model Formulation
18.3.4 Testing the Model
18.3.5 Policy Analysis
18.3.6 Implementation: Sequential Debottlenecking
18.3.7 Results
18.4 Summary
19. The Labor Supply Chain and the Origin of Business Cycles
19.1 The Labor Supply Chain
19.1.1 Structure of Labor and Hiring
19.1.2 Behavior of the Labor Supply Chain
19.2 Interactions of Labor and Inventory Management
Challenge: Mental Simulation of Inventory Management with Labor
19.2.1 Inventory—Workforce Interactions: Behavior
19.2.2 Process Point: Explaining Model Behavior
Challenge: Explaining Oscillations
19.2.3 Understanding the Sources of Oscillation
Challenge: Policy Design to Enhance Stability
19.2.4 Adding Overtime
19.2.5 Response to Flexible Workweeks
Challenge: Reengineering a Manufacturing Firm for Enhanced Stability
19.2.6 The Costs of Instability
Challenge: The Costs of Instability
Challenge: Adding Training and Experience
19.3 Inventory—Workforce Interactions and the Business Cycle
19.3.1 Is the Business Cycle Dead?
19.4 Summary
20. The Invisible Hand Sometimes Shakes: Commodity Cycles
20.1 Commodity Cycles: From Aircraft to Zinc
20.2 A Generic Commodity Market Model
20.2.1 Production and Inventory
20.2.2 Capacity Utilization
20.2.3 Production Capacity
20.2.4 Desired Capacity
Challenge: Intended Rationality of the Investment Process
20.2.5 Demand
20.2.6 The Price-Setting Process
20.3 Application: Cycles in the Pulp and Paper Industry
Challenge: Sensitivity to Uncertainty in Parameters
Challenge: Sensitivity to Structural Changes
Challenge: Implementing Structural Changes–Modeling Livestock Markets
Challenge: Policy Analysis
20.4 Summary
21.Truth and Beauty: Validation and Model Testing
21.1 Validation and Verification are Impossible
21.2 Questions Model Users Should Ask–But Usually Don’t
21.3 Pragmatics and Politics of Model Use
21.3.1 Types of Data
21.3.2 Documentation
21.3.3 Replicability
21.3.4 Protective versus Reflective Modeling
21.4 Model Testing in Practice
21.4.1 Boundary Adequacy Tests
21.4.2 Structure Assessment Tests
21.4.3 Dimensional Consistency
21.4.4 Parameter Assessment
21.4.5 Extreme Condition Tests
Challenge: Extreme Condition Tests
21.4.6 Integration Error Tests
21.4.7 Behavior Reproduction Tests
21.4.8 Behavior Anomaly Tests
21.4.9 Family Member Tests
21.4.10 Surprise Behavior Tests
21.4.11 Sensitivity Analysis
21.4.12 System Improvement Tests
Challenge: Model Testing
21.5 Summary
22. Challenges for the Future
22.1 Theory
22.2 Technology
22.3 Implementation
22.4 Education
22.5 Applications
Challenge: Putting System Dynamics Into Action
Appendix A Numerical Integration
Challenge: Choosing a Time Step
Appendix B Noise
Challenge: Exploring Noise
References
Index