Discovering Causal Structure
Artificial Intelligence, Philosophy of Science, and Statistical Modeling
Free Global Shipping
No minimum orderDescription
Discovering Causal Structure: Artificial Intelligence, Philosophy of Science, and Statistical Modeling provides information pertinent to the fundamental aspects of a computer program called TETRAD. This book discusses the version of the TETRAD program, which is designed to assist in the search for causal explanations of statistical data. or alternative models. This text then examines the notion of applying artificial intelligence methods to problems of statistical model specification. Other chapters consider how the TETRAD program can help to find god alternative models where they exist, and how it can help detect the existence of important neglected variables. This book discusses as well the procedures for specifying a model or models to account for non-experimental or quasi-experimental data. The final chapter presents a description of the format of input files and a description of each command. This book is a valuable resource for social scientists and researchers.
Table of Contents
About This Book
Foreword
Acknowledgments
Part I : Artificial Intelligence and Nonexperimental Science
1. The Problems of Science without Experiments
1.1 The Limits of Experimentation
1.2 The Limits of Human Judgment
1.3 The Artificial Intelligence Solution
2. The Case against Causal Modeling
2.1 The Critical Reaction
2.2 Making Sense of Causality
2.3 Causes, Indicators, and the Interpretation of Latent Variables
2.4 The Importance of Experiment
2.5 Justifying Assumptions
2.6 Linear Theories Are Literally False
2.7 Conclusion
3. Objections to Discovery by Computer
3.1 Introduction
3.2 The General Objections
3.3 No Peeking
Part II : The TETRAD Program
4. Causal and Statistical Models
4.1 Introduction
4.2 Directed Graphs and Causal Models
4.3 Statistical Models from Causal Models
4.4 Treks and Coordinating Path Effects
4.5 Constraints on Correlations
4.6 Correlated Errors Are Not Equivalent to Direct Effects
4.7 Statistical Issues, Briefly Considered
5. The Structure and Method of TETRAD
5.1 The Methodological Principles That Underlie TETRAD
5.2 How the Methodological Principles Are Realized in TETRAD
5.3 Search Strategies for Finding Good Causal Models
5.4 A Sketch of the TETRAD Program
5.5 How to Use TETRAD'S Output
5.6 TETRAD and Other Search Procedures
5.7 Future Developments
6. What TETRAD Can Do
6.1 Alienation
6.2 A Problem Using Simulated Data
6.3 Causal Order from Correlations
6.4 Kohn's Study and Temporal Order among Interview Questions
7. Simulation Studies
7.1 A Simulated Case
7.2 Distinguishing Correlation from Causation
7.3 Locating Connected Variables
8. Case Studies
8.1 Introduction
8.2 Industrial and Political Development
8.3 Measuring the Authoritarian Personality
8.4 Alternatives to Regression Models
8.5 Introducing Latent Variables: Longitudinal Data with SAT Scores
8.6 Roll Call Voting
8.7 The Effects of Summer Head Start
8.8 Achievement, Ability, and Approval
8.9 The Stability of Alienation
9. A Brief History of Heuristic Search in Applied Statistics
10. Mathematical Foundations
10.1 The Algorithm
10.2 Proofs of Correctness of the Algorithms Employed by TETRAD
Part III : Using TETRAD, EQS, and LISREL
11. Using TETRAD with EQS and LISREL
11.1 LISREL and Its Restrictions
11.2 Overcoming the Restrictions
11.3 EQS
12. Running TETRAD
12.1 Installing TETRAD
12.2 Entering and Exiting TETRAD
12.3 Getting Help
12.4 Input Files
12.5 Output Files
12.6 View and Edit
12.7 The Run Command and Menus
12.8 User Interrupts
12.9 Errors
12.10 TETRAD Commands
12.11 Running TETRAD in Batch Mode
12.12 List of Commands
12.13 Command Summaries
Appendix: The Grammar of the Input
About the Authors
References
Index
Product details
- No. of pages: 412
- Language: English
- Copyright: © Academic Press 1987
- Published: August 24, 1987
- Imprint: Academic Press
- eBook ISBN: 9781483265797