Buch, Englisch, 352 Seiten, Format (B × H): 174 mm x 246 mm, Gewicht: 453 g
Buch, Englisch, 352 Seiten, Format (B × H): 174 mm x 246 mm, Gewicht: 453 g
ISBN: 978-1-032-65778-3
Verlag: Taylor & Francis Ltd
This third edition of our book provides an updated introduction to event history modeling along with many instructive Stata examples. Using the latest Stata software, each of these practical examples develops a research question, points to useful contextual background information, gives a brief account of the underlying statistical concepts, describes the organization of input data and the application of Stata statistical procedures, and assists the reader in interpreting the content of the results obtained. Emphasizing the strengths and limitations of continuous-time event history analysis in different fields of social science applications, this book demonstrates that event history models provide a useful approach to uncover causal relation- ships or to map a system of causal relationships. In particular, this book demonstrates how long-term processes can be studied, how different forms of duration dependencies can be estimated using nonparametric, parametric and semiparametric models, and how parallel and interdependent dynamic systems can be analyzed from a causal-analytical point of view using the method of episode splitting. The book also shows how changing contextual information at the micro, meso and macro levels can be easily integrated into a dynamic analysis of longitudinal data. Finally, the book addresses the problem of unobserved heterogeneity of time-constant and time-dependent omitted variables and makes suggestions for dealing with these sometimes difficult methodological problems.
Causal Analysis with Event History Data Using Stata is an invaluable resource for both novice and experienced researchers from a variety of fields (e.g. sociology, economics, political science, education, psychology, demography, epidemiology, management research and organizational studies, as well as medicine and clinical applications) who need an introductory text- book on continuous-time event history analysis and who are looking for a practical handbook for their longitudinal research.
Zielgruppe
Postgraduate and Undergraduate Advanced
Autoren/Hrsg.
Fachgebiete
- Wirtschaftswissenschaften Betriebswirtschaft Wirtschaftsmathematik und -statistik
- Sozialwissenschaften Psychologie Psychologie / Allgemeines & Theorie Psychologische Forschungsmethoden
- Medizin | Veterinärmedizin Medizin | Public Health | Pharmazie | Zahnmedizin Medizin, Gesundheitswesen Epidemiologie, Medizinische Statistik
- Mathematik | Informatik Mathematik Stochastik
- Wirtschaftswissenschaften Volkswirtschaftslehre Volkswirtschaftslehre Allgemein Wirtschaftsstatistik, Demographie
Weitere Infos & Material
Start vii
Preface ix
1 Introduction 1
1.1 Causal Modeling and Observation Plans 5
1.1.1 Cross-Sectional Data 6
1.1.2 Panel Data 14
1.1.3 Event History Data 21
1.2 Event History Analysis and Causal Modeling 23
1.2.1 Causal Explanations 23
1.2.2 Transition Rate Models 37
2 Event History Data Structures 47
2.1 Basic Terminology 47
2.2 Event History Data Organization 51
3 Nonparametric Descriptive Methods 70
3.1 Life Table Method 70
3.2 Product-Limit Estimation 84
3.3 Comparing Survivor Functions 88
4 Exponential Transition Rate Models 103
4.1 The Basic Exponential Model 104
4.1.1 Maximum Likelihood Estimation 105
4.1.2 Models without Covariates 108
4.1.3 Time-Constant Covariates 111
4.2 Models with Multiple Destinations 119
4.3 Models with Multiple Episodes 129
5 Piecewise Constant Exponential Models 135
5.1 The Basic Model 135
5.2 Models without Covariates 137
5.3 Models with Proportional Covariate Effects 143
5.4 Models with Period-Specific Effects 144
6 Exponential Models with Time-Dependent Covariates 149
6.1 Parallel and Interdependent Processes 149
6.2 Interdependent Processes: The System Approach 152
6.3 Interdependent Processes: The Causal Approach 156
6.4 Episode Splitting with Qualitative Covariates 158
6.5 Episode Splitting with Quantitative Covariates 172
6.6 Application Examples.v 178
vi contents
7 Parametric Models of Time Dependence 208
7.1 Interpretation of Time Dependence 209
7.2 Gompertz Models 212
7.3 Weibull Models 222
7.4 Log-Logistic Models 230
7.5 Log-Normal Models 236
8 Methods for Testing Parametric Assumptions 242
8.1 Simple Graphical Methods 242
8.2 Pseudoresiduals 244
9 Semiparametric Transition Rate Models 250
9.1 Partial Likelihood Estimation 251
9.2 Time-Dependent Covariates 256
9.3 The Proportionality Assumption 261
9.4 Stratification with Covariates and for Multiepisode Data 266
9.5 Baseline Rates and Survivor Functions 271
9.6 Application Example 274
10 Problems of Model Specification 278
10.1 Unobserved Heterogeneity 278
10.2 Models with a Mixture Distribution 284
10.2.1 Models with a Gamma Mixture 287
10.2.2 Exponential Models with a Gamma Mixture 290
10.2.3 Weibull Models with a Gamma Mixture 292
10.2.4 Random Effects for Multiepisode Data 296
10.3 Discussion 300
11 Sequence Analysis 305
Brendan Halpin
11.1 What is Sequence Analysis? 305
11.2 Defining Distances 307
11.3 Doing Sequence Analysis in Stata 310
11.4 Unary Summaries 313
11.5 Intersequence Distance 315
11.6 What to Do with Sequence Distances? 317
11.7 Optimal Matching Distance 321
11.8 Special Topics 322
11.9 Conclusion 333
Appendix: Exercises 335
References 348
About the Authors 380
