Analysis of Microdata

Transcription

1 Rainer Winkelmann Stefan Boes Analysis of Microdata Second Edition 4u Springer

2 1 Introduction What Are Microdata? Types of Microdata Qualitative Data Quantitative Data Why Not Linear Regression? Common Elements of Microdata Models Examples Determinants of Fertility Secondary School Choice Female Hours of Work and Wages Overview of the Book 20 2 From Regression to Probability Models Introduction Conditional Probability Functions Definition Estimation ' Interpretation Probability and Probability Distributions Axioms of Probability Univariate Random Variables Multivariate Random Variables Conditional Probability Models Further Exercises 41 3 Maximum Likelihood Estimation Introduction Likelihood Function Score Function and Hessian Matrix Conditional Models 52

3 VIII Maximization Properties of the Maximum Likelihood Estimator Expected Score Consistency Information Matrix Asymptotic Distribution Covariance Matrix Normal Linear Model Further Aspects of Maximum Likelihood Estimation Invariance and Delta Method Numerical Optimization Identification Quasi Maximum Likelihood Testing Introduction Restricted Maximum Likelihood Wald Test Likelihood Ratio Test Score Test Model Selection Goodness-of-Fit Pros and Cons of Maximum Likelihood Further Exercises 93 4 Binary Response Models Introduction Models for Binary Response Variables General Framework Linear Probability Model Probit Model Logit Model Interpretation of Parameters Discrete Choice Models Estimation Maximum Likelihood Perfect Prediction Properties of the Estimator Endogenous Regressors in Binary Response Models Estimation of Marginal Effects Goodness-of-Fit Non-Standard Sampling Schemes Stratified Sampling Exogenous Stratification Endogenous Stratification Flexible Specification of Binary Response Models 132

4 IX 4.8 Further Exercises 135 Multinomial Response Models Introduction Multinomial Logit Model Basic Model Estimation Interpretation of Parameters Conditional Logit Model Introduction General Model of Choice Modeling Conditional Logits Interpretation of Parameters Independence of Irrelevant Alternatives Generalized Multinomial Response Models Multinomial Probit Model Mixed Logit Models Nested Logit Models Further Exercises 170 Ordered Response Models Introduction Standard Ordered Response Models General Framework :> Ordered Probit Model Ordered Logit Model Estimation Interpretation of Parameters Single Indices and Parallel Regression Generalized Threshold Models Generalized Ordered Logit and Probit Models Interpretation of^parameters Sequential Models Modeling Conditional Transitions Generalized Conditional Transition Probabilities Marginal Effects Estimation Interval Data Further Exercises 206 Limited Dependent Variables Introduction Corner Solution Outcomes Sample Selection Models Treatment Effect Models 214

5 X 7.2 Tobin's Corner Solution Model Introduction Tobit Model Truncated Normal Distribution Inverse Mills Ratio and its Properties Interpretation of the Tobit Model Comparing Tobit and OLS Further Specification Issues Sample Selection Models Introduction Censored Regression Model Estimation of the Censored Regression Model Truncated Regression Model Incidental Censoring Example: Estimating a Labor Supply Model Treatment Effect Models Introduction Endogenous Binary Variable Switching Regression Model Further Exercises Event History Models Introduction Duration Models Introduction Basic Concepts Discrete Time Duration Models Continuous Time Duration Models.. b Key Element: Hazard Function Duration Dependence Unobserved Heterogeneity Count Data Models Poisson Regression Model Unobserved Heterogeneity Efficient versus Robust Estimation Censoring and Truncation Hurdle and Zero-Inflated Count Data Models Further Exercises 298 References 301 Solutions to Selected Exercises 311 Index 339