2.1. Introduction -- 2.2. Continuous/interval scale data -- 2.3. Simple and multiple linear regression models -- 2.4. Checking assumptions in linear regression models -- 2.5. Likelihood: multiple linear regression -- 2.6. Comparing model likelihoods -- 2.7. Application of a multiple linear regression model -- 2.8. Exercises on linear models -- 3.1. Binary data -- 3.1.1. 3.1.2. Logistic regression -- 3.1.3. Logit and probit transformations -- 3.1.4. General logistic regression -- 3.1.5. Likelihood -- 3.1.6. Example with binary data -- 3.2. Ordinal data -- 3.2.1. 3.2.2. The ordered logit model -- 3.2.3. Dichotomization of ordered categories -- 3.2.4. 3.2.5. Example with ordered data -- 3.3. Count data -- 3.3.1. 3.3.2. Poisson regression models -- 3.3.3. 3.3.4. Example with count data -- 3.4. Exercises -- 4.1. 4.2. The linear model
4.3. The binary response model -- 4.4. The Poisson model -- 4.5. Likelihood -- 5.1. Introduction -- 5.2. Linear mixed model -- 5.3. The intraclass correlation coefficient -- 5.4. Parameter estimation by maximum likelihood -- 5.5. Regression with level-two effects -- 5.6. Two-level random intercept models -- 5.7. General two-level models including random intercepts -- 5.8. 5.9. Residuals -- 5.10. Checking assumptions in mixed models -- 5.11. Comparing model likelihoods -- 5.12. Application of a two-level linear model -- 5.13. Two-level growth models -- 5.13.1. A two-level repeated measures model -- 5.13.2. A linear growth model -- 5.13.3. A quadratic growth model -- 5.14. 5.15. Example using linear growth models -- 5.16. Exercises using mixed models for continuous/interval scale data -- 6.1. 6.2. The two-level logistic model -- 6.3. General two-level logistic models -- 6.4. Intraclass correlation coefficient -- 6.5. 6.6. Example using binary data -- 6.7. Exercises using mixed models for binary data
7.1. Introduction -- 7.2. The two-level ordered logit model -- 7.3. Likelihood -- 7.4. Example using mixed models for ordered data -- 7.5. Exercises using mixed models for ordinal data -- 8.1. 8.2. The two-level Poisson model -- 8.3. 8.4. Example using mixed models for count data -- 8.5. Exercises using mixed models for count data -- 9.1. 9.2. The mixed linear model -- 9.3. The mixed binary response model -- 9.4. The mixed Poisson model -- 9.5. 10.1. 10.2. Three-level random intercept models -- 10.3. Three-level generalized linear models -- 10.4. Linear models -- 10.5. Binary response models -- 10.6. 10.7. Example using three-level generalized linear models -- 10.8. Exercises using three-level generalized linear mixed models -- 11.1. 11.2. Multivariate two-level generalized linear model -- 11.3. Bivariate Poisson model: example -- 11.4. Bivariate ordered response model: example -- 11.5. Bivariate linear-probit model: example -- 11.6. Multivariate two-level generalized linear model likelihood
11.7. Exercises using multivariate generalized linear mixed models -- 12.1. Introduction -- 12.1.1. Left censoring -- 12.1.2. Right censoring -- 12.1.3. Time-varying explanatory variables -- 12.1.4. Competing risks -- 12.2. Duration data in discrete time -- 12.2.1. Single-level models for duration data -- 12.2.2. Two-level models for duration data -- 12.2.3. Three-level models for duration data -- 12.3. Renewal data -- 12.3.1. 12.3.2. Example: renewal models -- 12.4. Competing risk data -- 12.4.1. 12.4.2. Likelihood -- 12.4.3. Example: competing risk data -- 12.5. Exercises using renewal and competing risks models -- 13.1. 13.2. Mover-stayer model -- 13.3. Likelihood incorporating the mover-stayer model -- 13.4. Example 1: stayers within count data -- 13.5. Example 2: stayers within binary data -- 13.6. Exercises: stayers -- 14.1. Introduction to key issues: heterogeneity, state dependence and non-stationarity -- 14.2. Example -- 14.3. Random effects models -- 14.4. Initial conditions problem -- 14.5. Initial treatment
14.6. Example: depression data -- 14.7. Classical conditional analysis -- 14.8. Classical conditional model: example -- 14.9. Conditioning on initial response but allowing random effect uol to be dependent on z3 -- 14.10. Wooldridge conditional model: example -- 14.11. Modelling the initial conditions -- 14.12. Same random effect in the initial response and subsequent response models with a common scale parameter -- 14.13. Joint analysis with a common random effect: example -- 14.14. Same random effect in models of the initial response and subsequent responses but with different scale parameters -- 14.15. Joint analysis with a common random effect (different scale parameters): example -- 14.16. Different random effects in models of the initial response and subsequent responses -- 14.17. Different random effects: example -- 14.18. Embedding the Wooldridge approach in joint models for the initial response and subsequent responses -- 14.19. Joint model incorporating the Wooldridge approach: example -- 14.20. Other link functions -- 14.21. Exercises using models incorporating initial conditions/state dependence in binary data
15.1. Introduction -- 15.2. Fixed effects treatment of the two-level linear model -- 15.3. Dummy variable specification of the fixed effects model -- 15.4. Empirical comparison of two-level fixed effects and random effects estimators -- 15.5. Implicit fixed effects estimator -- 15.6. Random effects models -- 15.7. Comparing two-level fixed effects and random effects models -- 15.8. Fixed effects treatment of the three-level linear model -- 15.9. Exercises comparing fixed effects and random effects -- A.1. SabreR installation -- A.2. SabreR commands -- A.2.1. The arguments of the SabreR object -- A.2.2. The anatomy of a SabreR command file -- A.3. Quadrature -- A.3.1. Standard Gaussian quadrature -- A.3.2. Performance of Gaussian quadrature -- A.3.3. Adaptive quadrature -- A.4. Estimation -- A.4.1. Maximizing the log likelihood of random effects models -- A.5. Fixed effects linear models -- A.6. Endogenous and exogenous variables -- B.1. Getting started with R -- B.1.1. Preliminaries -- B.1.1.1. Working with R in interactive mode -- B.1.1.2. Basic functions -- B.1.1.3. Getting help
B.1.1.4. Stopping R -- B.1.2. Creating and manipulating data -- B.1.2.1. Vectors and lists -- B.1.2.2. Vectors -- B.1.2.3. Vector operations -- B.1.2.4. Lists -- B.1.2.5. Data frames -- B.1.3. Session management -- B.1.3.1. Managing objects -- B.1.3.2. Attaching and detaching objects -- B.1.3.3. Serialization -- B.1.3.4. R scripts -- B.1.3.5. Batch processing -- B.1.4. R packages -- B.1.4.1. Loading a package into R -- B.1.4.2. Installing a package for use in R -- B.1.4.3. R and Statistics -- B.2. Data preparation for SabreR -- B.2.1. Creation of dummy variables -- B.2.2. Missing values -- B.2.3. Creating lagged response covariate data