| |
| |
| |
| Introduction | |
| |
| |
| |
| Longitudinal studies | |
| |
| |
| |
| Examples | |
| |
| |
| |
| Notation | |
| |
| |
| |
| Merits of longitudinal studies | |
| |
| |
| |
| Approaches to longitudinal data analysis | |
| |
| |
| |
| Organization of subsequent chapters | |
| |
| |
| |
| Design considerations | |
| |
| |
| |
| Introduction | |
| |
| |
| |
| Bias | |
| |
| |
| |
| Efficiency | |
| |
| |
| |
| Sample size calculations | |
| |
| |
| |
| Continuous responses | |
| |
| |
| |
| Binary responses | |
| |
| |
| |
| Further reading | |
| |
| |
| |
| Exploring longitudinal data | |
| |
| |
| |
| Introduction | |
| |
| |
| |
| Graphical presentation of longitudinal data | |
| |
| |
| |
| Fitting smooth curves to longitudinal data | |
| |
| |
| |
| Exploring correlation structure | |
| |
| |
| |
| Exploring association amongst categorical responses | |
| |
| |
| |
| Further reading | |
| |
| |
| |
| General linear models for longitudinal data | |
| |
| |
| |
| Motivation | |
| |
| |
| |
| The general linear model, with correlated errors | |
| |
| |
| |
| The uniform correlation model | |
| |
| |
| |
| The exponential correlation model | |
| |
| |
| |
| Two-stage least-squares estimation and random effects models | |
| |
| |
| |
| Weighted least-squares estimation | |
| |
| |
| |
| Maximum likelihood estimation under Gaussian assumptions | |
| |
| |
| |
| Restricted maximum likelihood estimation | |
| |
| |
| |
| Robust estimation of standard errors | |
| |
| |
| |
| Parametric models for covariance structure | |
| |
| |
| |
| Introduction | |
| |
| |
| |
| Models | |
| |
| |
| |
| Pure serial correlation | |
| |
| |
| |
| Serial correlation plus measurement error | |
| |
| |
| |
| Random intercept plus serial correlation plus measurement error | |
| |
| |
| |
| Random effects plus measurement error | |
| |
| |
| |
| Model-fitting | |
| |
| |
| |
| Formulation | |
| |
| |
| |
| Estimation | |
| |
| |
| |
| Inference | |
| |
| |
| |
| Diagnostics | |
| |
| |
| |
| Examples | |
| |
| |
| |
| Estimation of individual trajectories | |
| |
| |
| |
| Further reading | |
| |
| |
| |
| Analysis of variance methods | |
| |
| |
| |
| Preliminaries | |
| |
| |
| |
| Time-by-time ANOVA | |
| |
| |
| |
| Derived variables | |
| |
| |
| |
| Repeated measures | |
| |
| |
| |
| Conclusions | |
| |
| |
| |
| Generalized linear models for longitudinal data | |
| |
| |
| |
| Marginal models | |
| |
| |
| |
| Random effects models | |
| |
| |
| |
| Transition (Markov) models | |
| |
| |
| |
| Contrasting approaches | |
| |
| |
| |
| Inferences | |
| |
| |
| |
| Marginal models | |
| |
| |
| |
| Introduction | |
| |
| |
| |
| Binary responses | |
| |
| |
| |
| The log-linear model | |
| |
| |
| |
| Log-linear models for marginal means | |
| |
| |
| |
| Generalized estimating equations | |
| |
| |
| |
| Examples | |
| |
| |
| |
| Counted responses | |
| |
| |
| |
| Parametric modelling for count data | |
| |
| |
| |
| Generalized estimating equation approach | |
| |
| |
| |
| Sample size calculations revisited | |
| |
| |
| |
| Further reading | |
| |
| |
| |
| Random effects models | |
| |
| |
| |
| Introduction | |
| |
| |
| |
| Estimation for generalized linear mixed models | |
| |
| |
| |
| Conditional likelihood | |
| |
| |
| |
| Maximum likelihood estimation | |
| |
| |
| |
| Logistic regression for binary responses | |
| |
| |
| |
| Conditional likelihood approach | |
| |
| |
| |
| Random effects models for binary data | |
| |
| |
| |
| Examples of logistic models with Gaussian random effects | |
| |
| |
| |
| Counted responses | |
| |
| |
| |
| Conditional likelihood method | |
| |
| |
| |
| Random effects models for counts | |
| |
| |
| |
| Poisson-Gaussian random effects models | |
| |
| |
| |
| Further reading | |
| |
| |
| |
| Transition models | |
| |
| |
| |
| General | |
| |
| |
| |
| Fitting transition models | |
| |
| |
| |
| Transition models for categorical data | |
| |
| |
| |
| Indonesian children's study example | |
| |
| |
| |
| Ordered categorical data | |
| |
| |
| |
| Log-linear transition models for count data | |
| |
| |
| |
| Further reading | |
| |
| |
| |
| Likelihood-based methods for categorical data | |
| |
| |
| |
| Introduction | |
| |
| |
| |
| Notation and definitions | |
| |
| |
| |
| Generalized linear mixed models | |
| |
| |
| |
| Maximum likelihood algorithms | |
| |
| |
| |
| Bayesian methods | |
| |
| |
| |
| Marginalized models | |
| |
| |
| |
| An example using the Gaussian linear model | |
| |
| |
| |
| Marginalized log-linear models | |
| |
| |
| |
| Marginalized latent variable models | |
| |
| |
| |
| Marginalized transition models | |
| |
| |
| |
| Summary | |
| |
| |
| |
| Examples | |
| |
| |
| |
| Crossover data | |
| |
| |
| |
| Madras schizophrenia data | |
| |
| |
| |
| Summary and further reading | |
| |
| |
| |
| Time-dependent covariates | |
| |
| |
| |
| Introduction | |
| |
| |
| |
| An example: the MSCM study | |
| |
| |
| |
| Stochastic covariates | |
| |
| |
| |
| Estimation issues with cross-sectional models | |
| |
| |
| |
| A simulation illustration | |
| |
| |
| |
| MSCM data and cross-sectional analysis | |
| |
| |
| |
| Summary | |
| |
| |
| |
| Lagged covariates | |
| |
| |
| |
| A single lagged covariate | |
| |
| |
| |
| Multiple lagged covariates | |
| |
| |
| |
| MSCM data and lagged covariates | |
| |
| |
| |
| Summary | |
| |
| |
| |
| Time-dependent confounders | |
| |
| |
| |
| Feedback: response is an intermediate and a confounder | |
| |
| |
| |
| MSCM data and endogeneity | |
| |
| |
| |
| Targets of inference | |
| |
| |
| |
| Estimation using g-computation | |
| |
| |
| |
| MSCM data and g-computation | |
| |
| |
| |
| Estimation using inverse probability of treatment, weights (IPTW) | |
| |
| |
| |
| MSCM data and marginal structural models using IPTW | |
| |
| |
| |
| Summary | |
| |
| |
| |
| Summary and further reading | |
| |
| |
| |
| Missing values in longitudinal data | |
| |
| |
| |
| Introduction | |
| |
| |
| |
| Classification of missing value mechanisms | |
| |
| |
| |
| Intermittent missing values and dropouts | |
| |
| |
| |
| Simple solutions and their limitations | |
| |
| |
| |
| Last observation carried forward | |
| |
| |
| |
| Complete case analysis | |
| |
| |
| |
| Testing for completely random dropouts | |
| |
| |
| |
| Generalized estimating equations under a random missingness mechanism | |
| |
| |
| |
| Modelling the dropout process | |
| |
| |
| |
| Selection models | |
| |
| |
| |
| Pattern mixture models | |
| |
| |
| |
| Random effect models | |
| |
| |
| |
| Contrasting assumptions: a graphical representation | |
| |
| |
| |
| A longitudinal trial of drug therapies for schizophrenia | |
| |
| |
| |
| Discussion | |
| |
| |
| |
| Additional topics | |
| |
| |
| |
| Non-parametric modelling of the mean response | |
| |
| |
| |
| Further reading | |
| |
| |
| |
| Non-linear regression modelling | |
| |
| |
| |
| Correlated errors | |
| |
| |
| |
| Non-linear random effects | |
| |
| |
| |
| Joint modelling of longitudinal measurements and recurrent events | |
| |
| |
| |
| Multivariate longitudinal data | |
| |
| |
| Appendix Statistical background | |
| |
| |
| |
| Introduction | |
| |
| |
| |
| The linear model and the method of least squares | |
| |
| |
| |
| Multivariate Gaussian theory | |
| |
| |
| |
| Likelihood inference | |
| |
| |
| |
| Generalized linear models | |
| |
| |
| |
| Logistic regression | |
| |
| |
| |
| Poisson regression | |
| |
| |
| |
| The general class | |
| |
| |
| |
| Quasi-likelihood | |
| |
| |
| Bibliography | |
| |
| |
| Index | |