Tu sei qui:

95727 - LONGITUDINAL AND TIME-TO-EVENT DATA ANALYSIS

Anno Accademico 2021/2022

                
                        Docente:
                        Jeanne Jacobine Duistermaat
                    
                        Crediti formativi:
                        6
                    
                        SSD:
                        SECS-S/05
                    
                        Lingua di insegnamento:
                        Inglese
                    
                        Modalità didattica:
                        Convenzionale - Lezioni in presenza
                        
                            Campus:
                            Bologna
                        
                            Corso:
                            Laurea Magistrale in
                            Statistical sciences (cod. 9222)

                            Risorse didattiche su Virtuale

Conoscenze e abilità da conseguire

By the end of the course the student knows the classical statistical models for dependent responses based on random components and the principal methods for time to event data analysis. The student acquires the skills required to solve real-world research problems.

Contenuti

This course will introduce analysis tools and models for data collected in follow-up studies. These include longitudinally measured data such as blood levels (glucose, cholesterol, lead) and health scores, and time to event data such as death or disease recurrence in cancer patients.

The first weeks will be focussed on models for longitudinal data. We will cover topics such as models for the mean, models for the covariance, mixed effects models, model diagnostics and missing data.

Then we will proceed with proportional hazard models for survival data. After a brief rehearsal of the basics of survival analysis we will consider time dependent covariates.

We will finish the course with joint modelling of time to event and longitudinal data. When the focus is on the longitudinal outcome joint models might provide proper parameter estimates when the data are non-missing at random by modelling the missingness mechanism. When the focus is on the event data, joint models provide proper estimates when the time dependent covariate is endogenous and measured with an error.

We will analyse different datasets using R.

Testi/Bibliografia

Applied longitudinal analysis. Second Edition. GM Fitzmaurice, NM Laird, JH Ware. Wiley series in probability and statistics. Chapter 1 to 10, 17 and 18.

Joint models for longitudinal and time-to-event data, with applications in R. D Rizopoulos. Chapman & Hall/CRC Biostatistics Series. Chapter 1, 3, 4 and 7.

Metodi didattici

Lectures and class exercises including R.

Modalità di verifica e valutazione dell'apprendimento

Written exam

Strumenti a supporto della didattica

Materials will be made available on Virtuale

Orario di ricevimento

Consulta il sito web di Jeanne Jacobine Duistermaat