Statistical Methods for Longitudinal Data

IDEAR Early career training program

Statistical Methods for Longitudinal Data

University College London, United Kingdom, April 16-20 2018

The aim of the course is to provide an introduction to and hands on experience of the analysis of longitudinal studies using:

  • Random effects models for continuous outcomes
  • Growth curve models
  • Random effects models for binary data
  • Event history models
  • Multiple imputation for longitudinal data

Learning objectives

By the end of the course students will be able to:

  • Use methods to identify between and within individual variation in outcomes
  • Use and interpret models for longitudinal outcomes
  • Understand the implications of missing data
  • Perform multiple imputation for longitudinal data
  • Propose, evaluate and select models
  • Interpret and communicate results

Pre-requisites are:

  • Experience of STATA or similar package
  • Experience of linear and logistic regression
  • Experience of data analysis
  • Some knowledge of multiple imputation is desirable

There will be five sessions, each consisting of a 90-minute lecture followed by a 90-minute computer practical session. There will also be seminars, networking time and social events. Attendance at all sessions is compulsory.

Schedule Statistical Methods for Longitudinal Data

 Morning Afternoon
MondayRandom effects models for continuous outcomes
TuesdayGrowth curve modelsSeminar: Introduction to missing data
WednesdayEvent history analysisSeminar: To be confirmed
ThursdayRandom effects models for binary outcomesSeminar: To be confirmed
FridayMultiple imputation in longitudinal studies
Arrival Monday @1:30pm. Finish Friday @1:00pm