Longitudinal Data Analysis in Stata, 3-4 December 2015, 9am to 4.30pm,

NISRA Boardroom, McAuley House, 14 Castle Street, Belfast

Instructor: Prof. Fiona Steele (London School of Economics & Political Science)

Studies using longitudinal data are generally concerned with either the change over time in a response variable (e.g. physical measures such as height and weight), or the timing of events (e.g. death or births). This course will introduce multilevel (random effects) models for investigating both types of question.Topics will include multilevel growth curve models for repeated measures on continuous outcomes, and discrete-time methods for event history data.

The emphasis will be on the practical application of methods for longitudinal data analysis and interpreting the results. Lectures will be combined with hands-on practical exercises in Stata using social science datasets.

Prerequisites:

Familiarity with multilevel modelling and logistic regression analysis to the level of knowledge obtained by completing Modules 5 and 6 of the Centre for Multilevel Modelling’s online course (

Although prior knowledge of Stata is not essential, working through the Statapracticals for Module 5 and 6 of the online course is recommended as preparation for the course.

Technical Prerequisites:

We ask participants to bring their own laptops with STATA (version 12 or above) installed. Unfortunately, we cannot provide a computer cluster room or STATA for the event. However, those who do not have their own STATA on their laptops can download a free 30-day trial version of STATA 14. -> ( ). This trial version is the full STATA 14 and it expires after 30 days. If you are interested in this short-course, please download the trial version from the above website onto your laptop a few days before the event by filling out the online application form. The form asks, which STATA features you'd like with your trial copy. Please make sure to tick all the boxes for longitudinal analysis, multilevel/mixed effects, survival and time series analysis and survey module, so that all the necessary longitudinal features will be included in your trial copy. Note that it will take around one to two working days for STATA to check your application and to email you the license code.

Recommended reading:

Rabe-Hesketh S, Skrondal A. (2012) Multilevel and Longitudinal Modeling Using Stata, Third Edition. Volume I: Continuous Responses. College Station, Texas: Stata Press. Chapters 2 and 3.

Singer, J.D. and Willet, J.B. (1993) “It’s about time: Using discrete-time survival analysis to study duration and the timing of events.” Journal of Educational Statistics, 18: 155-195.

This event is free of charge to NILS users and students. Non-NILS users are asked to bring a course fee of £ 10.00. Places will be allocated on a first come first serve basis, however, NILS-users and students take preference.

Please send booking requests to: .

Program:

DAY 1: Growth Curve Models
09:00 – 09:30 / Registration
09:30 – 10:30 / Lecture 1
Repeated measures data; random intercept models; choice of time metric
10:30 – 11:00 / Practical 1
11:00 – 11:15 / BREAK
11:15 – 12:15 / Lecture 2
Random slope models
12:15 – 13:00 / Practical 2
13:00 – 14:00 / LUNCH
14:00 – 15:00 / Lecture 3
Modelling nonlinear growth; covariates and fitting group-specific growth curves
15:00 – 15:30 / Practical 3
15:30 – 15:45 / BREAK
15:45 – 16:45 / Lecture 4
More advanced topics
16:45 – 17:30 / Practical 4
DAY 2: Event history (survival) analysis
09:00 – 09:30 / Review and Q & A from Day 1
09:30 – 10:30 / Lecture 5
Event history data; introduction to event history analysis
10:30 – 10:45 / BREAK
10:45 – 11:45 / Lecture 6
Discrete-time event history analysis
11:45 – 12:30 / Practical 5
12:30 – 13:30 / LUNCH
13:30 – 15:00 / Lecture 7
Analysis of recurrent events (multilevel event history analysis)
Q & A