Causal analysis with longitudinal data using fixed and random effects models



Organised by:

University of Manchester


Prof Ian Plewis


Intermediate (some prior knowledge)


Claire Spencer, 0161 275 1980,


View in Google Maps  (M13 9PL)


Basement lab
Cathie Marsh Institute
Humanities Bridgeford street
University of Manchester



It is intended that students attending the course will be able to specify and estimate causal models for observational longitudinal data.  Students new to longitudinal data should consider taking the CMIST short course running before this one.

The course begins with a general consideration of, and practical work with models for data collected on two occasions, including differences in differences and Lord’s Paradox. It is followed by extensions to more than two occasions, focussing on the different regression lines that can be estimated. After lunch, random and fixed effects models are considered in detail, again linked to a practical session.


  1. Review regression models for two occasions, differences in differences models and the problems generated by arbitrary scales, and interpretational issues arising from Lord’s Paradox.
  2. Provide hands-on training for 1 using real data.
  3. Explain how to specify and estimate models for data on more than two occasions with reference to recent literature and focusing on both fixed and random effects approaches.
  4. Provide hands-on training for 3 using real data and showing how to use STATA and MLwiN for related questions.




This course will run from 10am- 4pm


£30 per day for UK/EU registered students
£60 per day for staff at UK/EU academic institutions, UK/EU Research Councils researchers, UK/EU public sector staff and staff at UK/EU registered charity organisations and recognised UK/EU research institutions.
£220 per day for all other participants.

Website and registration:


North West


Longitudinal Research , Longitudinal Data Analysis, Stata, Lord’s Paradox

Related publications and presentations:

Longitudinal Research
Longitudinal Data Analysis

Back to archive...