8. Random Coefficients (Slopes) Models

Introduction to Analysing Repeated Measures Data
Training session with Dr Helen Brown, Senior Statistician, at The Roslin Institute, March 2016.
************************************************
These training sessions were given to staff and research students at the Roslin Institute. The material is also used for the Animal Biosciences MSc course taught at the Institute.
************************************************
*Recommended YouTube playback settings for the best viewing experience: 1080p HD
************************************************
Content:
Random coefficients (slopes) model
- Interest lies with relationship over time rather than at individual time points
- Time points may be unevenly spaced

Inappropriate Regression Approach
- Ignore individuals and perform regression on full data (eg black line below)
- Measurements wrongly treated as independent

Two-stage Fixed Effects Approach
- Fit slopes to each individual and analyse slopes
- Satisfactory but takes no account of number of observations on each individual

Random coefficient (slopes) model
- Slope estimates are shrunken and take into account number of observations per individual