METHODS - Discuss best approach to longitudinal study
There are basically two approaches to this longitudinal study:
- Re-do similar PCA and hierarchical clustering as original study for each year and then compare each (not actually each year but by 5 year groups - say 1988-1993 together, 1994-1999, and so on)
- Do PCA and clustering of the evolution of each countries and compare instead
Discuss which is most pertinent with available data and why?
Maybe run both and compare?