If you modelled attrition between waves 1 and 2 in the previous step, try fitting the same model to predict absence by wave 11 of those present at wave 1. Compare the results with those for the one-step transition.
Alternatively, model the attrition pattern for the Scottish and/or Welsh samples for 1999/2000 (waves 1 and 2 for those subsamples) and compare the results with your earlier ones. To identify sample origin, use wMEMORIG, which has values 5 for the new Scottish and 6 for the new Welsh samples.