We have two sets (each composed of 6 variations) of HLM version 3 runs using the artificial data (n=100) described in the Reply (pp. 236-7), with each individual having different times of observation: