Time-varying covariates II

My division runs a reading group.  The subject is always methodological – this semester it’s a textbook on longitudinal data analysis (“Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence” by Singer and Willett, for those who are interested).  The reading group won a university award for innovation in teaching (or something along those lines) so it’s quite interesting and useful just from the point of view of teaching methods.  It has also answered one of the questions I raised in my last post on time-varying covariates.  There are, it emerges, a number of ways of modelling the effect of change.  The one that struck me as particularly neat involved creating a new variable for TIME SINCE CHANGE (e.g. years in work since GED achieved, to half-inch Singer and Willett’s example).  It allows you to model the fact the changes in such a time-varying covariates occur at different times for different people.

This entry was posted in Methods and tagged , , . Bookmark the permalink.