Good paper on estimating a user's literacy from low-level system signals. I had a project a few years ago where we used a similar software setup to measure foc

Good paper on estimating a user's literacy from low-level system signals. I had a project a few years ago where we used a similar software setup to measure focus/concentration versus distraction (measuring switching programs/focus of windows etc.). Users included lawyers so great care had to be taken to do the logging in a way that is not too detailed (to log the program name that is running is okay, but logging the open document's file name is NOT).