Hi Statalist,

I am working with a dataset regarding students and their exam scores. I have data on what days those exams were taken, and also how many exams they take during any given examination period.

For each student, I wish to take the time delay between each exam for each examination period, and use that time delay to create some form of concentration index which indicates how concentrated a student's exams are for a particular session.

I have attempted to use the Herfindel index, but it is a little unwieldy: how exactly would I obtain an exam analogue of market share?

Regards, Kevin