Unless you are looking at trends over time, the annual variation doesn't matter - you simply lump all the data from 5, 10 years or whatever into one not-so-small database. The problems come with all the other things that have been touched on in this thread - what is a rescue (ie what about the ones where a call out has been avoided), what is a student, what is a student trip, what proportion of cavers in general are students, etc. As with any statistical analysis, the most important thing is to ask lots and lots of questions rather than to rush to gather data.