Since last October, we’ve been having what amounts to the most gigantic spring clean. Noticing that some teachers were re-uploading the same group of students each time they wanted to assign work (slap wrist!) we started cautiously de-duplicating the student list. And we went very cautiously; many schools have two or more students of the same name, so we took the approach that if there was any ambiguity at all, we would not de-duplicate.
And it worked well. Over two months, we better than halved the number of student records. The big benefit comes from having better continuity of data for each student. In many cases we now have years-worth of data from which to show progression, where previously we only had isolated data points that meant very little in themselves.
We did not stop there. Some student sets could not be merged but were clearly no longer in use. These we have started simply deleting after they have lain dormant for 13 months (a full year, plus a bit of leeway for teachers who only run one annual test). This is particularly useful for our most active schools; their student set lists were running to several pages, but with much of the content irrelevant to users.
What I am particularly proud of with this project is that we have achieved some very significant improvements without a single report of needed data going missing. A few people have asked what is going on (thus prompting this post) but, so far at least, it has all gone 100% as planned.