Assessment design studio

The idea is a tool that sits with you as you build an assessment, not after it. It asks what the task is really for, breaks it into every markable part, and benchmarks each one against a learning progression, so you can see what a question is actually asking of a student, not just how many marks it is worth.

It would flag where a paper won't give you the evidence you wanted, or doesn't sample widely enough to locate students rather than just score them. It drafts an assessment guide and moderation exemplars, and once you have marked, it returns item statistics, feedback sheets and a class report from a single entry of the marks. The numbers are always calculated, never guessed.

This is the big one, and right now it is just a concept. I am scoping it out: I will probably try it inside our Microsoft tools first, but it could just as easily be built elsewhere.