Open waaaaargh opened 7 years ago
Just writing down some thoughts I had on this:
This way, we can assess whether a learner actually has profitted from a lesson/project or whether they just already knew the stuff being taught. The random partitioning prevents us from accidentally putting harder questions in one of the subsets. If we also want to assess long-term retainment of the learnt material, we can simply partition into three subsets and let participants answer the last subset a couple of weeks later.
In a meatspace-based discussion yesterday, this came up. I assigned @CruzR because essentially, it was his idea.
Acceptance Criteria