In the last few years, I have experimented with different types of assessment strategies. In particular, I used something that would be recognizable as Standards Based Grading, and something that would be recognizable as Specifications Grading.
But this term, for at least one course, I am abandoning both. My best course is Euclidean Geometry. Standards based Grading didn’t work because the learning goals are big, process-oriented things. This made it seem like Specs Grading would be a decent fit, especially because a typical assignment (“find a proof of this theorem, present it to the class, and write a paper about it”) is a complicated, professional activity. In the end, the quality might vary, but you either did it well enough, or you didn’t. But the explicit statements like “do seven of these to earn an A” broke my spring class. Students rushed to their desired grade and then stopped. So, I won’t do either of them in the canonical way for this course again.
I have learned a lot from trying SBG and Specs. I think that a knowledgeable person can still see evidence of each in my work. I now have a clear statement of what the (previously nebulous) learning goals are (SBG). And I have much better language to describe what acceptable work looks like (Specs).
But I won’t use explicit promises about grade conversions anymore. Instead, I have described what typical achievement looks like for each grade in looser terms.
If you want to peek, go here.