Specifications grading?

In our most recent "how can we help you?" thread, a reader asks:

I have heard many teachers starting to use specification grading as a way of maintaining high standards without stressing the students out too much. Has anyone tried this and liked it/not liked it?

One reader submitted the following reply:

[T]he downside of this sort of approach is that students' answers can become very formulaic. They try to hit the points, if you know what I mean, and do not think creatively about their academic assignments. It perpetuates the worse vices of the American high school experience - teaching to the test. I hope this helps.

I honestly don't know that much about the practice, so I am curious to hear from people who have tried it on how they do it and how well they've found it to work!

Comments

[DS]

I haven't tried it, but people at my institution have, with mixed results. You can incorporate elements of specs grading into a class without fully embracing it. I would recommend checking out some of the literature. There are articles in the journal Teaching Philosophy by philosophers who have tried out specifications grading and how it went for them. There are also entire books on implementing specifications grading.

[Daniel Weltman]

I'm trying it this semester for the first time. I like it a lot but I don't feel confident enough saying much more because this semester isn't even over, so I don't have a lot of experience. I would say with respect to the reply about students becoming formulaic that if you write your rubric with care, I think this can be avoided. In fact you can include "assignment is not formulaic" as one of the points on the rubric, so long as you can effectively articulate what a formulaic assignment looks like. (If you can't, then your complaint is too inchoate for me to take it seriously!)

[Another DS]

I am pretty baffled by the submitted reply mentioned in the post. I have been using specifications grading for years and it no more 'teaches to the test' than most forms of teaching. '

Specifications grading does not need to involve setting specifications that are formulaic or rote. You can even have a specification for an assignment be connected to demonstrating creativity! Specifications are simply whatever you think will adequately demonstrate sufficient competency in skills and knowledge being assessed by the assignment.

I've generally had success with it and like to use it. However, it was a rocky start implementing it, and you need to have a very transparent teaching style with students. Partly because it can be confusing to them at first, and partly because it does not leave much room for 'fudging' behind the scenes (e.g. post-hoc grade curves). I usually devote the first class to explaining the grading approach and rationale behind it.

On my end, I am pretty satisfied with it. Rather than having a large portion of students muddle through with mediocre work, never improving, I have far fewer C students. Many that start there, eventually rise to the occasion and end up with Bs or As. One flipside of this is the stubborn muddlers do tend to end up with Ds and Fs, so grade distributions are often bimodal.

Students seem to like it too, once they understand it. I generally get at least a few comments on evals about feeling like it helped them be more comfortable making mistakes, partly due to not having to worry about points penalties and partly due to the opportunities for revision/retakes. Pedagogically, I consider those cases pretty clear wins!