The section of American Constitutional Foundations that I’m teaching in the fall will be my first fully online course. If you've done any reading about online and blended learning, you’ll have heard talk about the sizable amount of work that goes into shifting to this new format. For me, though, the biggest change and the main focus of my development time right now has to do with grading and assessments. After twelve years of teaching with a fairly standard grading structure, I’ve decided to try something new: specifications grading (also known as “specs grading”).
The system of specifications grading owes its existence to Linda Nilson’s book of the same name, subtitled “Restoring Rigor, Motivating Students, and Saving Faculty Time.” In it, Nilson lays out a grading system that represents a radical departure from how most faculty assess student work. In my opinion, the best way to get a handle on Nilson’s argument is with her description of the ideal grading system (you can read a summary of her 15 ideal features here).
I’d imagine that for most faculty, grading systems are an afterthought. While they may tweak the details of their particular assessments over time, they generally use the same grading structure they always have. Personally, I’ve always used a points-based system. Every gradable item in the course, from attendance and participation down to exams, has a point value attached to it. The relative weight of each assessment is tied to its point value. Students earn the full or a partial amount of points on each assessment and then, at the end of the semester, their total number of points is divided by the total number of points possible in order to arrive at their final letter grade.
The problem with this system is that it may have almost nothing to do with what students actually learn. At the end of my class, if a student has received a B, what does that mean? If the minimum percentage of points required to earn a B is 84%, does that mean the student is 84% of the way towards what I hoped they would learn? Does it mean they mastered 5/6 of the content but failed at 1/6? What if participation and attendance are worth 16% of their grade and they skipped every class session while completing everything else perfectly?
Despite the absolute hegemony of percentages and letter grades in education, they communicate almost nothing of value about student learning. And that’s only one of fifteen points that Nilson makes. Not to downplay the rest of the list, but for me the disconnect between grades and learning is the most troubling way in which traditional grading systems fall short of the grading ideal—and is reason enough to reconsider how I grade my students.
Recognizing a problem is the not the same thing as agreeing to a solution. The rest of Nilson’s book, however, convinced me to give specs grading a try. Describing the solution she offers is a bit difficult, because the final shape of a specs grading system can vary a lot in the implementation. At the heart, though, are the following principles:
Nilson isn’t wedded to Bloom’s Taxonomy, and she does outline a number of different systems for categorizing cognitive achievement. But the foundation for the rest of the system is clarity about where students will be at the end of a successful learning process. I mentioned this briefly in my previous post in this series.
Once you’ve identified the assessments you’ll use to measure students’ mastery of learning outcomes, it’s important to provide a complete and robust description of what qualifies as acceptable work. For those familiar with rubrics, these assessment specifications will seem like rubrics on steroids.
In a traditionally graded course, I would likely use a rubric to determine what percentage of points a student’s work deserved. In specs grading, the list of specifications serves to define a line of acceptable performance that a student either does or does not meet. A few notes on this point. First, meeting all an assignment’s specs does not mean that a student has to do absolutely perfect work; Nilson suggests that you write an assignment’s specs targeting the type of work that you would deem, informally, worthy of a B. Second, while pass/fail grading may seem unusually harsh, Nilson makes a good argument for how the availability of partial credit on an assignment serves to deter students from submitting quality work.
To counter the fear that pass/fail grading may seem too draconian, specs grading encourages instructors to build-in opportunities for students to revise their work in order to show that they have in fact mastered an outcome. Nilson talks about giving students a fixed number of “tokens” that can be redeemed for opportunities to revise assignments (among other purposes). Robert Talbert, a specs grading early adopter, uses the “making progress” grade to allow some students free passes at revision when their work is close but not quite at the acceptable level. Tokens and free passes are only part of the story, though. Nilson argues that simplifying the grading process to pass/fail will allow faculty to spend more time providing helpful, meaningful feedback that points students towards improving the quality of their work.
Since most institutions won’t allow instructors to dispense with letter grades altogether, Nilson describes two general strategies for how to map students’ work to a final grade. On the one hand, higher grades can be connected to students completing more assessments (which would reflect reaching mastery on more learning outcomes). On the other hand, higher grades could be tied to higher levels of performance or completion of more complex and challenging work (e.g. assessments that tap the higher levels of Bloom’s Taxonomy). My guess is that most faculty who implement specs grading will use a combination of the two strategies.
Several of Nilson’s attributes of the ideal grading system point out how students feel disempowered by the status quo. While specs grading is designed to be more rigorous, it is also fundamentally supportive of learners. Faculty are encouraged to build choice into the assessments that students must complete. Authentic and personally meaningful assessments are encouraged as well. The final outcome of a specs grading system is also presented to students as a choice. Rather than complete all the work for a course and then wonder how that work will be appraised, students can identify in advance what grade they would like to receive and then can work towards that goal from a clear and specific set of specifications of what the grade will entail.
There’s a lot more that I want to say about specs grading. For many, the system will feel strange and perhaps even untenable. For those folk I want to share some of my own thoughts about the system, including some of my fears going forward. I also want to lay out how I’ll be implementing specs grading in my own course. In the mean time, if your curiosity has been piqued I would encourage you to get a copy of Nilson’s book and to also read my colleague Robert Talbert’s blog posts about his own adoption of specs grading.