Specifications Grading
Nicole James, Chemistry
When I joined Reed last Fall, I spent a lot of time thinking about how I would ensure the feedback I provided gave students a clear sense of their standing, without traditional grades. After reading some literature and speaking with some friend-of-a-friends who used alternative grading methods, I took the plunge into specifications grading. In “specs” grading, the student’s grade is based on mastery of clearly articulated learning goals, with multiple opportunities for students to demonstrate mastery.
There are many ways to implement specs grading. For me, this meant looking over the course content and syllabus learning objectives, and writing a list of every single Thing The Students Should Be Able To Do, aka my learning goal list. For my introductory chemistry course, the full list has 72 items. I make it available to students on moodle; here’s a snapshot:
| Atom-atom interactions | |
|---|---|
| 1.5 | Construct an atomic-level explanation for why 2 isolated atoms attract upon approach and repel when too close, involving a graph of kinetic and potential energy as a function of internuclear distance |
| 1.6 | Explain what LDFs are and how they arise from electron densities |
| 1.7 | Predict/rank LDFs between atoms of different sizes |
| 1.8 | Relate strength of LDFs to relative melting/boiling points of noble gases |
| 1.9 | Explain how energy is transferred through atomic collisions, using a diagram/picture |
| 1.10 | Identify and describe nuclear reactions in terms of forces and potential energy |
| 1.11 | Differentiate between nuclear fusion, fission, and radioactive decay |
Each problem set and exam question is associated with one or more of these learning goals, and students receive feedback on their mastery of learning goals. Behind the scenes, these learning goals directly drive their course grade based on which and how many they mastered.
Last Fall, I gave narrative comments with a binary feedback “code”: 0 = not yet mastered, 1 = sufficiently mastered to be successful in future courses. I struggled with how this didn’t necessarily make clear to students if they were “doing well, but with some room to push farther.” Consequently, this upcoming year I’m piloting a 3-step system: 0 = not yet mastered; 1 = sufficient; 2 = excellent.
One of the things I particularly like about this approach is that it doesn’t penalize students for not knowing something early on, it simply recognizes and rewards what they have learned by the end of the course. If a student initially fails to demonstrate mastery of a learning goal, but as the class goes on their later work does demonstrate mastery, that overwrites their earlier learning goal record. In this way, the grade at the end of the term reflects what students know and can do, without “averaging in” earlier mistakes made as they were learning.