I’ve become interested in an approach to grading that I have learned is called Specifications Grading. This tries to solve problems with the common approach of calculating module grades as the average across assessments, each rated 0-100%. The ones that I feel most strongly about are:
- Aggregating performance across several assessments results in numbers that are difficult to interpret. A student earning 90%+ probably achieved all of the module outcomes, a student earning less than 10% probably achieved none of them. A “pass” tells us nothing about whether a student understood our teaching.
- It’s not clear to students what they need to do to earn a mark that they want to achieve. To struggling students, all assessments can appear difficult because they describe the expectations of first-class students. To pass such assessments just means not failing to be first-class too badly. Even success feels like failure.
- Marks can be earned for half-heated attempts. Submitting anything for an assignment will probably earn you 20%. Teachers therefore mark and provide feedback on work that was only submitted to grab a couple more points. We can spend longer marking work, than students spending doing it.
I became aware that there are alternatives, firstly through the work of Prof. Jeremy Levesley who is attempting to clarify the pass/fail boundary for maths students through skills tests, then during our department’s periodic review in which it was suggested we introduce “driving test” style assessments to increase the possibility that students can earn a pass on a module by making the pass-level requirements explicit and accessible.
I’m currently reading Linda Nilson’s book on Specifications Grading. It’s fascinating in its analysis of the problems associated with rigour in universities as well as problems with existing grading schemes from the perspective of employers, students and teachers. I’m considering how I could incorporate specifications grading into CO2201, my project management and software engineering project module next year. I’ll write again as my thoughts on this develop.