A couple of days ago a posted a few recommendations for assessment, with a view to being practical. This list, retrieved from a now defunct blog, is on the other end of the spectrum—it is perhaps less pragmatic, but I think the ideas are quite important.
This is a list of propositions (not facts) on the topic of what makes for a robust educational assessment practice—it was originally produced to help structure a discussion about assessment in university music degrees. It is to some extent a very direct summary of principles boiled down from my exposure to Sadler’s work. To enact these principles properly is very hard in the current higher education environment.
A key point is that the traditional requirement for the use of grades (beyond an end-of-degree pass/fail) has a very strong flow-on effect on what comprises a basis for legitimate assessment.
1. If grades cannot be done away with, they must have integrity to satisfy ethical and legislative requirements
2. For grades to have integrity, standards must be robust within the context
3. To make that work there must be group consensus about those standards
4. For those standards to be useful over time, the group consensus must be maintained
5. For that to be possible, there must be an available bank of exemplars, consulted by the relevant team on a regular basis
6. Assessment of this kind does not apply to learning, it is an appraisal of quality of work completed, performed, and/or submitted
7. Grades are only valid for the work assessed, in that time and place, in which they were awarded—aggregating grades creates a spurious representation of cumulative learning and achievement