Testing Against Borderline Cases

Once you have a draft rubric, test it against borderline cases—work that's good but not excellent, or that does some things well and others poorly. This is where the rubber meets the road. If your rubric can distinguish between excellent and good-but-not-excellent work, it's actually useful. If it treats them the same, your criteria need to be more specific.

Take a proposal that's strategically sound but poorly communicated. Does your rubric catch that? Take a design that has beautiful visual hierarchy but where the design choices don't actually reinforce the message. Does your rubric catch it? If it doesn't, you need to refine the criteria.

Also test it by applying it to your own work—work you've done that you're proud of, and work you know is adequate but not excellent. Can your rubric explain the difference to someone who didn't produce the work? If it can, you're getting somewhere.