Every assessment measures two things: the skill you meant to test and the format you chose to test it in. When a student struggles, you have to ask which one they struggled with. A timed multiple-choice exam on subnetting measures subnetting. It also measures reading speed, working memory under pressure, and test anxiety. If speed was never a course objective, the format is adding noise to the signal.
This is not an argument for making things easy. It is an argument for making them valid.
Start with clear criteria. Students should know before they begin what a strong response contains, in language a person can actually use to check their own work. “Demonstrates understanding” is not a criterion. “Explains why the packet was dropped and names the line in the config that caused it” is. The criteria protect students who cannot infer what I want from tone, and they protect me from grading on vibes.
Make feedback timely and actionable. Feedback that arrives three weeks later teaches nothing, because the student has moved on and cannot connect it to what they did. Feedback that says “good job” or “needs work” teaches nothing either. One or two sentences naming the specific next step, returned within the week, does more than a page of comments returned after the next assignment is due. I keep a comment bank of the twenty things I say most, which is how I keep the turnaround short.
Offer an equivalent response format where the format is not the skill. If the objective is “configure and verify a VLAN,” a two-minute screen recording with narration shows me that as well as a written report does, and for a student with dysgraphia or severe writing anxiety it may show me far more accurately. If the objective is “explain a design decision,” a written paragraph, a short recording, or a live conversation in office hours can all meet it. I decide what the objective is and then accept any format that demonstrates it.
But retain the required skill when the format is itself the objective. If the course outcome says students will present a technical recommendation to a non-technical audience, the presentation stays. Removing it would mean certifying a skill I never saw. What I can do is scaffold it: post the prompt early, allow a practice run, permit notes, and be clear about what is being scored. That is support, not substitution.
A last note on accommodations. Designing assessments this way reduces how often a student needs a formal accommodation, but it does not replace their right to one. Extended time, a distraction-reduced setting, and other approved arrangements come through the accessibility office and are not mine to negotiate. Good design and formal accommodation work together. One lowers the barrier for the whole room; the other guarantees the individual door stays open.
Leave a Reply