Rubrics and assessments beyond multiple choice

Multiple-choice quizzes are easy to build and easy to mark, which is why so many online courses rely on them. But they have a limit. They tell you whether a learner can recognize a good answer when they see one. They do not tell you whether the learner can produce that answer in real life, under real conditions, without a list of options in front of them.

If your course promises that members will be able to do something, such as write, handle a situation, make a product or solve a problem, your assessments should ask them to do it. That means going beyond multiple choice to tasks that show real skill, and it means using rubrics, which make judging those tasks fair, clear and manageable.

What multiple choice can and cannot tell you

Multiple choice is useful for checking knowledge: facts, definitions, the right sequence of steps, spotting the correct approach. It works well as practice and for quick checks during a course.

It struggles with anything that involves producing work, making judgment calls, communicating or performing a skill. A learner can pick the right answer about how to calm an upset customer and still freeze when an upset customer is in front of them. For those skills, you need to see the learner in action.

Assessment types that show real skill

Imagine a hypothetical membership that trains front-desk staff at dental practices, run by a former practice manager named Farida. Her course on handling difficult conversations uses several types of assessment:

• Performance tasks. Learners do the real thing and submit evidence. Farida asks them to record a role-play of a patient questioning a bill.
• Scenario responses. Learners read or watch a realistic situation and write or record what they would do and why.
• Work samples. Learners submit something they have made for real use, such as a rewritten reminder message for their own practice.
• Explain-it-back. Learners record a short explanation of a concept in their own words, as if teaching a new colleague.
• Self-assessment with evidence. Learners rate themselves against criteria and point to an example that supports each rating.

Scenario-based tasks pair well with teaching with case studies and real examples, since the same realistic cases can be used for teaching and for assessment.

Rubrics: what they are and why they help

A rubric is a scoring guide. It lists the criteria you are judging and describes what each level of quality looks like for each one. Instead of a vague overall grade, the learner sees exactly where they are strong and what to improve.

Rubrics help everyone involved. Learners know what good looks like before they start. Reviewers, whether you, an assistant or other learners, judge consistently. And feedback becomes faster, because much of it is already written into the rubric's descriptions.

There are two common styles. An analytic rubric scores each criterion separately, which gives detailed feedback. A holistic rubric gives one overall level based on a general description, which is quicker but less informative. For most course assessments, an analytic rubric with three or four criteria is the best balance.

A sample rubric

Here is part of Farida's rubric for the billing role-play. Each criterion has three levels.

Acknowledges the patient's concern

• Not yet: Moves straight to explaining the bill without recognizing the patient's frustration.
• Getting there: Acknowledges the concern with a general phrase, such as “I understand.”
• Confident: Names the specific concern in the patient's terms before explaining anything.

Explains clearly

• Not yet: Uses billing jargon or gives an explanation the patient would struggle to follow.
• Getting there: Explains accurately, but with more detail than the patient needs.
• Confident: Explains the charge in plain words, in two or three sentences, and checks the patient has understood.

Offers a next step

• Not yet: Ends the conversation without a resolution or next action.
• Getting there: Offers a next step, but it is vague, such as “someone will look into it.”
• Confident: Offers a specific next step, who will take it and when the patient will hear back.

Writing descriptions that work

The quality of a rubric depends on its descriptions. A few rules help:

• Describe what you can observe. “Names the specific concern” can be seen; “shows empathy” is open to interpretation.
• Avoid empty labels. “Good” and “excellent” tell a learner nothing. Say what the difference actually is.
• Keep levels few. Three or four levels are enough. More makes consistent judging harder.
• Use encouraging level names. “Not yet” suggests progress is coming; “Fail” does not.
• Test it on real work. Score two or three sample submissions. If you hesitate between levels, rewrite the description.

Always share the rubric with learners before they start the task, not just with their results.

Judging at scale

Performance tasks take longer to review than quizzes, but a clear rubric makes several approaches workable:

• Self-assessment first. Learners score their own work with the rubric before submitting it, which often leads them to improve it.
• Peer review with the rubric. Learners score each other using the same descriptions, as in peer review assignments that actually help.
• Selective expert review. You or a trained assistant review a sample, or every submission on a higher tier.

Rubrics are especially valuable for larger pieces of work such as final projects and capstones, where learners need clear expectations over several weeks.

Adding a skill-based assessment to your course

1. Choose one course outcome that describes something learners should be able to do.
2. Pick an assessment type that asks them to do it: a performance task, scenario, work sample or explain-it-back.
3. Write three or four criteria that matter most for that outcome.
4. Describe three levels for each criterion in observable terms.
5. Test the rubric on sample work and rewrite any unclear descriptions.
6. Share the rubric with learners along with the task.
7. Decide who will review submissions and how, then review the rubric after the first round.