How to Write a Continuing Education Post-Test That Passes Review

A continuing education post-test has to do two things at once: measure whether the learner met the stated objectives, and satisfy the accrediting body that the assessment was tied to those objectives rather than bolted on afterward. Most rejected post-tests fail on the second point. The questions are fine in isolation, but nothing connects them back to what the activity said it would teach.

If you provide CE for nurses, CPAs, engineers, social workers, or any of the other licensed professions in the US, the mechanics differ by accreditor but the underlying logic is remarkably consistent. This walks through that logic and the specific mistakes that cause a post-test to come back for revision.

Start from the objectives, not the content

The order matters more than anything else here. Write the learning objectives first, make each one measurable, then write questions that measure exactly those objectives. If you write the course, then the test, then reverse engineer objectives to fit, the mismatch is visible to anyone reviewing it, and it is the most common reason a post-test gets sent back.

Measurable is the operative word. An objective that begins with understand, appreciate, or be aware of cannot be tested, because there is no observable behavior attached to it. Rewrite it around a verb you could actually score.

A practical check before you write a single item: put your objectives in a column and the question numbers beside them. Every objective needs at least one question. Every question needs to map to an objective. An orphan question, one that tests something interesting from the content but corresponds to no stated objective, is the thing a reviewer circles.

How many questions should a CE post-test have?

Scale it to the credit hours. A widely used working ratio in CE is a small number of items per contact hour, commonly around five to ten, with a minimum that gives a defensible pass mark. The reason is arithmetic as much as pedagogy: on a five question test, one wrong answer costs 20 percentage points, so an 80 percent pass mark means a single misread question fails an otherwise competent learner. Below ten items, your pass threshold becomes very coarse.

Check your specific accreditor's requirements before settling on numbers, because several set their own minimums and a few specify the pass mark outright. The ratios above are the general shape, not a substitute for reading your provider agreement.

What pass mark should you set?

Somewhere between 70 and 80 percent is the norm across US CE, with 80 percent the more common requirement where the accreditor specifies one. Set it in your policy, apply it identically to everyone, and document your retake rule at the same time. The retake question tends to get decided ad hoc the first time someone fails, which is exactly when you least want to be inventing policy.

If you allow unlimited retakes with the same ten questions, you have built a test that everyone eventually passes by elimination, and you should expect a reviewer to say so. Two reasonable fixes: cap the attempts, or hold a larger pool and draw a different subset each attempt. The pool approach is better learning and better evidence, and once you have a question bank built it costs nothing per attempt.

What makes a CE question weak?

The failure modes are consistent, and most of them are visible on a quick read of the item without knowing the subject matter at all.

• The answer is in the stem. Wording that echoes the correct option lets a learner match text without knowing the content.
• Throwaway distractors. If two of the four options are obviously absurd, it is a coin flip dressed as a four option item. Every distractor should be something a person who half learned the material might genuinely pick.
• Trivia over application. Testing the publication year of a guideline instead of what the guideline tells you to do on Monday morning.
• All of the above. Recognizing two correct options gives the answer away, and it rarely measures anything you wanted to measure.
• Negatives and double negatives. Which of the following is not an exception to the rule tests reading comprehension under time pressure, not clinical or professional judgment.
• Two defensible answers. The one that generates complaints. It usually appears when the source material states something loosely and the item writer did not notice.

The single highest yield habit is writing the distractors from real misconceptions. If you teach this material, you already know the three wrong things people believe about it. Those are your distractors, and an item built that way tells you something diagnostic when learners miss it.

Does the post-test have to be tied to a specific revision of the content?

Practically, yes. When the activity content is updated, the post-test needs to be reviewed against it, and your records should make clear which version of the assessment a given learner completed. This is the same discipline as version controlling any other controlled document. It matters most when guidelines change mid cycle and a question that was correct in January is wrong by September, which happens more often than most providers plan for.

How do you build the item pool without spending a week on it?

Generating a first draft from the activity material itself is far faster than authoring from a blank page, and the subject matter expert's time is much better spent editing than composing. A workable sequence:

1. Finalize the measurable objectives.
2. Upload the handout, slide deck, or manuscript to a continuing education quiz generator and generate more items than you need, roughly double.
3. Have the subject matter expert cut anything that tests trivia, fix any item with two defensible answers, and rewrite weak distractors into real misconceptions.
4. Map the survivors to objectives and confirm every objective is covered.
5. Keep the extras as the pool for retakes and for the next cycle.

Generating double and cutting half is the part people skip, and it is what makes the difference. Editing down from a surplus produces a much better test than trying to make exactly twenty generated items work, because you can afford to discard the mediocre ones.

Do learners need to see which questions they missed?

Feedback after the attempt is good educational practice and some accreditors expect it. The tension is that showing the full answer key immediately makes any retake meaningless. A common compromise: report which objectives the learner missed and point back to the relevant section of the material, rather than revealing the correct option item by item. That preserves the retake, and it sends the learner back to the content, which is the point of the exercise.

The short version

Write measurable objectives first, one to two items per objective, scale the item count to the credit hours so your pass mark is not absurdly coarse, build distractors from real misconceptions, and keep a pool larger than the test so retakes are not a memory exercise. Review the test whenever the content changes, and record which version each learner completed. Most of the work is in the objectives, and if those are right the questions largely write themselves.