How to Create a Practice Test from a Lecture or Training Video

How to create a practice test from a lecture or training video

To create a practice test from a lecture video, pull the testable claims out of the recording, write two to three practice questions for every question the real exam will have, mark the answer key, attach a short explanation to each item, and tie every question back to the timestamp it came from so a wrong answer tells the learner exactly what to rewatch. You can do this by hand in roughly one to three hours per module, or generate the draft from the video and spend the time reviewing instead.

Why the practice test is the one that gets cut

Ask anyone who runs a course or a training program where their assessment time goes and you will hear the same thing: the graded exam gets written, and the practice test does not. That is not laziness. It is arithmetic. A practice test only works if learners can take it more than once, and that means writing three or four times as many questions as the graded exam needs. Nobody has an afternoon per module to spend on the low stakes version of a thing they already built.

So the practice test gets replaced by a slide deck and the instruction to "review the material." Which is the worst possible substitute, because rereading is the study method that feels the most productive and works the least well.

The research is unusually clear on this

Retrieval practice, often called the testing effect, is one of the better replicated findings in learning science. Pulling an answer out of memory strengthens that memory more than rereading the same material for the same amount of time. The effect gets larger when the practice test gives feedback rather than just a score, and larger again when attempts are spaced out over days instead of stacked into one session.

Three design consequences fall out of that, and they are the whole reason this guide exists:

• It has to be effortful. A practice test easy enough that nobody has to retrieve anything is a reading exercise wearing a quiz costume.
• It has to explain. A bare score is close to useless. "You got 14 out of 20" tells a learner nothing they can act on.
• It has to be retakeable. Which means more than one form, or the third attempt just measures whether someone remembers that option C was right.

Step 1: decide what the test is actually for

Practice tests and graded exams are different instruments and mixing them up causes most of the downstream pain. A graded exam is a measurement taken once, and it usually needs security, controlled conditions, and a score you can defend. A practice test is a study tool: unlimited attempts, immediate feedback, no consequence for being wrong.

The mistake worth naming: do not use your graded exam as the practice test. The moment learners have seen it, you have spent it, and now you need to write a replacement under time pressure. Generating a separate practice set from the same source material is far cheaper than rebuilding an exam you burned.

Step 2: mine the video for testable points

If you are doing this by hand, watch the recording with a document open and write down every claim that a learner could reasonably be asked to recall or apply. Not every fact, every testable fact. The distinction matters: a lecture contains hundreds of statements and perhaps twenty things that a competent learner should be able to retrieve a week later.

Note the timestamp next to each one as you go. This feels like busywork and it is the single highest leverage thing in the process, for the reason covered in step 5.

The honest problem with doing it manually is coverage. Attention fades, so questions cluster in whatever was covered in the first fifteen minutes, and the back half of a two hour training session ends up barely tested. If you have ever noticed that your own quizzes skew early, that is why.

Step 3: write questions that discriminate

A good multiple choice item separates people who understood from people who did not. A bad one separates people who are good at multiple choice from everyone else. The rules that matter most:

• The stem should stand alone. A learner should be able to read the question and know what is being asked before seeing the options.
• Distractors should be plausible. Wrong options that are obviously silly turn a four option question into a two option coin flip.
• Avoid trick phrasing. Testing whether someone noticed the word "not" is testing their reading speed, not their knowledge.
• Keep option lengths similar. The longest option being correct is the oldest tell in the business and test takers know it.

For a practice test specifically, err toward slightly harder than the real exam. Rehearsal that is easier than the performance builds false confidence, which is the exact failure mode you are trying to prevent.

Step 4: write the explanations

This is the step that turns a quiz into a study tool, and it is the step most often skipped. Every question needs one line explaining why the correct answer is correct. Not a citation, not a paragraph, one sentence that closes the loop while the learner is still thinking about it.

Feedback delivered immediately, at the moment of being wrong, is worth several times the same information delivered in a review session next week. This is not a small effect and it is the main reason a practice test beats a practice reading.

Step 5: tie every question to a timestamp

Here is where a score becomes a study plan. If question 7 carries a timestamp of 12:40, then getting it wrong does not mean "review the module." It means "watch 40 seconds starting at 12:40." That is the difference between feedback a learner acts on and feedback a learner nods at.

It also protects you. When someone disputes a question, a timestamp lets you check the item against the exact moment it came from in seconds rather than rewatching the recording hunting for the claim.

Step 6: build a second form

A working rule of thumb is two to three practice items per real exam item. For a twenty question exam, a bank of forty to sixty questions lets someone attempt it three times without seeing the same set twice.

That ratio is precisely why hand written practice tests do not happen, and why generating them changes the calculus: the marginal cost of a second and third form drops to seconds. This is the same reason learners preparing for competitive exams lean on unlimited mock tests rather than a single past paper. Volume is what makes rehearsal honest.

Doing it from the video automatically

If your source material is a recording, the drafting step is the one worth handing off. VidQuiz is a practice test maker that reads the video itself: paste a lecture capture, course module, compliance training, or webinar link, and it writes multiple choice questions with four options, the correct answer marked, a one line explanation, and a timestamp chip on every item. You do not transcribe anything first.

Coverage is the underrated part. Because it reads the whole recording, questions spread across the full two hours instead of clustering where your attention was still fresh. Run a second pass over the same video and you get a different set drawn from the same content, which is your second form.

The caveat stated plainly: what comes back is a draft written from your material, not a validated instrument. Its difficulty has not been calibrated and it has not been piloted. For a low stakes practice test that trade is easy, because the cost of a mediocre practice question is small and the learner sees the explanation regardless. Read it before it counts for anything. Every question is editable before you share.

Step 7: deliver it where learners already are

Share a browser link, or export the set into whatever already runs your assessments: a Google Forms ready CSV, QTI for your LMS, PDF for a paper practice exam, or plain CSV. The drafting tool and the delivery platform do not have to come from the same vendor, and for most teams they should not.

One practical note on cost. Some quiz platforms meter by active quiz taker, which means a practice test that becomes popular gets more expensive precisely because it worked. If you are choosing a tool for practice specifically, check whether unlimited retakes by a full cohort is priced as a feature or as a bill.

How long this takes

By hand, budget one to three hours per module once you count rewatching the recording, writing the items, and building the answer key. Generated, the draft takes seconds and the review takes as long as reading the questions, usually a few minutes for a twenty item set.

The compounding win is at course scale. A ten module course needs ten end of module checks, and that is exactly the point where hand writing practice tests quietly stops happening and modules ship with no rehearsal at all. Turning each course video into a quiz is what keeps that from happening.

The short version

Pull the testable claims from the recording with timestamps attached. Write two to three practice items per real exam item, harder rather than easier, with plausible distractors. Explain every answer in one line. Build a second form so retaking means something. Then deliver it wherever your learners already work, and make sure a wrong answer points at 40 seconds of video rather than at a whole module.