Skip to content

Study methods

How to use practice questions without memorising the answers

Practice questions are the most efficient way to prepare for a multiple-choice exam, and also the easiest way to fool yourself. Answer the same set a few times and your score climbs because you recognise the questions, not because you understand the topic. On exam day the wording is different and the score falls back. This guide explains how to tell the two apart and seven habits that keep practice honest.

By Ibrahim YesilayUpdated 6 min read

Key takeaways

  • A rising score on questions you have seen before measures memory of the questions, not readiness.
  • Track first-attempt accuracy on new questions; it is the number that predicts the exam.
  • Say the answer and the reason before you look at the options, then check both.
  • Read why every wrong option is wrong; distractors are built from the misconceptions you need to remove.
  • Avoid anything claiming to be real exam content: the Linux Foundation agreement prohibits disclosing it, and it trains recognition, not understanding.
On this page

Recognition is not understanding

When you see a question a second time, your memory of the question helps you answer it. That is recognition, and it is fast and easy. The exam asks about the same topics with different wording, different scenarios and different distractors, so recognition does not transfer. Understanding does: if you know why the answer is right, you can answer a new question about the same idea.

SignRecognitionUnderstanding
How you pick the answerIt looks familiarYou can state the rule it follows
Wrong optionsIgnoredYou can say why each is wrong
A reworded questionAccuracy dropsAccuracy holds
A week laterMostly forgottenMostly retained, faster with spaced reviews

Habit 1: answer before you look at the options

Read the stem, cover the options and say the answer and the reason. Then look. If your answer is there, check that the reason matches the explanation. If it is not, you have found a gap that a multiple-choice format would otherwise hide. This is the single most effective change most people can make, because it turns recognition into recall.

Habit 2: track first attempts, by domain

Keep one number per domain: the share of questions you answered correctly the first time you saw them. That number predicts the exam far better than your overall score, because it excludes repeats. Domain weights matter here too: a weak domain worth a large share of the exam is the most common way to fail.

  1. Count only questions you have never seen before.
  2. Record the result by the official domain of the exam.
  3. Compare each domain with the pass mark, not with your average.
  4. Spend the next session on the lowest domain that also carries a large weight.

Habit 3: read every option, not just the answer

A good wrong option is built from a real misconception: a field that exists on a different object, a default that changed, a term that sounds close. Reading why each one is wrong teaches you the boundaries of the idea, which is exactly what the next reworded question tests. When an explanation only covers the right answer, write one line for each wrong option yourself.

Habit 4: turn every miss into a note, then a repeat

A miss is the most useful thing that can happen during practice, because it points to a specific gap. Write one line about the concept, not the answer: "Preconditions run after match" is useful; "the answer was B" is not. Then make sure the question, or another question on the same idea, comes back after a delay. Spaced repeats, an hour later, a day later and a few days later, are what make the fix last.

Habit 5: mark guesses as misses

A guess that lands on the right answer looks like knowledge in your score but leaves the gap in place. Mark every question where you hesitated between two options, read its explanation as if you had missed it and let it come back for review. Your score will look lower; your exam score will be higher.

Habit 6: build something for the hard topics

Some topics stay abstract until you have done them: a NetworkPolicy, a PromQL aggregation, a Kyverno rule, a GitOps reconciler fixing drift. For those, a short hands-on exercise on a local cluster turns a fuzzy idea into a memory of something you did. Multiple-choice options are much easier to separate when you have seen the real behaviour.

Habit 7: keep timed mocks for measurement

A timed mock in the published format, 60 questions in 90 minutes for most CNCF multiple-choice exams, measures pace and readiness. Use one in the middle of your preparation to find weak domains and one near the end to decide whether to book. Between them, practise untimed and slowly, because that is when you learn. Repeating mocks back to back mostly trains recognition of the mock.

On GoldenKube, the smallest multiple-choice bank is CBA with 126 questions, and every bank is large enough to keep a set of unseen questions aside for those readiness checks.

What not to do

A few common shortcuts make practice scores look better without making you more ready.

  • Retaking the same full set until you score 100%.
  • Reading the answer first and the question second.
  • Skipping explanations for questions you got right by luck.
  • Using material that claims to be real exam questions: the Linux Foundation agreement prohibits disclosing or reproducing exam content.
  • Measuring readiness on an overall score that mixes new and repeated questions.

A 30-minute practice session

The habits above fit into a short daily session. Thirty minutes done most days does more than a long session once a week, because the repeats of missed questions need days to pass between them.

If a session goes badly, that is useful information, not a failure. A low first-attempt score on a domain tells you exactly where the next few sessions should go.

  1. First five minutes: clear any questions due for review, answering each before looking at the options.
  2. Next fifteen minutes: new questions from your weakest domain, one at a time, with the reason said before checking.
  3. Next eight minutes: read the explanation of every miss and guess, and write one line about the concept behind it.
  4. Last two minutes: note the session’s first-attempt accuracy for the domain and pick tomorrow’s focus.

Getting the most out of explanations

An explanation is only useful if you read it actively. Before you read it, predict what it will say. After reading, close it and restate the rule in your own words. If the explanation cites a documentation page, open it once for anything you got wrong twice: the documentation usually explains the idea more fully and shows the surrounding context the question was built from.

  • Predict the explanation before reading it.
  • Restate the rule without looking.
  • Open the cited documentation for repeat misses.
  • Write the rule, not the answer, into your notes.

Frequently asked questions

How do I know if I am memorising answers?

Compare two numbers: your accuracy on questions you have seen before and your accuracy on questions you have never seen. If the first is much higher, you are recognising questions. Another sign is that you remember the position or the look of the right option but cannot say why the others are wrong.

Is it bad to repeat questions?

Repeating a question you missed is useful if you can explain the answer the second time, and spaced repeats over days help the knowledge last. Repeating a whole set until the score is perfect is not useful, because the score then measures the set, not you.

What score on practice questions means I am ready?

Use first-attempt accuracy on new questions, by domain. When every domain stays above the exam pass mark (75% for most CNCF multiple-choice exams) on questions you have not seen, across more than one session, you are close. A high score on repeated questions says little.

Are exam dumps a shortcut?

No. The Linux Foundation certification agreement prohibits disclosing or reproducing exam questions and answers, so material claiming to be real exam content is a breach of that agreement, and it trains exactly the recognition this guide warns against.

How many practice questions should I do?

There is no official number. What matters is enough new questions per domain to measure first-attempt accuracy, and enough spaced repeats of your misses to fix them. A few dozen questions a day with careful review usually beats a few hundred skimmed.

Should I read the explanation when I got the answer right?

Yes, if you were not sure. A lucky guess counts as a correct answer but leaves the gap in place. Mark guessed questions and treat them like misses.

Do these habits work for hands-on exams too?

The ideas transfer. For hands-on exams, the equivalent of answering before looking is doing a task without notes, and the equivalent of first-attempt accuracy is the share of new tasks you finish correctly within the time. Repeating the same lab until it is fast trains recall of that lab, not the skill.

Sources

Every exam fact on this page comes from these pages, read on the date shown.

Independent community project, not affiliated with or endorsed by CNCF or The Linux Foundation.