How to use a practice test so it's worth the three hours
Most students take practice tests the way they weigh themselves: repeatedly, hoping the number moves. The number moves because of what you do between tests.
A practice test is a measuring instrument, not a training method. Sitting one produces data. The improvement comes from the two hours you spend afterwards working out what the data says — and most students skip that part, which is why the fourth practice test so often scores the same as the first.
Before: make the conditions real
- Full length, both sections, one sitting. A section sat on its own does not test the thing that breaks at question 40, which is concentration.
- Timed, with the real limits. An untimed paper measures your knowledge. A timed one measures your score.
- Phone in another room. Not face down. Another room.
- Take the 10 minute break, and take it the way you would on test day.
- Morning, ideally. The real test is in the morning, and if you have only ever practised at 10pm you do not know how you perform at 9am.
During: work the paper, don't just answer it
Strike out choices you rule out. Highlight the line that decides the answer. Flag what you guessed on. This is not busywork — it is the only record of your reasoning that will still exist in an hour, and it is what makes the review possible. A paper answered cleanly with no marks on it is a paper you cannot learn from.
After: the part that actually works
Leave at least an hour. Then go through every question you got wrong and every question you guessed on, including the guesses you got right — a lucky guess is a gap that the score hid from you.
For each one, write a single sentence: what did I believe when I answered? Not the correct answer. What you thought at the time.
Sort every miss into one of four buckets
- I did not know the content. → Study the topic. This is the only bucket that calls for learning material.
- I knew it but misread the question. → A reading habit, not a knowledge gap. Slow down on stems.
- I knew it and ran out of time. → A pacing problem. See where the minutes went.
- I had it right and changed it. → A confidence problem. Usually the cheapest points on the whole paper.
Count the buckets. Whichever is biggest is what to work on this week — and for most students it is not bucket 1, even though bucket 1 is what everyone studies.
How often?
Once every week or two while you are actively preparing. More often than that and you are measuring instead of improving; you will not have had time to change anything between tests, so the second score tells you nothing the first did not.
Between mocks, drill the specific topics the review identified. Twenty questions on one skill beats a whole section of mixed practice for fixing a named weakness.
Questions people ask
- How many practice tests should I take?
- One every week or two while preparing, with focused topic practice in between. Six to eight properly reviewed papers beats twenty unreviewed ones.
- Should I review questions I got right?
- Review the ones you guessed on or were unsure about. A lucky guess is a gap the score hid from you. There is no need to review questions you answered confidently and quickly.
- Do untimed practice tests help?
- They help you learn content. They do not tell you your score, and they do not train pacing. Use them deliberately, not by default.
2,079 questions, full-length adaptive mocks, and a report on how you worked.
Start practising — free, no account