How to review IMAT mistakes: build an error log that helps

An error log should change your next study session. If it only stores question numbers and the correct letters, it records what happened without explaining what to do about it.

Before opening the solution, reconstruct the decision

Write one sentence about the rule or calculation you used. If you guessed, say so. If time expired before you reached the question, record “not reached”; that is different from trying and failing to solve it.

This step preserves the evidence you are about to lose. Once you see the answer, a solution can look obvious even though you could not produce it independently.

Then compare your reasoning with a trustworthy explanation. Find the earliest step where they diverge, rather than only the final arithmetic difference.

Match the repair to the cause

The categories below are a practical sorting tool. An answer can have more than one cause; record the one you need to address first.

Cause	Evidence to look for	A repair to try
Knowledge gap	You cannot state the relevant principle even without a timer	Learn the idea and its prerequisites, then explain it without notes
Misconception	You can state a rule, but it is the wrong rule for this situation	Write the contrast between the two ideas and test both
Reading error	Your work answers a different quantity or misses a condition	Restate the requested quantity and condition before solving
Execution error	The method is sound, but units, signs or algebra fail	Practise that specific operation with a short check at the end
Time pressure	You can solve an unfamiliar equivalent calmly but repeatedly stall under timing	Try a pacing change in a timed set

Do not label every incorrect calculation “careless”. That label cannot distinguish an arithmetic slip from a method you never understood. For recurring slow answers, use the separate time-management guide.

A worked example: a chemistry answer with a maths cause

Suppose a practice exercise gives the relation pH = −log₁₀[H⁺] and asks you to use [H⁺] = 10⁻³ mol/L. You answer −3.

The useful correction is not “the answer is 3”. It is: “I found log₁₀(10⁻³) = −3, but forgot the minus sign outside the logarithm.”

Your next check could use 10⁻⁵ instead. Explain both steps before writing 5. Then change the coefficient as well, using the logarithms for pH guide, so the check is not just a memorised pattern.

This is an original learning example, not a reproduced exam question. The exercise supplies the concentration approximation so the review can isolate the sign error.

Use a short log you will actually revisit

For that example, one entry might be:

Field	Entry
Topic and question reference	Logarithms; practice set A, question 4
My reasoning	I stopped after finding the logarithm
Cause	Missed the outer minus sign
Repair	Write the logarithm result first, then apply the negative sign
Later check	Solve a changed example without notes and explain the sign

Keep the reference so you can return to the question. Keep the correction short enough that the important contrast remains visible.

Review correct guesses too

A correct answer with weak reasoning still deserves attention. Note your confidence before checking the answer; otherwise knowing you were right can change your recollection of how certain you felt.

Butler, Karpicke and Roediger found that feedback also improved later retention of low-confidence correct answers. That finding supports checking uncertain successes as well as failures; it does not mean every correct response needs a long written analysis.

Close the log and test the repair

After reviewing, explain the corrected idea without the solution visible. Return later with a different example that requires the same distinction.

Roediger and Karpicke found benefits of retrieval testing for delayed retention in their experiments. Applying that principle here means attempting the reasoning again, not repeatedly admiring a clear explanation.

If the error returns, reopen the entry and revise the diagnosis. If it does not, reduce its priority while keeping it available for occasional review.

The mistake-review traps

Copying the whole solution: it can conceal the exact step you need to repair.

Repeating only the same question: recognising its answer is weaker evidence than solving a changed example.

Calling everything a knowledge gap: more reading will not necessarily fix a unit conversion or pacing habit.

Logging everything and reviewing nothing: choose a small number of recurring problems to address before the next mock exam.