After a few weeks of practice, most candidates discover that their three Core scores are not level. One subtest feels almost comfortable, one is workable, and one is a reliable disappointment. The advice people hear at this point is simple: work on your weakest area. That advice is often right, but not always, and following it blindly can absorb weeks in exchange for very little. A weakness caused by not knowing a method is worth attacking. A weakness caused by running out of time looks identical from the outside but responds to a completely different fix. This article sets out how to tell which one you have, and how to divide your remaining study time between shoring up the floor and protecting what already works.
An uneven profile is the normal case
The Core module asks for three genuinely different things. Figure Sequences is visual rule detection, Mathematical Equations is mental algebra, and Latin Squares is constraint management — the structure is set out on our syllabus page. These draw on habits you have been building for years without meaning to, so arriving with one strong subtest and one weak one is the expected outcome, not a sign that something has gone wrong with your preparation.
What makes the balance question worth taking seriously is that the subtests sit inside one module. The official sources describe results at module level rather than publishing a separate mark for each subtest, so the three trade against one another instead of standing alone. Confirm the current reporting format on those sources before you plan around it. In practice this means you are not trying to be good at all three. You are trying to make the module total as high as your remaining weeks allow, which is a different and more forgiving target. Our guide to reading your score and percentile covers how that total is expressed.
Separate a technique gap from a speed gap
Before you reallocate a single hour, find out what kind of weakness you actually have. It takes one session and is worth more than a fortnight of guessing.
Take a set of your weakest subtest and do it untimed. Work carefully, take as long as you want, and mark it honestly.
- Accuracy is poor even untimed. This is a technique gap. You do not yet have a reliable first move, so each item is being solved from scratch by improvisation. This is the weakness that rewards concentrated work, and it is the case where "spend more time on your worst subtest" is straightforwardly correct.
- Accuracy is fine untimed but collapses under the clock. This is a speed gap. You know how to do the questions; you cannot yet do them fast enough. Grinding more question sets will not fix this, because the thing that is broken is pacing and fluency, not understanding.
The two fixes barely overlap. A technique gap wants a repeatable opening move for each task type, which is what our Core module preparation guide lays out subtest by subtest. A speed gap wants short timed sets and a firm rule for abandoning a question, which is the ground covered in the guide to pacing under pressure.
There is a third possibility worth naming. Because the Core subtests are designed to be solved in your head rather than on paper, a subtest can look like a reasoning weakness when it is really a holding-things-in-mind weakness — you lose the partial result before you can use it. If that description fits, training working memory will do more for you than another block of practice questions.
Where the next hour is worth most
Once you know what kind of gap you have, the allocation follows a simple principle: early points are cheap and late points are expensive. Moving a subtest from poor to competent usually means learning two or three methods and applying them consistently. Moving one from strong to excellent means chasing the handful of hardest items in the set. The first costs less and returns more, which is why the weakest subtest is usually where the next hour belongs.
That principle has a limit, and the limit is the reason "always work on your weakest" fails some people. A strength is only a strength if it survives the clock. If your best subtest is best only when you have time to spare, it is not banked, and neglecting it for a month will quietly cost you more than the weak subtest gains. So the split is a majority, not a monopoly. Give the weak subtest the larger share of new learning, and give each strong subtest a short maintenance run — one timed set, reviewed properly — often enough that it stays warm.
Let your error log decide the exact ratio rather than your instinct about which subtest you dislike. Instinct systematically over-weights the subtest that feels unpleasant, which is not always the one leaking the most marks. A timed run of the free dMAT practice test under realistic conditions gives you a cleaner profile than a series of isolated sets, because it shows how each subtest performs when you are already tired.
When to leave a weakness alone
Time changes the answer. If you have weeks to work with, rebuilding a weak subtest's technique from the ground up is a sound investment. If the exam is close, it usually is not: a method learned in the final days tends to be the first thing that deserts you under pressure. Late on, the better play is consolidation — sharpen what you already do reliably, keep all three subtests in rotation so none goes cold, and accept the profile you have. Our guide to the final week sets out what that looks like in practice.
Leaving a weakness alone is not the same as abandoning it. Even a subtest you have given up improving still contributes to the module, and walking into it having not touched it for a month will cost you marks you would otherwise have kept. Maintenance is cheap. Reconstruction is not.
The short version
Diagnose before you reallocate. Do an untimed set of your weakest subtest: poor accuracy means a technique gap worth real investment, while accuracy that only fails under time means a pacing problem that more questions will not solve. Give the weak subtest the larger share of your new learning and every strong subtest a short regular run to keep it live. And as the date approaches, stop rebuilding and start maintaining — a balanced, warm profile beats a lopsided one that has been half-repaired.
