Review system guide
A Forty-Minute Protocol for Reviewing a Mock
Applicants who finish a practice set, glance at the score, and immediately start another set without extracting anything.
What this guide gives you
- Review time should roughly equal attempt time for a short set.
- Classify every item, including the ones you answered correctly by luck.
- End with one rule change and one scheduled retake, not a list of eleven wishes.
Why the review is the actual practice
Taking a practice set produces data. Reviewing it produces skill. Candidates who take six sets without reviewing have six measurements of the same ability plus six sets of memorised items. Candidates who take two sets and review them properly have two measurements and a corrected method. The second group improves faster with less work, which is why review is not an optional courtesy at the end of practice.
The asymmetry is easy to explain. During an attempt you are producing answers under time pressure, which means you are executing your current habits. During review you are allowed to inspect those habits and change one. Change happens in review; attempts only reveal what needs changing.
Forty minutes is enough for a ten-item set if you work in a fixed order. Without an order, review becomes an anxious re-reading of the whole set that ends with a resolution to try harder. The protocol below replaces that with five bounded steps.
Step one: reopen every item, not just the misses
Start with the items you answered correctly. For each, state the rule or reason you used. If the reason is sound, the item is banked. If the reason is wrong, invalid, or missing, the item goes into the review pile even though it was correct. These lucky corrects are the most dangerous items in any attempt, because they inflate the score and hide the gap.
A practical test for a lucky correct: could you explain the answer to someone else in one sentence without referring to the options? If you cannot, mark it. Most candidates find one or two lucky corrects in a ten-item set, which is enough to change how the score should be read.
Also reopen items you answered slowly. Slow corrects are not failures, but they consume budget. Write the time beside them. A pattern of slow corrects in one topic tells you where the pacing risk lives, even though nothing was marked wrong.
Step two: classify, step three: rewrite the rule
For every item in the review pile, assign one cause: knowledge, misread, method, or time. Do not assign two. If two seem to apply, choose the one that came first in the chain, because that is the one that decided the outcome. A misread followed by a wrong computation is a misread.
Then write the corrected rule as a single imperative sentence. Not review figure sequences, but check parity before checking rotation. Not be careful with units, but write the unit next to every letter before substituting. The sentence must be small enough to execute in the next attempt without conscious effort.
Keep the sentences. The set of rule corrections that recurs across attempts is your personal checklist, and it is more useful than any generic strategy list because it was derived from your own failure modes. Three recurring sentences is a typical result after a week of honest review.
Step four: fix the timing story
Look at the clock data before you look at the score. Were items abandoned? Did the last three items get less than half the average time? Did one item consume a fifth of the section? Each of those is a pacing symptom with a specific fix, and none of them is fixed by studying more content.
If you abandoned items, the fix is the three-pass method and a written exit rule. If the final items were rushed, the fix is checkpoint discipline. If one item consumed a fifth of the section, the fix is a hard per-item cap with a forced guess. Write the fix into the next attempt as a constraint, not as an intention: next attempt, no item goes past ninety seconds.
Timing review is also where most candidates discover that their score is closer to their potential than they thought. An attempt with seven correct, two abandoned and one rushed is often a nine-correct attempt with a broken clock rather than a seven-correct attempt with weak knowledge.
Run the protocol with a checklist
A twelve-item review checklist in the five steps above, with progress saved in this browser so you can close the tab and return to it.
This guide is independent study material. It is not affiliated with g.a.s.t., APS India, or any German university, and it does not publish official dMAT questions.
Step five: schedule the retake with one variable changed
Finish the forty minutes by writing the date of the next attempt and the single variable you will change. One variable: the pacing method, the reading protocol, or the topic mix. Then leave at least one study session between the two attempts, and make that session exercise the corrected rule rather than reviewing the same items again.
Do not retake immediately to feel better. An immediate retake measures recognition of the items you just reviewed, which will rise regardless of skill. That rise feels like progress and teaches nothing, and it is the most common self-deception in short-set practice.
Finally, record the attempt in a single line: date, correct, abandoned, dominant cause, change made. After three attempts the line becomes a trend, and the trend is what should influence any decision about buying more practice material. A flat trend with a changing dominant cause means the practice is working even if the score has not moved yet.
There is a temptation to compress this protocol into a quick glance once you are tired, and the compression is almost always step one, the correct answers. Guard against it with a timer. Twenty minutes for the items, ten for the classification, ten for the rule rewriting and the schedule. A timer turns review from an open-ended obligation into a bounded session you will repeat tomorrow.
Reviewing with another applicant helps if you hold the protocol and hurt if you do not. The useful version is to explain your rule for a correct answer and let the other person challenge it; the unhelpful version is to compare scores, which teaches nothing and introduces a comparison that is not evidence about either of you. If you review alone, explaining the rule out loud to an empty room still works, because the act of producing a sentence is what exposes a missing reason.
Keep the artefact. One page per attempt, five steps, dated. Three pages side by side show whether the same cause keeps appearing, and that comparison is the only reliable argument for changing your study focus. Without the pages, every decision about what to study next is a memory of how the last attempt felt, and feelings about attempts are systematically worse than the records of them.
Official sources
This guide defers to these official pages on format, rules, dates, fees and eligibility. They are external and may change, so check the version that applies to your own administration.
- Official dMAT websiteTest structure, preparation materials, fees, test centres and the 2027 India test dates.
- APS India dMAT noticeWho needs the dMAT for APS documentation, exemptions, and how the certificate reaches your APS file.
Questions about this guide
How long should review take for a longer set?
Scale it with the attempt. A ten-item set fits in forty minutes; a forty-item set needs a longer block or two sessions, with the same five steps in order.
Should I review with the explanations open?
Write your own rule first, then compare. Opening explanations first replaces your diagnosis with someone else's, which is why the review stops teaching.
What if I cannot find a cause for a miss?
Label it method and write the smallest possible next action, such as re-deriving the answer from scratch without a clock. An unresolved miss should still produce one action.
Put the method to work
The free diagnostic shows which of these failure modes is actually costing you marks. Ten questions, fifteen minutes, instant topic feedback, no account.