NMAT Logical Reasoning: How Scoring Differs From CAT

DILR

Two aspirants finish the same NMAT Logical Reasoning section having answered the same number of questions correctly. Their scorecards come back with different numbers. Neither of them made an error reading the result, and nothing went wrong with the exam.

That is scaled scoring working as designed, and it is the part of NMAT Logical Reasoning that CAT preparation leaves an aspirant least equipped to understand. CAT counts marks: three for a correct answer, minus one for a wrong one, and the arithmetic is yours to do. NMAT reports a scaled score between 12 and 120 that reflects how you performed against the difficulty of the specific questions you were served, and no third party can reconstruct the conversion.

How Scaled Scoring Actually Works

GMAC scales each NMAT section onto a 12 to 120 range, producing a total between 36 and 360 across the three sections. The stated purpose is to place raw performance on a common scale that removes discrepancies associated with different levels of difficulty.

That last phrase is doing the real work. Because the exam is adaptive and serves a personalised sequence of questions, two candidates do not sit the same paper. Raw counts would therefore not be comparable, and scaling is what makes them comparable. A candidate who answered harder questions correctly is credited accordingly.

The consequence for preparation is narrow but important: your raw accuracy on a third-party mock is a genuine measure of your reasoning reliability, and a poor predictor of a scaled number. No mock provider holds GMAC's equating tables, so any predicted scaled score you are shown is an estimate built on assumptions rather than a conversion.

NMAT Logical Reasoning Next to CAT DILR

The two sections share a name and very little else structurally. Every row below changes something about how you should behave.

Factor	CAT DILR	NMAT Logical Reasoning
Questions and time	22 in 40 minutes	36 in 40 minutes
Structure	Set based caselets	Largely individual questions
Marking	Plus 3, minus 1 on MCQs	No negative marking
Scoring reported as	Raw marks then percentile	Scaled score from 12 to 120
Revisiting a question	Allowed within the section	Not possible

The first row is the one that reframes everything else. Thirty-six questions in the same forty minutes CAT gives to twenty-two means roughly 67 seconds per question against CAT's 109. There is no room in that budget for a fifteen minute caselet, which is itself the clearest evidence about what the section contains.

The Scoring Audit: 3 Shifts From CAT DILR

Run your DILR instincts against these three. Each is rational on CAT and costly here.

Shift 1: Answer Every Single Question

This is the cheapest marks available on the entire paper and the one CAT preparation most actively trains you out of. Every question you leave blank is a free attempt declined. Across 36 questions, even five blind guesses carry real expected value, and there is no downside case at all.

Shift 2: Replace Selection With a Time Ceiling

On CAT the skill is choosing which sets to attempt. Here the only lever is when to stop working on the question in front of you. Ninety seconds is a sensible ceiling at a 67 second average. When you hit it, eliminate what you can, answer, and move on permanently.

Shift 3: Judge Yourself on Accuracy by Type

Since scaled scores are not reconstructable, the useful feedback loop is accuracy within each question format. "I get syllogisms right eight times in ten and input-output four times in ten" is actionable. A predicted total is not.

What the Section Actually Contains

GMAC's published syllabus splits Logical Reasoning into two halves, and only one of them overlaps meaningfully with CAT DILR.

• Verbal reasoning, covering critical reasoning, conclusions, inferences, assumptions, strong and weak arguments, and deductions.
• Analytical reasoning, covering syllogisms, analytical puzzles, input-output, seating arrangement, decision making and Venn diagrams.

The verbal reasoning half is the part CAT aspirants tend to underestimate. CAT tests critical reasoning only obliquely, inside VARC, and rarely as standalone assumption or inference questions. Here it is explicitly half the syllabus, and it is a different skill from puzzle solving.

The Honest Limit of What Is Known About Structure

You will find confident claims that NMAT Logical Reasoning is entirely standalone with no sets at all. GMAC does not characterise it either way. What can be said from the published design is that item-level adaptivity, one question served at a time and no ability to review make large CAT-style caselets structurally difficult to deliver. Prepare for mostly individual questions while staying capable of handling a compact set, rather than assuming one or the other.

Why the Verbal Reasoning Half Decides Most Scores

If the syllabus is genuinely half verbal reasoning, then roughly eighteen of your thirty-six questions sit in territory CAT never drilled you on directly. That is not a minor gap, and it is the single most under-prepared part of this section for a CAT-focused aspirant.

The good news is that verbal reasoning is a closed, learnable skill with a small number of recurring traps. An assumption question is asking what must be true for the argument to hold, not what would strengthen it. An inference question is asking what necessarily follows, not what seems reasonable. A strong-versus-weak argument question is asking about relevance and scope, not about whether you agree.

Each of those distinctions can be learned in an afternoon and then drilled. Aspirants who lose marks here almost never lose them to complexity. They lose them to answering a slightly different question from the one asked, which is a fixable error rather than a capability gap.

Practise these separately from puzzles. They use a different mental mode, and mixing them in the same session tends to leave both half-trained.

What Happens When You Run Out of Clock

On CAT, running out of time in DILR is survivable in a specific way: the questions you never reached were probably in sets you had already decided to skip, so the loss is partly planned. NMAT removes that consolation.

Because questions arrive in an order you do not control, the ones left unanswered when the clock stops are simply the last ones in your personal sequence. They are not the hard ones you chose to abandon. They might have been the easiest questions of your entire section, and you will never know.

That asymmetry is worth internalising, because it changes what good pacing means. On CAT, finishing the section is optional and often correct to forgo. Here, reaching the end is close to mandatory, and a slightly lower accuracy across all thirty-six beats high accuracy across twenty-eight with eight unseen.

Build the habit in practice: if a session ends with unanswered questions, treat it as a pacing failure regardless of how well you did on the ones you reached.

What Your CAT DILR Months Still Buy You

The differences are real, and it would be wrong to conclude your DILR preparation does not transfer. Most of the underlying capability does.

• Condition reading. Extracting exactly what a constraint says, and not more, is the central skill in both papers.
• Arrangement technique. Seating arrangement appears in the published syllabus and your CAT work covers it comfortably.
• Venn logic. Directly named in the syllabus and directly built by CAT set theory work.
• Working under a section clock. The budget is tighter, but the discipline is one you already have.
• Reviewing errors by cause. Transfers unchanged and remains the fastest route to improvement.

What genuinely needs adding is the verbal reasoning half and the habit of answering everything. Our guide on which DILR sets are worth skipping covers the selection skill CAT still rewards, which stays the higher priority while CAT is the bigger paper in your season.

Fitting NMAT Reasoning Into a CAT Week

Three additions cover it, and none of them requires displacing your CAT DILR work.

One Critical Reasoning Session a Week

Thirty minutes on assumptions, inferences and strong versus weak arguments. This is the genuine gap in a CAT-trained profile, and it is a closed topic that responds quickly to deliberate practice.

One Timed Block With No Going Back

Thirty-six mixed reasoning questions in forty minutes, with a rule that you may not return to anything. Rehearsing the constraint is the point; the questions themselves are secondary.

Track Blanks as a Separate Error Category

Count unanswered questions separately from wrong ones in every session. On a paper with no negative marking, a blank is a self-inflicted error rather than a neutral outcome, and it needs its own line in your log. A personalised CAT preparation plan that tracks pattern rather than score will keep this visible week to week.