CMAT common mistakes
CMAT · Understand the traps and how to avoid them.
Spending a fixed 2–3 minutes on every Quantitative Techniques question regardless of difficulty, so the easy arithmetic and percentage questions get the same time as a multi-step time-and-work problem and the section ends with several untouched questions.
Candidates treat the section as a uniform test and rehearse one pace for all questions, often because they have only practised full-length sets rather than timed drills with per-question cut-offs.
What to do instead
Do 40-question timed sets in three passes: first pass solve only questions you can finish in 60–75 seconds, second pass take the medium ones, third pass attempt leftovers. Log the pass in which each question was solved to see whether your cut-off is calibrated.
Approximating in Data Interpretation (rounding 18.7% to 20%, 4,890 to 5,000) on every step and then picking the answer closest to the estimate, which lands on a distractor placed deliberately near the rounded value.
DI sets are built with close options, and candidates under time pressure round early and repeatedly, compounding error until the option set becomes ambiguous.
What to do instead
Round only once and in a direction you note down, keep the ratio structure intact (compare 47/89 rather than 0.5/0.9), and verify the chosen option by checking whether the nearest rival option would require a different sign or a different percentage-point gap. Practise percentage-change questions where two options differ by under one percentage point.
Reading a Logical Reasoning arrangement or grouping puzzle and immediately drawing a full grid before extracting the constraints that actually bind, then running out of time with a half-filled diagram and no answer.
Candidates memorise diagram templates from practice and apply them mechanically, without first classifying the puzzle type or spotting the constraint that fixes a position outright.
What to do instead
Before drawing anything, spend 20 seconds listing constraints and mark the one that is most restrictive (a fixed position, an either/or, a negative clue). Build the diagram from that constraint outward. If after 90 seconds no cell is forced, abandon and return later - track abandon rate in practice to make this a habit.
Treating 'assumption', 'inference', 'strengthen' and 'weaken' questions with the same method - choosing the option that sounds most reasonable or most closely repeats the passage - instead of applying the question-type-specific test.
The four types look similar in wording and candidates prepare them as one 'critical reasoning' block, so they never internalise that an assumption must be necessary (negate it and the argument collapses) while an inference must be fully supported by the passage.
What to do instead
For each type, apply one explicit test: assumption - negate the option and check the argument breaks; inference - check the passage states it, not merely permits it; strengthen/weaken - ask whether the option changes the likelihood of the conclusion given the stated evidence. Drill 15 questions of a single type in a row before mixing.
Answering Language Comprehension questions from prior knowledge or from the topic's general plausibility rather than from the passage, choosing an option that is true in the world but not stated in the text.
Under time pressure the passage is skimmed once and options are judged on how familiar or sensible they sound; familiar topics (economy, environment, technology) trigger strong prior beliefs.
What to do instead
For every option, point to the line or paraphrase in the passage that supports it, and reject options you cannot locate. Practise with passages on topics you know well to expose the bias, and mark each chosen answer with the paragraph number it came from.
Losing marks on para-jumbles and summary questions by anchoring on the sentence that 'sounds like a good opening' instead of finding mandatory pairs - pronoun and demonstrative links, connectors such as however/therefore, and the definite article referring back.
Candidates read for overall flow rather than for grammatical cohesion markers, and the opening-sounding sentence is often a distractor that begins with a broad general statement.
What to do instead
Build mandatory pairs first: identify what each pronoun, 'this', 'such' and 'the' refers to, and which sentence must precede it. Then fix the opener as the sentence with no back-reference, and confirm the closing sentence resolves the argument. Time-box to 90 seconds per jumble.
Preparing General Awareness by reading monthly news compilations cover to cover without retaining static anchors - constitutional articles, currency and capital pairings, award categories, core economic terms - so current-affairs questions are answered but static and conceptual ones are not.
Compilations feel productive and are easy to consume passively; candidates mistake recognition ('I have seen this') for recall under exam conditions.
What to do instead
Split preparation: a small static core (polity structures, economy definitions such as repo rate, fiscal deficit, GDP vs GNP, major institutions and their roles) revised by active recall, plus current affairs tied to that core. Convert news into question form and self-test weekly rather than re-reading.
Guessing on General Awareness and Logical Reasoning items with no elimination and no time cost, so a run of wrong guesses erases the gains from correctly solved questions in a negative-marking paper.
Candidates believe guessing is free or marginally positive, and do not track their actual accuracy on guessed items, so the expected value is never measured.
What to do instead
Set a guessing rule and follow it: attempt only if at least two options are eliminated or you have a specific partial recollection. Record guesses separately in practice tests and compute guess accuracy; if it sits near 25%, stop guessing blind.
Writing Innovation and Entrepreneurship answers in abstract generalities ('innovation is important for growth', 'entrepreneurs create jobs') without naming a stage, framework, or example, so answers earn no credit for application.
The domain is often prepared by reading headlines rather than frameworks, and candidates lack a stock of concrete Indian and global venture examples to anchor a point.
What to do instead
Prepare a small toolkit and use it explicitly: idea-to-scale stages, minimum viable product and validation, design thinking steps, funding stages (bootstrapping, angel, seed, Series A), and business-model components. Memorise six to eight specific ventures with the problem they solved and the model they used, and cite one per answer.
Treating Innovation and Entrepreneurship as pure recall and ignoring the numerical and reasoning content embedded in it - break-even quantity, margin and mark-up, payback period, market-sizing arithmetic - and losing those marks to calculation slips.
Candidates categorise the domain as 'theory' and skip the quantitative practice, so they meet break-even and market-sizing questions cold in the exam.
What to do instead
Practise a fixed set of business calculations until automatic: break-even = fixed cost ÷ (price − variable cost per unit), contribution margin, mark-up versus margin, payback period, and top-down market sizing (population × penetration × frequency × price). Do ten such calculations a week and check units and whether the answer is a quantity or a value.