IIT JAM MA common mistakes
IIT JAM MA · Understand the traps and how to avoid them.
In a multiple-correct (MSQ) item, selecting only the answer that first looks right and stopping, because the paper awards no partial credit and full marks require every correct option and no wrong option.
Candidates trained on single-answer MCQ papers treat MSQs as MCQ and stop at the first true statement; they also skip checking whether a second option is also true under a different reading of the hypothesis.
What to do instead
Before locking an MSQ, test each option independently as a true/false proposition and write T/F beside it; only options that are true under the exact stated hypotheses get selected, and any option you cannot justify is left out.
Producing a proof or counterexample answer that only handles the generic case, omitting the degenerate or boundary cases (n = 1, empty set, zero vector, equality in an inequality, endpoint of an interval).
Under time pressure the mind jumps to the typical picture; boundary cases feel like technicalities, yet JAM MA questions on sequences, continuity, and linear algebra are frequently decided exactly there.
What to do instead
Keep a fixed checklist to run after every proof: smallest index, empty/trivial object, equality case, and endpoints; explicitly write one line covering each before moving on.
Losing marks on the numerical-answer (NAT) section by entering an unrounded or wrongly rounded value, or by writing a fraction where a decimal is required.
The answer box accepts only a number in the specified format; candidates compute a correct exact value such as 7/3 and then type 2.33 or 2.3 when the tolerance expects 2.333, or they misread the required number of decimal places.
What to do instead
Read the format instruction first, keep the exact value until the final step, round only once at the end to the stated precision, and sanity-check magnitude and sign before submitting.
Applying a convergence test or a theorem without verifying its hypotheses (for example using the ratio test when the limit is 1, or interchanging limit and integral without uniform convergence).
Named tests are memorised as recipes, and the hypothesis line in the statement is skipped; the resulting argument looks correct and is hard to spot under exam pressure.
What to do instead
For every named theorem invoked, write the hypothesis it needs in one short phrase next to the use; if a hypothesis fails, switch to a different tool rather than forcing the test.
Setting up a linear algebra computation (rank, eigenvalue, diagonalisability) by expanding a determinant fully instead of using row operations or the characteristic polynomial's structure.
Determinant expansion is the first method taught and feels safe, but it is slow and error-prone for 3x3 and larger matrices, consuming time needed for later questions.
What to do instead
Default to row reduction for rank and to trace/determinant shortcuts for eigenvalues; expand a determinant only when the matrix is triangular or sparse.
Treating a 'which of the following is true' item as asking for a proof of the first plausible option rather than eliminating options with a single counterexample.
Candidates over-invest in verifying one option in full, when a cheap counterexample would eliminate three options faster and with less risk of algebra slips.
What to do instead
For each option, hunt for the quickest counterexample first (small n, simple function, 2x2 matrix); eliminate until one option survives, then verify that survivor briefly.
Mixing up the direction of an implication or the meaning of 'necessary' versus 'sufficient' when answering logic and real analysis items.
Everyday language treats 'if' and 'only if' loosely, and the quantifier order in statements like 'for every epsilon there exists delta' is easy to read backwards.
What to do instead
Rewrite each statement in strict form with explicit quantifiers in order (for all ... there exists ...), then test the converse separately; never rely on the informal phrasing.
Spending disproportionate time on one hard problem in the first pass, leaving easy later questions unattempted.
The paper mixes difficulty without warning; the sunk-cost feeling of a nearly finished proof keeps candidates on a single item while straightforward marks expire.
What to do instead
Do two passes: first pass answer everything solvable within about a minute, mark the rest, and return only after all easy marks are secured; cap any single problem at a preset time.
Making silent algebra slips in limits, series, and integration (sign errors when substituting, dropping a factor from the chain rule, mis-summing a geometric series).
These steps are done mentally at speed, and with no partial credit a single sign flip turns a correct method into a wrong final answer.
What to do instead
Write substitution steps explicitly, substitute a simple numerical check back into the result, and verify series sums against the first two terms before finalising.
Misreading the question's quantifier scope or the exact object asked for (value of the limit versus existence of the limit, sum versus number of terms).
Exam wording is compact and the final phrase carries the actual demand; candidates solve the problem they expected rather than the one printed.
What to do instead
Underline the requested object and any qualifier before starting, and restate the goal in one line at the top of the rough work; check the final answer against that line.