IISER IAT common mistakes
IISER IAT · Understand the traps and how to avoid them.
In Biology, treating the IAT's assertion–reason and 'which of the following is correct/incorrect' items as simple recall, picking the option that is a true statement rather than the one that answers the exact logical relation asked (e.g., marking 'both true, R explains A' when R is true but independent of A).
Most school practice trains statement-memorisation, so candidates match on truth value alone and skip testing the causal link between assertion and reason; the IAT deliberately mixes one true and one false statement to punish this.
What to do instead
Before answering, write A and R separately and label each T/F, then explicitly ask 'does R causally explain A?'. Practise 20 assertion–reason items per week from NCERT chapters and justify the link in one written sentence before choosing.
In Biology, answering genetics and molecular-biology numericals (dihybrid ratios, recombination frequency, DNA base composition, pedigree probability) by eyeballing familiar ratios instead of computing from the given numbers.
Candidates memorise 9:3:3:1 and Chargaff's 1:1 rules and apply them even when the question states linkage, epistasis, or single-stranded RNA where those ratios do not hold.
What to do instead
For every such item, set up the cross or the base-pairing table on paper before looking at options. Drill Chargaff problems with single-stranded data, and practise pedigree questions by assigning genotypes to each individual explicitly.
In Biology, confusing plant physiology and diversity-of-life terminology that look similar - apoplast vs symplast, C3 vs C4 vs CAM anatomy, monocot vs dicot vascular bundles, gymnosperm vs angiosperm reproduction - and choosing the option with the familiar-sounding word.
NCERT chapters are read as lists, so visually similar terms blur together under time pressure; the IAT uses these near-miss pairs as distractors.
What to do instead
Build a two-column contrast table for each confusable pair (feature vs pathway), and self-test by covering one column. Redraw the C4 and CAM carbon-fixation diagrams from memory weekly.
In Chemistry, attempting stoichiometry, equilibrium, or electrochemistry numericals mentally and picking the nearest option without tracking units and significant figures (moles vs millimoles, g vs mg, molarity vs molality, kJ vs J).
The IAT is time-pressured with no partial credit, so candidates shortcut the arithmetic; unit slips then make two options look plausible.
What to do instead
Do a dimensional check on every calculation: write the unit at each step and cancel. Convert all quantities to SI at the start, and estimate the answer's order of magnitude before matching options.
In Chemistry, predicting organic reaction products from the reagent name alone without checking substrate class, leaving-group ability, solvent, and temperature - for example applying SN1 to a primary halide or expecting Markovnikov addition where a peroxide is present.
Candidates memorise reagent–product pairs from a list rather than the mechanism (carbocation stability, steric access, radical initiation), so condition changes are ignored.
What to do instead
For each reaction, annotate the mechanism first (SN1/SN2/E1/E2, electrophilic addition, radical), then derive the product. Maintain a one-page decision flowchart for substitution vs elimination and review it before each mock.
In Chemistry, treating periodic trends and bonding concepts as fixed rules (electronegativity always predicts polarity, 'ionic means conducts', bond order always tracks stability) and ignoring anomalies like HF hydrogen bonding, N vs P basicity, or O2 paramagnetism.
Simplified school rules are taught as absolutes; the IAT tests the exceptions and the underlying orbital/energetic reasoning.
What to do instead
Keep an 'exceptions log' (anomalous trends, MO diagrams for O2, F2, N2, back-bonding in CO) and revise it weekly. For each trend question, justify with the governing factor (effective nuclear charge, orbital overlap, solvation) rather than a memorised rule.
In Mathematics, jumping to a formula on calculus and coordinate-geometry items without checking domain, differentiability, and boundary conditions - e.g., applying L'Hôpital where the limit form is not indeterminate, or finding a local extremum and reporting it as global.
Practice is dominated by routine 'apply the formula' problems, so candidates skip the hypothesis-checking step that the IAT's twisted versions require.
What to do instead
Adopt a fixed checklist before solving: domain of each function, continuity/differentiability at breakpoints, and whether the question asks local or global. Verify final answers by substitution or a quick graph sketch.
In Mathematics, losing marks on combinatorics, probability, and complex numbers by double-counting cases, forgetting to subtract overcounted arrangements, or mishandling the argument/principal-value branch of complex numbers.
Candidates enumerate cases informally without a systematic partition, and complex-number arguments are treated as arctan without quadrant checks.
What to do instead
For counting, always define the partition (e.g., by number of items in a box) and check the sum of cases equals the total. For complex numbers, plot the point to fix the quadrant before computing the argument, and practise de Moivre with roots explicitly.
In Physics, substituting into equations without drawing a free-body, ray, or circuit diagram, so sign conventions (direction of g, current, induced emf, lens distances) and force components are wrong.
Under time pressure candidates pattern-match the equation to the keywords and skip the setup that fixes signs and geometry.
What to do instead
Make a diagram mandatory for every mechanics, optics, and EMI question - mark axes, chosen positive direction, and known/unknown quantities. Practise 10 sign-convention problems (projectile, lens, Lenz's law) per session until the convention is automatic.
In Physics, accepting an answer that is dimensionally or physically implausible (speed greater than c, efficiency above 100%, negative absolute temperature) or ignoring approximation validity in error analysis and SHM/wave problems.
Candidates stop at the first option matching their arithmetic and do not run a plausibility or limiting-case check; small-angle and ideal-gas approximations are applied outside their range.
What to do instead
Reserve 15 seconds per numerical to check dimensions and one limiting case (e.g., set a variable to zero or infinity). In error analysis, propagate percentage errors explicitly and state whether the small-angle approximation applies.