TECHNIQUE: Dimensional Shortcuts on MCQs

JEE-Main gives you, on average, 180 seconds per question. NEET gives you barely 60. Now the uncomfortable arithmetic: a typical physics MCQ takes 2–3 minutes to solve honestly. Multiply by 45 questions and you are out of time before you are out of questions. Toppers do not beat this clock by writing faster — they beat it by not solving a large fraction of questions at all. The single highest-paying skill in that arsenal is the dimension audit: before doing any physics, read the options and ask what dimensions each one carries. A formula for a time period that doesn't have dimensions of time is not "probably wrong" — it is dead, with certainty, no matter how plausible it looks. In a well-set paper this one move routinely executes two or three options in fifteen seconds — sometimes the mark for the price of a glance. You already know the machinery — dimensional formulae, homogeneity, balancing powers of M, L and T. This module turns that knowledge into speed craft: five named shortcut moves, a drill ladder that shows each one against the clock, and — just as important — an honest map of where the method fails. Because the second skill of the dimension auditor is knowing when two options are dimensional twins, and the audit must hand over to real physics.

Key takeaways
  • Read the options first: any option whose dimensions disagree with the asked quantity is eliminated with certainty — no physics needed.
  • Arguments of sin⁡\sin, cos⁡\cos, exe^x, log⁡\log must be dimensionless; a dimensional argument kills an option on sight.
  • Dimensional analysis predicts the form of an answer (T∝ℓ/gT \propto \sqrt{\ell/g}) but is blind to pure numbers like 2π2\pi — the skeleton, never the coefficient.
  • Same-dimension twins (energy vs torque, impulse vs momentum) are where paper-setters hide distractors — the audit cannot separate them.
  • The full audit costs 15–30 seconds; make it the first move on every formula-shaped MCQ, then spend the saved minutes where real physics is needed.
  • Key idea: dimensions can prove an option wrong with certainty, but can never prove an option right — eliminate ruthlessly, then confirm the survivor.
The core insight

Every valid physics equation must be dimensionally homogeneous, so any MCQ option that fails a 15-second M–L–T audit is wrong with certainty — dimensional analysis is an elimination weapon, not a derivation of truth.

Where this lives today

The exam-hall version of this skill is a miniature of how working engineers operate. Wind-tunnel testing exists because of the backfire table above: you cannot test a full-size aircraft, so you test a scale model at the same dimensionless numbers (Reynolds, Mach) — Buckingham's π theorem industrialized Move 2 into the design method behind every plane, dam and ship hull. And the newest habitat is quality control for AI-generated physics: the first automated sanity check run on a machine-proposed formula is — literally — a dimension audit, the one test that needs no understanding of where the formula came from.

The physical picture

Picture the moment of contact with a question you have never seen: "The terminal velocity of a raindrop of radius rr falling through air of viscosity η\eta is…" followed by four candidate formulas. The untrained reflex is to reconstruct Stokes' law from memory. The trained reflex is colder: ignore the physics, tag each option with its dimensions, and watch the field thin out. An MCQ is not one question but four claims, each of which must pass a non-negotiable entrance test before it may compete. Dimensions are that test — the bouncer at the door.

Figure
The dimension audit in action — three options dead before any physics
Q. The period T of a simple pendulum is: need: [T] = time (A) T = 2π ℓ / g [L] / [L T⁻²] = [T²] — not time ✗ DEAD in 5 s (B) T = 2π √(ℓ / g) √([L] / [L T⁻²]) = [T] — survives ✓ the only survivor = the answer (C) T = 2π √(g / ℓ) √([T⁻²]) = [T⁻¹] — a frequency ✗ DEAD in 5 s (D) T = 2π √(m ℓ / g) stray [M½] — nothing cancels mass ✗ DEAD in 5 s Total cost: about 15 seconds. Physics used: none.

The question in the figure was answered — full mark, zero risk — without recalling a single fact about pendulums. That is not rare: examiners must make wrong options wrong somehow, and dimensional damage is the easiest wrongness to manufacture. But look at the same audit from the examiner's chair. If I know you will run this audit, I will build distractors that pass it: T=2π2ℓ/gT = 2\pi\sqrt{2\ell/g}, or T=πℓ/gT = \pi\sqrt{\ell/g} — dimensionally perfect, physically wrong. The audit is a sieve, not an oracle, which is why this module spends as much time on the method's failure modes as on its kills.

The toolkit: dimensions you must read at sight

Speed is the whole point, so the base-dimension bookkeeping has to be automatic. Everything in mechanics reduces to mass, length and time: a derived quantity's dimensional formula is the recipe

[Q]=Ma Lb Tc(1)[Q] = \mathrm{M}^a\,\mathrm{L}^b\,\mathrm{T}^c \qquad (1)

where: aa, bb, cc — the powers with which mass, length and time enter QQ. The formula records how the quantity is built: force is mass times acceleration, so it carries one power of M, one of L, minus two of T — and any expression claiming to be a force must carry exactly that signature. In an exam you do not derive these signatures; you read them at sight. This is the table to overlearn:

quantitydimensional formulasame-dimension twins (trap fuel)
velocity vvL T−1\mathrm{L\,T^{-1}}—
acceleration aa, ggL T−2\mathrm{L\,T^{-2}}—
force FFM L T−2\mathrm{M\,L\,T^{-2}}weight, tension, thrust
energy, work EE, WWM L2 T−2\mathrm{M\,L^2\,T^{-2}}torque, heat, PE, KE
power PPM L2 T−3\mathrm{M\,L^2\,T^{-3}}—
momentum ppM L T−1\mathrm{M\,L\,T^{-1}}impulse FΔtF\Delta t
pressure, stressM L−1 T−2\mathrm{M\,L^{-1}\,T^{-2}}energy density, Young's modulus
frequency ff, ω\omegaT−1\mathrm{T^{-1}}decay constant, velocity gradient
gravitational GGM−1 L3 T−2\mathrm{M^{-1}\,L^3\,T^{-2}}—
Planck's hhM L2 T−1\mathrm{M\,L^2\,T^{-1}}angular momentum LL
surface tension σ\sigmaM T−2\mathrm{M\,T^{-2}}spring constant kk, force gradient
viscosity η\etaM L−1 T−1\mathrm{M\,L^{-1}\,T^{-1}}—

The third column is your danger map: those pairings are the "same dimensions as…" questions NEET and JEE-Main recycle nearly every year — and exactly where a dimension audit alone cannot choose.

The five shortcut moves

The audit is five distinct moves. Learn them by name so that under pressure you reach for the right one deliberately.

Move 1 — The option audit. Tag the dimensions of the asked quantity, then tag each option. Kill every mismatch. This is the workhorse — it answers the pendulum question in the figure above outright.

Move 2 — Form prediction. Before looking at options at all, predict the answer's shape by balancing powers. If a period can only depend on ℓ\ell and gg, then T∼ℓ/gT \sim \sqrt{\ell/g} is forced — no other combination has dimensions of time. Strongest on "derive-flavoured" MCQs where all four options are algebraically messy.

Move 3 — The last-line check. The audit turned inward: after your own algebra sprint, spend five seconds checking the dimensions of your final line before matching options. A dropped square or a swapped ratio announces itself immediately — cheap insurance against the most expensive exam mistake: correct method, careless last line, confidently wrong bubble.

Move 4 — The dimensionless-argument kill. Whatever sits inside sin⁡(…)\sin(\ldots), cos⁡(…)\cos(\ldots), e(…)e^{(\ldots)}, ln⁡(…)\ln(\ldots) must be a pure number. An option like x=asin⁡(kt2)x = a\sin(kt^2) survives only if [k]=T−2[k] = \mathrm{T^{-2}} — and by forcing arguments to be dimensionless you can read off unknown constants' dimensions, a question type JEE has recycled for decades.

Move 5 — The constants discipline. Pure numbers (12\tfrac12, 2π2\pi, ee, trig values) are dimensionless and invisible to the audit; dimensional constants (GG, hh, kBk_B, ε0\varepsilon_0) carry M–L–T baggage that must balance like any variable. You can never use dimensions to fix a 2π2\pi, and you can never ignore the dimensions of GG. One story before the derivations make Move 2 concrete: assuming the speed of sound depends on pressure and density, the balance forces v=kP/ρv = k\sqrt{P/\rho} — Newton got exactly this far in 1687; the coefficient (k=γk = \sqrt{\gamma}, Laplace) took another 129 years. The skeleton in seconds, the flesh never.

Derivation AThe pendulum period by pure dimensional balance

We derive the form of the period of a simple pendulum using nothing but dimensional homogeneity — the canonical demonstration of Move 2, and the derivation every exam-taker should be able to reproduce in under a minute. The strategy: assume TT depends on the length ℓ\ell, the gravitational acceleration gg, and the bob's mass mm as a product of powers, then force both sides to carry the same dimensions. Watch two things: how the mass is expelled by the balance itself, and where the method silently gives up (the constant).

Step 1. Write the trial form as a product of powers: T=k ℓ xg ym zT = k\,\ell^{\,x} g^{\,y} m^{\,z}, with kk dimensionless. Why a product of powers? Because dimensional formulas are themselves products of powers — sums or more exotic dependences cannot be captured, a genuine limitation of the method.

Step 2. Insert dimensions on both sides:

T=(L)x(L T−2)y(M)z\mathrm{T} = \left(\mathrm{L}\right)^{x}\left(\mathrm{L\,T^{-2}}\right)^{y}\left(\mathrm{M}\right)^{z}

Step 3. Balance each base dimension. Mass: z=0z = 0 — the balance forbids mass from appearing, since nothing else could cancel it (and indeed the real pendulum's period is mass-independent; this is the fastest way to see it must be). Length: x+y=0x + y = 0. Time: −2y=1-2y = 1, so y=−12y = -\tfrac12 and x=+12x = +\tfrac12.

Step 4. Assemble: T=kℓ/gT = k\sqrt{\ell/g}. The method stops here. The value k=2πk = 2\pi comes only from solving the actual equation of motion — dimensional analysis is constitutionally blind to it. Verified: with ℓ=1 m\ell = 1\ \mathrm{m}, g=9.8 m/s2g = 9.8\ \mathrm{m/s^2}, ℓ/g=0.32 s\sqrt{\ell/g} = 0.32\ \mathrm{s} and 2π×0.32=2.0 s2\pi \times 0.32 = 2.0\ \mathrm{s} — the familiar metre-pendulum result, confirming both the form and the size of the invisible constant.

Go deeper · Derivation B · 1 minWave speed on a string — solving the exponent system at speed
Derivation BWave speed on a string — solving the exponent system at speed

We derive the speed of a transverse wave on a stretched string from the tension FF and the linear mass density μ\mu (mass per unit length) — the second-most-quoted Move 2 balance in Indian exam archives, and the template for the "solve the exponent system" drill. Strategy: product-of-powers ansatz, then solve the small linear system. The target with practice: under 45 seconds end to end.

Step 1. Ansatz: v=k Faμbv = k\,F^{a}\mu^{b}. Dimensions: [v]=L T−1[v] = \mathrm{L\,T^{-1}}, [F]=M L T−2[F] = \mathrm{M\,L\,T^{-2}}, [μ]=M L−1[\mu] = \mathrm{M\,L^{-1}}.

Step 2. Balance: L T−1=Ma+b La−b T−2a\mathrm{L\,T^{-1}} = \mathrm{M}^{a+b}\,\mathrm{L}^{a-b}\,\mathrm{T}^{-2a}.

Step 3. Read off the system. M: a+b=0a + b = 0. T: −2a=−1⇒a=12-2a = -1 \Rightarrow a = \tfrac12. Hence b=−12b = -\tfrac12, and the L equation, a−b=1a - b = 1, checks out automatically: 12+12=1\tfrac12 + \tfrac12 = 1. ✓ Why check the leftover equation? Free error-detection: with three base dimensions and two unknowns, one equation is redundant — if it fails, you mis-tagged a dimension upstream.

Step 4. Assemble: v=kF/μv = k\sqrt{F/\mu}. (The full wave-equation treatment gives k=1k = 1 — this time the invisible constant happens to be trivial, which the method could not have told you.) Verified: a steel guitar string with F=100 NF = 100\ \mathrm{N}, μ=5 g/m=5×10−3 kg/m\mu = 5\ \mathrm{g/m} = 5\times10^{-3}\ \mathrm{kg/m} gives v=2×104≈141 m/sv = \sqrt{2\times10^4} \approx 141\ \mathrm{m/s} — the right order for audible string instruments.

When it backfires

An honest speed-craft module tells you where the shortcut loses. Route every question through this five-second decision tree first:

Figure
The five-second routing decision — when to audit, when to switch weapons
Are the options algebraic expressions in symbols? no — numbers, ratios, angles audit useless solve honestly, move on yes run the audit tag target, kill mismatches how many survive? one bubble it — done 15-30 s total, full mark two+ (twins) switch weapons limiting case picks it pure-number gap: derive or recall the formula

The same failure map as a table:

situationdoes the audit work?what to do instead
options differ in M–L–T signatureyes — fullyaudit kills all mismatches; if one survivor, done
options differ only by pure numbers (12\tfrac12, 22, 2π2\pi, 2\sqrt{2})no — completely blindreal physics: derive, or test a limiting case
same-dimension twins offered (torque for energy, impulse for momentum)norecall the defining relation, not the formula
options are pure numbers, ratios, anglesno — everything is dimensionlesssolve honestly; audit wastes time here
dimensionally correct but physically wrong (e.g. ρ+σ\rho + \sigma where ρ−σ\rho - \sigma belongs)nolimiting/special cases: set ρ=σ\rho = \sigma, θ→0\theta \to 0, m→0m \to 0
more than 3 independent quantities in an M–L–T problempartiallya balance pins at most 3 exponents; combine with physics (this is why fluids needs the Reynolds number)

Worked examples — the drill ladder

Each drill below shows the same MCQ fought twice — the slow honest way, then the shortcut way — so you can feel the seconds being bought.

Drill 1 · Board level — the period audit

Which of the following can be the time period of a simple pendulum of length ℓ\ell? (A) 2πℓ/g2\pi\ell/g (B) 2πℓ/g2\pi\sqrt{\ell/g} (C) 2πg/ℓ2\pi\sqrt{g/\ell} (D) 2πmℓ/g2\pi\sqrt{m\ell/g}

The slow way. Recall the pendulum formula, hope the memory is right, match — but recall under pressure is exactly what fails, and a mis-remembered g/ℓ\sqrt{g/\ell} matches option (C) beautifully.

The shortcut way. Need dimensions of time. Step 1. (A): [ℓ/g]=L/L T−2=T2[\ell/g] = \mathrm{L}/\mathrm{L\,T^{-2}} = \mathrm{T^2} — a squared time, dead. Step 2. (C): inverts to T−1\mathrm{T^{-1}} — a frequency, dead. Step 3. (D): a naked M1/2\mathrm{M^{1/2}} with nothing to cancel it, dead. Step 4. (B) is the sole survivor: (B)\boxed{\text{(B)}} — and note you never recalled anything; the audit reconstructed the formula's shape for you. Timing note: 15 seconds by audit versus about a minute by recall-plus-doubt — and the same audit is the 3-mark "check by dimensions" step CBSE and ISC award explicitly.

Drill 2 · Board level — Newton's sound formula

The speed of sound in a medium of pressure PP and density ρ\rho may be expressed as (A) Pρ\sqrt{P\rho} (B) P/ρ\sqrt{P/\rho} (C) P/ρP/\rho (D) ρ/P\sqrt{\rho/P}

The slow way. Try to remember whether it is P/ρ\sqrt{P/\rho} or ρ/P\sqrt{\rho/P} — the two look equally plausible at 9 a.m. on exam day.

The shortcut way. Step 1. Target: L T−1\mathrm{L\,T^{-1}}. Step 2. [P/ρ]=M L−1 T−2/M L−3=L2 T−2[P/\rho] = \mathrm{M\,L^{-1}\,T^{-2}} / \mathrm{M\,L^{-3}} = \mathrm{L^2\,T^{-2}} — its square root is exactly a speed: (B) survives. Step 3. Spot-check the tempting twin (D): it is the inverse of a speed. (A) carries M\mathrm{M}; (C) is L2 T−2\mathrm{L^2\,T^{-2}}, a speed squared. (B)\boxed{\text{(B)}} Timing note: 20 seconds. The γ\gamma in the true v=γP/ρv = \sqrt{\gamma P/\rho} is dimensionless — invisible to the audit and irrelevant here, since no option carries it.

Drill 3 · Board level — which is a pressure?

Which of the following has the dimensions of pressure? (A) energy per unit volume (B) momentum per unit area (C) force per unit length (D) energy per unit area

The slow way. Convert each phrase to a formula, expand all four, compare — 90 seconds of clerical work.

The shortcut way. Step 1. Pressure: M L−1 T−2\mathrm{M\,L^{-1}\,T^{-2}}. Step 2. Energy/volume: M L2 T−2⋅L−3=M L−1 T−2\mathrm{M\,L^2\,T^{-2}} \cdot \mathrm{L^{-3}} = \mathrm{M\,L^{-1}\,T^{-2}} — match, and you should know this twin from the table (energy density ↔ pressure). Step 3. Quick kills: (B) is M L−1 T−1\mathrm{M\,L^{-1}\,T^{-1}}, (C) is M T−2\mathrm{M\,T^{-2}} (surface tension!), (D) is M T−2\mathrm{M\,T^{-2}} too. (A)\boxed{\text{(A)}} Timing note: 25 seconds if you expand; 5 seconds if the twin table is overlearned — this question IS row 7 of the table.

Drill 4 · Competitive level — the Planck length

A length can be constructed from the gravitational constant GG, Planck's constant hh, and the speed of light cc as (A) Gh/c3\sqrt{Gh/c^3} (B) hc3 ⁣/G\sqrt{hc^3\!/G} (C) Gc/h3\sqrt{Gc/h^3} (D) Gh/c2Gh/c^2

The slow way. There is no slow way — you either balance exponents or guess. A pure Move 2 question, JEE-Advanced's favourite flavour.

The shortcut way. Set ℓ=Gahbcd\ell = G^a h^b c^d and balance. Step 1. Signatures: [G]=M−1 L3 T−2[G] = \mathrm{M^{-1}\,L^3\,T^{-2}}, [h]=M L2 T−1[h] = \mathrm{M\,L^2\,T^{-1}}, [c]=L T−1[c] = \mathrm{L\,T^{-1}}. Step 2. Mass must vanish: −a+b=0-a + b = 0, so b=ab = a — GG and hh enter together, as the product GhGh. That single fact already kills (B) and (C), where they appear in ratio. Step 3. Time must vanish: d=−3ad = -3a; length must total 1: 3a+2a−3a=2a=13a + 2a - 3a = 2a = 1, so a=12a = \tfrac12. Step 4. Assemble: ℓ=Gh/c3\ell = \sqrt{Gh/c^3}. (A)\boxed{\text{(A)}} Numerically ∼4×10−35 m\sim 4\times10^{-35}\ \mathrm{m} — the Planck length, where quantum gravity lives. Timing note: 90 seconds for the full balance — but Step 2 alone took 20 seconds and left only (A) vs (D); then (D) fails the length count. Partial balances kill fast.

Drill 5 · Competitive level — dimensions kill three, a limit picks the survivor

A small sphere of radius rr and density ρ\rho falls at terminal velocity vtv_t through a fluid of density σ\sigma and viscosity η\eta. Then vtv_t equals (A) 2r2(ρ−σ)g9η\dfrac{2r^2(\rho-\sigma)g}{9\eta} (B) 2r(ρ−σ)g9η\dfrac{2r(\rho-\sigma)g}{9\eta} (C) 2r2(ρ+σ)g9η\dfrac{2r^2(\rho+\sigma)g}{9\eta} (D) 2r2(ρ−σ)g29η\dfrac{2r^2(\rho-\sigma)g^2}{9\eta}

The slow way. Re-derive Stokes drag balance from scratch: weight − buoyancy = 6πηrv6\pi\eta r v. Two to three minutes, error-prone under pressure.

The shortcut way — a two-weapon kill. Step 1. Audit. [η]=M L−1 T−1[\eta] = \mathrm{M\,L^{-1}\,T^{-1}}; check (A): L2⋅M L−3⋅L T−2M L−1 T−1=L T−1\dfrac{\mathrm{L^2}\cdot\mathrm{M\,L^{-3}}\cdot\mathrm{L\,T^{-2}}}{\mathrm{M\,L^{-1}\,T^{-1}}} = \mathrm{L\,T^{-1}} ✓ a velocity. Step 2. (B) is (A) ÷ L — dead. (D) is (A) × L T−2\mathrm{L\,T^{-2}} — dead. Two kills, 20 seconds. Step 3. But (C) has identical dimensions to (A) — ρ+σ\rho+\sigma and ρ−σ\rho-\sigma are twins; the audit is done, hand over to physics. Step 4. Limiting case: a neutrally buoyant sphere (ρ=σ\rho = \sigma) floats motionless, so vtv_t must vanish. (A) gives 0 ✓; (C) does not ✗. (A)\boxed{\text{(A)}} Timing note: 45–60 seconds total. This is the master pattern for JEE-Advanced: dimensions clear the field, one physical limit picks the champion.

Exam traps

The audit itself has mark-losing failure modes — every one below appears in JEE/NEET archives.

Bubbling a twin on dimensions alone. The costliest trap. Torque vs energy, impulse vs momentum, ρ+σ\rho+\sigma vs ρ−σ\rho-\sigma — same signature, one right answer. With JEE-Main and NEET at −1-1 per wrong bubble, a dimensions-only guess between twins loses marks on average; switch weapons first.

Forgetting that arguments must be bare numbers. Anything inside sin⁡\sin, e ⋅e^{\,\cdot} or log⁡\log is dimensionless — sin⁡(ωt2)\sin(\omega t^2) and e−txe^{-tx} die instantly, and examiners plant exactly these. The reverse slip: treating an angle as dimensional and "killing" a correct option.

Prefix and system slips. Auditing with gg in grams, η\eta in CGS poise, or rr in centimetres breaks the balance you are trusting; convert to SI before the audit.

Mis-tagging the target. Asked for energy density, tagged energy — the audit kills the right answer with full confidence. Read the question, tag the target, then read options.

Spending the winnings carelessly. The audit buys 60–90 seconds; students donate it back by skipping the final-line dimension check on questions they solved honestly. Run Move 3 on your own algebra — the cheapest mark insurance on the paper.

Origins · 1 minLord Rayleigh and how this was discovered
LR
Lord Rayleigh
1842–1919 · Cambridge and Terling

John William Strutt, third Baron Rayleigh, made "the method of dimensions" a working physicist's everyday weapon. In The Theory of Sound (1877) and a stream of papers he attacked hard problems — why the sky is blue, how drops vibrate — by first writing down the only dimensionally possible form of the answer; his blue-sky 1/λ41/\lambda^4 law falls out essentially by dimensional balance. He also embodied the method's honesty — it gives forms, not coefficients — and chased a tiny nitrogen density discrepancy the method could never explain until it yielded argon and the 1904 Nobel Prize.

History
1822
— Joseph Fourier's Théorie analytique de la chaleur introduces dimensions: physical quantities carry them, and equations must balance.
1877
— Rayleigh's The Theory of Sound turns dimensional balance into a systematic research method, "the method of dimensions."
1899
— Max Planck combines GG, hh and cc into natural units of length, mass and time — the Drill 4 calculation, done for the first time.
1914
— Edgar Buckingham publishes the π theorem, the general statement of how many independent dimensionless groups a problem possesses.
1922
— Percy Bridgman's Dimensional Analysis, the first book-length treatment, fixes the modern logic of the method.

Going further

Texts. NCERT Physics XI, Ch. 1 (Units and Measurement) · H.C. Verma, Concepts of Physics Vol. 1, Ch. 1 · Bridgman, Dimensional Analysis (1922) · Rayleigh, The Theory of Sound §I for the method in the wild.

Provenance. Fresh module (exam-technique class); the table and drill MCQs are standard exam-archive material. A verifier should recompute: the pendulum balance (x=12x = \tfrac12, y=−12y = -\tfrac12, z=0z = 0); the string-wave system with its redundant-equation check; the Planck-length exponents (a=b=12a = b = \tfrac12, d=−32d = -\tfrac32, ∼4×10−35 m\sim 4\times10^{-35}\ \mathrm{m} with hh); the Stokes drill and its ρ=σ\rho = \sigma limit; [B]=L[B] = \mathrm{L} in the quiz.

Prerequisites. Dimensions & Dimensional Formulae · Dimensional Analysis I: Checking Equations · Dimensional Analysis II: Deriving Relations & Limits. Leads to. Error-Question Patterns · Limiting & Extreme Cases · Option Elimination & Answer Checking (the track capstone).

How do I memorize dimensional formulas fast?

Don't memorize them as arbitrary strings — rebuild them from one defining relation each. Force is mama: M⋅L T−2\mathrm{M \cdot L\,T^{-2}}. Energy is force × distance: add an L. Power is energy ÷ time: subtract a T. Pressure is force ÷ area: remove two L's. Momentum is mvmv; impulse is force × time — compute both and discover they match, which is one exam question pre-answered. Rebuild each row of the twelve-row table twice from its definition and you own it. The only ones worth brute-memorizing are the constants: [G]=M−1L3T−2[G] = \mathrm{M^{-1}L^3T^{-2}} and [h]=ML2T−1[h] = \mathrm{M L^2 T^{-1}}.

Is this safe in NEET where wrong answers cost me a mark?

Safer than solving, for the kills. A dimensional elimination is a certainty, not a guess — homogeneity is a law of physics, so an option with the wrong dimensions cannot be the answer, ever. The negative-marking risk comes from the other side: bubbling a survivor you haven't confirmed when two options were dimensional twins. The discipline: only bubble when one option survives, or you've separated the twins with real physics — a limiting case usually does it in seconds; otherwise solve or skip on merits.

What if two options have exactly the same dimensions?

Then the audit has done all it can, and you switch weapons — the most important honesty rule of the technique. Same-dimension survivors differ either by a pure number (22, 12\tfrac12, 2π2\pi — only a real derivation or memory separates those) or by physical structure (ρ−σ\rho - \sigma versus ρ+σ\rho + \sigma — a limiting case separates those instantly: imagine the sphere as dense as the fluid and ask which formula correctly predicts it hangs motionless). Drill 5 is the full template. Never stare at dimensions hoping they'll break the tie; they can't.

Do angles and radians have dimensions? What about trig functions of them?

Angles are dimensionless — an angle in radians is arc length ÷ radius, a length divided by a length. That's exactly why they're allowed inside sin⁡\sin and cos⁡\cos, and why ωt\omega t and kxkx are legal arguments: each multiplies out to a pure number. The output of a trig function is dimensionless too, so in y=Asin⁡(kx−ωt)y = A\sin(kx - \omega t) all dimensional content lives in the amplitude: [y]=[A][y] = [A], always. Questions asking "dimensions of Asin⁡(…)A\sin(\ldots)" are testing exactly this.

Will the examiner accept an answer I got by elimination instead of solving?

In an MCQ, nobody sees your method — a mark earned by a 15-second audit is identical to a mark earned by a 3-minute derivation; JEE and NEET reward deciding correctly, not showing work. The place to be careful is the written board paper: if a question says "derive" or "show that", a dimensional argument earns only the marks allotted to checking, not deriving. There, use the audit as your private verification layer — CBSE and ISC do award explicit steps for "checking dimensional consistency".

Why exactly must the argument of an exponential or sine be dimensionless?

Because those functions are secretly infinite sums: eu=1+u+u22+⋯e^u = 1 + u + \tfrac{u^2}{2} + \cdots. If uu carried dimensions — say, seconds — the series would try to add a pure number to seconds to seconds-squared, violating homogeneity in every term. The only way the sum makes sense is if uu is a pure number. This one idea is an exam factory: given N=N0e−λtN = N_0 e^{-\lambda t} you instantly know [λ]=T−1[\lambda] = \mathrm{T^{-1}}; given x=asin⁡(bt2)x = a\sin(bt^2), [b]=T−2[b] = \mathrm{T^{-2}}. In any unfamiliar formula, the arguments are free dimension-information the paper-setter had no choice but to leave for you.

Can dimensional analysis derive any formula in physics?

No — and knowing the boundary is part of the skill. It fails for dimensionless constants (the 2π2\pi in the pendulum — invisible forever); for sums like s=ut+12at2s = ut + \tfrac12 at^2, which no product ansatz can find; for problems with more independent quantities than M, L, T can constrain; and for dimensionless physics expressed in ratios or angles. What it always does is constrain: even where it can't finish, it eliminates impossible forms — which for MCQ purposes is usually enough.

Before this

Where this leads