Measure honestly: units, dimensions, uncertainty and instruments tell different parts of the story.
JEE Main · JEE Advanced · NEET UG65 min study guide9 indexed sections
BY THE END, YOU CAN
Convert compound units and test equations dimensionally.
Derive scaling relations and state their limits.
Propagate stated uncertainty models without confusing worst-case bounds and random spread.
Read vernier and screw-gauge scales with signed zero corrections.
BEFORE YOU START
Powers, scientific notation and elementary algebra.
Arithmetic means and percentages.
Original educational lesson. Work through the examples before revealing the practice answers. Check the current official syllabus separately; historical question frequency does not establish eligibility or predict future questions.
CONNECT THE CONCEPT TO THE EXAM
Selected exam questions & worked solutions
These selected examples are separate from the original practice below. Each identifies its exam, year and wording. They do not represent every paper or every selected year.
Available in this lesson: NEET UG 2021. This is a limited lesson sample, not a complete exam-question bank.
1 sourced sample question available across the listed exams and years.
Matched topics: Dimensional Analysis.
NEET UG · 2021 · 12 September · Booklet M1 · Physics · Q25
Paraphrased exam question
Answer checked against the official final key. Worked solution written for this lesson.
Determine the dimensions of E/G, where E represents energy and G is the universal gravitational constant.
Reveal answer & worked solution
Answer: [E/G] = M²L⁻¹T⁰ (original option 1).
Energy has dimensions force × distance: [E] = ML²T⁻².
From F = Gm₁m₂/r², [G] = [F][r]²/[m]² = M⁻¹L³T⁻².
Divide by subtracting exponents: [E/G] = M¹⁻⁽⁻¹⁾L²⁻³T⁻²⁻⁽⁻²⁾ = M²L⁻¹T⁰. Time cancels.
Source: NEET (UG) 2021, 12 September, M1, Q25; official final answer key. Solution written for this lesson.
01 / UNDERSTAND & APPLY
1. A number becomes a measurement only with a unit
A physical quantity is the product of a numerical value and a reference unit. The same length can be 0.025 m, 2.5 cm or 25 mm without changing the object. The number changes inversely with the size of the chosen unit. Writing units at every line makes conversions visible rather than trusting a remembered power of ten.
The SI base quantities are length, mass, time, electric current, thermodynamic temperature, amount of substance and luminous intensity, with units metre, kilogram, second, ampere, kelvin, mole and candela. Derived quantities combine these. A newton is kg m s⁻², a joule is kg m² s⁻² and a pascal is kg m⁻¹ s⁻². Symbols are case-sensitive: m is metre, while M as a prefix means mega.
Convert by multiplying by ratios equal to one. For powers of length the entire conversion factor is raised to that power. A square centimetre is 10⁻⁴ square metres, not 10⁻². For compound units, convert numerator and denominator separately. Distinguish the mass unit kilogram from gram when using SI force and energy equations.
Temperature differences in kelvin and degrees Celsius have equal numerical magnitude, but absolute temperatures differ by 273.15. A gas-law ratio requires kelvin temperatures. Angular measure in radians is an arc-length/radius ratio and is dimensionless, although keeping the unit rad clarifies physical meaning. Dimensional equivalence never means that all quantities of those dimensions have the same physical interpretation.
ORIGINAL WORKED EXAMPLE
Convert before substituting
A liquid density is 1.20 g cm⁻³. Express it in SI, then find the mass of 2.50 × 10² cm³.
One gram is 10⁻³ kg and one cubic centimetre is (10⁻² m)³ = 10⁻⁶ m³.
Thus 1 g cm⁻³ = 10³ kg m⁻³, so density is 1.20 × 10³ kg m⁻³.
The volume is 250 × 10⁻⁶ m³ = 2.50 × 10⁻⁴ m³. Multiply density and volume: mass = 0.300 kg.
Answer Density 1.20 × 10³ kg m⁻³; mass 0.300 kg.
02 / UNDERSTAND & APPLY
2. Dimensions test the structure of an equation
A dimension names the kind of quantity independently of the unit. Let M, L and T denote mass, length and time dimensions, and I denote current. Speed has dimension LT⁻¹ and acceleration LT⁻². Build unfamiliar dimensions from a physical definition rather than maintaining an isolated formula list: pressure is force divided by area, so it has dimension ML⁻¹T⁻².
Every term being added must share dimensions. In an equation for displacement, a speed term must be multiplied by a time and an acceleration term by a time squared. Derivatives and integrals also carry units: differentiating energy with respect to distance gives force, while integrating power over time gives energy. A graph slope and its area therefore answer different physical questions.
The argument of an exponential, logarithm, sine or cosine must be dimensionless. In x = A exp(−bt), b must have inverse-time dimensions. In x = A sin(ωt + φ), both ωt and φ are dimensionless, while A carries the length dimension. One cannot repair an expression by assigning units to the output of an ordinary sine function.
Dimensional consistency is necessary but insufficient. The equations K = mv² and K = mv²/2 both pass; only a dynamical derivation determines the factor one half. Torque and energy share dimensions, but torque is an axial vector and energy a scalar. A dimensionless quantity can still depend on geometry, angle or another dimensionless ratio.
[F] = MLT⁻²; [E] = ML²T⁻²; [P] = ML²T⁻³
Here P denotes power. Write the meaning of a symbol because P is also commonly used for pressure.
[G] = M⁻¹L³T⁻²; [h] = ML²T⁻¹
Obtain G from F = Gm₁m₂/r² and Planck’s constant h from E = hν.
ORIGINAL WORKED EXAMPLE
A variable force law
A force is written F = a√x + b v². Find the dimensions of a and b, where x ≥ 0 is displacement and v is speed.
Each term separately has force dimensions MLT⁻².
For a√x, divide by L¹ᐟ²: [a] = ML¹ᐟ²T⁻².
For bv², divide by L²T⁻²: [b] = ML⁻¹. The two constants need not have the same dimensions.
Answer [a] = ML¹ᐟ²T⁻² and [b] = ML⁻¹.
03 / UNDERSTAND & APPLY
3. Use dimensional reasoning to discover a dependence
Suppose a small-amplitude simple pendulum period depends only on its length l, gravitational acceleration g and bob mass m. Write Tₚ = C lᵃgᵇmᶜ, where C is dimensionless. Matching mass powers gives c = 0; matching time powers gives −2b = 1; matching length powers gives a + b = 0. Thus Tₚ is proportional to √(l/g). This predicts the scaling, not the numerical coefficient.
The phrase “depends only on” is the hidden modelling assumption. Real pendulum periods can also depend on amplitude and air drag. Amplitude as an angle supplies a dimensionless variable, so dimensional analysis cannot rule out a multiplying function of angle. Small-amplitude dynamics, rather than dimensions, explains why that function approaches a constant.
When the number of unknown exponents exceeds the independent base dimensions, dimensional matching may not determine a unique formula. This is useful information: it warns that the physics is missing. Also distinguish a change of units from a change in the system. Measuring the same period in milliseconds changes its numerical value; lengthening the pendulum changes the actual period.
ORIGINAL WORKED EXAMPLE
Wave-speed scaling
Assume the speed c of a wave on a taut ideal string depends only on tension F and linear mass density μ. Find its form and predict the effect of quadrupling F.
Write c = C Fᵃ μᵇ. Dimensions give LT⁻¹ = (MLT⁻²)ᵃ(ML⁻¹)ᵇ.
Matching time gives a = 1/2; matching mass gives b = −1/2; the length equation is then satisfied.
Hence c = C√(F/μ). Dimensional analysis cannot determine C, but for unchanged string density quadrupling tension doubles speed.
Answer c ∝ √(F/μ); the speed doubles.
04 / UNDERSTAND & APPLY
4. What does the uncertainty actually mean?
A measured result is better thought of as an estimate supported by a measurement procedure than as a perfect number. Random variation changes unpredictably between repeats. A systematic effect, such as a shifted instrument zero, can displace every reading in the same direction. Repeating readings reduces the influence of some random effects but does not by itself remove a calibration bias.
Accuracy concerns closeness to an accepted reference; precision concerns repeatability or spread. A tightly clustered set can be precise but biased. Resolution is the smallest displayed or distinguishable increment. It is not automatically the complete uncertainty: alignment, calibration, object shape, temperature and operator judgement may contribute more than the scale increment.
In the common school-level mean-absolute-error convention, average the observations, find each absolute deviation from that mean, and average those deviations. The result is an estimate of spread, not a universal statistical confidence interval. Use the uncertainty convention stated in the problem. Do not silently replace maximum-error bounds by standard deviations or combine the two as if they meant the same thing.
Write a measured value and its absolute uncertainty in matching units and decimal places, for example (12.40 ± 0.05) cm. Relative uncertainty divides absolute uncertainty by the magnitude of the value; multiply by 100 for percentage uncertainty. Near zero, relative uncertainty is not useful even when the absolute uncertainty is perfectly meaningful.
ORIGINAL WORKED EXAMPLE
Repeated length readings
Three corrected readings are 8.12, 8.16 and 8.14 cm. Use the mean-absolute-error convention to report the mean and spread.
The arithmetic mean is (8.12 + 8.16 + 8.14)/3 = 8.14 cm.
Absolute deviations are 0.02, 0.02 and 0.00 cm. Their average is 0.0133… cm.
At a suitable rounded precision, report about (8.14 ± 0.01) cm under this convention. Before rounding, the relative spread is 0.0133/8.14 ≈ 0.00164, or 0.164%.
This calculation does not test for an unknown systematic error in the scale.
Answer Mean 8.14 cm; unrounded mean absolute deviation 0.0133 cm; approximately 0.16% relative spread.
05 / UNDERSTAND & APPLY
5. Predict which measurement limits the result
For small maximum-error bounds, a sum or difference has an absolute uncertainty bounded by the sum of the input absolute uncertainties. Subtraction can be especially sensitive: subtracting two nearly equal lengths gives a small difference while the absolute uncertainty need not shrink. A precise-looking calculator answer then overstates what the experiment established.
For products and quotients, first-order relative maximum errors add. For a power, multiply the fractional error by the magnitude of the exponent. Thus a radius measurement contributes twice its percentage uncertainty to a circular area and three times its percentage uncertainty to a spherical volume. Exact numerical constants such as π do not create a measurement uncertainty.
These rules arise by differentiating a function. For y = aᵖbᑫ, small signed changes obey dy/y ≈ p da/a + q db/b. A conservative worst-case bound replaces each contribution by its absolute magnitude. Independent random standard uncertainties instead often combine in quadrature. The correct rule depends on what the uncertainty numbers represent, and correlations matter.
First-order propagation is an approximation. If an uncertainty is large, compute the endpoint bounds or use a more appropriate uncertainty model. If two quantities share the same calibration error, treating them as independent may overestimate or underestimate the result. In a ratio measured on one scale, a common multiplicative calibration factor may cancel.
Δ(a ± b) ≤ Δa + Δb
Maximum absolute uncertainty bound; the uncertainties are nonnegative.
Δy/|y| ≈ |p| Δa/|a| + |q| Δb/|b|
First-order worst-case relative bound for y = C aᵖbᑫ with nonzero a, b and exact C.
ORIGINAL WORKED EXAMPLE
Density of a wire
A uniform cylindrical wire has mass measured to 1%, length to 0.5% and diameter to 2%. Estimate the maximum percentage uncertainty in density.
For a cylinder, ρ = 4m/(πd²l). The diameter enters squared because it determines cross-sectional area.
For small maximum errors, Δρ/ρ ≈ Δm/m + 2Δd/d + Δl/l.
Insert the percentages: 1 + 2(2) + 0.5 = 5.5%. The diameter measurement is the dominant contribution.
Halving the diameter uncertainty to 1% would reduce the bound to 3.5%, whereas eliminating the length uncertainty entirely would improve it only to 5%.
Answer Approximately 5.5%; improving diameter measurement is most effective.
06 / UNDERSTAND & APPLY
6. Read a vernier or screw gauge as two coordinated scales
A vernier resolves a fraction of a main-scale division by comparing two slightly different spacings. For a standard direct vernier with n vernier divisions equal to n − 1 main-scale divisions, one vernier division is smaller by one nth of a main division. The least count is therefore main-scale division divided by n. More generally use the actual difference between the scale spacings, not a memorised denominator.
Read the main-scale value immediately before the vernier zero, then add the index of the coincident vernier mark multiplied by the least count. View the marks perpendicularly to avoid parallax. Make gentle contact with the object and check that the jaw faces are appropriate for the dimension: outside diameter, inside diameter and depth use different contacts.
A micrometer screw gauge converts rotation to translation. Pitch is the axial distance moved in one complete turn; divide pitch by the number of circular-scale divisions to obtain least count. Add the sleeve reading, including any visible half-millimetre mark, to circular-scale reading times least count. The ratchet helps make the contact force reproducible; overtightening can compress an object.
Zero error is the reading when the true quantity should be zero. Corrected reading equals observed reading minus signed zero error. A positive zero error must be subtracted; a negative zero error must be added in magnitude. Backlash occurs when reversing a screw leaves a small rotation without translation. Approach the final reading consistently from one direction when the procedure requires it.
Full-size scientific diagram. Swipe or scroll horizontally; keyboard users can focus the diagram and use arrow keys.
Quantitative scale geometry. Ten vernier divisions equal nine main-scale divisions. No zero error is included.
ORIGINAL WORKED EXAMPLE
Signed zero correction
A vernier has 20 divisions spanning 19 mm. The main reading is 12 mm and the seventh vernier mark coincides. Closed jaws read +0.10 mm. What is the corrected length?
One main division is 1 mm; one vernier division is 19/20 mm. Their difference is 0.05 mm.
Observed reading = 12 + 7(0.05) = 12.35 mm.
Subtract the signed zero error: 12.35 − 0.10 = 12.25 mm.
Answer 12.25 mm.
ORIGINAL WORKED EXAMPLE
A screw gauge with negative zero error
A screw advances 0.5 mm per turn and has 50 circular divisions. The sleeve shows 3.5 mm and the circular reading is 28. Its signed zero error is −0.03 mm.
Least count = 0.5/50 = 0.01 mm.
Observed thickness = 3.5 + 28(0.01) = 3.78 mm.
Corrected thickness = 3.78 − (−0.03) = 3.81 mm.
Answer 3.81 mm.
07 / UNDERSTAND & APPLY
7. Report the information the measurement supports
Significant figures encode meaningful digits in a stated value, though an explicit uncertainty is clearer. Leading zeros locate the decimal point and are not significant. Zeros between nonzero digits count. Trailing zeros after a decimal point count. An integer such as 1500 is ambiguous without context; 1.50 × 10³ states three significant figures unambiguously.
For elementary reporting rules, addition and subtraction are limited by decimal place, while multiplication and division are limited by the least number of significant figures among measured inputs. Exact counts and defined conversions do not restrict precision. Keep guard digits through intermediate calculations and round at the end; repeated early rounding can change the final result.
Instrument least count and significant figures are connected but not interchangeable. A display with many digits may have poor calibration, and averaging may support an estimate between adjacent scale marks without eliminating systematic uncertainty. Do not report seven decimal places merely because a calculator displays them. When uncertainty is supplied, it should control the final rounding.
08 / UNDERSTAND & APPLY
8. Original practice: choose the rule before calculating
These problems are independently written practice. State the unit convention or uncertainty model before calculating, then use the explanation to check both the method and the numerical result.
ORIGINAL PRACTICE · 1
Convert 72 km h⁻¹ to m s⁻¹.
Reveal answer & reasoning
20 m s⁻¹.
Multiply by 1000 m/km and divide by 3600 s/h: 72 × 1000/3600 = 20.
ORIGINAL PRACTICE · 2
Find the dimensions of surface tension, defined as force per unit length.
Reveal answer & reasoning
MT⁻².
Divide MLT⁻² by L. The same dimensions follow from energy per area.
ORIGINAL PRACTICE · 3
In E = A exp(−x/b), E is energy and x is position. What are [A] and [b]?
Reveal answer & reasoning
[A] = ML²T⁻²; [b] = L.
The exponential is dimensionless; its argument x/b must be dimensionless, and A supplies the energy dimension.
ORIGINAL PRACTICE · 4
A square side is (10.0 ± 0.1) cm. Estimate the first-order maximum percentage uncertainty in area.
Reveal answer & reasoning
2%.
The side uncertainty is 1%; A = l² doubles the relative bound. Exact upper and lower areas are 102.01 and 98.01 cm².
ORIGINAL PRACTICE · 5
Two lengths are (20.0 ± 0.1) cm and (19.5 ± 0.1) cm. Find their difference and maximum uncertainty.
Reveal answer & reasoning
(0.5 ± 0.2) cm; 40% relative uncertainty.
Subtract central values but add maximum absolute errors. The small difference is much less well determined fractionally.
ORIGINAL PRACTICE · 6
A vernier has 50 divisions spanning 49 main divisions; each main division is 1 mm. Find its least count.
Reveal answer & reasoning
0.02 mm.
The spacing difference is 1 − 49/50 = 1/50 mm.
ORIGINAL PRACTICE · 7
A micrometer observed reading is 2.46 mm and signed zero error is +0.04 mm. What is the corrected reading?
Reveal answer & reasoning
2.42 mm.
Correction is the negative of the zero error: corrected = observed − error.
ORIGINAL PRACTICE · 8
Can dimensional analysis distinguish torque from work? Can it determine the coefficient in K = Cmv²?
Reveal answer & reasoning
No in both cases.
Torque and work both have ML²T⁻² dimensions, although their vector/scalar meanings differ. C is dimensionless and must come from physics or experiment.
ORIGINAL PRACTICE · 9
For g = 4π²l/T², l has 1% maximum error and T has 0.4%. Estimate the maximum percentage error in g.
Reveal answer & reasoning
1.8%.
The length contributes 1%; the squared period contributes 2 × 0.4%. Add magnitudes for a first-order maximum bound.
ORIGINAL PRACTICE · 10
A set of repeated measurements is tightly clustered but all exceed a calibrated reference by 4%. What improvement is most directly needed?
Reveal answer & reasoning
Investigate and correct the systematic bias.
Additional repeats alone cannot eliminate a persistent zero or scale error. The data show good repeatability but poor agreement with the reference.
09 / UNDERSTAND & APPLY
9. A measurement checklist that prevents lost marks
Before trusting a calculation, inspect four layers: the physical model, the dimensional structure, the measured inputs and the reported precision. A dimensionally correct formula with the wrong geometry is still wrong. A physically correct formula with a centimetre-to-metre mistake can fail by six orders of magnitude when volume is involved.
For an instrument problem, write the least count explicitly, separate observed and corrected readings, and keep the zero-error sign visible. For propagation, identify whether the question asks a maximum bound, mean absolute error or standard uncertainty. For scaling, state which quantities are held fixed. These short declarations expose the assumptions an examiner is testing.
Can you derive unfamiliar dimensions from a definition rather than recall?
Can you explain why a logarithm needs a dimensionless argument?
Can you distinguish resolution, precision, accuracy and uncertainty?
Can you correct a negative zero error without reversing its sign twice?
Can you predict which measurement dominates a product or power?
Can you explain what dimensions cannot tell you?
KEEP THE BIG IDEAS
Quick revision
Units describe a chosen standard; dimensions describe the kind of quantity.
Consistency rejects impossible equations but does not prove a model.
Uncertainty must name a convention and include relevant systematic effects.
A zero correction has the opposite sign to the zero error.
The teaching notes, worked explanations and original practice are written for this lesson. Separately labeled selected exam questions identify their exam, year, source and whether wording is paraphrased or an exact excerpt. Educational references do not certify current exam coverage.