Significant Figures and Error Analysis

Why Every Measurement Is Uncertain

No measuring instrument is perfect. When you measure the length of a pencil with a ruler, the measurement might be 12.3 cm — but is it exactly 12.3? Could it be 12.27 or 12.34? The answer depends on the precision of the ruler.

Significant figures tell us how precisely a number is known. Error analysis tells us how uncertainties in individual measurements combine when we calculate derived quantities.

This chapter appears in Class 11 Physics and is foundational for JEE. Errors account for 1–2 marks in JEE Main and form the basis for experimental section questions in NEET practical-based MCQs.

Key Terms & Definitions

Significant figures (sig figs): All the digits in a measurement that are known with certainty, plus one estimated (uncertain) digit. They reflect the precision of the measuring instrument.

Absolute error: The magnitude of the difference between the true (mean) value and the measured value. Unit: same as the measured quantity.

Relative error: Absolute error divided by the mean value. Dimensionless.

Percentage error: Relative error multiplied by 100. Expressed as a percentage.

Accuracy: How close a measurement is to the true value (relates to systematic errors).

Precision: How close repeated measurements are to each other (relates to random errors).

Significant Figures — Rules

Counting Sig Figs

The rules seem arbitrary until you understand the logic: each rule exists because of how we write numbers, not because of the physics.

Rule 1 — All non-zero digits are significant. Example: 12.34 has 4 sig figs. 739 has 3 sig figs.

Rule 2 — Zeros between non-zero digits are significant. Example: 1005 has 4 sig figs. 30.07 has 4 sig figs.

Rule 3 — Leading zeros (before the first non-zero digit) are NOT significant. Example: 0.0045 has 2 sig figs (the leading zeros just fix the decimal point). 0.00309 has 3 sig figs.

Rule 4 — Trailing zeros in a number WITH a decimal point ARE significant. Example: 3.00 has 3 sig figs. 12.50 has 4 sig figs. (The trailing zero tells us we measured to the hundredths place.)

Rule 5 — Trailing zeros in a whole number WITHOUT a decimal point are AMBIGUOUS. Example: 1200 could have 2, 3, or 4 sig figs. To remove ambiguity, use scientific notation: $1.2 \times 10^3$ (2 sig figs), $1.20 \times 10^3$ (3 sig figs), $1.200 \times 10^3$ (4 sig figs).

The cleanest way to specify sig figs in a whole number is scientific notation. Whenever you write “1200 m” in an answer, your teacher/examiner doesn’t know if it’s 2, 3, or 4 sig figs. Writing $1.20 \times 10^3$ m makes it unambiguous — 3 sig figs.

Sig Figs in Calculations

Addition/Subtraction: The result should have the same number of decimal places as the measurement with the fewest decimal places.

Example: $12.1 + 0.034 + 4.35 = 16.484$ . But $12.1$ has only 1 decimal place, so the answer rounds to $16.5$ .

Multiplication/Division: The result should have the same number of sig figs as the measurement with the fewest sig figs.

Example: $3.14 \times 2.6 = 8.164$ . But $2.6$ has only 2 sig figs, so the answer is $8.2$ .

Types of Errors

Systematic Errors

These occur consistently in one direction — always too high or always too low. They cannot be reduced by repeating measurements.

Sources:

Instrumental error: Zero error in a vernier calliper, least count error in a screw gauge
Personal error: The habit of always reading a scale from the same angle (parallax)
Environmental error: Temperature change affecting a spring balance

Systematic errors affect accuracy. They can be detected by comparing with a standard instrument.

Random Errors

These fluctuate randomly around the true value — sometimes too high, sometimes too low. They can be reduced by taking multiple measurements and averaging.

Sources: slight variations in experimental conditions, random fluctuations in the observer’s judgement.

Random errors affect precision. Taking $n$ measurements reduces random error by a factor of $\sqrt{n}$ .

Gross Errors (Blunders)

Mistakes in reading or recording. Example: writing 1.3 instead of 13. These are eliminated by careful observation and checking.

Error Propagation — The Key Formulas

\text{Absolute error} = |\bar{x} - x_i|

\text{Mean absolute error} = \frac{1}{n}\sum_{i=1}^n |\bar{x} - x_i|

\text{Relative error} = \frac{\Delta \bar{x}}{\bar{x}}

\text{Percentage error} = \frac{\Delta \bar{x}}{\bar{x}} \times 100\%

For addition/subtraction ( $Z = A \pm B$ ):

\Delta Z = \Delta A + \Delta B

(absolute errors add)

For multiplication/division ( $Z = \frac{A \times B}{C}$ ):

\frac{\Delta Z}{Z} = \frac{\Delta A}{A} + \frac{\Delta B}{B} + \frac{\Delta C}{C}

(relative errors add)

For powers ( $Z = A^n$ ):

\frac{\Delta Z}{Z} = n \cdot \frac{\Delta A}{A}

(relative error multiplies by the exponent)

Key principle: Errors ALWAYS add (never cancel). We consider worst-case scenario.

The reason errors always add in multiplication/division: we are asking “what is the maximum possible error?” The worst case is when all errors push the result in the same direction.

Solved Examples

Easy — CBSE Level

Q: In measuring the refractive index of glass, a student gets values 1.44, 1.42, 1.45, 1.43, 1.46. Calculate the mean value, mean absolute error, and percentage error.

Solution:

Mean: $\bar{\mu} = \frac{1.44 + 1.42 + 1.45 + 1.43 + 1.46}{5} = \frac{7.20}{5} = 1.44$

Absolute errors: |1.44−1.44|=0.00, |1.42−1.44|=0.02, |1.45−1.44|=0.01, |1.43−1.44|=0.01, |1.46−1.44|=0.02

Mean absolute error: $\Delta\mu = \frac{0.00+0.02+0.01+0.01+0.02}{5} = \frac{0.06}{5} = 0.012$

Percentage error: $\frac{0.012}{1.44} \times 100\% = 0.83\%$

Result: $\mu = 1.44 \pm 0.01$ (rounded to the place of the error)

Medium — JEE Main Level

Q: The resistance $R$ of a wire is calculated from $R = V/I$ where $V = (100 \pm 5)$ V and $I = (10 \pm 0.2)$ A. Find the percentage error in $R$ .

Solution:

$\frac{\Delta R}{R} = \frac{\Delta V}{V} + \frac{\Delta I}{I} = \frac{5}{100} + \frac{0.2}{10} = 0.05 + 0.02 = 0.07$

Percentage error in $R = 7\%$ .

Mean value: $R = V/I = 100/10 = 10\ \Omega$ . So $R = (10 \pm 0.7)\ \Omega$ .

Hard — JEE Advanced Level

Q: The period of oscillation of a simple pendulum is $T = 2\pi\sqrt{L/g}$ . The length is measured with an error of 2% and the time for 20 oscillations is measured with an error of 1%. Find the error in $g$ .

Solution:

From $T = 2\pi\sqrt{L/g}$ , we get $g = \frac{4\pi^2 L}{T^2}$ .

\frac{\Delta g}{g} = \frac{\Delta L}{L} + 2\cdot\frac{\Delta T}{T} = 2\% + 2 \times 1\% = 4\%

(Note: the error in $T^2$ is $2 \times$ error in $T$ , by the power rule. The $4\pi^2$ is a constant with no error.)

Exam-Specific Tips

CBSE Class 11: Focus on types of errors, formulas for combination of errors, and the procedure for finding mean value and mean absolute error from a data set.

JEE Main: Percentage error questions are very common. Know the rules for addition, multiplication, and powers cold. A 2-mark MCQ will often give you a formula like $\rho = \frac{m}{V}$ and ask for the percentage error given errors in $m$ and $V$ .

NEET: Least count of common instruments (vernier calliper: 0.01 mm, screw gauge: 0.001 mm) and the concept of least count error appear in practical-based questions.

JEE Main 2024 asked: “If the percentage errors in measuring length and time period are 2% and 1% respectively, what is the percentage error in the calculated value of $g$ from $g = 4\pi^2 L/T^2$ ?” This is a direct application of the propagation formula: $\Delta g/g = \Delta L/L + 2(\Delta T/T) = 2\% + 2\% = 4\%$ .

Common Mistakes to Avoid

Mistake 1: In powers, students write $\frac{\Delta(A^n)}{A^n} = n^2 \cdot \frac{\Delta A}{A}$ (squaring the exponent). The correct rule is multiplication: $\frac{\Delta(A^n)}{A^n} = n \cdot \frac{\Delta A}{A}$ .

Mistake 2: In addition ( $Z = A + B$ ), trying to add relative errors. Relative errors should only be added for multiplication/division. For addition/subtraction, add the absolute errors: $\Delta Z = \Delta A + \Delta B$ .

Mistake 3: Counting zeros incorrectly. The number 0.003 does NOT have 3 sig figs — it has only 1. The leading zeros are just placeholders.

Mistake 4: Writing the answer to more decimal places than justified. If your least precise input has 3 sig figs, your answer should also have 3 sig figs. A calculated answer of 12.34567 from data with 3 sig figs should be written as 12.3.

Mistake 5: Confusing accuracy and precision. A dartboard analogy: all darts clustered together but far from the bullseye = precise but not accurate (systematic error). Darts scattered randomly around the bullseye = accurate but not precise (random error).

Practice Questions

1. How many significant figures are in 0.00502?

3 significant figures. The leading zeros (0.00) are not significant — they are just placeholders. The digits 5, 0, and 2 are significant. Note that the zero between 5 and 2 IS significant because it is between two non-zero digits.

2. Add 13.5 + 1.234 + 1.05 and express the answer to correct sig figs.

Numerical sum: 13.5 + 1.234 + 1.05 = 15.784. For addition, we keep the fewest decimal places. 13.5 has 1 decimal place; 1.234 has 3; 1.05 has 2. Answer: 15.8 (1 decimal place).

3. The diameter of a sphere is measured as $(4.0 \pm 0.2)$ cm. Find the percentage error in its volume.

$V = \frac{4}{3}\pi r^3 = \frac{\pi}{6} d^3$ . So $\frac{\Delta V}{V} = 3 \cdot \frac{\Delta d}{d} = 3 \times \frac{0.2}{4.0} = 3 \times 0.05 = 0.15 = 15\%$ .

4. What is the difference between least count error and zero error?

Least count error is the minimum error in any measurement, equal to the least count (smallest division) of the measuring instrument. It is a random error and cannot be avoided — you can only minimise it by using an instrument with a smaller least count. Zero error is a systematic error that occurs when the instrument does not read zero when it should (e.g., vernier calliper jaws closed but reading ≠ 0). Zero error is constant and must be subtracted from every reading.

5. In the formula $g = 4\pi^2 l/T^2$ , if the error in $l$ is $a\%$ and error in $T$ is $b\%$ , what is the error in $g$ ?

$\frac{\Delta g}{g} = \frac{\Delta l}{l} + 2\cdot\frac{\Delta T}{T} = a\% + 2b\%$ . The factor of 2 comes because $T$ appears as $T^2$ — the power rule multiplies the relative error by the exponent.

FAQs

Q: Why do errors always add and never cancel in error propagation? Error propagation assumes worst-case scenario: all individual errors conspire to push the result in the same direction. In practice, errors partially cancel, but we cannot guarantee this. So we take the conservative (maximum possible) estimate.

Q: Can percentage error exceed 100%? Yes. If a measurement has a very large uncertainty relative to its value, percentage error can exceed 100%. For example, measuring a very small length with an instrument that has a large least count can give percentage errors >100%.

Q: How do I report a measurement correctly? Write it as: (mean value) ± (mean absolute error), with both numbers having the same number of decimal places. Example: $(1.44 \pm 0.01)$ . The error tells the reader the last meaningful digit.