Errors in Measurement

Error analysis is the single highest-yield sub-topic of this chapter — boards regularly set a 3-mark numerical asking for the combined error in a derived quantity (like density or resistance). Mastering it also stops you over-stating precision all year.

Why it matters

Every measurement has some uncertainty. Accuracy is closeness to the true value; precision is closeness of repeated readings to each other. The two are independent — a reading can be precise but inaccurate.

Types of error

• Systematic errors — one-directional, from instrument zero-error, imperfect technique or the environment. Reduced by correcting the instrument.
• Random errors — unpredictable in size and sign; reduced by taking many readings and averaging.
• Least count error — the smallest reading an instrument can resolve.

Quantifying error

Quantity	Formula
Mean value	a_mean = (a₁+a₂+…+aₙ)/n
Absolute error	Δaᵢ = mod(a_mean − aᵢ)
Mean absolute error	Δa_mean = (Σ mod(Δaᵢ))/n
Relative error	Δa_mean / a_mean
Percentage error	(Δa_mean / a_mean) × 100%

Combination rules (learn cold):

• Sum/difference Z = A ± B → ΔZ = ΔA + ΔB (absolute errors add).
• Product/quotient Z = AB or A/B → ΔZ/Z = ΔA/A + ΔB/B (relative errors add).
• Power Z = A^p·B^q / C^r → ΔZ/Z = p(ΔA/A) + q(ΔB/B) + r(ΔC/C).

Exam Tricks & Tips

• 🎯 In products/quotients, relative errors add; in sums/differences, absolute errors add. Never mix them.
• 🎯 A power multiplies the relative error — a term like r² contributes 2×(Δr/r), so the most-powered quantity needs the most careful measurement.
• 🎯 Even for a difference the errors ADD, never subtract — this is heavily tested.
• 🎯 Systematic error → reduce by better instrument/zero correction; random error → reduce by averaging repeats.
• 🎯 Percentage error of a result = sum of percentage errors of its factors (weighted by powers) — fast shortcut.
• ❌ Common mistake: subtracting errors for Z = A − B. Errors always add, so a small difference of two large numbers has a huge relative error.

Expected exam pattern

3-mark numerical: given Δ in measured quantities, find % error in a formula like ρ = m/V or R = V/I or T = 2π√(L/g). 1-mark: accuracy vs precision, or systematic vs random.

Quick recap

Accuracy ≠ precision. Systematic (correctable) vs random (average out). Sum/difference → add absolute errors; product/quotient/power → add relative errors weighted by the power.

Cover the answer, recall, then check. 12 cards on errors in measurement.

Q1. What is the difference between accuracy and precision?
A1. Accuracy = closeness to the true value; precision = closeness of repeated readings to one another. They are independent.

Q2. What causes systematic errors and how are they reduced?
A2. Instrument zero-error, faulty technique or environment; reduced by correcting/calibrating the instrument.

Q3. How are random errors reduced?
A3. By taking many readings and averaging — they are equally likely positive or negative.

Q4. How do errors combine for a sum or difference (Z = A ± B)?
A4. Absolute errors add: ΔZ = ΔA + ΔB.

Q5. How do errors combine for a product or quotient (Z = AB or A/B)?
A5. Relative errors add: ΔZ/Z = ΔA/A + ΔB/B.

Q6. How does a power affect the error, e.g. Z = A²?
A6. ΔZ/Z = 2(ΔA/A) — the exponent multiplies the relative error.

Q7. Define relative error and percentage error.
A7. Relative error = Δa_mean/a_mean; percentage error = that × 100%.

Q8. What is the least count error?
A8. The uncertainty equal to the smallest division (resolution) an instrument can measure.

Q9. Why is a difference of two nearly equal quantities risky?
A9. Absolute errors add while the result is small, so the relative error becomes very large.

Q10. How do you find the absolute error of a single reading?
A10. Δaᵢ = |a_mean − aᵢ| — magnitude of the deviation from the mean.

Q11. For Z = A^p B^q / C^r, write the relative error.
A11. ΔZ/Z = p(ΔA/A) + q(ΔB/B) + r(ΔC/C) — all terms added.

Q12. Can a measurement be precise but not accurate?
A12. Yes — repeated readings can agree closely yet all be off from the true value (e.g. uncorrected zero error).

No measurement is ever perfect — the honest scientist reports not just a value but its uncertainty. Knowing how errors combine is what separates a lucky guess from a reliable result.

Definition / core idea

An error is the difference between the measured and true value. Errors are of two families: systematic (consistent bias — same direction every time) and random (scatter — different each trial).

Deep explanation

Beginner — types of error

• Systematic: instrumental (zero error), imperfect technique, personal bias. Reduce by calibration.
• Random: unpredictable fluctuations; reduce by taking many readings and averaging.
• Gross: human blunders (misreading).

Intermediate — quantifying error

For n readings a₁…aₙ:

• Mean: a̅ = (Σaᵢ)/n
• Absolute error of each: |Δaᵢ| = |a̅ − aᵢ|
• Mean absolute error: Δa̅ = (Σ|Δaᵢ|)/n
• Relative error = Δa̅ / a̅
• Percentage error = (Δa̅ / a̅) × 100%

Advanced — combination (propagation) of errors

This is the exam favourite. If quantities have errors, the result's error is found by rules:

• Sum/Difference (Z = A ± B): absolute errors add → ΔZ = ΔA + ΔB.
• Product/Quotient (Z = AB or A/B): relative errors add → ΔZ/Z = ΔA/A + ΔB/B.
• Power (Z = A^p B^q / C^r): ΔZ/Z = p(ΔA/A) + q(ΔB/B) + r(ΔC/C). Powers multiply the fractional error — the biggest source of error is the quantity raised to the highest power.

Worked example

Density ρ = m/V. Mass m = 5.00 ± 0.01 g, volume V = 2.0 ± 0.1 cm³. Find % error in ρ.
Relative errors add: Δρ/ρ = Δm/m + ΔV/V = 0.01/5.00 + 0.1/2.0 = 0.002 + 0.05 = 0.052.
Percentage error = 5.2%. Notice the volume (5%) dominates — measure it more carefully to improve ρ.

Real-world application

Engineers add a safety factor precisely because measured loads carry error; a bridge is designed to survive its worst-case, not just its average, measurement.

Exam tricks & shortcuts

• "Add absolute errors for + and −, add relative errors for × and ÷."
• Powers count: g in T = 2π√(L/g) means error in g gets doubled (since g ∝ 1/T², and T is squared).
• The quantity with the highest power contributes most — improve that measurement first.

Key formulas

• Absolute error: Δaᵢ = |a_mean − aᵢ|.
• Mean absolute error: Δa_mean = (Σ|Δaᵢ|)/n.
• Relative error: Δa_mean / a_mean.
• Percentage error: (Δa_mean / a_mean) × 100%.
• Sum/difference Z = A ± B: ΔZ = ΔA + ΔB (absolute errors add).
• Product/quotient Z = AB or A/B: ΔZ/Z = ΔA/A + ΔB/B (relative errors add).
• Powers Z = Aᵖ·Bᵠ/Cʳ: ΔZ/Z = p·(ΔA/A) + q·(ΔB/B) + r·(ΔC/C).

Errors are classified as systematic (consistent bias) and random (scatter). Least count of the instrument sets the smallest measurable value.

Worked Example

Problem: The length of a rod is measured as l = 5.00 cm with an absolute error of ±0.01 cm, and its cross-sectional radius as r = 0.20 cm with an absolute error of ±0.01 cm. The volume is V = πr²l. Find the percentage error in the calculated volume.

Solution:
Step 1 — Write how each relative error contributes. For a product/power formula V = πr²l, the relative errors add, with the power of each quantity as its weight:
ΔV/V = 2(Δr/r) + (Δl/l).

Step 2 — Compute the relative error in radius (weighted by its power 2):
Δr/r = 0.01 / 0.20 = 0.05, so 2(Δr/r) = 2 × 0.05 = 0.10.

Step 3 — Compute the relative error in length:
Δl/l = 0.01 / 5.00 = 0.002.

Step 4 — Add the contributions:
ΔV/V = 0.10 + 0.002 = 0.102.

Step 5 — Convert to a percentage:
Percentage error = 0.102 × 100 = 10.2%.

Answer: The percentage error in the volume is about 10.2% (dominated by the radius term).