Significant Figures in Uncertainty

The Two Golden Rules

RULE 1

Round the uncertainty to 1 or 2 significant figures

The uncertainty itself should be reported to 1 significant figure in most cases, or 2 significant figures when the leading digit is 1 or 2. Never report an uncertainty to 4 or 5 significant figures - false precision in the uncertainty is meaningless.

RULE 2

Round the result to match the uncertainty

The last digit of your result must align with the last digit of your uncertainty. If your uncertainty is ±0.3, your result should be reported to the nearest 0.1. If your uncertainty is ±20, your result should be rounded to the nearest 10.

These two rules apply universally - to every measurement, every propagated uncertainty, and every final result in experimental physics. Violating them is one of the most penalised errors in university lab reports.

What is a Significant Figure?

A significant figure is any digit that carries meaning about the precision of a measurement. The rules for counting them:

1.All non-zero digits are significant (e.g. 3.45 has 3 sig figs)
2.Zeros between non-zero digits are significant (e.g. 3.05 has 3 sig figs)
3.Leading zeros are never significant (e.g. 0.0045 has 2 sig figs)
4.Trailing zeros after a decimal point are significant (e.g. 3.50 has 3 sig figs)
5.Trailing zeros in a whole number are ambiguous - use scientific notation to be clear (e.g. 3500 could be 2, 3, or 4 sig figs; write 3.5 × 10³ for 2 sig figs)

Step-by-Step Rounding Procedure

Step 1: Compute the uncertainty

Calculate the propagated uncertainty Δf using the appropriate propagation rule. Do not round yet - keep all digits during intermediate calculations.

Step 2: Round the uncertainty

Identify the first significant figure of Δf. Round to 1 significant figure (or 2 if the leading digit is 1 or 2). This gives you Δf_rounded.

Step 3: Match the result

Round the result f to the same decimal place as Δf_rounded. Do not round to a fixed number of significant figures - round to match the uncertainty.

Step 4: Write the final result

Report as: f ± Δf (units). Both numbers must have the same last decimal place.

Worked Examples

Example 1 - Basic rounding

Computed result: f = 23.4872 J Computed uncertainty: Δf = 1.3284 J

Step 1: Round uncertainty → 1.3284 rounds to 1 (1 sig fig) or 1.3 (2 sig figs, since leading digit is 1) Step 2: Match result → round 23.4872 to nearest 0.1 → 23.5

f = 23.5 ± 1.3 J ✓

Example 2 - Small uncertainty

Computed result: f = 9.81374 m/s² Computed uncertainty: Δf = 0.02341 m/s²

Step 1: Round uncertainty → 0.02 (1 sig fig) or 0.023 (2 sig figs, since leading digit is 2) Step 2: Match result → round 9.81374 to nearest 0.001 → 9.814

f = 9.814 ± 0.023 m/s² ✓

Example 3 - Large uncertainty

Computed result: f = 4523.7 kg Computed uncertainty: Δf = 342.8 kg

Step 1: Round uncertainty → 300 (1 sig fig) Step 2: Match result → round 4523.7 to nearest 100 → 4500

f = 4500 ± 300 kg ✓
Or in scientific notation: (4.5 ± 0.3) × 10³ kg ✓

Example 4 - The leading 1 rule

Computed result: f = 0.08734 s Computed uncertainty: Δf = 0.01823 s

Step 1: Leading digit of uncertainty is 1 → keep 2 sig figs → 0.018 Step 2: Match result → round 0.08734 to nearest 0.001 → 0.087

f = 0.087 ± 0.018 s ✓

Right and Wrong

Wrong ✕	Right ✓	Problem
20.0 ± 1.0823 J	20.0 ± 1.1 J	Too many sig figs in uncertainty
20.0000 ± 1.1 J	20.0 ± 1.1 J	Result has too many decimal places
20 ± 1.1 J	20.0 ± 1.1 J	Result rounded too aggressively
9.81 ± 0.1000 m/s²	9.81 ± 0.10 m/s²	Trailing zeros imply false precision
4500 ± 342 kg	4500 ± 300 kg	Uncertainty has too many sig figs
0.0873 ± 0.02 s	0.087 ± 0.020 s	Decimal places don't match

When to Use Scientific Notation

Scientific notation is strongly recommended when:

The result and uncertainty span very different orders of magnitude
Trailing zeros in a whole number create ambiguity
The leading zeros in a small number make the significant figures unclear

For example, instead of writing 4500 ± 300 kg (ambiguous - how many sig figs in 4500?), write (4.5 ± 0.3) × 10³ kg. This is unambiguous and cleaner.

The standard form is: (result ± uncertainty) × 10ⁿ, where both the result and uncertainty are expressed with the same power of 10.

✓ (1.23 ± 0.04) × 10⁵ Pa

✓ (9.11 ± 0.02) × 10⁻³¹ kg

✓ (2.998 ± 0.001) × 10⁸ m/s

Common Mistakes to Avoid

Mistake 1: Rounding the result before the uncertainty

Always round the uncertainty first, then match the result to it. If you round the result first, you may round to the wrong decimal place. The uncertainty sets the precision - the result follows.

Mistake 2: Reporting uncertainty to too many significant figures

Writing ± 1.0823 J implies you know the uncertainty to 5 significant figures, which is never the case in experimental physics. Round to 1 or 2 significant figures. The uncertainty itself has uncertainty.

Mistake 3: Mismatched decimal places

If Δf = ± 0.3 J, then f must be reported to the nearest 0.1 J - not 20.47 ± 0.3 J (too many decimal places in result) or 20 ± 0.3 J (too few). The last significant digit of the result must align with the uncertainty.

Mistake 4: Rounding intermediate results

Round only at the very end. Rounding partial derivatives, intermediate sums, or individual uncertainty contributions introduces compounding rounding errors. Keep full precision throughout and round only the final result.

Mistake 5: Confusing significant figures with decimal places

Significant figures and decimal places are different concepts. 0.0023 has 2 significant figures but 4 decimal places. Always think in terms of which figures carry meaning, not how many digits appear after the decimal point.

Quick Reference

Situation	Rule
Uncertainty leading digit is 3–9	Round to 1 sig fig
Uncertainty leading digit is 1 or 2	Round to 2 sig figs
Rounding the result	Match last decimal place of rounded uncertainty
Intermediate calculations	Keep all digits - do not round
Ambiguous trailing zeros	Use scientific notation
Percentage uncertainty	Δf/f × 100% - useful sanity check

Frequently Asked Questions

Q1How many significant figures should I use for uncertainty?

Typically 1 significant figure. Use 2 significant figures when the leading digit of the uncertainty is 1 or 2, since rounding would otherwise cause a large relative change (e.g. rounding 0.019 to 0.02 is a 5% change in the uncertainty itself). Your lab manual may specify - always follow local guidelines.

Q2Should I round during the calculation or only at the end?

Only at the very end. Keep all significant figures during intermediate steps to avoid compounding rounding errors. This is called carrying guard digits. Round only when writing the final reported result.

Q3What if my uncertainty is larger than my result?

Report it honestly. For example, f = 0.5 ± 0.8 J is a valid result - it simply means the measurement is consistent with zero, and the experiment needs improvement. Rounding both to 1 sig fig: f = 0.5 ± 0.8 J is correct.

Q4My result is 12345.6789 and my uncertainty is 0.4. How do I round?

Round the uncertainty to 0.4 (1 sig fig, already done). Then round the result to the nearest 0.1: 12345.7. Report as 12345.7 ± 0.4. The large number of digits before the decimal is fine - what matters is that the last decimal place matches.

Q5Is it wrong to report too few significant figures?

Yes. Under-reporting is just as wrong as over-reporting. Writing f = 20 ± 1.1 J loses a meaningful digit in the result. The correct form is f = 20.0 ± 1.1 J - the trailing zero is significant and must be shown.

The calculator automatically shows results with correct significant figures and rounding - so you can check your own lab report against the proper format instantly.

Open the Calculator →