Approximation and Error (AI HL)

An approximate value differs from the true value by an error, and the percentage error expresses that error as a percentage of the true value. A rounded number also has upper and lower bounds, such as 51.5≤m<52.551.5\le m<52.5 when m=52m=52 to the nearest whole number. The topic covers percentage error, bounds, and how errors grow when a measurement is used in a calculation. It's part of the broader Standard Form & Approximation topic.

21 questions on this sub-topic.

Syllabus

CodeSyllabus content
SL1.6Approximation: decimal places, significant figures - choosing an appropriate degree of accuracy for given data. Upper and lower bounds of rounded numbers: if x=4.1x=4.1 to one decimal place, then 4.05≤x<4.154.05\le x<4.15. Percentage errors, including measurement and rounding errors. Estimation, and recognising whether a calculated result is reasonable.

Official IB source: Mathematics: applications and interpretation subject brief (PDF, ibo.org).

Skills practised on this page:

Percentage error and compounding

Percentage error

ε=|vA−vEvE|×100%\varepsilon = \left|\dfrac{v_A-v_E}{v_E}\right|\times100\%

Always divide by the exact (true) value vEv_E, never the approximate value vAv_A.

✓ In the formula booklet

Compounding error

When a measured quantity is raised to a power in a formula (e.g. V∝r3V\propto r^3), the percentage error roughly multiplies by that power - a small error in radius becomes a larger error in volume.

Not in the formula booklet - consequence of the formula

Need bounds notation and the full syllabus wording? See Standard Form & Approximation.

Key terms

What are the bounds of a rounded number?

The lower and upper bounds are the smallest and largest values that would round to the given number. They lie half a unit of the rounding accuracy either side of it, and the upper bound is not included.

e.g. A mass of 52 kg to the nearest kg satisfies 51.5≤m<52.551.5\le m<52.5.

What is the size of an error?

The error is the difference between the approximate value and the exact value. Its size is taken as positive and is measured in the same units as the quantity.

e.g. A length of exactly 8.0 cm that is measured as 8.2 cm has an error of 0.2 cm.

What is a rounding error?

A rounding error is the error that comes from replacing a number by a rounded version of it. The more decimal places or significant figures kept, the smaller it is.

e.g. Rounding 3.141593.14159 to 3.143.14 gives a rounding error of 0.001590.00159.

How do you check that a result is reasonable?

Round each number in the calculation to one significant figure, or to an easy value, and do the estimate mentally. The estimate should be close to the calculated answer.

e.g. 49.7×2.1≈50×2=10049.7\times2.1\approx50\times2=100, so a calculated answer of 104.37104.37 is reasonable.

Worked examples

1
Easy
Calculator
[4 marks]

A quantity has true value 246000.246\,000.

(a) Write it in the form a×10ka \times 10^{k}.

(b) It is measured as 250000.250\,000. Find the percentage error, to 3 significant figures.

Worked solution

(a) 246000246\,000 lies between 10510^{5} and 10610^{6}, so k=5.k=5. M1
246000=2.46×105.246\,000 = 2.46\times10^{5}. A1

(b) ε=|vA−vEvE|×100%=250000−246000246000×100%.\varepsilon = \left|\dfrac{v_A-v_E}{v_E}\right|\times100\% = \dfrac{250\,000-246\,000}{246\,000}\times100\%. M1

ε=4000246000×100%≈1.63%.\varepsilon = \dfrac{4000}{246\,000}\times100\% \approx 1.63\%. A1

M1 Identify the power of 10 A1 State the value in standard form M1 Set up the percentage error formula A1 ≈1.63%\approx1.63\%
2
Medium
Calculator
[5 marks]

A square tile is measured as having side 2020 cm, but the true side is 19.619.6 cm.

(a) Find the percentage error in the side measurement.

(b) Find the percentage error in the calculated area, to 3 significant figures.

Worked solution

(a) ε=|vA−vEvE|×100%=20−19.619.6×100%.\varepsilon = \left|\dfrac{v_A-v_E}{v_E}\right|\times100\% = \dfrac{20-19.6}{19.6}\times100\%. M1

ε=0.419.6×100%≈2.04%.\varepsilon = \dfrac{0.4}{19.6}\times100\% \approx 2.04\%. A1

(b) Measured area =202=400=20^2=400 cm2^2; true area =19.62=384.16=19.6^2=384.16 cm2^2. M1
ε=400−384.16384.16×100%.\varepsilon = \dfrac{400-384.16}{384.16}\times100\%. A1
ε=15.84384.16×100%≈4.12%.\varepsilon = \dfrac{15.84}{384.16}\times100\% \approx 4.12\%. A1

M1 Set up the percentage error formula A1 ≈2.04%\approx2.04\% M1 Find the measured and true areas A1 Set up the percentage error in the area A1 ≈4.12%\approx4.12\%
3
Hard
GDC
[4 marks]

A value 1.0051.005 is rounded to 1.01.0 and used 200 times in a sum.
Estimate the total absolute error.

Worked solution

Each term loses |1.005−1.0|.|1.005-1.0|. M1
|1.005−1.0|=0.005.|1.005-1.0| = 0.005. A1
Total error =200×0.005.=200\times0.005. M1
200×0.005=1.000≈1.0.200\times0.005 = 1.000 \approx 1.0. A1On the GDC. Estimate the total absolute error. The calculator shows about 1.00.

TI-84 Plus CE: 2nd mode then type 200*0.005 enterCasio fx-CG50: MENU Run-Matrix then type 200*0.005 EXECasio fx-CG100: HOME Calculate then type 200*0.005 EXETI-Nspire CX: home Calculator then type 200*0.005 enterFull GDC guide: Enter a calculation on the home screen Topic page

M1 Find the error per term A1 =0.005=0.005 M1 Set up the total over 200 terms A1 Correct answer of ≈1.0\approx1.0

Common mistakes

  • Dividing by the wrong value in the percentage error formula. The denominator must always be the exact (true) value, never the approximate one - swapping them gives a different, incorrect percentage.
  • Forgetting that error compounds through a power. When a linear measurement feeds into an area or volume formula, the resulting percentage error is not the same size as the original - a 2% error in a side length gives roughly double that in the area, not 2% again.
  • Halving the tolerance instead of using it directly. A value rounded to the nearest 0.10.1 has bounds ±0.05\pm0.05, not ±0.1\pm0.1 - mixing these up shifts every upper/lower bound calculation that follows.

Using your GDC

Every step below is a real button sequence, not a vague "use your calculator" hint - covering the TI-84 Plus, TI-Nspire, and Casio fx-9860/fx-CG50. Pick your model to filter down to just the steps that apply to you.

Show steps for:
Enter scientific notation (standard form)

For very large or very small numbers - avoids typing long strings of zeros and prevents rounding errors.

  1. Scientific notation means a×10na\times10^n, e.g. 3.2×1083.2\times10^{8} or 4.5×10−34.5\times10^{-3}.
  2. Use 2nd → , (EE) to enter the ×10 part: type 3.2 2nd , 8 to enter 3.2×1083.2\times10^{8}. Do NOT type ×10^ separately.TI-84
  3. Use the EE key (or type ×10^ from the keyboard template) to enter scientific notation. Or just type 3.2×1083.2\times10^{8} using the ^ key.Nspire
  4. Use the ×10ˣ key (EXP key) - type 3.2 then EXP then 8. Do NOT type ×10^ manually.Casio
  5. To display answers in scientific notation: on TI-84 press MODE and choose SCI; on Casio set the display mode in SET UP.

Tip: A common mistake is typing ×10^ instead of using the EE/EXP key - this gives ×10×... (multiplication, then a power) rather than proper scientific notation.

Bounds and percentage-error questions don't need a special calculator routine - they're ordinary arithmetic once you've set up the expression. Type the working directly on the home screen (or its equivalent), and switch the display mode to Sci or Norm using the steps above if the question specifically asks for standard form.

See the full GDC guide for more calculator models and topics.

Quick questions

How do you calculate percentage error?

ε=|vA−vEvE|×100%\varepsilon = \left|\dfrac{v_A-v_E}{v_E}\right|\times100\%, where vAv_A is the approximate value and vEv_E is the exact value. Always divide by the exact value.

Why does error compound when a measurement is raised to a power?

If a formula raises a measured quantity to a power, such as V∝r3V\propto r^3, a small percentage error in the measured quantity roughly multiplies by that power in the final answer - a 2%2\% error in radius becomes roughly a 6%6\% error in volume.

Is the upper bound of a rounded number included?

No. The upper bound is not included because it would round up to the next value. For x=4.1x=4.1 to one decimal place, x<4.15x<4.15, since 4.154.15 rounds to 4.24.2.

Can a percentage error be more than 100%?

Yes, when the error is larger than the exact value. If the exact value is 22 and the approximate value is 55, the percentage error is |5−22|×100%=150%\left|\dfrac{5-2}{2}\right|\times100\%=150\%.

Should I round before or after finding the percentage error?

Keep full values throughout the working and round only the final answer to the accuracy asked for. Rounding early can change the last digits of the answer.

Ready to practise properly?

21 approximation-and-error questions, marked instantly like the real exam.