When a length is given as "12 cm to the nearest cm", it is not exactly 12 — it is anything that rounds to 12, i.e. from 11.5 up to (but not including) 12.5. Those two edges are the lower bound and upper bound. Real measurements are always rounded, so knowing how far off they might be is genuine, examinable maths.
1. Finding the bounds — half the last unit
- Halve the degree of rounding and add/subtract it. Rounded to the nearest 1 → ±0.5; nearest 10 → ±5; nearest 0.1 → ±0.05.
- Lower bound = value − half unit; upper bound = value + half unit.
- The upper bound is the value the measurement reaches but never equals — we still write it as the neat number.
| Measurement | Lower bound | Upper bound |
|---|---|---|
| 12 cm (nearest cm) | 11.5 cm | 12.5 cm |
| 250 g (nearest 10 g) | 245 g | 255 g |
| 4.6 s (nearest 0.1 s) | 4.55 s | 4.65 s |
2. Calculating with bounds
To get the extremes of a calculation, pick the bound of each measurement that pushes the answer highest or lowest. Addition and multiplication are intuitive; division and subtraction flip.
| Operation | Maximum uses | Minimum uses |
|---|---|---|
| a + b | a_upper + b_upper | a_lower + b_lower |
| a − b | a_upper − b_lower | a_lower − b_upper |
| a × b | a_upper × b_upper | a_lower × b_lower |
| a ÷ b | a_upper ÷ b_lower | a_lower ÷ b_upper |
Exam technique: the command word Calculate the upper bound wants the bound stated and then substituted — write "UB length = 8.5" before multiplying. Method marks reward the correct bound choice even if the arithmetic slips.