Secondary 2 G3 Mathematics Tutor Bukit Timah | Measurement Error Bounds and Accuracy Intervals. Did you know? A length written as 5.2 cm to the nearest tenth is not a promise that the true length is exactly 5.2 cm. Under the usual school rounding convention, the underlying measurement can be as low as 5.15 cm and must be less than 5.25 cm. That tiny interval changes how we reason about area, perimeter and the limits of a calculated answer.
This article is an applied Mathematics guide for Secondary 2 G3 learners and Bukit Timah parents, with a clear distinction between the foundational skills of rounding and the more advanced language of lower bounds, upper bounds, intervals and compounded measurement uncertainty. Some schools cover formal bounds later than Secondary 2, so use the child’s current teaching scope to select between basic rounding practice and this extension.
I want pupils to recognise that there are three separate things: the real quantity, the reported rounded number, and the level of precision at which the report was made. Knowing only the rounded number without its precision is often insufficient to identify a tight interval.
Find the accuracy interval you need
- Quick answer: half a rounding unit on each side
- What ‘to the nearest’ says about an actual measurement
- Decimal places versus significant figures
- Why an interval is not the same as an exact equality
- Adding and subtracting measurements
- Multiplication with positive measurements
- Division can reverse which bound you use
- Real measurements and model assumptions
- A checklist for reliable lower and upper bounds
- Forty-six original accuracy-bound clinics
- A two-dimensional enlargement bound
- A twenty-minute workshop on measuring a drawn rectangle
- A two-week accuracy revision pathway
- Frequently asked: why is the upper bound not included?
- Frequently asked: are bounds the same as instrument uncertainty?
- Frequently asked: can a bound question have two different correct interval formats?
- Frequently asked: should I round an upper bound again?
- Frequently asked: does Secondary 2 G3 always cover formal error bounds?
- Frequently asked: how should a tutor differentiate this topic?
- Bukit Timah teaching benchmark and official progression
Quick answer: half a rounding unit on each side
If a positive measurement m is given to the nearest unit u, the standard half-up school model gives m−u/2 ≤ actual value < m+u/2. For 5.2 cm to the nearest 0.1 cm, the interval is 5.15 ≤ x < 5.25 cm. For 240 cm to the nearest 10 cm, the interval is 235 ≤ x < 245 cm.
The symbol ≤ on the lower boundary and < on the upper boundary reflects the usual positive-number half-up rounding convention: the midpoint below rounds up to the reported figure, while the midpoint above rounds to the next figure. Some digital or scientific rounding rules break exact ties differently; a problem must state or assume its convention. Most of our examples use positive measurements and the school half-up rule.
What ‘to the nearest’ says about an actual measurement
A reported length of 7 cm to the nearest centimetre means actual values from 6.5 cm up to but not including 7.5 cm can produce that report. This is not because the measurement is known to have exactly ±0.5 cm random error; it is the mathematical rounding interval implied by the reporting unit. Measurement uncertainty from the instrument can be a separate issue.
For a positive height rounded to one decimal place, the rounding unit is 0.1 and half-unit is 0.05. For two decimal places, the unit is 0.01 and half-unit is 0.005. For a nearest metre report, the half-unit is 0.5 metre. These scales should never be guessed from the number of digits alone.
Decimal places versus significant figures
Rounding 8.65 to one decimal place produces 8.7 under ordinary half-up rounding, with the nearest 0.1 interval interpreted around the reported value. Significant figures instead start counting at the first nonzero digit. A large number such as 6,300 to two significant figures often implies a rounding step of 100 rather than 1, giving a much wider possible range.
A reported 6.3 × 10³ to two significant figures makes its precision clearer than writing 6,300 with no explanatory context. A student should identify which positional unit was rounded—not simply how many zeros appear at the end.
Why an interval is not the same as an exact equality
If 5.15 ≤ x < 5.25, writing x=5.20 can be a useful central estimate, but it is not a logically necessary exact statement. The actual length might be 5.18 or 5.24. School questions sometimes ask for ‘lower bound’, ‘upper bound’, ‘maximum’ or ‘minimum’; the careful answer distinguishes an included minimum boundary from an excluded upper limit.
In the interval 5.15 ≤ x < 5.25, the greatest allowable real value does not exist: values can get arbitrarily close to 5.25 from below but never equal it. We refer to 5.25 as an upper bound or least upper bound, not a necessarily attained maximum. This distinction is especially useful for strong students working on strict inequalities.
Adding and subtracting measurements
If positive measurements A and B satisfy 4.5 ≤ A < 5.5 and 2.5 ≤ B < 3.5, then 7 ≤ A+B < 9. The smallest sum is obtained from both included lower endpoints; the upper limit is approached when both measurements approach their upper bounds. The same arithmetic should be applied to the actual interval endpoints, not to the displayed central values alone.
For subtraction A−B, the lowest potential difference comes from the smallest A and the largest B, while the upper potential difference comes from largest A and smallest B. With these example ranges, A−B is strictly greater than 1 and strictly less than 3; it can approach both limits but cannot attain them because the relevant upper bounds are excluded. Never subtract lower from lower and upper from upper without considering which combination minimises or maximises.
Multiplication with positive measurements
For strictly positive length and width, area grows when either dimension grows. If 3.5 ≤ L < 4.5 cm and 1.5 ≤ W < 2.5 cm, the rectangular area satisfies 3.5×1.5 ≤ A < 4.5×2.5, so 5.25 ≤ A < 11.25 cm². These bounds are much wider than rounding the central area 4×2=8 cm² alone.
This monotonic reasoning is simple because both measurements are positive. If an algebraic interval crosses zero, multiplying endpoints requires more careful consideration of all combinations. The school examples here intentionally use positive physical lengths unless a later extension states otherwise.
Division can reverse which bound you use
Suppose distance d is in [90,110) metres and time t is in [9,11) seconds, with t positive. The smallest possible distance/time value is approached when d is smallest and t is largest: it is greater than 90/11 m/s. The largest possible rate is approached when distance is largest and time is smallest: it is less than 110/9 m/s. The range is approximately 8.18 < speed < 12.23 m/s, with the precise limits stated through fractions.
Students often divide lower distance by lower time and upper distance by upper time. That pairs the same direction of uncertainty instead of searching for the smallest and largest possible ratio. Explain which variable should increase and which should decrease to produce each extreme.
Real measurements and model assumptions
A rounding interval is conditional on how a value was recorded. An independently specified limit, such as a box being exactly 10 cm long by design, is not automatically the same as a ruler measurement reported to the nearest centimetre. Likewise, a digital display may use truncation or a special rounding convention. Read the stated precision, not an imagined property of the device.
These numerical bounds are classroom illustrations rather than guaranteed engineering safety tolerances. Actual construction or medical measurements require instrument precision, calibration, method, regulatory standards and uncertainty analysis beyond this school-focused model. Use school questions to practise reasoning, not to approve a real manufactured fit.
A checklist for reliable lower and upper bounds
- Identify the reported value and the rounding step—nearest whole unit, tenth, hundredth, ten or hundred.
- Calculate half the rounding step, then write the lower-inclusive and upper-exclusive interval using the stated convention.
- Convert units only after tracking whether the quantity is linear, squared or cubic.
- For sums of positive quantities, combine both lower endpoints for the lower bound and upper endpoints for the upper limit.
- For differences, smallest-first minus largest-second controls the lower limit; largest-first minus smallest-second controls the upper.
- For positive products, lower×lower and upper×upper give the area or volume limits; for positive quotients, lower numerator/upper denominator and vice versa.
- Explain whether an endpoint is included; an excluded upper limit should not be called an attained maximum.
Forty-six original accuracy-bound clinics
One decimal place
Precision problem. Length reported 5.2 cm to nearest 0.1 cm.
Worked interval or calculation. 5.15 ≤ x < 5.25 cm.
Why these endpoints matter. Half of0.1 is0.05, added or subtracted around5.2. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Using ±0.1 doubles the interval width.
Independent check. 5.24 rounds to5.2, while5.25 ordinarily rounds to5.3. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Nearest whole centimetre
Precision problem. Length reported 7 cm to nearest cm.
Worked interval or calculation. 6.5 ≤ x < 7.5 cm.
Why these endpoints matter. Rounding unit1 cm has half-unit0.5. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Treating 7 as exact discards possible values.
Independent check. 6.5 rounds to7; 7.5 rounds to8 under half-up convention. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Nearest tenth of a metre
Precision problem. Distance reported 2.4 m to nearest0.1m.
Worked interval or calculation. 2.35 ≤ x < 2.45 m.
Why these endpoints matter. Half tenth is0.05m. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Writing 2.3 to2.5 is too broad.
Independent check. 2.35m is235cm and belongs at lower edge. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Two decimal places
Precision problem. A positive reading is3.47 to nearest0.01.
Worked interval or calculation. 3.465 ≤ x < 3.475.
Why these endpoints matter. Hundredth rounding step0.01 has half0.005. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Using ±0.05 confuses first and second decimal places.
Independent check. 3.474 remains a value rounding to3.47. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Nearest ten metres
Precision problem. Distance is240m to nearest10m.
Worked interval or calculation. 235 ≤ x < 245m.
Why these endpoints matter. Half ten is five. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Writing 239.5 to240.5 treats nearest metre instead.
Independent check. 244.9m still rounds to240m. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Nearest hundred
Precision problem. A population estimate is1500 to nearest100.
Worked interval or calculation. 1450 ≤ x < 1550, under positive rounding convention.
Why these endpoints matter. Half of100 is50. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. 1500±0.5 incorrectly assumes nearest integer.
Independent check. Exact interpretation requires the reporting precision. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Nearest thousand
Precision problem. A quantity is6000 to nearest1000.
Worked interval or calculation. 5500 ≤ x < 6500.
Why these endpoints matter. Half the thousand-unit step is500. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. 6,000 to nearest unit would be a very different interval.
Independent check. The value6,499 rounds to6000. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
A small decimal
Precision problem. A length is0.38cm to nearest0.01cm.
Worked interval or calculation. 0.375 ≤ x < 0.385cm.
Why these endpoints matter. Half-hundredth is0.005cm. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Leading zeros do not make the rounding step0.1.
Independent check. 0.384cm is allowed. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
A millimetre report
Precision problem. An object measures32mm to nearest millimetre.
Worked interval or calculation. 31.5 ≤ x < 32.5mm.
Why these endpoints matter. Rounding step1mm has half0.5mm. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Changing to cm without changing bounds is wrong.
Independent check. In cm, 3.15 ≤ x < 3.25. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
One decimal length in centimetres
Precision problem. A measured side is8.0cm to nearest0.1cm.
Worked interval or calculation. 7.95 ≤ x < 8.05cm.
Why these endpoints matter. Displayed trailing decimal0 signals tenth precision. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Treating 8.0 as nearest whole centimetre gives too wide an interval.
Independent check. An exact8.05 would round to8.1. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Two decimals and trailing zero
Precision problem. A positive length is5.20cm to nearest0.01cm.
Worked interval or calculation. 5.195 ≤ x < 5.205cm.
Why these endpoints matter. The trailing0 indicates hundredth reporting precision. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Interpreting as nearest0.1 would be wider.
Independent check. 5.199cm is included. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Nearest half unit
Precision problem. A measurement is8.5 units to nearest0.5 unit.
Worked interval or calculation. 8.25 ≤ x < 8.75.
Why these endpoints matter. Half of0.5 equals0.25. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. The half-unit rounding rule is not necessarily0.5 error.
Independent check. 8.74 belongs,8.75 ordinarily rounds to9.0. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Rounding to tens
Precision problem. Reported time is70s to nearest10s.
Worked interval or calculation. 65 ≤ t < 75 seconds.
Why these endpoints matter. Half-step5 seconds. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. 70±10 gives interval twice as broad.
Independent check. At75s, nearest ten is80s. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Find stated rounding step
Precision problem. Lower bound is5.15, upper limit5.25 around reported5.2.
Worked interval or calculation. Interval width0.10, so the rounding step is0.1.
Why these endpoints matter. Half-width0.05 each side. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Width0.1 does not mean half-width0.1.
Independent check. The report was to nearest tenth. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Bound from significant figures
Precision problem. 6300 is explicitly to two significant figures.
Worked interval or calculation. 6250 ≤ x < 6350.
Why these endpoints matter. The last retained digit is hundreds; rounding step100. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Unqualified number6300 could have ambiguous implied precision.
Independent check. Writing6.3×10³ makes two significant digits clear. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Small number two significant figures
Precision problem. 0.0047 is reported to two significant figures.
Worked interval or calculation. 0.00465 ≤ x < 0.00475 under half-up positive rounding.
Why these endpoints matter. The last significant digit is ten-thousandths: step0.0001. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Counting leading zeros as significant corrupts precision.
Independent check. 0.00474 rounds to0.0047. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Sum of two rounded integers
Precision problem. A rounded5 to nearest whole and B rounded3 similarly.
Worked interval or calculation. A∈[4.5,5.5), B∈[2.5,3.5); 7≤A+B<9.
Why these endpoints matter. Both lower endpoints can occur while neither upper endpoint can. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Simply saying sum8 exactly is unsupported.
Independent check. Values5.4 and3.4 sum8.8. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Difference of rounded lengths
Precision problem. A≈5 nearest whole; B≈3 nearest whole. Find difference limits.
Worked interval or calculation. Since A∈[4.5,5.5) and B∈[2.5,3.5), 1<A−B<3.
Why these endpoints matter. Smallest difference approaches4.5−3.5 but cannot attain it. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Reporting1≤A−B≤3 ignores excluded upper endpoints.
Independent check. A=4.5,B=3.49 produces1.01. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Rectangle area lower bound
Precision problem. Length4cm nearest whole, width2cm nearest whole.
Worked interval or calculation. 3.5≤L<4.5;1.5≤W<2.5; area5.25≤A<11.25cm².
Why these endpoints matter. Positive area increases with both dimensions. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Multiplying nominal4×2=8 is central estimate only.
Independent check. The full possible interval is considerably wider. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Rectangle perimeter bounds
Precision problem. L=4cm and W=2cm to nearest whole.
Worked interval or calculation. Perimeter2(L+W) lies 10≤P<14cm.
Why these endpoints matter. Both dimensions contribute twice. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Using 2×(4+2)=12 as exact discards bounds.
Independent check. At included lower values P=10. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Square area with measured side
Precision problem. Side is3.0cm to nearest0.1cm.
Worked interval or calculation. 2.95≤s<3.05; 8.7025≤s²<9.3025cm².
Why these endpoints matter. Squaring positive endpoints finds area limits. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Nominal9cm² need not be exact area.
Independent check. Upper limit9.3025cm² is not attained. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Cube volume bounds
Precision problem. Side is2cm to nearest whole.
Worked interval or calculation. 1.5≤s<2.5, so3.375≤V<15.625cm³.
Why these endpoints matter. Cube of positive length grows with side. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. A nominal8cm³ can differ greatly from extreme possibilities.
Independent check. Both endpoints are cube of the corresponding length limits. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Two measured side product
Precision problem. Positive sides6.0cm and2.0cm to nearest0.1cm.
Worked interval or calculation. L∈[5.95,6.05), W∈[1.95,2.05); area11.6025≤A<12.4025cm².
Why these endpoints matter. Positive product extremes pair matching lower or upper endpoints. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Using ±0.1 on each side doubles the uncertainty.
Independent check. Nominal area12cm² is central estimate. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Subtracting lengths with different precision
Precision problem. A=10cm to nearestcm, B=2.0cm to nearest0.1cm.
Worked interval or calculation. A∈[9.5,10.5), B∈[1.95,2.05); difference7.45<A−B<8.55cm.
Why these endpoints matter. For minimum subtract B upper from A lower; both upper ends are excluded where relevant. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Subtracting 9.5−1.95 and10.5−2.05 pairs wrong extremes.
Independent check. Nominal difference8cm. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Multiply length by an exact count
Precision problem. A side is4cm nearest whole; there are exactly3 identical strips end to end.
Worked interval or calculation. Each length ∈[3.5,4.5); combined three copies 10.5≤total<13.5cm if all strips share same measured side length.
Why these endpoints matter. Exact count3 multiplies both interval endpoints. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Do not interpret count3 as itself rounded.
Independent check. Nominal total12cm. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Positive distance/time interval
Precision problem. Distance∈[90,110)m, time∈[9,11)s.
Worked interval or calculation. Speed satisfies90/11 < v < 110/9m/s when each interval is independently feasible.
Why these endpoints matter. To minimise ratio, minimise numerator and maximise denominator. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Dividing lower by lower does not give true lower limit.
Independent check. Limits≈8.18 and12.22m/s. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
A velocity-like sign warning
Precision problem. Suppose an algebraic interval for x crosses zero, and y is negative.
Worked interval or calculation. Do not blindly multiply lower×lower and upper×upper; test all endpoint combinations.
Why these endpoints matter. Multiplication monotonicity depends on signs. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. The positive-length formula is not universal for signed variables.
Independent check. Use actual bounds and sign logic before extension. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Fixed whole count
Precision problem. A rounded measurement4.0cm to nearest0.1, repeated in five identical components.
Worked interval or calculation. Total lies19.75≤T<20.25cm under the identical-length model.
Why these endpoints matter. Five is an exact count; scale interval limits by five. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Rounding each component again causes extra error.
Independent check. Nominal total20cm. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Exact area and rounded dimension
Precision problem. Area is exactly24cm²; positive width rounds4cm nearest whole.
Worked interval or calculation. 3.5≤w<4.5, so exact length l=24/w satisfies24/4.5<l≤24/3.5cm.
Why these endpoints matter. Positive reciprocal reverses order and endpoint inclusion. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Lower length limit not attained because w upper excluded.
Independent check. Upper length24/3.5 is attainable at w=3.5. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Rounding a price display
Precision problem. A fictional ticket has a stated exact price of$12.00 in the problem.
Worked interval or calculation. Do not automatically attach a ±$0.005 measurement interval: the story treats price as exact unless it says rounded.
Why these endpoints matter. Money displayed to cents can be an exact stipulated charge. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Rounding theory is not a licence to change exact data.
Independent check. If explicitly approximate to nearest cent, the interval model could differ. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Nearest minute report
Precision problem. A travel time is12min to nearest minute.
Worked interval or calculation. 11.5≤t<12.5min.
Why these endpoints matter. Half a minute is30 seconds. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Using nearest second limits is too narrow.
Independent check. In seconds,690≤t<750. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Two time intervals sum
Precision problem. A is12min nearest minute, B is8min nearest minute.
Worked interval or calculation. Together19≤T<21min.
Why these endpoints matter. Both positive intervals add at matching limits. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Nominal20min is not necessarily exact.
Independent check. A=11.5,B=7.5 gives lower19. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Percentage measured to nearest whole percent
Precision problem. A share is reported37% to nearest percentage point.
Worked interval or calculation. 36.5%≤p<37.5%, if p is an underlying continuous percentage.
Why these endpoints matter. Percentage-point rounding unit is1 point. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. 37%±1 percentage point is twice intended width.
Independent check. An underlying37.49% rounds to37%. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Scale conversion of bounds
Precision problem. A measured side is5.0m nearest0.1m. Convert bounds to cm.
Worked interval or calculation. 4.95≤s<5.05m becomes495≤s<505cm.
Why these endpoints matter. Exact unit conversion applies to both limits. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Using ±0.5m would assume nearest whole metre.
Independent check. Nominal5m=500cm. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
One measurement with zero lower
Precision problem. A positive quantity reports0.1cm nearest0.1cm.
Worked interval or calculation. Under nonnegative domain and usual half-up rounding, 0.05≤x<0.15cm.
Why these endpoints matter. An input of exactly0.05 rounds to0.1. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. A negative lower value is not needed here.
Independent check. 0.15 ordinarily rounds to0.2. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
A rounded area itself
Precision problem. A rectangle’s area is reported12m² to nearest square metre.
Worked interval or calculation. 11.5≤A<12.5m².
Why these endpoints matter. The reported area unit is square metres; the rounding step is 1m². Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Do not square the one-metre rounding unit again; area is already measured.
Independent check. This differs from calculating area from rounded sides. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
A measured volume itself
Precision problem. A container capacity is given4.0L to nearest0.1L.
Worked interval or calculation. 3.95≤V<4.05L.
Why these endpoints matter. Half of0.1 litre is0.05 litre. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. This is a volume report already, not a length needing cubing.
Independent check. In mL, 3950≤V<4050. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
A lower bound for filled water
Precision problem. A fill volume is2.5L nearest0.1L; a withdrawal is0.5L nearest0.1L.
Worked interval or calculation. Remaining volume lies strictly above1.9L and strictly below2.1L (subject to nonnegative physical model).
Why these endpoints matter. Subtract largest possible withdrawal from smallest fill and vice versa. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Using2.5−0.5=2.0 as exact remaining volume loses uncertainty.
Independent check. One limit2.45−0.55=1.90 is not attained. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Midpoint of a rounding interval
Precision problem. Length in [5.15,5.25). Find reported nearest tenth.
Worked interval or calculation. Reported centre5.2cm.
Why these endpoints matter. Midpoint of the two interval bounds is5.2. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Do not call interval midpoint the actual measurement necessarily.
Independent check. Many values in interval round to5.2. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Bounded ratio of positive measures
Precision problem. Positive A∈[4,6), B∈[2,3).
Worked interval or calculation. A/B has lower limit approaching4/3 and upper limit approaching3, with 4/3<A/B<3.
Why these endpoints matter. Small numerator/large denominator minimises ratio; reverse maximises. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. 4/2 and6/3 are not extremes.
Independent check. A=4,B=2.99 gives ratio close to4/3. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
An exact count has no rounding interval
Precision problem. A group contains exactly12 pupils.
Worked interval or calculation. The count is12, not an interval11.5 to12.5.
Why these endpoints matter. Discrete exact count differs from an estimated continuous measurement. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Treating every printed integer as rounded is unsupported.
Independent check. A reported approximate crowd count would need explicit wording. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Relative error and nominal reading
Precision problem. A reported length10.0cm nearest0.1cm has maximum absolute rounding deviation approaching0.05cm.
Worked interval or calculation. Relative error magnitude can approach0.05/actual or roughly0.5% using10cm as nominal reference.
Why these endpoints matter. The half-step controls rounding deviation; relative comparison needs a stated base. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Claiming every value has exactly0.05cm error is false.
Independent check. Actual reading can be exactly10.0 with zero rounding error. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
Maximum or upper limit language
Precision problem. A measurement x satisfies5.15≤x<5.25.
Worked interval or calculation. 5.25 is an upper bound, but no largest permitted real x exists.
Why these endpoints matter. Values can approach5.25 from below arbitrarily closely. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Saying x=5.25 is the maximum incorrectly includes excluded endpoint.
Independent check. The minimum x=5.15 is permitted under this convention. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
A contradiction in intervals
Precision problem. School conditions require x≥8 and x<7.
Worked interval or calculation. No real value satisfies both; feasible set empty.
Why these endpoints matter. Intersecting contradictory intervals produces no solution. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. Averaging7 and8 to7.5 does not satisfy either pair.
Independent check. Sketch both on one number line. A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
A two-dimensional enlargement bound
Precision problem. Each side of a square is reported4cm nearest whole.
Worked interval or calculation. The side lies from 3.5 cm (inclusive) to 4.5 cm (exclusive), so the square’s area satisfies 12.25≤A<20.25 cm².
Why these endpoints matter. Square the correct positive side limits. Record the rounding unit first, explain which extreme values lead to each bound, and distinguish strict limits from included measurements.
Incorrect shortcut to avoid. A numerical12.25 lower arises from3.5², not from a product with width2.
Independent check. Nominal square area16cm². A pupil should test one boundary value against the original reporting convention rather than trusting an unexplained interval.
A twenty-minute workshop on measuring a drawn rectangle
Give each of three pupils an imaginary rectangle with positive reported length 6.0 cm and width 2.0 cm, both to the nearest tenth. One calculates the length interval [5.95,6.05), one the width interval [1.95,2.05), and one computes the possible area limits. Together the students should find 11.6025 ≤ area < 12.4025 cm² and a central nominal calculation of12cm².
Exchange roles and ask what changes when only the length is reported to the nearest whole centimetre. The area interval becomes wider because the permitted length range is broader. The exercise builds an important principle: a multiplication cannot create greater certainty than is justified by the stated input measurements.
A two-week accuracy revision pathway
Week one: Practise identifying rounding steps from nearest whole number, tenth, hundredth, ten and hundred. Write lower-inclusive, upper-exclusive intervals and ask students to test borderline values. Connect decimals to the precise place value of the last retained digit. End with a simple unit-conversion check on an interval.
Week two: Combine positive lengths in perimeter, area and volume problems. Include a difference and a positive quotient so the child must choose opposing endpoints correctly. Ask each learner to explain why an upper bound may not be attainable and to make a separate nominal estimate before comparing the permitted interval.
Frequently asked: why is the upper bound not included?
Under ordinary positive half-up rounding, a value exactly halfway to the next displayed number rounds up to that next number. For a value displayed as5.2 to nearest0.1, an actual5.25 would round to5.3 rather than5.2. That is why the upper endpoint is normally written with a strict less-than sign.
Frequently asked: are bounds the same as instrument uncertainty?
No. A rounding interval describes all exact values that could have produced a reported rounded result under a specified rounding rule. Real instrument uncertainty can include calibration, resolution, user technique and systematic effects. Those factors may require a more sophisticated model than this school exercise.
Frequently asked: can a bound question have two different correct interval formats?
Equivalent inequalities and interval notation can describe the same set: 5.15≤x<5.25 is often written [5.15,5.25). A pupil should use the school’s preferred notation and make clear which endpoints are included. The opening and closing symbols carry mathematical meaning.
Frequently asked: should I round an upper bound again?
Usually keep bounds exact to the given precision where possible. Rounding an upper limit can make the statement too narrow or too broad depending on the direction and method. If a school explicitly requests an approximate bound at a stated accuracy, show the unrounded calculation first and explain the final presentation.
Frequently asked: does Secondary 2 G3 always cover formal error bounds?
No. Rounding, significant figures and estimation are foundational syllabus skills, but formal bound calculations may be enrichment or later-year teaching depending on the school. Ask for the actual assessed chapter list and do not replace core fraction, algebra or basic mensuration practice with advanced uncertainty work unnecessarily.
Frequently asked: how should a tutor differentiate this topic?
A learner who mistakes nearest0.1 for nearest0.01 needs a place-value repair. Another who writes correct intervals but multiplies lower numerator by lower denominator needs a monotonicity explanation. A pupil who calculates bounds accurately but calls an excluded upper limit an attained maximum needs conceptual language practice. These are distinct teaching needs.
Bukit Timah teaching benchmark and official progression
The unchanged eduKateSG three-pupil Mathematics reference describes small-group tutorials of around1.5 hours weekly with guided corrections at 8 Fourth Avenue, Singapore 268674, near Sixth Avenue MRT. This article serves Bukit Timah families and does not imply additional teaching locations or a vacant group.
Under Full Subject-Based Banding, G3 denotes a Mathematics subject level. The SEAB 2027 G3 syllabus index and K310 Mathematics syllabus discuss approximation and estimation; an individual school’s Secondary 2 term scope determines which bounds extensions should be assigned.
For supporting instruction, see rounding and significant figures, area, volume and capacity units, and inequality intervals. Each page supplies a related prerequisite rather than repeating the present article’s interval calculations.
For a tuition discussion, use the original programme reference and bring a marked accuracy or measurement question, the required school rounding convention and the upcoming chapters. Confirm group fit, actual times and availability with eduKateSG directly.
Editorial review: 28 October 2026. All measurements are fictional classroom data. Interval examples assume positive measurements and the usual half-up rounding convention unless otherwise stated; boundary attainability is discussed explicitly. This is not a professional measurement-certification guide.