Did you know? A student can understand an entire Mathematics question yet lose accuracy on the last line by rounding too early. A calculator may show six or nine digits, but the assessment may ask for three significant figures, two decimal places, or a sensible estimate. The skill is not to make every number short. It is to explain what the answer needs to communicate.
Quick answer: decimal places versus significant figures
- Decimal places (d.p.): count digits to the right of the decimal point. For example, 17.486 becomes 17.49 to 2 d.p.
- Significant figures (s.f.): count from the first nonzero significant digit. For example, 0.00467 becomes 0.0047 to 2 s.f.
- Estimation: replace numbers with convenient nearby values to predict a plausible magnitude.
- Checking: compute with adequate precision, then round the final non-exact answer as instructed.
The published 2027 G1 SEC Mathematics K110 syllabus includes approximation and estimation, including decimal places and significant figures. Its assessment instructions state that non-exact numerical answers should ordinarily be given to 3 s.f., and non-exact angles in degrees to 1 d.p., unless the question specifies something else. Those rules refer to the 2027 graduating examination, not every Secondary 2 class worksheet; always follow the actual school and question instructions.
Find the answer you need
- One-minute accuracy check
- Three principles
- 33 worked accuracy clinics
- Errors to repair
- A small-group lesson
- Practice at home
- FAQs
- Sources
A one-minute test for understanding
Ask the student to round 0.00637 to 2 significant figures, then 18.456 to 2 decimal places, and finally estimate 49.8×3.9. The expected answers are 0.0064, 18.46 and approximately 200. If the student gives 0.00 for the first, they may be counting leading zeros. If 18.45 appears in the second, they may be dropping digits without rounding. If the estimate is close to 20 or 2000, place value is the missing prerequisite.
Why accuracy questions test mathematical judgment
Keep exact values exact when possible
An exact fraction such as 3/7 can be retained in a calculation until the result is required as a decimal. Similarly, retain the calculator value of π or an exact expression such as 49π rather than replacing it with 3.1 early. Rounding intermediate steps repeatedly can create a misleading final answer even when the formula is correct.
Choose the requested accuracy—not the longest display
A calculator can show 17.142857142, yet ‘to 3 significant figures’ requests 17.1. The same number to 3 decimal places is 17.143. These are different instructions, and the number of digits shown after the decimal point is not a general substitute for the number of significant digits. Zeroes can also matter: 8.0 explicitly displays two significant figures, while 8 commonly displays one.
An estimate catches scale errors
Before multiplying 39.6×5.1, think ‘about 40×5’, or roughly 200. If a calculator entry returns 20 or 2000, the estimate signals a problem worth investigating. An estimate need not be the marked final answer. It is a quick independent check against an order-of-magnitude mistake.
33 original worked accuracy clinics
Try one cluster at a time. Cover the worked answer, decide whether the task asks for a decimal place, significant figure, exact answer or estimate, then calculate and explain the choice.
1. Round to two decimal places
Try it. Round 17.486 to 2 decimal places.
Reasoning. Keep two digits after the decimal point (48); the next digit is six, so the answer is 17.49.
The trap. Dropping the six without checking whether it changes the retained hundredths.
Check. The rounded answer differs from the original by 0.004.
2. One decimal place may need a written zero
Try it. Round 4.032 to 1 decimal place.
Reasoning. Keep the tenths place. The hundredths digit is three, so report 4.0 to show one decimal place.
The trap. Reporting 4 with no indication of the requested display precision.
Check. Explain that 4.0 and 4 are numerically equal, but the question requested 1 dp.
3. A close-to-ten value may carry
Try it. Round 9.996 to 2 decimal places.
Reasoning. The thousandths digit is six, so 9.99 rounds up across the boundary to 10.00.
The trap. Reporting 9.100 or 9.99 without carrying through.
Check. The rounded result has two digits after the decimal point.
4. A small number’s leading zeros are not significant
Try it. Round 0.00467 to 2 significant figures.
Reasoning. The first two significant digits are 4 and 6; the third is 7, so the result is 0.0047.
The trap. Counting the zeros before 4 as significant figures.
Check. 0.0047 is close to 0.00467 and has exactly two significant digits (4 and 7).
5. Round a four-digit whole number to three significant figures
Try it. Round 5892 to 3 significant figures.
Reasoning. Keep 5,8,9 and inspect the fourth digit 2. The result is 5890.
The trap. Writing 589, which changes the magnitude by a factor of ten.
Check. 5890 is within two units of the original 5892.
6. A number near 2500 to the nearest hundred
Try it. Round 2490 to the nearest hundred.
Reasoning. The nearest hundreds are 2400 and 2500; 2490 is ten away from 2500, so it rounds to 2500.
The trap. Reporting 2400 because the thousands digit was not considered.
Check. Compare distances: 2490−2400=90 versus 2500−2490=10.
7. Round a number to the nearest ten
Try it. Round 154 to the nearest ten.
Reasoning. The ones digit is four, so the answer is 150.
The trap. Rounding to 200 as if the instruction said nearest hundred.
Check. 150 is four away from 154; 160 is six away.
8. Rounded zeros may matter for significant-figure reporting
Try it. Round 7.996 to 2 significant figures.
Reasoning. The first two significant digits are 7 and 9; the next digit is 9 and rounding causes a carry. Report 8.0 to indicate 2 significant figures.
The trap. Writing just 8 if the requested precision must be made explicit.
Check. Check 8.0 has two significant figures, while 8 alone ordinarily indicates one.
9. Rounding a near-one value to two significant figures
Try it. Round 0.995 to 2 significant figures.
Reasoning. Under the usual school rounding convention, 0.995 rounds to 1.0 to 2 significant figures.
The trap. Giving 0.10, which changes the magnitude.
Check. 1.0 has two significant figures (1 and trailing zero).
10. Small numbers can contain internal significant zeros
Try it. Round 0.0008026 to 3 significant figures.
Reasoning. The significant digits are 8,0,2; the next digit 6 rounds 2 to 3. Answer: 0.000803.
The trap. Deleting the zero between eight and two. That zero is significant.
Check. Multiply by a million to compare 802.6 and rounded 803.
11. The requested place differs from significant figures
Try it. Round 103.49 to the nearest whole number.
Reasoning. Since the tenths digit is four, the result is 103.
The trap. Interpreting ‘nearest whole’ as one significant figure and answering 100.
Check. Check 103 is 0.49 away; 104 is 0.51 away.
12. Negative values can be rounded
Try it. Round −2.66 to 1 decimal place.
Reasoning. The magnitude 2.66 rounds to 2.7; retain the negative sign to give −2.7.
The trap. Using −2.6 because the number appears on the negative side of the number line.
Check. −2.7 differs from −2.66 by 0.04; −2.6 differs by 0.06.
13. Show required significant figures with an appropriate trailing zero
Try it. Round 5.004 to 3 significant figures.
Reasoning. The first three significant digits are 5,0,0; the next digit is 4, so report 5.00.
The trap. Writing 5 and not indicating the requested 3 significant figures.
Check. 5.00 and 5.004 are close, and 5.00 has three significant figures.
14. Estimate a multiplication
Try it. Estimate 39.6 × 5.1.
Reasoning. Use 40 × 5 ≈ 200. The exact product is 201.96, so the estimate is reasonable.
The trap. Giving 20 by losing a place value.
Check. Both factors are close to the convenient numbers used.
15. Estimate a division before entering a calculator
Try it. Estimate 598 ÷ 19.9.
Reasoning. Use 600 ÷ 20 ≈ 30. Exact value is about 30.05.
The trap. An answer near 300 is an order-of-magnitude warning.
Check. Multiply 20×30=600 to justify the mental estimate.
16. Spot a misplaced decimal point from scale alone
Try it. A calculation of 19.8 × 4.9 is reported as 9.702. Is that plausible?
Reasoning. No. Roughly 20×5=100. The exact answer is 97.02, so the reported value is ten times too small.
The trap. Assuming a precise-looking decimal must be right.
Check. Test the size before using the calculator output.
17. Use 3 significant figures for a non-exact fraction result when instructed
Try it. Express 3 ÷ 7 to 3 significant figures.
Reasoning. 3/7=0.428571…, so to 3 significant figures it is 0.429.
The trap. Reporting 0.428 because the next digit five was not used for rounding.
Check. Multiplying 0.429×7 gives approximately 3.003, close to three.
18. An exact result does not need arbitrary extra digits
Try it. Evaluate 2.5 × 4.
Reasoning. The exact product is 10. It is acceptable to write 10 unless a particular display accuracy is required.
The trap. Treating every result as non-exact and automatically reporting 10.0.
Check. Check 10/4=2.5.
19. Round a division answer appropriately
Try it. Find 120 ÷ 7 to 3 significant figures.
Reasoning. The unrounded quotient is 17.142857…, so to 3 significant figures give 17.1.
The trap. Writing 17.14, which is four significant figures.
Check. 17.1×7=119.7, consistent with the approximate quotient.
20. Pi should not be replaced too early
Try it. A circle has radius 7 cm. Find its area to 3 significant figures using the calculator value of π.
Reasoning. Area = π×7² = 49π ≈ 153.938, which rounds to 154 cm².
The trap. Using 3.1 instead of calculator π and rounding the intermediate value unnecessarily.
Check. Area slightly below 50×3.14≈157 is plausible; use the exact expression until the final step.
21. Circumference depends on one length dimension
Try it. Find the circumference for a circle with radius 3.2 cm to 3 significant figures.
Reasoning. C=2πr=6.4π≈20.1062, giving 20.1 cm.
The trap. Using the area formula πr² or attaching square units.
Check. Check the answer exceeds the diameter 6.4 cm several times.
22. Angles can have their own specified rounding
Try it. An angle is calculated as 34.567°. Express it to 1 decimal place.
Reasoning. The hundredths digit is six, so give 34.6°.
The trap. Automatically giving 34.6 with no degree symbol or 34.57 despite the one-decimal instruction.
Check. The 2027 K110 scheme states 1 dp for non-exact angles unless otherwise specified; always read the question.
23. Distinguish exact money totals from calculator rounding
Try it. Three items cost $6.99 each. Find the exact total before any separate charges.
Reasoning. 3×$6.99=$20.97. This is an exact cent-valued total under the stated prices.
The trap. Rounding $6.99 to $7 before computing and presenting $21 as an exact bill.
Check. $21 is a reasonable estimate but the exact result is $20.97.
24. Keep units in an estimate of area
Try it. A rectangle is 2.4 m by 1.5 m. Find its area.
Reasoning. 2.4×1.5=3.6 m². A quick estimate of 2.5×1.5≈3.75 m² agrees in size.
The trap. Reporting 3.6 m, a length rather than an area.
Check. Check the multiplication uses two metre quantities.
25. Use an estimate to check a time answer
Try it. A journey of 151 km at an average speed of 49 km/h takes approximately how long?
Reasoning. Time=151/49≈3.0816 h, about 3.08 h to 3 significant figures. Roughly 150/50=3 h.
The trap. Multiplying speed by distance or reporting 308 h after a decimal error.
Check. 49 km/h×3.08 h≈150.9 km.
26. Significant figures are not the same as decimal places
Try it. Round 12.3456 to 3 significant figures and separately to 3 decimal places.
Reasoning. Three significant figures gives 12.3; three decimal places gives 12.346.
The trap. Using 12.346 for both because both instructions contain the number three.
Check. Count from the first nonzero digit for s.f., and after the decimal point for d.p.
27. A rounded mean may not equal any data value
Try it. Find the mean of 12, 13, 15 and express it to 3 significant figures.
Reasoning. Sum 40; divide by three to get 13.333…, so 13.3 to 3 significant figures.
The trap. Rounding each input unnecessarily or writing 13 without the requested precision.
Check. Three copies of 13.333… total 40.
28. Use measurement precision only as instructed
Try it. A student measures 2.437 m and reports it to 2 decimal places. What should appear?
Reasoning. 2.44 m: keep hundredths (43) and round up because the thousandths digit is seven.
The trap. Reporting 2.43 m by truncating the last digit.
Check. The change of 0.003 m is smaller than the 0.007 m difference to 2.43.
29. Convert units before applying a requested accuracy
Try it. Convert 2350 mm to metres and give the answer to 2 decimal places.
Reasoning. 2350 mm=2.35 m exactly; reported to 2 dp, it remains 2.35 m.
The trap. Rounding to 2.4 m because the learner treats 2 dp as 2 significant figures.
Check. Reverse with 2.35×1000=2350 mm.
30. An estimate and an exact answer have different jobs
Try it. Four packets contain 198 stickers each. Estimate, then find total.
Reasoning. Estimate 4×200≈800. Exact calculation 4×198=792 stickers.
The trap. Presenting the estimate 800 as exact when the question requests an exact total.
Check. Exact result is only eight stickers below the estimate.
31. Read the requested unit before rounding
Try it. A calculation gives 2.75 hours. Express this as hours and minutes.
Reasoning. 0.75 h×60=45 min. The exact duration is 2 h 45 min.
The trap. Reading 2.75 as 2 h 75 min, which is not normal clock notation.
Check. Check 2+45/60=2.75 h.
32. Find a percentage with sensible precision
Try it. What percent of 37 is 11, to 3 significant figures?
Reasoning. 11/37×100%≈29.7297%, so 29.7%.
The trap. Rounding 11/37 to 0.3 before multiplying to produce a misleading 30%.
Check. Check 29.7% of 37 is about 10.989, near eleven.
33. State that a rounded value is an approximation
Try it. An area calculation gives 36.742 cm², but the final answer is required to 1 decimal place.
Reasoning. Report approximately 36.7 cm², not exactly 36.7 as the full underlying value.
The trap. Treating the rounded number as if it preserves every original decimal digit.
Check. The approximation loses 0.042 cm² of displayed precision.
Five common mistakes and the right interventions
Leading zeros are counted as significant
If a pupil rounds 0.00467 to two significant figures as 0.00, return to place value: the first meaningful nonzero digit is 4. Have the student mark the first two significant digits before making a rounding decision. Switch to 0.000638 and see if the method transfers.
The student truncates instead of rounding
If 17.486 becomes 17.48 to two decimal places, the student kept the right number of digits but did not inspect the next digit. Ask them to underline the retained hundredths place, circle the following digit and explain whether the retained value should increase. Separate how many digits to keep from which direction to round.
The answer is made less accurate several times
A learner approximates π as 3.1, then approximates a radius, then rounds the product again. Work with exact stated lengths and calculator π until the final step, unless the task directs otherwise. Compare both methods so the child can see how small early changes accumulate.
A wrong unit accompanies the right rounded number
Writing 36.7 cm for an area of 36.742 cm² still loses the meaning of the calculation. Ask the learner to identify whether the measure describes a boundary, covered surface, filled space or rate. The units tell part of the answer and can reveal formula-selection errors.
The pupil treats an estimate as an exact result
An estimate of about 800 stickers is useful for checking four packets of 198, but the exact result is 792. Teach students to label ‘approximately’ versus ‘exactly’ and to follow the final wording. A fast estimate should support, not replace, a required exact multiplication.
Inside a focused three-pupil Mathematics tutorial
The unchanged eduKateSG Clementi tutorial reference describes 90-minute sessions with three pupils near Sixth Avenue MRT. A useful G1 accuracy lesson can compare three distinct students: one counting significant figures incorrectly, one rounding at every stage, and one calculating correctly but reporting the wrong unit. Each needs a different correction; a shared worksheet alone cannot diagnose that difference.
- First 10 minutes: two quick place-value and estimate questions.
- Next 20 minutes: link place value with significant-figure and decimal-place rules.
- Next 25 minutes: individually solve selected questions with first-wrong-line feedback.
- Next 15 minutes: apply final-answer accuracy to area, time, rate and money contexts.
- Next 10 minutes: unseen mixed precision instructions.
- Last 10 minutes: reattempt a corrected error without the example visible.
Five short home activities
- Monday: round five decimals with a specified number of decimal places.
- Tuesday: mark the first two or three significant digits in numbers with leading or internal zeros.
- Wednesday: estimate four arithmetic results before using a calculator.
- Thursday: solve an area or speed question without prematurely rounding intermediate results.
- Friday: attempt a mixed set and revisit the first accuracy error after a short delay.
Parents can ask one simple question: ‘Is that an exact answer, a rounded answer, or an estimate?’ A confident child should be able to say which kind it is, what precision the question required, and whether the size makes sense.
G1 subject level and the right exam-year instruction
SEAB publishes the 2027 Mathematics K110 syllabus for G1 SEC candidates (earlier subject-code reference 4046). The final paper instructions on non-exact answers are specific and worth reading carefully. A Secondary 2 student in 2026 should instead follow their school marking instructions now and consult the official syllabus for the student’s actual graduation year when available. Nothing here promises a particular question type on the next school paper.
Frequently asked questions
Is 0.00470 a two-significant-figure number?
No. In ordinary mathematical notation, the digits 4, 7 and the trailing zero after the decimal are all significant; 0.00470 displays three significant figures.
Can a whole-number answer show significant figures?
Yes, but the notation may require care. Standard form can make significant figures explicit where trailing zeros in a whole number are ambiguous.
Should I use 3.14 for π in every question?
No. Follow the question’s exact instruction or the examination’s general guidance. The 2027 K110 scheme says use calculator π or π=3.142 unless the question requires an answer in terms of π.
What is the difference between an estimate and a rounded answer?
An estimate often uses simplified input numbers to judge size before computing. A rounded final answer comes from the calculation and is then expressed to the requested precision.
Does my child need to memorise the 2027 K110 accuracy note today?
It is useful background, but the immediate Secondary 2 task is to follow the current school’s and teacher’s specified accuracy for each question.
What does a good tutor do when the calculator output is wrong?
Ask the student to predict magnitude, inspect entered brackets and units, then decide whether the error is a keystroke, operation choice or rounding rule.
Related Secondary 2 G1 Mathematics guides
- Fractions, decimals and signed numbers
- Formula substitution and units
- Revision and exam strategy
- Inequality signs and number lines
- Map scales and floor plans
Official sources and teaching note
- SEAB: 2027 SEC G1 school-candidate syllabuses
- SEAB K110 2027 G1 Mathematics syllabus and accuracy instructions
- eduKateSG: immutable Clementi three-pupil tutorial programme
Reader note: These 33 worked questions are original teaching examples, not past exam items or school assessment predictions. All hypothetical distances and prices are labelled as examples. Actual class suitability, timetable, location and fees need direct confirmation; a tutoring format cannot guarantee a mark.
Bring an example where your child rounded too early and an example where the number was correct but the unit was not. Enquire about eduKate Singapore Mathematics tutorials and ask how independent accuracy checks are taught.