PSLE Learner’s Guide · Vol 0094 · Mathematics · Primary 6
When two totals describe some of the same pupils, addition is not the whole answer
A class survey finds that 18 pupils enjoy basketball and 15 enjoy badminton. Can you say that 33 different pupils enjoy at least one of the sports? Not yet. Some pupils may enjoy both, and those pupils appear in both reported numbers. If 7 pupils enjoy both, the number enjoying at least one is 18 + 15 − 7 = 26. The subtraction is not a trick introduced to confuse you; it removes the extra copy of each pupil counted twice.
This guide focuses on one examination-performance job: decide whether two groups overlap before combining their totals. It is different from merely listing possibilities or solving a one-group fraction. You need a reliable way to organise people or objects into ‘first only’, ‘both’, ‘second only’ and ‘neither’, then answer the actual question without counting someone twice or leaving them out.
The problem is particularly tempting when the data comes from separate questions, charts, lists or paragraphs. One table reports pupils who selected Art; another reports pupils who selected Music. Adding the counts feels reasonable until you notice that the two categories were not stated to be exclusive. The word ‘and’ can mean a shared membership, while ‘or’ in survey questions often includes pupils who chose either or both. You must read the intended language rather than guess.
A trustworthy first move is to ask: ‘Could one person legitimately appear in both groups?’ If yes, do not add the group totals to get the number of different people without further information. When a number for ‘both’ is supplied, remove that duplicate once. When the total and ‘neither’ are given, use them to find how many belong to at least one. When the overlap cannot be determined, sometimes the correct conclusion is that there is not enough information for one exact total.
This is a PSLE learning guide, not a claim that a particular diagram style or notation is mandated in a specific examination year. Official information for 2026 is available from SEAB’s PSLE format list, which identifies the revised Mathematics syllabus for the year. Use your school syllabus and question instructions for the exact expectations. All surveys below are invented for learning.
1. Four regions inside a two-group story
Imagine two circles drawn on scrap paper, one named A and one named B. Their shared middle represents pupils in both. The part inside A but outside B represents ‘A only’; the corresponding part inside B represents ‘B only’. Outside both is ‘neither’. You do not need to draw a formal diagram for every sum, but this four-region mental map helps the words keep their meaning.
Suppose 18 pupils like basketball, 15 like badminton and 7 like both. ‘Basketball only’ is 18 − 7 = 11. ‘Badminton only’ is 15 − 7 = 8. The shared group contains 7. Therefore the pupils in at least one sport are 11 + 7 + 8 = 26. This agrees with 18 + 15 − 7 = 26. When two methods agree for a clearly stated reason, you gain a useful independent check.
If the class has 30 pupils, the number liking neither is 30 − 26 = 4. You may not infer that all 30 chose a sport unless the question says so. ‘At least one’ and ‘all pupils’ are different totals. If the survey reported 30 pupils and no separate count of the ‘neither’ group, you must be careful whether the categories cover everyone.
The order matters when finding exclusive parts. A’s reported total already contains everyone in the overlap. To find ‘A only’, subtract the overlap once: 18 − 7. But to find the overall union, add the totals and subtract the overlap once: 18 + 15 − 7. Do not subtract overlap twice from the sum of the group totals; that would incorrectly remove the shared pupils altogether.
Try a physical model. Place eleven blue counters into an A-only area, seven into an overlapping area, and eight into B-only. Counting all counters once gives 26. Counting A gives eighteen; counting B gives fifteen. The sum of group counts is 33 because each of the seven middle counters has been counted two times. Exactly one extra count per shared counter must be removed. This is the reason the arithmetic works.
2. Translate survey words into exact membership
1. Both A and B
This is the intersection or shared group. A pupil who likes both sports appears in both original group totals but should appear once in the combined count of distinct pupils. When told ‘7 chose both’, put 7 in the shared region before calculating the exclusive regions.
Quick diagnosis: rewrite the phrase as ‘A only’, ‘shared’, ‘B only’ or ‘outside both’ before calculating. If the phrase refers to more than one region, say which ones. You are translating language into quantities, not hunting for a signal word that always triggers the same operation.
2. A only
This means in A and not in B. Subtract the shared count from A’s total. ‘A only’ is not the same as ‘all of A’. Check the word ‘only’ before you use the full group count in a final answer.
Quick diagnosis: rewrite the phrase as ‘A only’, ‘shared’, ‘B only’ or ‘outside both’ before calculating. If the phrase refers to more than one region, say which ones. You are translating language into quantities, not hunting for a signal word that always triggers the same operation.
3. B only
The same reasoning applies in the opposite direction. B’s total contains the middle; the exclusive B region does not. You can use B total minus shared to isolate the region.
Quick diagnosis: rewrite the phrase as ‘A only’, ‘shared’, ‘B only’ or ‘outside both’ before calculating. If the phrase refers to more than one region, say which ones. You are translating language into quantities, not hunting for a signal word that always triggers the same operation.
4. At least one
This includes A only, both, and B only. It excludes neither. The union is A total + B total − both. ‘At least one’ does not mean ‘exactly one’.
Quick diagnosis: rewrite the phrase as ‘A only’, ‘shared’, ‘B only’ or ‘outside both’ before calculating. If the phrase refers to more than one region, say which ones. You are translating language into quantities, not hunting for a signal word that always triggers the same operation.
5. Exactly one
This means A only or B only, excluding the shared middle. Calculate (A − both) + (B − both), or at least one minus both. Explain why the middle is not part of the requested answer.
Quick diagnosis: rewrite the phrase as ‘A only’, ‘shared’, ‘B only’ or ‘outside both’ before calculating. If the phrase refers to more than one region, say which ones. You are translating language into quantities, not hunting for a signal word that always triggers the same operation.
6. Neither
This is outside both groups. If the total population is known, subtract the number in at least one group. Do not place ‘neither’ in the overlap or in one of the exclusive regions.
Quick diagnosis: rewrite the phrase as ‘A only’, ‘shared’, ‘B only’ or ‘outside both’ before calculating. If the phrase refers to more than one region, say which ones. You are translating language into quantities, not hunting for a signal word that always triggers the same operation.
7. Everyone chose one or both
This states that nobody is in the neither region. Then the overall population equals the union. If the totals and population are known, the overlap can be found by A + B − population.
Quick diagnosis: rewrite the phrase as ‘A only’, ‘shared’, ‘B only’ or ‘outside both’ before calculating. If the phrase refers to more than one region, say which ones. You are translating language into quantities, not hunting for a signal word that always triggers the same operation.
8. One or the other
Read the full context. Ordinary inclusive ‘or’ may allow both; wording such as ‘exactly one of the two’ or ‘but not both’ excludes the overlap. Do not assume every ‘or’ is exclusive.
Quick diagnosis: rewrite the phrase as ‘A only’, ‘shared’, ‘B only’ or ‘outside both’ before calculating. If the phrase refers to more than one region, say which ones. You are translating language into quantities, not hunting for a signal word that always triggers the same operation.
9. Different pupils
This is a warning that you need a count of distinct individuals. Group totals may repeat the same names. Confirm the basis before converting reported preferences into a count of people.
Quick diagnosis: rewrite the phrase as ‘A only’, ‘shared’, ‘B only’ or ‘outside both’ before calculating. If the phrase refers to more than one region, say which ones. You are translating language into quantities, not hunting for a signal word that always triggers the same operation.
10. Selected two activities
The phrase may describe exactly two choices if the survey lists several activities. In this two-group guide, ‘selected both A and B’ gives the shared group. With more than two categories, new overlap information may be needed; do not apply the two-circle method unchanged.
Quick diagnosis: rewrite the phrase as ‘A only’, ‘shared’, ‘B only’ or ‘outside both’ before calculating. If the phrase refers to more than one region, say which ones. You are translating language into quantities, not hunting for a signal word that always triggers the same operation.
3. A six-move method for reliable overlap questions
Move 1: Find the population
Underline the exact population: pupils in one class, respondents in a survey, visitors at a museum, or objects inside one collection. Do not combine survey totals drawn from different populations unless the question defines a shared basis.
Put a brief note on your practice working if needed. ‘A 18, B 15, both 7’ is much safer than memorising a number that lacks a label. A small labelled diagram or table is worthwhile when it makes the boundaries of each group visible.
Move 2: Identify each category
Write the names of the two sets and what membership means. ‘Bought fruit’ and ‘bought vegetables’ can overlap; ‘left-handed’ and ‘right-handed’ may not be the intended exhaustive alternatives if the survey defines them differently. Follow the question rather than general assumptions.
Put a brief note on your practice working if needed. ‘A 18, B 15, both 7’ is much safer than memorising a number that lacks a label. A small labelled diagram or table is worthwhile when it makes the boundaries of each group visible.
Move 3: Locate the shared part first
If given the number in both, place that number in the centre. If given the total of everyone in at least one, derive the overlap from the two reported category totals. A missing overlap is not automatically zero.
Put a brief note on your practice working if needed. ‘A 18, B 15, both 7’ is much safer than memorising a number that lacks a label. A small labelled diagram or table is worthwhile when it makes the boundaries of each group visible.
Move 4: Find exclusive counts
Subtract shared membership from each reported group total. Confirm that neither result is negative. If the overlap exceeds either individual group, the stated quantities are inconsistent and need rereading.
Put a brief note on your practice working if needed. ‘A 18, B 15, both 7’ is much safer than memorising a number that lacks a label. A small labelled diagram or table is worthwhile when it makes the boundaries of each group visible.
Move 5: Answer the requested region
Distinguish exactly one, both, at least one, A only, B only and neither. The numbers needed for each answer are different even when they come from the same survey.
Put a brief note on your practice working if needed. ‘A 18, B 15, both 7’ is much safer than memorising a number that lacks a label. A small labelled diagram or table is worthwhile when it makes the boundaries of each group visible.
Move 6: Audit the full count
If a population total is given, add A only + both + B only + neither. The sum should equal the total. If it does not, find the earliest misread figure or mistaken double-count rather than blindly adjusting the final answer.
Put a brief note on your practice working if needed. ‘A 18, B 15, both 7’ is much safer than memorising a number that lacks a label. A small labelled diagram or table is worthwhile when it makes the boundaries of each group visible.
4. Worked survey cases: a different decision each time
Worked case 1: Two school sports
Given situation: 18 enjoy basketball, 15 enjoy badminton, and 7 enjoy both in a class of 30. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: Basketball only = 11; badminton only = 8; both = 7; at least one = 26; neither = 4. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: The class size is not 33; seven members were counted in both reported category totals. Checking 11 + 7 + 8 + 4 = 30 verifies all four regions. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Change the overlap from 7 to 10 while keeping the two group totals and class size. Recalculate each region. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 2: Reading two kinds of books
Given situation: Twenty pupils read adventure stories, fourteen read mysteries, and six read both. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: Exactly one genre = (20 − 6) + (14 − 6) = 22; at least one = 28. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: A pupil may answer 28 when asked ‘exactly one’. The shared six readers belong in ‘at least one’ but must be excluded from ‘exactly one’. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Write a second question asking how many read only mysteries. Answer it without confusing 14 with 8. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 3: A club where everyone joins something
Given situation: A group of 32 pupils joins either chess or drama or both. Twenty join chess; nineteen join drama. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: The shared group is 20 + 19 − 32 = 7; chess only = 13; drama only = 12. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: The clause ‘everyone joins something’ makes neither equal zero. Without it, 32 would not automatically equal the union and the overlap would be uncertain. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Add that three pupils join neither instead. Recompute the union and the overlap. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 4: Finding the unreported category
Given situation: A survey covers 40 pupils. Twelve choose Science Club only, nine choose Reading Club only, and five choose both. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: At least one = 12 + 9 + 5 = 26; neither = 40 − 26 = 14; Science total = 17, Reading total = 14. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: The exclusive figures are not the total membership of each club. Add the overlap to each only group when the question asks for the club’s full membership. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Make the survey total 45 and keep the three regions. State which counts change. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 5: When the overlap is unknown
Given situation: Twenty-four pupils like oranges and nineteen like apples; the total class has 40 pupils. There is no count for both or neither. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: One exact overlap cannot be determined from these data alone. There could be three or more in both, depending on how many like neither. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: The sum 43 exceeds class size 40, so at least three must appear in both, but the data do not prove exactly three unless everyone likes at least one. ‘At least three’ is a bound, not the exact answer. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Add ‘every pupil likes at least one fruit’ and find the overlap exactly. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 6: A question about neither
Given situation: In a class of 35, 17 pupils take part in art, 21 in music, and 8 in both. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: At least one = 17 + 21 − 8 = 30; neither = 35 − 30 = 5. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: A wrong solution subtracts 17 and 21 from 35 without returning the shared eight. This incorrectly removes shared pupils twice and may even produce a negative count. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Make the class size 38, then change only the overlap to 10. Compare the separate effects. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 7: Children choosing snacks
Given situation: Twenty-two pupils buy fruit, sixteen buy yoghurt, and seven buy both. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: Fruit only = 15; yoghurt only = 9; exactly one = 24; at least one = 31. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: A ‘both’ buyer purchases two kinds but represents one child. The question may be about the number of purchases or the number of distinct customers. Here it is about pupils; count each once. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Change the question to total snack items if each pupil buys one item of each selected kind. Explain why 22 + 16 now has another interpretation. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 8: A graph showing category totals
Given situation: A bar chart shows 14 walkers and 17 cyclists, while a note says 5 pupils use both modes on different days. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: Distinct pupils using at least one mode = 14 + 17 − 5 = 26. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: The bars show category participation, not an exclusive partition. A graph alone can tempt a learner to assume every bar contains different people. The note changes how totals may be combined. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Keep bar heights but change the overlapping group to eight pupils; recompute the union. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 9: Two tests passed
Given situation: Twenty-eight pupils passed an English quiz; twenty-four passed a Mathematics quiz; eighteen passed both. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: At least one pass = 28 + 24 − 18 = 34; English only = 10; Maths only = 6. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: ‘Both passed’ is not a third separate group to add on top of the two reported pass totals. It is already inside each. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: If there are 38 pupils, calculate how many passed neither test and discuss what that number means. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 10: Two languages spoken
Given situation: Thirty-one pupils speak Language A, seventeen speak Language B, and eleven speak both. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: Language A only = 20, Language B only = 6, both = 11, union = 37. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: A question asking how many speak Language B is answered by 17, not 6. The six is B only. Read whether ‘only’ is actually present. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Change the question from ‘Language B only’ to ‘at least one language’; do not discard the middle. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 11: A class announcement with opt-ins
Given situation: Fifteen pupils opt into a museum trip, twelve opt into a science workshop, and four opt into both. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: Unique participants in at least one option = 15 + 12 − 4 = 23. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: This does not mean there are 27 different pupils; it means there were 27 recorded activity selections from 23 people. If the question asks for selected activities instead of pupils, the answer could be different. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Assume each participant chooses exactly one ticket per selected activity. Find the number of activity selections. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 12: Discover the shared group from only counts
Given situation: Nine choose robotics only, seven choose coding only, and the two activities have a combined membership of 25 distinct pupils. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: Both = 25 − 9 − 7 = 9; robotics total = 18, coding total = 16. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: The information is about exclusive regions, so it is wrong to subtract an unknown overlap from them again. Build the regions in the order the data specifies. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: If 30 distinct pupils are in at least one activity with the same only counts, find the new overlap. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 13: Everyone in the class versus respondents
Given situation: Forty pupils are in a school class; thirty-six complete a survey. Among respondents, 20 select football and 18 select swimming, with 8 selecting both. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: Among survey respondents, at least one = 30 and neither = 6. The four nonrespondents are not automatically pupils who selected neither. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: Population definitions matter. ‘Did not respond’ is different from ‘responded and chose neither’. Combining them silently changes the group whose choices were measured. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Find how many of all forty pupils are known to have selected at least one based on the responses, and explain what remains unknown. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 14: The same person counted in two time slots
Given situation: A library records 25 pupils at a morning activity and 23 at an afternoon activity. Nine attend both. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: Unique pupils attending at least one session = 25 + 23 − 9 = 39. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: The activity-count total of 48 represents attendances, not unique people. Ask whether the problem requests events, registrations, seats or distinct pupils before choosing the result. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: If the nine shared pupils attend exactly both sessions, explain why 48 attendances can correspond to 39 people. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 15: Percentages of a known class
Given situation: In a class of 50, 60% join Art, 40% join Music, and 10 pupils join both. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: Art total = 30; Music total = 20; at least one = 30 + 20 − 10 = 40; neither = 10. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: Convert the percentages using the class as the whole before applying overlap. Do not subtract ten from percentages and label the result ‘pupils’. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Change the class size to 40 and keep percentages, then supply a new overlap of eight pupils. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 16: Two colours of stickers
Given situation: Among 28 pupils, 16 own a red sticker, 19 own a blue sticker, and 10 own both. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: Only red = 6, only blue = 9, both = 10, neither = 3. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: A number belonging to the red group can include a pupil who owns blue as well. The ‘only red’ region is smaller than the full red total. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Ask for the number who own exactly one colour and verify that it equals 15. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 17: Two kinds of revision resource
Given situation: A survey of 48 pupils finds that 29 use printed notes, 24 use digital notes, and 13 use both. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: At least one = 40; neither = 8; exactly one = (29 − 13) + (24 − 13) = 27. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: The subtraction once in the union counts everyone who uses both exactly once. Exactly one, by contrast, removes that entire shared region from the union. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Change only the ‘both’ value to sixteen, and say what happens to ‘exactly one’ and ‘neither’. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 18: When the numbers cannot be true
Given situation: A question states 12 pupils enjoy chess, 9 enjoy drama, and 14 enjoy both. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: These data are inconsistent: the shared group cannot exceed the smaller category total of nine. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: A method cannot repair impossible inputs by producing a negative only-group and carrying on. Read the values again; if the problem really states them, the claims cannot all describe the same population. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Replace the shared count with seven. Compute the union and check every exclusive region is nonnegative. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 19: A minimum overlap bound
Given situation: A class has 30 pupils. Twenty-two join choir and twenty-one join dance. Some may join neither. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: At least 22 + 21 − 30 = 13 pupils must be in both, but the exact overlap is not fixed. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: Even if there is no other information, a union cannot exceed the thirty pupils in the class. The overlap must remove enough duplicate counts to respect that ceiling. More overlap is possible if some pupils join neither. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Add that exactly two pupils join neither; the union is 28 and the overlap becomes 15. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
Worked case 20: Survey wording that changes the question
Given situation: Twenty pupils select History, twelve select Geography, and four select both. Compare ‘selected at least one’ with ‘selected exactly one’. Pause here and classify each number as ‘A total’, ‘B total’, ‘both’, ‘neither’, ‘only’ or overall population.
Worked answer: At least one = 20 + 12 − 4 = 28; exactly one = (20 − 4) + (12 − 4) = 24. Before moving on, point to every counted region and explain why each person is included the correct number of times.
Why the tempting route fails: The numerical data did not change. Only the requested region changed. A learner who reuses 28 for the second prompt did not fail subtraction; they failed to update the selection condition. A correct final number is useful only if the category labels were maintained throughout the working.
Changed-detail practice: Write a third question asking for Geography only, then a fourth asking for both. Identify the one region for each. Hide the earlier answer and rebuild the four regions. If a change makes the data inconsistent, say why rather than manufacturing a neat count.
5. How to spot an impossible or underdetermined answer
The overlap cannot be negative and cannot exceed either group total. If 12 pupils are in A and 9 in B, ‘both’ must be between zero and nine, subject to any other population constraints. The count in at least one group cannot exceed the population size when the survey describes one population. It also cannot be smaller than the larger individual group. These bounds are useful before and after exact arithmetic.
An underdetermined question does not supply a unique overlap. Suppose 18 pupils like reading and 15 like cycling in a class of 30, with no information about both or neither. There must be at least three in both because 18 + 15 = 33 group memberships but there are only thirty pupils. However, there could be more than three in both if some pupils like neither. Writing ‘three exactly’ invents the assumption that everyone chose at least one.
An impossible question is different. If 19 pupils are said to be in both groups but one reported group contains only 12, the input numbers conflict. Do not proceed to ’12 − 19 = −7 pupils’ and assume the negative count means something physical. Verify the wording, category, unit and population. If the numbers are still inconsistent, the model cannot describe that survey.
A valid result also needs whole-number people. If a computation suggests 6.5 pupils attended a club, check whether you accidentally used a percentage, rounded too soon or combined mismatched populations. The object being counted controls whether a fractional result makes sense. This is a context check, not a licence to round every impossible number up.
6. Three ways to verify a result without doing the same sum twice
Region reconstruction: Find A only, both, B only and neither, then add them. When the overall population is known, the sum should equal it. This reconstructs the survey from its nonoverlapping pieces. If a student first used A + B − both, the region method offers a different perspective on the same underlying count.
Membership accounting: Count activity memberships before counting people. In the basketball/badminton example there are 33 memberships: 18 basketball and 15 badminton. Seven pupils have two memberships. Removing one extra count per shared pupil gives 26 distinct people. This explains the subtraction and prevents blindly subtracting shared people twice.
Bounds: The union must be at least as large as each group and at most the sum of both groups; it also cannot exceed the population. In the case of 18 and 15 group members in a class of 30 with seven shared, the union of 26 satisfies all relevant bounds. A computed union of 33 would exceed the class and reveal a missing overlap adjustment.
Use a check that targets the likely failure. If you frequently confuse ‘exactly one’ and ‘at least one’, draw the regions. If you frequently misread survey language, rewrite the requested group in words. If you frequently make arithmetic slips, add the four region counts independently. Repeating the same mistaken calculation three times does not make it safer.
7. A progression from small counters to unseen word problems
Foundation: Draw two overlapping circles on paper and use physical counters or small numbers. Place 3 in A only, 2 in both and 4 in B only. Compute A total = 5, B total = 6, union = 9 and exactly one = 7. Then change just the middle from two to three. The learner should see how membership totals and distinct-person totals change for different reasons.
Recognition: Read ten short questions without calculating. Mark each request as A only, B only, both, at least one, exactly one or neither. Include one ‘not enough information’ case. This prevents a pupil who has memorised the union formula from applying it every time they see two group totals.
Direct calculation: Use six surveys with A total, B total and both supplied. Solve the four regions and verify the population when given. Switch the group names and contexts so the learner cannot rely on reading the larger number as group A. Introduce percentages only after absolute counts are secure.
Reverse problems: Give a population total and state that everyone selected at least one, then ask for the overlap. Later introduce a neither group. The learner should explicitly derive the union as total minus neither before finding the overlap. This is not merely a different algebraic arrangement; it is a changed route from words to the same four regions.
Mixed application: Combine bar charts, school registrations and two-choice notices. The representation may change but the membership logic should remain stable. Ask whether a bar count refers to unique people or activity selections. This is where reading skills and mathematics meet without treating the graph as a substitute for the language of the question.
Delayed transfer: Leave the subject for two days, then return with unfamiliar labels and one underdetermined case. One week later, mix two-group items with ordinary addition problems where categories really are disjoint. The learner must decide whether an overlap correction is necessary, not merely carry out subtraction whenever two groups appear.
8. Examination control: the shortest safe route
Read the question’s final request before calculating. Then draw two quick overlapping circles or write ‘A only | both | B only | neither’ in a small line. Place the shared count if known. If only the two category totals and the overall class count are given, check whether the problem says everyone selected one or more. This choice determines whether the overlap is exactly recoverable.
In a timed paper, a quick correct setup is usually more useful than a large decorative diagram. You need labels that will survive arithmetic. A small ‘both = 7’ in the middle and ‘A total = 18’ beside the left circle may be sufficient. Keep ‘only’ attached to exclusive regions and do not treat the whole class as the union unless the text says so.
If you receive an impossible result, do not immediately erase all your working. Check whether you subtracted the shared group twice, used a total from another population, or assumed that a nonrespondent chose neither. One focused check often finds the exact first wrong decision. If the sum remains underdetermined, do not invent a value just because the question appears to want one.
You can pair this method with Vol 0090 on reading graph scales when membership totals come from a graph. Vol 0090 handles extracting numerical values; this volume handles whether those values may be combined as distinct people. For more complete lessons on fractions, percentages, representation and word problems, continue at the Mathematics Learning Hub.
9. Parents and tutors: diagnose the relationship error
Do not begin a correction by saying ‘you forgot the formula’. Ask which pupils would be counted twice if you added the two reported categories. An answer such as ‘those who chose both’ proves the learner understands the source of the duplicate. If the child cannot identify those pupils, reconstruct the four regions with named counters and a smaller example before writing the subtraction.
A simple tutoring record separates three capacities: locating the shared group, calculating exclusive regions, and selecting the requested combination. A learner may complete the arithmetic perfectly but answer ‘at least one’ when the task says ‘exactly one’. Another learner may choose the right region but subtract seven from the wrong category. Those are different interventions; the next worksheet should target the observed decision.
For pupils who often get lucky with examples where nobody overlaps, vary the overlap deliberately. Start with zero shared pupils, then add some, then use a survey where all pupils belong to both categories. Ask the child to explain which totals remain possible. This creates a more robust understanding than repeating twenty identical ‘add and subtract the middle’ questions.
A parent can also reduce unnecessary pressure by asking a child to explain one correct region choice rather than turning every exercise into a speed challenge. A ten-minute discussion of why 18 + 15 does not represent 33 different pupils may remove a recurring error. The goal is independent reasoning, not a longer homework pile.
10. Frequently asked questions
Do I always subtract the overlap when two groups appear?
Only when you are combining their membership totals to count distinct people or objects. If the question asks for total activity selections, counting both memberships may be appropriate. Invent an unfamiliar survey example that demonstrates the rule, then explain which people are counted and which are not. This tests understanding more strongly than memorising the formula.
What does ‘both’ mean in a two-group survey?
It names people belonging to A and B together. It is already included inside each full group total, so adding the full totals counts shared people twice. Invent an unfamiliar survey example that demonstrates the rule, then explain which people are counted and which are not. This tests understanding more strongly than memorising the formula.
What is the difference between ‘exactly one’ and ‘at least one’?
Exactly one includes the A-only and B-only regions. At least one also includes the shared middle. Invent an unfamiliar survey example that demonstrates the rule, then explain which people are counted and which are not. This tests understanding more strongly than memorising the formula.
What if I know the class size but not the overlap?
You may be able to set a lower bound on the overlap, but an exact value requires enough information about the union or neither group. Invent an unfamiliar survey example that demonstrates the rule, then explain which people are counted and which are not. This tests understanding more strongly than memorising the formula.
Can an overlap be larger than either group?
No. Everyone in the shared group belongs to both groups, so that count cannot exceed the smaller full group total. Invent an unfamiliar survey example that demonstrates the rule, then explain which people are counted and which are not. This tests understanding more strongly than memorising the formula.
Why do you subtract the overlap once, not twice?
A shared pupil is counted two times when you add the group totals, but should be counted one time in the union. Removing one extra copy is sufficient. Invent an unfamiliar survey example that demonstrates the rule, then explain which people are counted and which are not. This tests understanding more strongly than memorising the formula.
Does a bar chart prove there is no overlap?
No. The chart’s categories and notes define whether the counts are exclusive. Two bars can report separate activities that some pupils attend together. Invent an unfamiliar survey example that demonstrates the rule, then explain which people are counted and which are not. This tests understanding more strongly than memorising the formula.
Is a two-circle diagram compulsory?
Not always. Use any valid representation required or permitted by the question. A small region table or well-labelled arithmetic can perform the same reasoning job. Invent an unfamiliar survey example that demonstrates the rule, then explain which people are counted and which are not. This tests understanding more strongly than memorising the formula.
Can I use percentages?
Yes, after identifying the common population and converting percentages to compatible counts or shares. Do not combine percentages from different whole groups without establishing a valid basis. Invent an unfamiliar survey example that demonstrates the rule, then explain which people are counted and which are not. This tests understanding more strongly than memorising the formula.
11. Your next route
Build your own survey with 36 pupils and two interests. Choose legal numbers for A only, both, B only and neither that add to 36. Then convert your region figures into the two reported membership totals. Ask another learner to reconstruct the regions using only those totals plus suitable information about the union. If they cannot determine the overlap, identify which missing fact your question must add.
When that exercise works even after the names and counts change, you are ready to continue to the next decision. Review the Vol 0093 English guide if the wording of a question itself causes misunderstanding; a precise relationship in language comes before safe arithmetic. Return to the PSLE Learning Guide for the broader English, Mathematics and Science performance route, and follow the Mathematics Learning Hub for dedicated topic teaching.