G1 Maths K110 | Cumulative Frequency, Quartiles and Percentiles

Independent educational guide, October 2026, based on SEAB’s official 2027 K110 syllabus. All datasets and graph readings described as fictional are original practice, not official questions or live statistics about Singapore businesses, schools or services. Follow the current syllabus and actual question’s quartile conventions and accuracy requirements.

The official 2027 K110 assessment context

Part of K110	Official content	What to remember
Paper 2	90 minutes, 50 marks, 50% of Maths; Number and Algebra with Statistics and Probability	All questions compulsory
Data representations	Tables, bar graphs, pictograms, line graphs, pie charts, dot diagrams, equal-width histograms and cumulative frequency diagrams	Each chart answers different questions
Data summaries	Mean, mode, median of ungrouped data; percentiles, quartiles, range, interquartile range	Calculate and interpret, not just name
Probability	Simple single-event situations, including listing possible outcomes	A separate Paper 2 focus
Paper 1	90 minutes, 50 marks, 50%; Number and Algebra with Geometry and Measurement	Do not substitute the Paper 2 statistics strand
Presentation	Approved calculator allowed; essential working required; 3 significant figures for non-exact results unless directed otherwise	Interpret chart units and real context

Understand cumulative frequency in three moves

Move 1: Read the total. The final cumulative frequency tells how many observations are represented. Move 2: Find the cumulative position. A median or percentile corresponds to a position within that total, according to the applicable convention. Move 3: Read across to the data value. Find that cumulative count on the vertical scale and trace to the curve, then down to the horizontal value. A frequency is a count; the value at a percentile may instead be minutes, centimetres, dollars or another measurement. They must not be swapped.

The same data can look different depending on its representation. A frequency table tells how many observations are in intervals. A cumulative-frequency table gives totals up to successive boundaries. A cumulative-frequency curve shows those accumulated totals visually. The best way to prepare is to convert one small fictional dataset among these forms and check that no count disappears or appears twice.

Navigate by difficulty

Read data and charts · Cumulative frequency · Quartiles and percentiles · Original worked datasets · Frequently asked questions

1. Understand what one observation is

A measurement is not the same thing as the person or object being measured. In a set of waiting times, one observation may be one recorded wait. Identifying what is counted helps decide whether the cumulative total represents people, journeys or transactions.

A table showing thirty journeys does not necessarily represent thirty unique commuters; the same traveller might contribute several trips. Use the context provided, not a guess about identities.

Ask the pupil to identify the unit, sample and number of observations before adding or comparing counts. This single step prevents misleading statements later.

2. Separate a population from a sample

A sample contains selected observations from a broader group. An average calculated from a sample describes those collected values. It does not automatically describe everyone in the population unless there is an appropriate basis for generalising.

If a classroom measures twenty students’ journey durations, the resulting median belongs to those twenty records. Calling it Singapore’s national median would be unsupported.

Use the scope of the question in the final sentence. Honest statistical interpretation is more valuable than an exaggerated conclusion that the graph cannot prove.

3. Identify numerical and categorical variables

Numerical variables represent amounts such as minutes, temperature or price. Categorical variables name groups such as transport type or preferred colour. A person who computes the mean of bus route numbers is usually treating identifiers as measurements.

If a survey lists routes 7, 10 and 15, the numbers are labels, not necessarily distances. By contrast, journey durations on those routes could be averaged when the problem and data support it.

Decide the variable’s nature before choosing a representation. A pie chart might show the distribution of route categories, whereas a cumulative curve needs ordered numerical values.

4. Read graph axes and scales exactly

A graph’s vertical axis might be frequency or cumulative frequency, while the horizontal shows numerical values. A curve that rises from left to right is not a time history unless the horizontal axis is time.

Imagine a cumulative graph of package mass measured in kilograms. A height of 50 refers to fifty observations accumulated up to a mass boundary, not fifty kilograms.

Tell the pupil to point to each axis label and state its unit aloud. Then ask what one plotted point actually represents before calculating.

5. Use frequency tables correctly

Frequency is the count of observations with a particular value or within an interval. Adding all frequencies gives the total represented by the table, assuming groups cover each observation once and without overlap.

For a fictional table with interval frequencies 3, 7 and 10, the total is twenty. A student who adds only the first and last has omitted the middle group.

Convert the numbers into one cumulative column: 3, 10, 20. The progression helps explain how accumulation differs from a stand-alone interval count.

6. Read a pictogram key before counting

A pictogram picture may represent several objects rather than one. A half picture may represent a fraction of the key’s value, if the diagram uses that convention. Students should not calculate from artwork without reading the stated scale.

If a school chart assigns one symbol to five books, three symbols represent fifteen books. The symbol count and book count differ.

Change the key to four books and ask how the total changes. Understanding the representation is more useful than simply recognising the book icon.

7. Interpret bar charts with appropriate comparisons

A bar chart compares category frequencies or values. The numerical scale controls the meaning of bar height; a truncated vertical axis can exaggerate small differences.

Suppose two fictional clubs have attendance 94 and 98. The difference is four, even if a graph beginning at ninety makes one bar look much taller.

A valid interpretation names the categories, values, difference and relevant sample context rather than merely reporting that one bar ‘looks huge’.

8. Understand line graphs as recorded patterns

A line graph shows values across ordered observations, often but not always time. A rise may describe an increasing quantity within the plotted range. It does not prove why the pattern occurred.

For hypothetical monthly visitors, values 50, 60 and 58 show first an increase and then a decrease. A claim that advertising caused the change requires other evidence.

Read the axes, find the values and describe the observed sequence before offering a qualified interpretation.

9. Translate pie-chart proportions into counts

Pie-chart sectors represent shares of a whole, with 360° for the complete circle. A quarter sector is 90° regardless of how many observations the full chart contains.

For a fictional group of forty, a 90° sector represents ten people. For a group of eighty with the same proportion, it represents twenty.

This is why students must read the total before calculating a count. The angle alone identifies the proportion, not the absolute number.

10. Count repeated observations in dot diagrams

A dot diagram uses marks for individual observations, including repeats. A learner who records one value per labelled position misses the actual number of data points.

If a dot diagram shows values 2, 2, 3, 3, 3 and 5, there are six observations, not three. Their mean is eighteen divided by six, which is three.

Ask the student to write the full ordered list before finding a median or mode. This simple transcription supports reliable statistical summaries.

11. Understand an equal-width histogram

K110 specifically includes histograms with equal class intervals. Grouped numerical data are displayed in touching bars, with the frequency attached to the relevant class under the given convention.

If bins represent durations 0–5, 5–10 and 10–15 minutes at equal widths, the bar height can be read as frequency in the specified arrangement. The exact individual duration within a bin is usually lost.

Do not claim every person in the 5–10 interval took precisely 7.5 minutes. A midpoint can be useful in a stated approximation, but it is not a recovered raw observation.

12. Recognise what cumulative means

Cumulative frequency is a running total. Each successive count includes the previous observations and those newly added up to the next boundary. It is not the number that occurred only in the final interval.

Consider frequencies 4, 6, 5 and 5. The cumulative totals are 4, 10, 15 and 20. Each total tells how many have been included through that point.

A non-decreasing cumulative graph makes sense for a properly ordered dataset. If the recorded cumulative total falls, review the construction or scale.

13. Construct a cumulative table from intervals

Begin with each interval frequency and add it to the preceding running total. State the class boundaries correctly. The total in the last row should equal the sum of all original frequency values.

For fictional interval counts 3, 7, 8 and 2, write running totals 3, 10, 18 and 20. Check 3+7+8+2=20.

Now change the third frequency from eight to nine. The final cumulative total should become twenty-one, and the later cumulative values must update. This tests actual understanding of accumulation.

14. Get one class frequency from two cumulative totals

When the cumulative values refer to corresponding successive boundaries, subtracting them gives the number of new observations in the intervening class. Students often report the later cumulative total instead of the interval’s count.

If cumulative frequency rises from twelve to twenty-five between two class boundaries, the interval contains thirteen additional observations.

Ask what the earlier twelve represents. It is not discarded; it is simply removed from the later running total when the question asks about the intervening range.

15. Recognise why a cumulative graph never needs negative frequencies

Cumulative totals count more or the same observations as boundaries move through sorted data. A drop would imply observations disappeared from the total, which is not what cumulative frequency means in a fixed dataset.

A graph may have flat segments where no observations lie in a range, but it should not fall as the boundary increases. The shape depends on the distribution, not on a sales-like trend over time.

Ask the learner to explain this property verbally before drawing points. It is a powerful reasonableness check.

16. Understand upper class boundaries

Grouped data often use a cumulative total at an upper boundary. The precise interpretation of inclusion or exclusion depends on the class definitions supplied. The candidate should preserve that convention rather than invent overlapping groups.

For a fictional interval 10–20 followed by 20–30, the boundary treatment determines where an exact value of twenty belongs. A consistent method ensures it is counted once.

In a class test, read the published table headings carefully. The diagram and question’s conventions govern the plotted positions.

17. Plot cumulative points with the correct coordinates

For a cumulative-frequency curve, the horizontal position represents the data boundary and the vertical position the accumulated count. Reversing these produces a graph with the wrong meaning even if all numbers have been copied accurately.

If the cumulative total is sixteen up to a boundary of thirty minutes, one point is (30 minutes, 16 observations), assuming the stated class construction.

Have the pupil label every coordinate in words, then check that the last plotted cumulative value equals the total sample size.

18. Recognise the total N from the curve

The final cumulative frequency gives the total number of observations represented. The total is needed for determining the median and percentile positions. A candidate should not take the largest horizontal data value as N.

If the curve ends at eighty on the cumulative-frequency axis and at thirty-five minutes on the value axis, the dataset contains eighty observations, not thirty-five.

Ask the learner what each number means. Unit recognition prevents a foundational mistake.

19. Find a value below a given cumulative count

A question may ask for the value below which a stated number of observations occur. Locate the count on the vertical axis, move to the curve and then read the corresponding horizontal value.

If a fictional curve shows cumulative count fifty at a waiting time of twelve minutes, that point describes approximately fifty observations up to the stated time boundary under its convention.

It does not automatically mean the mean wait is twelve minutes. The graph provides a positional relationship, not every average.

20. Find the count below a given value

The reverse question may specify a horizontal value and ask how many observations are up to that value. Start at the value, trace to the curve and read the cumulative count on the vertical axis.

For a fictional plot, if the curve is at forty observations when x=20 minutes, then roughly forty observations have durations at or below the relevant boundary.

Follow the actual line and scale. Do not guess the answer from the nearest labelled point if interpolation is needed and permitted.

21. Understand interpolation as an estimate

When the requested point lies between plotted coordinates, a graph reading may involve interpolation. The quality of the estimate depends on the curve and scale. It should not be presented as an exact original observation.

If the curve crosses the height fifty between fifteen and twenty minutes, the horizontal reading will be within that range, with precision limited by the graph.

Students should show how the reading was obtained and use the accuracy specified by the exam instead of inventing decimal places.

22. Recognise information that grouped curves cannot reveal

A cumulative curve summarises a dataset, but exact individual measurements may not be recoverable. A pupil should not claim to know that every participant in an interval had the same value.

A curve may show that twenty observations lie between ten and twenty minutes without indicating each person’s exact time.

This limitation matters if a question asks for an estimate versus a precise raw reading. Use the wording and available data honestly.

23. Compare two cumulative curves

If two curves use the same scale and comparable conditions, their horizontal positions at the same cumulative proportion may indicate differences in typical values. Sample size must also be considered before comparing raw counts.

For two groups of different sizes, comparing at cumulative count twenty may correspond to different percentiles. A proportion-based comparison could be more meaningful.

State the question’s target and verify that axes, units and sample definitions are compatible.

24. Distinguish the curve from an ordinary time-series graph

A cumulative-frequency curve does not necessarily show events unfolding over time. Even when the horizontal axis uses minutes, it may represent individual waiting durations rather than the clock time during an experiment.

The graph’s rise means increasing numbers of observations included by progressively larger data values; it does not mean more people necessarily arrived during a particular minute.

Identifying what the horizontal measurement represents is essential before writing an interpretation.

25. Find the median position

The median divides ordered observations into lower and upper portions. On a cumulative-frequency curve with total N, a standard graphical reading uses a position around half N; the corresponding horizontal data value is the estimated median.

For N=80, locate cumulative frequency forty, trace to the curve and read the data axis. That number is a waiting time or other measured value, not ‘forty’ by default.

The exact convention for a discrete, small, ungrouped dataset can differ from graphical estimation. Follow the method specified in the problem.

26. Find the first quartile position

The first quartile Q1 represents a lower-quarter position in the ordered distribution. A usual cumulative-frequency graphical estimate uses a count around one quarter of N.

For N=80, one quarter is twenty; identify the curve at cumulative count twenty and read its corresponding horizontal measurement.

If the curve’s scale is in minutes, the Q1 answer is in minutes. The mark position is a count; the quartile value is a data quantity.

27. Find the third quartile position

The third quartile Q3 corresponds to about three quarters of the ordered observations. It helps describe the upper boundary of the middle half of the dataset.

For N=80, three quarters is sixty. Trace the curve from cumulative frequency sixty to the horizontal axis and report the estimated value with suitable graph-reading accuracy.

Do not call the value ‘60%’ unless the horizontal measurement itself is a percentage. The sixty here is a position count.

28. Understand interquartile range

IQR is the difference Q3−Q1. It describes the spread of the middle half of an ordered distribution and is often less affected by extreme values than the full range.

If graphical readings give Q1=12 minutes and Q3=26 minutes, IQR=14 minutes. The unit is minutes, and the meaning concerns central spread rather than a duration actually observed for one person.

Use an appropriate level of precision for the graph readings; a calculated IQR cannot be more exact than its underlying estimates.

29. Understand the full range

Range is maximum minus minimum. It considers the extremes, while IQR focuses on the middle half. Different datasets can have the same median and widely different ranges.

For values 2,4,6,8,10 the range is eight. Replace ten with one hundred and the range becomes ninety-eight, even though the first four values remain unchanged.

Explain why a large extreme value can strongly affect the full range while other summaries may move less.

30. Distinguish percentiles from percentages

A percentile locates a position in a distribution. A percentage mark describes a proportion of marks earned. They are not equivalent: an 80% examination score is not automatically the 80th percentile.

A percentile rank depends on how the comparison group performed. If almost everyone scores above ninety, a mark of eighty could be below many peers despite being 80% of available marks.

Students should identify whether the question asks for a percentage of a quantity or a percentile position in a dataset.

31. Estimate a percentile value from the curve

To find an approximate pth percentile on a cumulative graph, identify a position near p/100 of total N under the stated convention, then read the corresponding x-value.

For N=80, the 75th percentile uses a graphical position around sixty. If the curve at that count corresponds to 26 minutes, the estimated percentile value is 26 minutes, not sixty.

This resembles reading Q3. A reader should still pay attention to graph conventions and any specific examination instructions.

32. Find an approximate percentile rank from a value

Some questions give a measured value and ask approximately what share of observations fall up to that value. Read the cumulative count at the stated horizontal point and compare with the total N.

If the curve gives cumulative count forty-eight out of eighty at a particular value, that represents 48/80=60% of the sample up to that boundary under the graph’s convention.

Avoid claiming that the person ‘scored 60%’ unless the horizontal axis actually represents examination marks.

33. Calculate quartiles from ungrouped records

For a small ordered dataset, quartiles may be obtained by dividing the observations into suitable halves and finding their medians, subject to the question’s convention. The learner should sort first and follow the stated method.

For ordered values 2,4,6,8,10,12,14,16, the median is nine. Using the median-of-halves method, Q1=5 and Q3=13, giving IQR eight.

Do not assume that every software package or textbook uses identical percentile conventions for every small dataset. Follow the school and the question.

34. Understand the mean as a different measure

The mean is total divided by number of observations for ungrouped data. It is not read from the fifty-percent cumulative point, which relates to the median. This difference is one of the most important conceptual boundaries.

For fictional values 1,2,3,4,20, the mean is six and the median is three. A cumulative curve would not automatically give the mean at its midpoint.

Explain why the large final value influences the mean while the central ordered observation remains three.

35. Recognise the mode

The mode is the most frequent value or category. It may coincide with the median or mean in some datasets, but there is no universal rule that they are equal.

For data 2,2,2,3,4,8, the mode is two because it appears most often. The median is halfway between the third and fourth values, or 2.5, under standard ungrouped conventions.

Ask which measure suits a question about the most commonly chosen item versus the middle waiting time.

36. Compare datasets fairly

Two datasets can have equal medians but different spread. A single number rarely tells the whole story. For service data, consistency can matter alongside typical duration.

Imagine two clinics’ fictional waiting-time lists with the same median but one set ranging widely. Comparing IQR or range provides more context, provided the data are genuinely comparable.

This is not a claim about actual healthcare services; the example teaches how to interpret spread in everyday situations.

37. Recognise what a sample cannot prove

An increase in sample counts does not always imply an entire community’s behaviour has changed. Sampling method, time, population and data quality influence how results can be generalised.

If a fictional shop surveys twenty customers during one lunch hour, its observed median wait does not establish every customer’s wait across all operating hours.

The stronger conclusion states what the data actually supports and which additional evidence could broaden it.

38. Read an unfamiliar statistical question strategically

In the examination, begin with the action word and relevant data source. If asked to calculate, show the relationship and units. If asked to compare, identify both values and the basis. If asked to interpret, explain what the statistic means in context.

A pupil who sees Q3 may immediately write N×0.75 without noticing the question asks for the actual data value at that position. Trace both axes and the requested output.

Reasoning is a sequence of decisions, not a collection of isolated formulas.

39. Use calculator output sensibly

An approved calculator is allowed in both K110 papers, but drawing conclusions from a graph requires reading and interpretation. A calculator cannot choose the correct axis, identify a distorted scale or decide whether a percentile is meaningful.

For a fictional IQR calculation, the subtraction may be simple, while locating Q1 and Q3 on the cumulative curve is the real challenge.

Show the relevant working and answer in the original measurement units. SEAB notes that essential working omissions can lose marks.

40. Build a six-week improvement routine

Week one covers tables, categories and sample size. Week two trains frequency versus cumulative frequency. Week three uses graph plotting and reading in both directions. Week four develops median and quartiles. Week five covers percentiles, IQR and distribution comparison. Week six mixes changed examples in timed practice.

A student with unstable fraction and percentage skills should repair those first. The schedule is not an official prescribed timetable or a guarantee of a grade.

Strong progress appears when the learner can read an unfamiliar graph and explain a result without the tutor supplying each arrow.

41. Connect data literacy with future choices

Interpreting data can support further technical, service, business and community learning. A teenager who can distinguish an observed count from an invented cause gains a useful real-world habit, even if statistics never becomes their favourite subject.

G1 identifies the Maths subject level; admissions depend on full SEC qualifications and programme rules. One graph question should not be treated as a lifetime forecast.

The immediate educational goal is one defensible reading of an unfamiliar curve, checked against axes, units and sample size.

Six original worked examples of distributions

Example 1: frequencies become cumulative totals

A hypothetical school records grouped counts of 5, 9, 11 and 5 participants across four successive equal-width categories. The total is thirty. Running totals are 5, 14, 25 and 30. Plot each total at the appropriate upper boundary on a cumulative-frequency diagram, according to the classes given.

To find the frequency in the third class, subtract cumulative fourteen from twenty-five, giving eleven. Reading twenty-five as that class’s frequency would count earlier classes again.

Example 2: median from a cumulative curve

A fictional curve has cumulative total eighty. The graphical midpoint is around the fortieth observation. Suppose reading the curve at that count gives approximately 18 minutes. Then the estimated median is 18 minutes.

The value forty is a count position, not a waiting time. The 18 minutes is read from the horizontal measurement scale. The approximation depends on the graph and the specified method.

Example 3: quartiles and IQR

An illustrative curve for eighty observations gives approximate x-values of 12 minutes at cumulative count twenty and 26 minutes at cumulative count sixty. Under the usual graphical convention, these are Q1 and Q3, giving IQR 26−12=14 minutes.

Explain that the middle half of the observations spans roughly fourteen minutes under that graph. It does not mean every individual in that half waited fourteen minutes.

Example 4: a percentile position

A fictional cumulative curve for one hundred measured parcels is read at cumulative frequency ninety. Suppose that crossing corresponds to a mass of 4.2 kg. Under the graphical interpretation, the approximate 90th percentile is 4.2 kg.

That is a position in the distribution, not a ninety-percent mass or a statement that every parcel weighs 4.2 kg. The exact interpretation should follow the axis and boundary conventions.

Example 5: ungrouped median versus mean

Original data: 2,4,6,8,10,12,14,16. The median is (8+10)/2=9; the mean is 72/8=9 as well in this symmetric example. Under the median-of-halves quartile convention, Q1=5 and Q3=13.

Now replace the final sixteen with fifty. The mean becomes 106/8=13.25, but the median remains nine. This shows why the mean is more sensitive to a large extreme value.

Example 6: count and percentile interpretation

An invented chart reports cumulative count 48 out of 80 at a waiting time of fifteen minutes. The observed share at or below fifteen minutes is 48/80=0.6, or 60%, under the plotted convention.

A statement that fifteen minutes is near the 60th percentile of this sample is supported as an approximate graphical interpretation. It does not establish a national standard or prove the source of any delay.

Twelve checks that reveal the exact misconception

1. The frequency column is confused with running total

One class has eight observations but its cumulative total is twenty. Which quantity should describe the interval alone?

Find the plotted quantity and unit, identify the mathematical decision that should follow, and make one supported correction. Then change the data or question target and retry without a memorised answer.

2. The curve drops as x increases

A student draws a cumulative curve that goes down between class boundaries. Why does this conflict with a fixed dataset’s running total?

3. The last x-value is treated as N

The curve ends at 40 minutes and eighty observations. Which is the sample size?

4. Q1 is reported as a count instead of a time

The learner chooses cumulative count twenty but reports ‘twenty minutes’ without reading across the curve. What step is missing?

5. IQR is given as a percentage without evidence

Both quartiles are waiting times, yet the answer is in percent. Which measurement unit should the difference have?

6. A percentile is confused with a percentage grade

A student obtains 70% on a test and claims to be at the 70th percentile. What distribution information is missing?

7. An interval boundary is counted twice

Two neighbouring classes both include a value at their common boundary. What grouping convention should be specified?

8. A histogram midpoint becomes an exact observation

A class interval contains many data values, but the learner assumes all equal the midpoint. Which raw information has been lost?

9. An outlier shifts the mean

One unusually large value raises the mean while leaving the median almost unchanged. Why is that mathematically possible?

10. Different sample sizes are compared using the same count

Two cumulative curves describe groups of 20 and 80 people. Why might comparing a count of ten produce different relative positions?

11. The graph is read as a clock-time trend

The x-axis represents individual waiting durations, but the pupil claims attendance increased over the day. What does the axis actually measure?

12. A weak graph result is treated as an identity

The student finds cumulative frequency difficult and assumes all future education routes are closed. Which exact prerequisite and institutional information matter instead?

Frequently asked cumulative-frequency questions

Is cumulative frequency included in 2027 G1 Mathematics?

Yes. The official K110 Statistics and Probability strand lists cumulative frequency diagrams.

Which Maths paper assesses it?

Paper 2 covers Statistics and Probability alongside Number and Algebra; it contributes 50% of G1 Maths.

Are quartiles and percentiles required?

Yes. K110 explicitly lists percentiles, quartiles, range and interquartile range.

Are histograms included?

Yes, specifically histograms with equal class intervals.

How do I find the median on a cumulative curve?

Use a cumulative position around half the total N, then read the corresponding data-axis value according to the question’s convention.

How do I calculate IQR?

Subtract Q1 from Q3, retaining the data’s measurement unit.

What is the difference between frequency and cumulative frequency?

Frequency counts one category or interval; cumulative frequency totals observations up to a given value or boundary.

Does the 75th percentile mean a mark of 75%?

No. Percentile describes a position in a distribution; a percentage mark is a portion of attainable marks.

Can I obtain exact raw observations from a grouped histogram?

Not generally. Grouping loses individual detail, so avoid claiming unsupported precision.

Will every book or software use identical quartile conventions?

Not always. Follow the method and conventions specified by your question and school.

Can an approved calculator be used in Paper 2?

Yes, but reading axes, showing essential working and interpreting results remain necessary.

Does G1 Maths permanently fix future pathways?

No. Further-study decisions depend on the learner’s complete results, current programme prerequisites and school guidance.

Official syllabus and related eduKateSG guides

Read chart title, axis labels, measurement unit and total observations before doing any calculation.
For cumulative frequency, locate the count on the vertical axis and read the corresponding data value from the horizontal axis.
For an interval’s frequency, subtract the cumulative totals at its two endpoints when the conventions support that.
Keep centre (mean/median/mode), total spread (range) and middle-half spread (IQR) conceptually separate.
For percentile positions, use the data total and the method indicated by the diagram or question.
Explain what the resulting number says about the given sample, not automatically all of Singapore.

The official 2027 K110 assessment context

Part of K110	Official content	What to remember
Paper 2	90 minutes, 50 marks, 50% of Maths; Number and Algebra with Statistics and Probability	All questions compulsory
Data representations	Tables, bar graphs, pictograms, line graphs, pie charts, dot diagrams, equal-width histograms and cumulative frequency diagrams	Each chart answers different questions
Data summaries	Mean, mode, median of ungrouped data; percentiles, quartiles, range, interquartile range	Calculate and interpret, not just name
Probability	Simple single-event situations, including listing possible outcomes	A separate Paper 2 focus
Paper 1	90 minutes, 50 marks, 50%; Number and Algebra with Geometry and Measurement	Do not substitute the Paper 2 statistics strand
Presentation	Approved calculator allowed; essential working required; 3 significant figures for non-exact results unless directed otherwise	Interpret chart units and real context

Understand cumulative frequency in three moves

Navigate by difficulty

Read data and charts · Cumulative frequency · Quartiles and percentiles · Original worked datasets · Frequently asked questions

1. Understand what one observation is

A table showing thirty journeys does not necessarily represent thirty unique commuters; the same traveller might contribute several trips. Use the context provided, not a guess about identities.

Ask the pupil to identify the unit, sample and number of observations before adding or comparing counts. This single step prevents misleading statements later.

2. Separate a population from a sample

If a classroom measures twenty students’ journey durations, the resulting median belongs to those twenty records. Calling it Singapore’s national median would be unsupported.

Use the scope of the question in the final sentence. Honest statistical interpretation is more valuable than an exaggerated conclusion that the graph cannot prove.

3. Identify numerical and categorical variables

If a survey lists routes 7, 10 and 15, the numbers are labels, not necessarily distances. By contrast, journey durations on those routes could be averaged when the problem and data support it.

Decide the variable’s nature before choosing a representation. A pie chart might show the distribution of route categories, whereas a cumulative curve needs ordered numerical values.

4. Read graph axes and scales exactly

Imagine a cumulative graph of package mass measured in kilograms. A height of 50 refers to fifty observations accumulated up to a mass boundary, not fifty kilograms.

Tell the pupil to point to each axis label and state its unit aloud. Then ask what one plotted point actually represents before calculating.

5. Use frequency tables correctly

For a fictional table with interval frequencies 3, 7 and 10, the total is twenty. A student who adds only the first and last has omitted the middle group.

Convert the numbers into one cumulative column: 3, 10, 20. The progression helps explain how accumulation differs from a stand-alone interval count.

6. Read a pictogram key before counting

If a school chart assigns one symbol to five books, three symbols represent fifteen books. The symbol count and book count differ.

Change the key to four books and ask how the total changes. Understanding the representation is more useful than simply recognising the book icon.

7. Interpret bar charts with appropriate comparisons

A bar chart compares category frequencies or values. The numerical scale controls the meaning of bar height; a truncated vertical axis can exaggerate small differences.

Suppose two fictional clubs have attendance 94 and 98. The difference is four, even if a graph beginning at ninety makes one bar look much taller.

A valid interpretation names the categories, values, difference and relevant sample context rather than merely reporting that one bar ‘looks huge’.

8. Understand line graphs as recorded patterns

A line graph shows values across ordered observations, often but not always time. A rise may describe an increasing quantity within the plotted range. It does not prove why the pattern occurred.

For hypothetical monthly visitors, values 50, 60 and 58 show first an increase and then a decrease. A claim that advertising caused the change requires other evidence.

Read the axes, find the values and describe the observed sequence before offering a qualified interpretation.

9. Translate pie-chart proportions into counts

Pie-chart sectors represent shares of a whole, with 360° for the complete circle. A quarter sector is 90° regardless of how many observations the full chart contains.

For a fictional group of forty, a 90° sector represents ten people. For a group of eighty with the same proportion, it represents twenty.

This is why students must read the total before calculating a count. The angle alone identifies the proportion, not the absolute number.

10. Count repeated observations in dot diagrams

A dot diagram uses marks for individual observations, including repeats. A learner who records one value per labelled position misses the actual number of data points.

If a dot diagram shows values 2, 2, 3, 3, 3 and 5, there are six observations, not three. Their mean is eighteen divided by six, which is three.

Ask the student to write the full ordered list before finding a median or mode. This simple transcription supports reliable statistical summaries.

11. Understand an equal-width histogram

K110 specifically includes histograms with equal class intervals. Grouped numerical data are displayed in touching bars, with the frequency attached to the relevant class under the given convention.

Do not claim every person in the 5–10 interval took precisely 7.5 minutes. A midpoint can be useful in a stated approximation, but it is not a recovered raw observation.

12. Recognise what cumulative means

Consider frequencies 4, 6, 5 and 5. The cumulative totals are 4, 10, 15 and 20. Each total tells how many have been included through that point.

A non-decreasing cumulative graph makes sense for a properly ordered dataset. If the recorded cumulative total falls, review the construction or scale.

13. Construct a cumulative table from intervals

Begin with each interval frequency and add it to the preceding running total. State the class boundaries correctly. The total in the last row should equal the sum of all original frequency values.

For fictional interval counts 3, 7, 8 and 2, write running totals 3, 10, 18 and 20. Check 3+7+8+2=20.

Now change the third frequency from eight to nine. The final cumulative total should become twenty-one, and the later cumulative values must update. This tests actual understanding of accumulation.

14. Get one class frequency from two cumulative totals

If cumulative frequency rises from twelve to twenty-five between two class boundaries, the interval contains thirteen additional observations.

Ask what the earlier twelve represents. It is not discarded; it is simply removed from the later running total when the question asks about the intervening range.

15. Recognise why a cumulative graph never needs negative frequencies

A graph may have flat segments where no observations lie in a range, but it should not fall as the boundary increases. The shape depends on the distribution, not on a sales-like trend over time.

Ask the learner to explain this property verbally before drawing points. It is a powerful reasonableness check.

16. Understand upper class boundaries

For a fictional interval 10–20 followed by 20–30, the boundary treatment determines where an exact value of twenty belongs. A consistent method ensures it is counted once.

In a class test, read the published table headings carefully. The diagram and question’s conventions govern the plotted positions.

17. Plot cumulative points with the correct coordinates

If the cumulative total is sixteen up to a boundary of thirty minutes, one point is (30 minutes, 16 observations), assuming the stated class construction.

Have the pupil label every coordinate in words, then check that the last plotted cumulative value equals the total sample size.

18. Recognise the total N from the curve

If the curve ends at eighty on the cumulative-frequency axis and at thirty-five minutes on the value axis, the dataset contains eighty observations, not thirty-five.

Ask the learner what each number means. Unit recognition prevents a foundational mistake.

19. Find a value below a given cumulative count

A question may ask for the value below which a stated number of observations occur. Locate the count on the vertical axis, move to the curve and then read the corresponding horizontal value.

If a fictional curve shows cumulative count fifty at a waiting time of twelve minutes, that point describes approximately fifty observations up to the stated time boundary under its convention.

It does not automatically mean the mean wait is twelve minutes. The graph provides a positional relationship, not every average.

20. Find the count below a given value

The reverse question may specify a horizontal value and ask how many observations are up to that value. Start at the value, trace to the curve and read the cumulative count on the vertical axis.

For a fictional plot, if the curve is at forty observations when x=20 minutes, then roughly forty observations have durations at or below the relevant boundary.

Follow the actual line and scale. Do not guess the answer from the nearest labelled point if interpolation is needed and permitted.

21. Understand interpolation as an estimate

If the curve crosses the height fifty between fifteen and twenty minutes, the horizontal reading will be within that range, with precision limited by the graph.

Students should show how the reading was obtained and use the accuracy specified by the exam instead of inventing decimal places.

22. Recognise information that grouped curves cannot reveal

A cumulative curve summarises a dataset, but exact individual measurements may not be recoverable. A pupil should not claim to know that every participant in an interval had the same value.

A curve may show that twenty observations lie between ten and twenty minutes without indicating each person’s exact time.

This limitation matters if a question asks for an estimate versus a precise raw reading. Use the wording and available data honestly.

23. Compare two cumulative curves

For two groups of different sizes, comparing at cumulative count twenty may correspond to different percentiles. A proportion-based comparison could be more meaningful.

State the question’s target and verify that axes, units and sample definitions are compatible.

24. Distinguish the curve from an ordinary time-series graph

The graph’s rise means increasing numbers of observations included by progressively larger data values; it does not mean more people necessarily arrived during a particular minute.

Identifying what the horizontal measurement represents is essential before writing an interpretation.

25. Find the median position

For N=80, locate cumulative frequency forty, trace to the curve and read the data axis. That number is a waiting time or other measured value, not ‘forty’ by default.

The exact convention for a discrete, small, ungrouped dataset can differ from graphical estimation. Follow the method specified in the problem.

26. Find the first quartile position

The first quartile Q1 represents a lower-quarter position in the ordered distribution. A usual cumulative-frequency graphical estimate uses a count around one quarter of N.

For N=80, one quarter is twenty; identify the curve at cumulative count twenty and read its corresponding horizontal measurement.

If the curve’s scale is in minutes, the Q1 answer is in minutes. The mark position is a count; the quartile value is a data quantity.

27. Find the third quartile position

The third quartile Q3 corresponds to about three quarters of the ordered observations. It helps describe the upper boundary of the middle half of the dataset.

For N=80, three quarters is sixty. Trace the curve from cumulative frequency sixty to the horizontal axis and report the estimated value with suitable graph-reading accuracy.

Do not call the value ‘60%’ unless the horizontal measurement itself is a percentage. The sixty here is a position count.

28. Understand interquartile range

IQR is the difference Q3−Q1. It describes the spread of the middle half of an ordered distribution and is often less affected by extreme values than the full range.

If graphical readings give Q1=12 minutes and Q3=26 minutes, IQR=14 minutes. The unit is minutes, and the meaning concerns central spread rather than a duration actually observed for one person.

Use an appropriate level of precision for the graph readings; a calculated IQR cannot be more exact than its underlying estimates.

29. Understand the full range

Range is maximum minus minimum. It considers the extremes, while IQR focuses on the middle half. Different datasets can have the same median and widely different ranges.

For values 2,4,6,8,10 the range is eight. Replace ten with one hundred and the range becomes ninety-eight, even though the first four values remain unchanged.

Explain why a large extreme value can strongly affect the full range while other summaries may move less.

30. Distinguish percentiles from percentages

A percentile rank depends on how the comparison group performed. If almost everyone scores above ninety, a mark of eighty could be below many peers despite being 80% of available marks.

Students should identify whether the question asks for a percentage of a quantity or a percentile position in a dataset.

31. Estimate a percentile value from the curve

To find an approximate pth percentile on a cumulative graph, identify a position near p/100 of total N under the stated convention, then read the corresponding x-value.

For N=80, the 75th percentile uses a graphical position around sixty. If the curve at that count corresponds to 26 minutes, the estimated percentile value is 26 minutes, not sixty.

This resembles reading Q3. A reader should still pay attention to graph conventions and any specific examination instructions.

32. Find an approximate percentile rank from a value

Some questions give a measured value and ask approximately what share of observations fall up to that value. Read the cumulative count at the stated horizontal point and compare with the total N.

If the curve gives cumulative count forty-eight out of eighty at a particular value, that represents 48/80=60% of the sample up to that boundary under the graph’s convention.

Avoid claiming that the person ‘scored 60%’ unless the horizontal axis actually represents examination marks.

33. Calculate quartiles from ungrouped records

For ordered values 2,4,6,8,10,12,14,16, the median is nine. Using the median-of-halves method, Q1=5 and Q3=13, giving IQR eight.

Do not assume that every software package or textbook uses identical percentile conventions for every small dataset. Follow the school and the question.

34. Understand the mean as a different measure

For fictional values 1,2,3,4,20, the mean is six and the median is three. A cumulative curve would not automatically give the mean at its midpoint.

Explain why the large final value influences the mean while the central ordered observation remains three.

35. Recognise the mode

The mode is the most frequent value or category. It may coincide with the median or mean in some datasets, but there is no universal rule that they are equal.

For data 2,2,2,3,4,8, the mode is two because it appears most often. The median is halfway between the third and fourth values, or 2.5, under standard ungrouped conventions.

Ask which measure suits a question about the most commonly chosen item versus the middle waiting time.

36. Compare datasets fairly

Two datasets can have equal medians but different spread. A single number rarely tells the whole story. For service data, consistency can matter alongside typical duration.

Imagine two clinics’ fictional waiting-time lists with the same median but one set ranging widely. Comparing IQR or range provides more context, provided the data are genuinely comparable.

This is not a claim about actual healthcare services; the example teaches how to interpret spread in everyday situations.

37. Recognise what a sample cannot prove

An increase in sample counts does not always imply an entire community’s behaviour has changed. Sampling method, time, population and data quality influence how results can be generalised.

If a fictional shop surveys twenty customers during one lunch hour, its observed median wait does not establish every customer’s wait across all operating hours.

The stronger conclusion states what the data actually supports and which additional evidence could broaden it.

38. Read an unfamiliar statistical question strategically

A pupil who sees Q3 may immediately write N×0.75 without noticing the question asks for the actual data value at that position. Trace both axes and the requested output.

Reasoning is a sequence of decisions, not a collection of isolated formulas.

39. Use calculator output sensibly

For a fictional IQR calculation, the subtraction may be simple, while locating Q1 and Q3 on the cumulative curve is the real challenge.

Show the relevant working and answer in the original measurement units. SEAB notes that essential working omissions can lose marks.

40. Build a six-week improvement routine

A student with unstable fraction and percentage skills should repair those first. The schedule is not an official prescribed timetable or a guarantee of a grade.

Strong progress appears when the learner can read an unfamiliar graph and explain a result without the tutor supplying each arrow.

41. Connect data literacy with future choices

G1 identifies the Maths subject level; admissions depend on full SEC qualifications and programme rules. One graph question should not be treated as a lifetime forecast.

The immediate educational goal is one defensible reading of an unfamiliar curve, checked against axes, units and sample size.

Six original worked examples of distributions

Example 1: frequencies become cumulative totals

To find the frequency in the third class, subtract cumulative fourteen from twenty-five, giving eleven. Reading twenty-five as that class’s frequency would count earlier classes again.

Example 2: median from a cumulative curve

The value forty is a count position, not a waiting time. The 18 minutes is read from the horizontal measurement scale. The approximation depends on the graph and the specified method.

Example 3: quartiles and IQR

Explain that the middle half of the observations spans roughly fourteen minutes under that graph. It does not mean every individual in that half waited fourteen minutes.

Example 4: a percentile position

That is a position in the distribution, not a ninety-percent mass or a statement that every parcel weighs 4.2 kg. The exact interpretation should follow the axis and boundary conventions.

Example 5: ungrouped median versus mean

Original data: 2,4,6,8,10,12,14,16. The median is (8+10)/2=9; the mean is 72/8=9 as well in this symmetric example. Under the median-of-halves quartile convention, Q1=5 and Q3=13.

Now replace the final sixteen with fifty. The mean becomes 106/8=13.25, but the median remains nine. This shows why the mean is more sensitive to a large extreme value.

Example 6: count and percentile interpretation

An invented chart reports cumulative count 48 out of 80 at a waiting time of fifteen minutes. The observed share at or below fifteen minutes is 48/80=0.6, or 60%, under the plotted convention.

Twelve checks that reveal the exact misconception

1. The frequency column is confused with running total

One class has eight observations but its cumulative total is twenty. Which quantity should describe the interval alone?

2. The curve drops as x increases

A student draws a cumulative curve that goes down between class boundaries. Why does this conflict with a fixed dataset’s running total?

3. The last x-value is treated as N

The curve ends at 40 minutes and eighty observations. Which is the sample size?

4. Q1 is reported as a count instead of a time

The learner chooses cumulative count twenty but reports ‘twenty minutes’ without reading across the curve. What step is missing?

5. IQR is given as a percentage without evidence

Both quartiles are waiting times, yet the answer is in percent. Which measurement unit should the difference have?

6. A percentile is confused with a percentage grade

A student obtains 70% on a test and claims to be at the 70th percentile. What distribution information is missing?

7. An interval boundary is counted twice

Two neighbouring classes both include a value at their common boundary. What grouping convention should be specified?

8. A histogram midpoint becomes an exact observation

A class interval contains many data values, but the learner assumes all equal the midpoint. Which raw information has been lost?

9. An outlier shifts the mean

One unusually large value raises the mean while leaving the median almost unchanged. Why is that mathematically possible?

10. Different sample sizes are compared using the same count

Two cumulative curves describe groups of 20 and 80 people. Why might comparing a count of ten produce different relative positions?

11. The graph is read as a clock-time trend

The x-axis represents individual waiting durations, but the pupil claims attendance increased over the day. What does the axis actually measure?

12. A weak graph result is treated as an identity

The student finds cumulative frequency difficult and assumes all future education routes are closed. Which exact prerequisite and institutional information matter instead?

Frequently asked cumulative-frequency questions

Is cumulative frequency included in 2027 G1 Mathematics?

Yes. The official K110 Statistics and Probability strand lists cumulative frequency diagrams.

Which Maths paper assesses it?

Paper 2 covers Statistics and Probability alongside Number and Algebra; it contributes 50% of G1 Maths.

Are quartiles and percentiles required?

Yes. K110 explicitly lists percentiles, quartiles, range and interquartile range.

Are histograms included?

Yes, specifically histograms with equal class intervals.

How do I find the median on a cumulative curve?

Use a cumulative position around half the total N, then read the corresponding data-axis value according to the question’s convention.

How do I calculate IQR?

Subtract Q1 from Q3, retaining the data’s measurement unit.

What is the difference between frequency and cumulative frequency?

Frequency counts one category or interval; cumulative frequency totals observations up to a given value or boundary.

Does the 75th percentile mean a mark of 75%?

No. Percentile describes a position in a distribution; a percentage mark is a portion of attainable marks.

Can I obtain exact raw observations from a grouped histogram?

Not generally. Grouping loses individual detail, so avoid claiming unsupported precision.

Will every book or software use identical quartile conventions?

Not always. Follow the method and conventions specified by your question and school.

Can an approved calculator be used in Paper 2?

Yes, but reading axes, showing essential working and interpreting results remain necessary.

Does G1 Maths permanently fix future pathways?

No. Further-study decisions depend on the learner’s complete results, current programme prerequisites and school guidance.

Official syllabus and related eduKateSG guides

To survive G1 Mathematics questions on cumulative frequency, quartiles and percentiles, first read what the data represents and how many observations were recorded. In Singapore’s 2027 SEC G1 Mathematics K110, Statistics and Probability is assessed in Paper 2 alongside Number and Algebra. Paper 2 lasts 90 minutes, is worth 50 marks and makes up 50% of the subject. Students need to understand tables, bar charts, pictograms, line graphs, pie charts, dot diagrams, histograms with equal class intervals and cumulative frequency diagrams. They must interpret centre, spread and relative positions instead of guessing from a colourful curve.

The quickest way to make cumulative frequency understandable is to distinguish an interval’s count from the total accumulated so far. If twelve observations have been recorded up to one boundary and twenty up to the next, eight lie between those cumulative cutoffs under the stated grouping convention. For a dataset of eighty observations, the halfway position is near the fortieth observation; the lower and upper quartile positions correspond approximately to the twentieth and sixtieth. The exact data values must be read from the plotted horizontal scale or ordered records. A student who looks only at the vertical axis will miss the question.

The official 2027 SEAB G1 Mathematics K110 syllabus includes cumulative frequency, quartiles, percentiles, mean, mode, median, range and interquartile range, plus the purposes and limitations of statistical representations. SEAB’s 2027 G1 subject directory confirms K110 as the correct exam code. This is an independent learning article, not an official specimen question. For details of interpolation or choosing quartile positions in a borderline dataset, follow the method specified by the question and the school’s current teaching conventions rather than assume every statistical definition uses the same algorithm.

Data literacy is useful well beyond the Mathematics paper. It helps people interpret waiting times, travel patterns, service data, workplace performance claims and consumer surveys. These abilities may support later technical, business, service and other learning paths without guaranteeing any particular course. G1 describes a subject level in the SEC, not the entire student’s capacity. Families should consider a learner’s complete qualifications and current institutional entry rules when future choices become relevant.

The five-minute survival routine

Read chart title, axis labels, measurement unit and total observations before doing any calculation.
For cumulative frequency, locate the count on the vertical axis and read the corresponding data value from the horizontal axis.
For an interval’s frequency, subtract the cumulative totals at its two endpoints when the conventions support that.
Keep centre (mean/median/mode), total spread (range) and middle-half spread (IQR) conceptually separate.
For percentile positions, use the data total and the method indicated by the diagram or question.
Explain what the resulting number says about the given sample, not automatically all of Singapore.

The official 2027 K110 assessment context

Part of K110	Official content	What to remember
Paper 2	90 minutes, 50 marks, 50% of Maths; Number and Algebra with Statistics and Probability	All questions compulsory
Data representations	Tables, bar graphs, pictograms, line graphs, pie charts, dot diagrams, equal-width histograms and cumulative frequency diagrams	Each chart answers different questions
Data summaries	Mean, mode, median of ungrouped data; percentiles, quartiles, range, interquartile range	Calculate and interpret, not just name
Probability	Simple single-event situations, including listing possible outcomes	A separate Paper 2 focus
Paper 1	90 minutes, 50 marks, 50%; Number and Algebra with Geometry and Measurement	Do not substitute the Paper 2 statistics strand
Presentation	Approved calculator allowed; essential working required; 3 significant figures for non-exact results unless directed otherwise	Interpret chart units and real context

Understand cumulative frequency in three moves

Navigate by difficulty

Read data and charts · Cumulative frequency · Quartiles and percentiles · Original worked datasets · Frequently asked questions

1. Understand what one observation is

A table showing thirty journeys does not necessarily represent thirty unique commuters; the same traveller might contribute several trips. Use the context provided, not a guess about identities.

Ask the pupil to identify the unit, sample and number of observations before adding or comparing counts. This single step prevents misleading statements later.

2. Separate a population from a sample

If a classroom measures twenty students’ journey durations, the resulting median belongs to those twenty records. Calling it Singapore’s national median would be unsupported.

Use the scope of the question in the final sentence. Honest statistical interpretation is more valuable than an exaggerated conclusion that the graph cannot prove.

3. Identify numerical and categorical variables

If a survey lists routes 7, 10 and 15, the numbers are labels, not necessarily distances. By contrast, journey durations on those routes could be averaged when the problem and data support it.

Decide the variable’s nature before choosing a representation. A pie chart might show the distribution of route categories, whereas a cumulative curve needs ordered numerical values.

4. Read graph axes and scales exactly

Imagine a cumulative graph of package mass measured in kilograms. A height of 50 refers to fifty observations accumulated up to a mass boundary, not fifty kilograms.

Tell the pupil to point to each axis label and state its unit aloud. Then ask what one plotted point actually represents before calculating.

5. Use frequency tables correctly

For a fictional table with interval frequencies 3, 7 and 10, the total is twenty. A student who adds only the first and last has omitted the middle group.

Convert the numbers into one cumulative column: 3, 10, 20. The progression helps explain how accumulation differs from a stand-alone interval count.

6. Read a pictogram key before counting

If a school chart assigns one symbol to five books, three symbols represent fifteen books. The symbol count and book count differ.

Change the key to four books and ask how the total changes. Understanding the representation is more useful than simply recognising the book icon.

7. Interpret bar charts with appropriate comparisons

A bar chart compares category frequencies or values. The numerical scale controls the meaning of bar height; a truncated vertical axis can exaggerate small differences.

Suppose two fictional clubs have attendance 94 and 98. The difference is four, even if a graph beginning at ninety makes one bar look much taller.

A valid interpretation names the categories, values, difference and relevant sample context rather than merely reporting that one bar ‘looks huge’.

8. Understand line graphs as recorded patterns

A line graph shows values across ordered observations, often but not always time. A rise may describe an increasing quantity within the plotted range. It does not prove why the pattern occurred.

For hypothetical monthly visitors, values 50, 60 and 58 show first an increase and then a decrease. A claim that advertising caused the change requires other evidence.

Read the axes, find the values and describe the observed sequence before offering a qualified interpretation.

9. Translate pie-chart proportions into counts

Pie-chart sectors represent shares of a whole, with 360° for the complete circle. A quarter sector is 90° regardless of how many observations the full chart contains.

For a fictional group of forty, a 90° sector represents ten people. For a group of eighty with the same proportion, it represents twenty.

This is why students must read the total before calculating a count. The angle alone identifies the proportion, not the absolute number.

10. Count repeated observations in dot diagrams

A dot diagram uses marks for individual observations, including repeats. A learner who records one value per labelled position misses the actual number of data points.

If a dot diagram shows values 2, 2, 3, 3, 3 and 5, there are six observations, not three. Their mean is eighteen divided by six, which is three.

Ask the student to write the full ordered list before finding a median or mode. This simple transcription supports reliable statistical summaries.

11. Understand an equal-width histogram

K110 specifically includes histograms with equal class intervals. Grouped numerical data are displayed in touching bars, with the frequency attached to the relevant class under the given convention.

Do not claim every person in the 5–10 interval took precisely 7.5 minutes. A midpoint can be useful in a stated approximation, but it is not a recovered raw observation.

12. Recognise what cumulative means

Consider frequencies 4, 6, 5 and 5. The cumulative totals are 4, 10, 15 and 20. Each total tells how many have been included through that point.

A non-decreasing cumulative graph makes sense for a properly ordered dataset. If the recorded cumulative total falls, review the construction or scale.

13. Construct a cumulative table from intervals

Begin with each interval frequency and add it to the preceding running total. State the class boundaries correctly. The total in the last row should equal the sum of all original frequency values.

For fictional interval counts 3, 7, 8 and 2, write running totals 3, 10, 18 and 20. Check 3+7+8+2=20.

Now change the third frequency from eight to nine. The final cumulative total should become twenty-one, and the later cumulative values must update. This tests actual understanding of accumulation.

14. Get one class frequency from two cumulative totals

If cumulative frequency rises from twelve to twenty-five between two class boundaries, the interval contains thirteen additional observations.

Ask what the earlier twelve represents. It is not discarded; it is simply removed from the later running total when the question asks about the intervening range.

15. Recognise why a cumulative graph never needs negative frequencies

A graph may have flat segments where no observations lie in a range, but it should not fall as the boundary increases. The shape depends on the distribution, not on a sales-like trend over time.

Ask the learner to explain this property verbally before drawing points. It is a powerful reasonableness check.

16. Understand upper class boundaries

For a fictional interval 10–20 followed by 20–30, the boundary treatment determines where an exact value of twenty belongs. A consistent method ensures it is counted once.

In a class test, read the published table headings carefully. The diagram and question’s conventions govern the plotted positions.

17. Plot cumulative points with the correct coordinates

If the cumulative total is sixteen up to a boundary of thirty minutes, one point is (30 minutes, 16 observations), assuming the stated class construction.

Have the pupil label every coordinate in words, then check that the last plotted cumulative value equals the total sample size.

18. Recognise the total N from the curve

If the curve ends at eighty on the cumulative-frequency axis and at thirty-five minutes on the value axis, the dataset contains eighty observations, not thirty-five.

Ask the learner what each number means. Unit recognition prevents a foundational mistake.

19. Find a value below a given cumulative count

A question may ask for the value below which a stated number of observations occur. Locate the count on the vertical axis, move to the curve and then read the corresponding horizontal value.

If a fictional curve shows cumulative count fifty at a waiting time of twelve minutes, that point describes approximately fifty observations up to the stated time boundary under its convention.

It does not automatically mean the mean wait is twelve minutes. The graph provides a positional relationship, not every average.

20. Find the count below a given value

The reverse question may specify a horizontal value and ask how many observations are up to that value. Start at the value, trace to the curve and read the cumulative count on the vertical axis.

For a fictional plot, if the curve is at forty observations when x=20 minutes, then roughly forty observations have durations at or below the relevant boundary.

Follow the actual line and scale. Do not guess the answer from the nearest labelled point if interpolation is needed and permitted.

21. Understand interpolation as an estimate

If the curve crosses the height fifty between fifteen and twenty minutes, the horizontal reading will be within that range, with precision limited by the graph.

Students should show how the reading was obtained and use the accuracy specified by the exam instead of inventing decimal places.

22. Recognise information that grouped curves cannot reveal

A cumulative curve summarises a dataset, but exact individual measurements may not be recoverable. A pupil should not claim to know that every participant in an interval had the same value.

A curve may show that twenty observations lie between ten and twenty minutes without indicating each person’s exact time.

This limitation matters if a question asks for an estimate versus a precise raw reading. Use the wording and available data honestly.

23. Compare two cumulative curves

For two groups of different sizes, comparing at cumulative count twenty may correspond to different percentiles. A proportion-based comparison could be more meaningful.

State the question’s target and verify that axes, units and sample definitions are compatible.

24. Distinguish the curve from an ordinary time-series graph

The graph’s rise means increasing numbers of observations included by progressively larger data values; it does not mean more people necessarily arrived during a particular minute.

Identifying what the horizontal measurement represents is essential before writing an interpretation.

25. Find the median position

For N=80, locate cumulative frequency forty, trace to the curve and read the data axis. That number is a waiting time or other measured value, not ‘forty’ by default.

The exact convention for a discrete, small, ungrouped dataset can differ from graphical estimation. Follow the method specified in the problem.

26. Find the first quartile position

The first quartile Q1 represents a lower-quarter position in the ordered distribution. A usual cumulative-frequency graphical estimate uses a count around one quarter of N.

For N=80, one quarter is twenty; identify the curve at cumulative count twenty and read its corresponding horizontal measurement.

If the curve’s scale is in minutes, the Q1 answer is in minutes. The mark position is a count; the quartile value is a data quantity.

27. Find the third quartile position

The third quartile Q3 corresponds to about three quarters of the ordered observations. It helps describe the upper boundary of the middle half of the dataset.

For N=80, three quarters is sixty. Trace the curve from cumulative frequency sixty to the horizontal axis and report the estimated value with suitable graph-reading accuracy.

Do not call the value ‘60%’ unless the horizontal measurement itself is a percentage. The sixty here is a position count.

28. Understand interquartile range

IQR is the difference Q3−Q1. It describes the spread of the middle half of an ordered distribution and is often less affected by extreme values than the full range.

If graphical readings give Q1=12 minutes and Q3=26 minutes, IQR=14 minutes. The unit is minutes, and the meaning concerns central spread rather than a duration actually observed for one person.

Use an appropriate level of precision for the graph readings; a calculated IQR cannot be more exact than its underlying estimates.

29. Understand the full range

Range is maximum minus minimum. It considers the extremes, while IQR focuses on the middle half. Different datasets can have the same median and widely different ranges.

For values 2,4,6,8,10 the range is eight. Replace ten with one hundred and the range becomes ninety-eight, even though the first four values remain unchanged.

Explain why a large extreme value can strongly affect the full range while other summaries may move less.

30. Distinguish percentiles from percentages

A percentile rank depends on how the comparison group performed. If almost everyone scores above ninety, a mark of eighty could be below many peers despite being 80% of available marks.

Students should identify whether the question asks for a percentage of a quantity or a percentile position in a dataset.

31. Estimate a percentile value from the curve

To find an approximate pth percentile on a cumulative graph, identify a position near p/100 of total N under the stated convention, then read the corresponding x-value.

For N=80, the 75th percentile uses a graphical position around sixty. If the curve at that count corresponds to 26 minutes, the estimated percentile value is 26 minutes, not sixty.

This resembles reading Q3. A reader should still pay attention to graph conventions and any specific examination instructions.

32. Find an approximate percentile rank from a value

Some questions give a measured value and ask approximately what share of observations fall up to that value. Read the cumulative count at the stated horizontal point and compare with the total N.

If the curve gives cumulative count forty-eight out of eighty at a particular value, that represents 48/80=60% of the sample up to that boundary under the graph’s convention.

Avoid claiming that the person ‘scored 60%’ unless the horizontal axis actually represents examination marks.

33. Calculate quartiles from ungrouped records

For ordered values 2,4,6,8,10,12,14,16, the median is nine. Using the median-of-halves method, Q1=5 and Q3=13, giving IQR eight.

Do not assume that every software package or textbook uses identical percentile conventions for every small dataset. Follow the school and the question.

34. Understand the mean as a different measure

For fictional values 1,2,3,4,20, the mean is six and the median is three. A cumulative curve would not automatically give the mean at its midpoint.

Explain why the large final value influences the mean while the central ordered observation remains three.

35. Recognise the mode

The mode is the most frequent value or category. It may coincide with the median or mean in some datasets, but there is no universal rule that they are equal.

For data 2,2,2,3,4,8, the mode is two because it appears most often. The median is halfway between the third and fourth values, or 2.5, under standard ungrouped conventions.

Ask which measure suits a question about the most commonly chosen item versus the middle waiting time.

36. Compare datasets fairly

Two datasets can have equal medians but different spread. A single number rarely tells the whole story. For service data, consistency can matter alongside typical duration.

Imagine two clinics’ fictional waiting-time lists with the same median but one set ranging widely. Comparing IQR or range provides more context, provided the data are genuinely comparable.

This is not a claim about actual healthcare services; the example teaches how to interpret spread in everyday situations.

37. Recognise what a sample cannot prove

An increase in sample counts does not always imply an entire community’s behaviour has changed. Sampling method, time, population and data quality influence how results can be generalised.

If a fictional shop surveys twenty customers during one lunch hour, its observed median wait does not establish every customer’s wait across all operating hours.

The stronger conclusion states what the data actually supports and which additional evidence could broaden it.

38. Read an unfamiliar statistical question strategically

A pupil who sees Q3 may immediately write N×0.75 without noticing the question asks for the actual data value at that position. Trace both axes and the requested output.

Reasoning is a sequence of decisions, not a collection of isolated formulas.

39. Use calculator output sensibly

For a fictional IQR calculation, the subtraction may be simple, while locating Q1 and Q3 on the cumulative curve is the real challenge.

Show the relevant working and answer in the original measurement units. SEAB notes that essential working omissions can lose marks.

40. Build a six-week improvement routine

A student with unstable fraction and percentage skills should repair those first. The schedule is not an official prescribed timetable or a guarantee of a grade.

Strong progress appears when the learner can read an unfamiliar graph and explain a result without the tutor supplying each arrow.

41. Connect data literacy with future choices

G1 identifies the Maths subject level; admissions depend on full SEC qualifications and programme rules. One graph question should not be treated as a lifetime forecast.

The immediate educational goal is one defensible reading of an unfamiliar curve, checked against axes, units and sample size.

Six original worked examples of distributions

Example 1: frequencies become cumulative totals

To find the frequency in the third class, subtract cumulative fourteen from twenty-five, giving eleven. Reading twenty-five as that class’s frequency would count earlier classes again.

Example 2: median from a cumulative curve

The value forty is a count position, not a waiting time. The 18 minutes is read from the horizontal measurement scale. The approximation depends on the graph and the specified method.

Example 3: quartiles and IQR

Explain that the middle half of the observations spans roughly fourteen minutes under that graph. It does not mean every individual in that half waited fourteen minutes.

Example 4: a percentile position

That is a position in the distribution, not a ninety-percent mass or a statement that every parcel weighs 4.2 kg. The exact interpretation should follow the axis and boundary conventions.

Example 5: ungrouped median versus mean

Original data: 2,4,6,8,10,12,14,16. The median is (8+10)/2=9; the mean is 72/8=9 as well in this symmetric example. Under the median-of-halves quartile convention, Q1=5 and Q3=13.

Now replace the final sixteen with fifty. The mean becomes 106/8=13.25, but the median remains nine. This shows why the mean is more sensitive to a large extreme value.

Example 6: count and percentile interpretation

An invented chart reports cumulative count 48 out of 80 at a waiting time of fifteen minutes. The observed share at or below fifteen minutes is 48/80=0.6, or 60%, under the plotted convention.

Twelve checks that reveal the exact misconception

1. The frequency column is confused with running total

One class has eight observations but its cumulative total is twenty. Which quantity should describe the interval alone?

2. The curve drops as x increases

A student draws a cumulative curve that goes down between class boundaries. Why does this conflict with a fixed dataset’s running total?

3. The last x-value is treated as N

The curve ends at 40 minutes and eighty observations. Which is the sample size?

4. Q1 is reported as a count instead of a time

The learner chooses cumulative count twenty but reports ‘twenty minutes’ without reading across the curve. What step is missing?

5. IQR is given as a percentage without evidence

Both quartiles are waiting times, yet the answer is in percent. Which measurement unit should the difference have?

6. A percentile is confused with a percentage grade

A student obtains 70% on a test and claims to be at the 70th percentile. What distribution information is missing?

7. An interval boundary is counted twice

Two neighbouring classes both include a value at their common boundary. What grouping convention should be specified?

8. A histogram midpoint becomes an exact observation

A class interval contains many data values, but the learner assumes all equal the midpoint. Which raw information has been lost?

9. An outlier shifts the mean

One unusually large value raises the mean while leaving the median almost unchanged. Why is that mathematically possible?

10. Different sample sizes are compared using the same count

Two cumulative curves describe groups of 20 and 80 people. Why might comparing a count of ten produce different relative positions?

11. The graph is read as a clock-time trend

The x-axis represents individual waiting durations, but the pupil claims attendance increased over the day. What does the axis actually measure?

12. A weak graph result is treated as an identity

The student finds cumulative frequency difficult and assumes all future education routes are closed. Which exact prerequisite and institutional information matter instead?

Frequently asked cumulative-frequency questions

Is cumulative frequency included in 2027 G1 Mathematics?

Yes. The official K110 Statistics and Probability strand lists cumulative frequency diagrams.

Which Maths paper assesses it?

Paper 2 covers Statistics and Probability alongside Number and Algebra; it contributes 50% of G1 Maths.

Are quartiles and percentiles required?

Yes. K110 explicitly lists percentiles, quartiles, range and interquartile range.

Are histograms included?

Yes, specifically histograms with equal class intervals.

How do I find the median on a cumulative curve?

Use a cumulative position around half the total N, then read the corresponding data-axis value according to the question’s convention.

How do I calculate IQR?

Subtract Q1 from Q3, retaining the data’s measurement unit.

What is the difference between frequency and cumulative frequency?

Frequency counts one category or interval; cumulative frequency totals observations up to a given value or boundary.

Does the 75th percentile mean a mark of 75%?

No. Percentile describes a position in a distribution; a percentage mark is a portion of attainable marks.

Can I obtain exact raw observations from a grouped histogram?

Not generally. Grouping loses individual detail, so avoid claiming unsupported precision.

Will every book or software use identical quartile conventions?

Not always. Follow the method and conventions specified by your question and school.

Can an approved calculator be used in Paper 2?

Yes, but reading axes, showing essential working and interpreting results remain necessary.

Does G1 Maths permanently fix future pathways?

No. Further-study decisions depend on the learner’s complete results, current programme prerequisites and school guidance.

How to survive G1 Mathematics? | Cumulative Frequency, Quartiles, Percentiles and Charts

The official 2027 K110 assessment context

Understand cumulative frequency in three moves

Navigate by difficulty

1. Understand what one observation is

2. Separate a population from a sample

3. Identify numerical and categorical variables

4. Read graph axes and scales exactly

5. Use frequency tables correctly

6. Read a pictogram key before counting

7. Interpret bar charts with appropriate comparisons

8. Understand line graphs as recorded patterns

9. Translate pie-chart proportions into counts

10. Count repeated observations in dot diagrams

11. Understand an equal-width histogram

12. Recognise what cumulative means

13. Construct a cumulative table from intervals

14. Get one class frequency from two cumulative totals

15. Recognise why a cumulative graph never needs negative frequencies

16. Understand upper class boundaries

17. Plot cumulative points with the correct coordinates

18. Recognise the total N from the curve

19. Find a value below a given cumulative count

20. Find the count below a given value

21. Understand interpolation as an estimate

22. Recognise information that grouped curves cannot reveal

23. Compare two cumulative curves

24. Distinguish the curve from an ordinary time-series graph

25. Find the median position

26. Find the first quartile position

27. Find the third quartile position

28. Understand interquartile range

29. Understand the full range

30. Distinguish percentiles from percentages

31. Estimate a percentile value from the curve

32. Find an approximate percentile rank from a value

33. Calculate quartiles from ungrouped records

34. Understand the mean as a different measure

35. Recognise the mode

36. Compare datasets fairly

37. Recognise what a sample cannot prove

38. Read an unfamiliar statistical question strategically

39. Use calculator output sensibly

40. Build a six-week improvement routine

41. Connect data literacy with future choices

Six original worked examples of distributions

Example 1: frequencies become cumulative totals

Example 2: median from a cumulative curve

Example 3: quartiles and IQR

Example 4: a percentile position

Example 5: ungrouped median versus mean

Example 6: count and percentile interpretation

Twelve checks that reveal the exact misconception

1. The frequency column is confused with running total

2. The curve drops as x increases

3. The last x-value is treated as N

4. Q1 is reported as a count instead of a time

5. IQR is given as a percentage without evidence

6. A percentile is confused with a percentage grade

7. An interval boundary is counted twice

8. A histogram midpoint becomes an exact observation

9. An outlier shifts the mean

10. Different sample sizes are compared using the same count

11. The graph is read as a clock-time trend

12. A weak graph result is treated as an identity

Frequently asked cumulative-frequency questions

Is cumulative frequency included in 2027 G1 Mathematics?

Which Maths paper assesses it?

Are quartiles and percentiles required?

Are histograms included?

How do I find the median on a cumulative curve?

How do I calculate IQR?

What is the difference between frequency and cumulative frequency?

Does the 75th percentile mean a mark of 75%?

Can I obtain exact raw observations from a grouped histogram?

Will every book or software use identical quartile conventions?

Can an approved calculator be used in Paper 2?

Does G1 Maths permanently fix future pathways?

Official syllabus and related eduKateSG guides

The official 2027 K110 assessment context