Sec 2 G1 Maths Tutor Bukit Timah

Reader note: Every numerical example, chart description, sample and survey in this guide is fictional. No values are presented as real Bukit Timah demographics, authentic student performance data or current business statistics. This is educational reasoning, not an official examination paper or a substitute for checking original data sources. Confirm current tutor groups, prices, venue and learning arrangements directly.

If a school worksheet contains a confusing graph, bring the original printed axes and the pupil’s interpretation to an eduKate Singapore Mathematics enquiry. Ask the tutor to show exactly which numerical claim can be supported and how it would be checked independently.

Official sources and accuracy note

First 10 minutes: retrieve fractions, percentages and one straightforward axis-reading item.
Next 20 minutes: compare accurate and misleading displays of the same fictional values.
Next 25 minutes: individually diagnose a scale, denominator or summary-statistic error.
Next 15 minutes: translate a numerical table into a graph and discuss what information grouping removes.
Next 10 minutes: write accurate interpretations of unfamiliar data without help.
Final 10 minutes: correct one overstated conclusion and plan a later independent retest.

G1 subject level and published SEC syllabus

MOE’s G1 Mathematics content by level includes Data Analysis in Secondary Two, with dot diagrams, histograms with equal class intervals and summaries such as mean, median and mode. SEAB lists the 2027 G1 Mathematics K110 syllabus under the new SEC, referencing the earlier 4046 code. These are curriculum references, not a promise that all the broader data-literacy extensions above will appear in a given 2026 Secondary 2 school test. Follow the school’s actual topic scope and the pupil’s eventual examination year.

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Yes. An axis that begins far above zero can emphasise small differences, and unclear units can encourage mistaken comparisons. Inspect the numerical labels and context before interpreting appearance.

Is median always better than mean?

No. They answer different questions. The mean uses every value, while the median describes the central position. Select the summary appropriate to the dataset and purpose.

What is a modal class?

It is the interval with the largest frequency in a grouped display. It generally does not tell you an exact most-common individual value.

Can a graph prove that one thing caused another?

A graph alone generally cannot establish causation. The way data were gathered, possible other influences and relevant evidence must be considered.

Are all 37 examples required for Secondary 2 G1?

No. The article includes G1-aligned graph reading and broader critical interpretation. The school’s current curriculum and actual question instructions take priority.

How will a tutor know my child has improved?

Have the learner read a different graph, calculate with the right denominator and rewrite a new claim accurately without copying a model explanation. Retest after a delay.

Official sources and accuracy note

First 10 minutes: retrieve fractions, percentages and one straightforward axis-reading item.
Next 20 minutes: compare accurate and misleading displays of the same fictional values.
Next 25 minutes: individually diagnose a scale, denominator or summary-statistic error.
Next 15 minutes: translate a numerical table into a graph and discuss what information grouping removes.
Next 10 minutes: write accurate interpretations of unfamiliar data without help.
Final 10 minutes: correct one overstated conclusion and plan a later independent retest.

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Yes. An axis that begins far above zero can emphasise small differences, and unclear units can encourage mistaken comparisons. Inspect the numerical labels and context before interpreting appearance.

Is median always better than mean?

No. They answer different questions. The mean uses every value, while the median describes the central position. Select the summary appropriate to the dataset and purpose.

What is a modal class?

It is the interval with the largest frequency in a grouped display. It generally does not tell you an exact most-common individual value.

Can a graph prove that one thing caused another?

A graph alone generally cannot establish causation. The way data were gathered, possible other influences and relevant evidence must be considered.

Are all 37 examples required for Secondary 2 G1?

No. The article includes G1-aligned graph reading and broader critical interpretation. The school’s current curriculum and actual question instructions take priority.

How will a tutor know my child has improved?

Have the learner read a different graph, calculate with the right denominator and rewrite a new claim accurately without copying a model explanation. Retest after a delay.

Official sources and accuracy note

Day 1: read axes, units and tick spacing from two graphs in the child’s current textbook.
Day 2: compute a percentage from a fictional survey count with a clearly stated sample size.
Day 3: compare mean and median in a five-value dataset containing one unusually large value.
Day 4: make a dot plot or equal-width frequency histogram and identify what exact information each display preserves.
Day 5: rewrite one overconfident chart claim in a precise sentence that names the measured group and limitation.

Use school-issued charts or reliable published examples with their context retained. Avoid asking the child to make up real statistics about local people. The productive parent question is ‘Where does that claim appear in the numbers?’ If the pupil can point to the relevant count, axis and reference group, their explanation is becoming defensible.

How a small-group G1 Mathematics lesson can teach this

The immutable eduKateSG Clementi small-group Mathematics programme describes 90-minute tutorials for three pupils near Sixth Avenue MRT. In an illustrative data-literacy lesson, one pupil may practise reading a graph scale, another may compare group percentages and a third may evaluate an overconfident inference. The tutor can ask each student to support a conclusion with the numbers and revise it after feedback. This does not confirm any current Secondary 2 G1 class opening.

First 10 minutes: retrieve fractions, percentages and one straightforward axis-reading item.
Next 20 minutes: compare accurate and misleading displays of the same fictional values.
Next 25 minutes: individually diagnose a scale, denominator or summary-statistic error.
Next 15 minutes: translate a numerical table into a graph and discuss what information grouping removes.
Next 10 minutes: write accurate interpretations of unfamiliar data without help.
Final 10 minutes: correct one overstated conclusion and plan a later independent retest.

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Yes. An axis that begins far above zero can emphasise small differences, and unclear units can encourage mistaken comparisons. Inspect the numerical labels and context before interpreting appearance.

Is median always better than mean?

No. They answer different questions. The mean uses every value, while the median describes the central position. Select the summary appropriate to the dataset and purpose.

What is a modal class?

It is the interval with the largest frequency in a grouped display. It generally does not tell you an exact most-common individual value.

Can a graph prove that one thing caused another?

A graph alone generally cannot establish causation. The way data were gathered, possible other influences and relevant evidence must be considered.

Are all 37 examples required for Secondary 2 G1?

No. The article includes G1-aligned graph reading and broader critical interpretation. The school’s current curriculum and actual question instructions take priority.

How will a tutor know my child has improved?

Have the learner read a different graph, calculate with the right denominator and rewrite a new claim accurately without copying a model explanation. Retest after a delay.

Official sources and accuracy note

Day 1: read axes, units and tick spacing from two graphs in the child’s current textbook.
Day 2: compute a percentage from a fictional survey count with a clearly stated sample size.
Day 3: compare mean and median in a five-value dataset containing one unusually large value.
Day 4: make a dot plot or equal-width frequency histogram and identify what exact information each display preserves.
Day 5: rewrite one overconfident chart claim in a precise sentence that names the measured group and limitation.

How a small-group G1 Mathematics lesson can teach this

First 10 minutes: retrieve fractions, percentages and one straightforward axis-reading item.
Next 20 minutes: compare accurate and misleading displays of the same fictional values.
Next 25 minutes: individually diagnose a scale, denominator or summary-statistic error.
Next 15 minutes: translate a numerical table into a graph and discuss what information grouping removes.
Next 10 minutes: write accurate interpretations of unfamiliar data without help.
Final 10 minutes: correct one overstated conclusion and plan a later independent retest.

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Yes. An axis that begins far above zero can emphasise small differences, and unclear units can encourage mistaken comparisons. Inspect the numerical labels and context before interpreting appearance.

Is median always better than mean?

No. They answer different questions. The mean uses every value, while the median describes the central position. Select the summary appropriate to the dataset and purpose.

What is a modal class?

It is the interval with the largest frequency in a grouped display. It generally does not tell you an exact most-common individual value.

Can a graph prove that one thing caused another?

A graph alone generally cannot establish causation. The way data were gathered, possible other influences and relevant evidence must be considered.

Are all 37 examples required for Secondary 2 G1?

No. The article includes G1-aligned graph reading and broader critical interpretation. The school’s current curriculum and actual question instructions take priority.

How will a tutor know my child has improved?

Have the learner read a different graph, calculate with the right denominator and rewrite a new claim accurately without copying a model explanation. Retest after a delay.

Official sources and accuracy note

Numerical check: calculate the difference, ratio or percentage with its correct denominator.
Source check: state the sample size, time window and categories actually shown.
Representation check: name whether the figure shows an individual value, an interval, a count, a rate or a summary.
Claim check: distinguish ‘the data show’ from ‘the model suggests’ or ‘we would need more evidence’.

This is not about distrusting every graph. A clear and accurate visual can be extremely helpful. The purpose is to give the learner specific reasons for accepting a conclusion, checking a possible problem, or asking for one missing piece of information. The best answer often identifies both the valid insight and the limitation.

A five-day family data-literacy practice plan

Day 1: read axes, units and tick spacing from two graphs in the child’s current textbook.
Day 2: compute a percentage from a fictional survey count with a clearly stated sample size.
Day 3: compare mean and median in a five-value dataset containing one unusually large value.
Day 4: make a dot plot or equal-width frequency histogram and identify what exact information each display preserves.
Day 5: rewrite one overconfident chart claim in a precise sentence that names the measured group and limitation.

How a small-group G1 Mathematics lesson can teach this

First 10 minutes: retrieve fractions, percentages and one straightforward axis-reading item.
Next 20 minutes: compare accurate and misleading displays of the same fictional values.
Next 25 minutes: individually diagnose a scale, denominator or summary-statistic error.
Next 15 minutes: translate a numerical table into a graph and discuss what information grouping removes.
Next 10 minutes: write accurate interpretations of unfamiliar data without help.
Final 10 minutes: correct one overstated conclusion and plan a later independent retest.

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Yes. An axis that begins far above zero can emphasise small differences, and unclear units can encourage mistaken comparisons. Inspect the numerical labels and context before interpreting appearance.

Is median always better than mean?

No. They answer different questions. The mean uses every value, while the median describes the central position. Select the summary appropriate to the dataset and purpose.

What is a modal class?

It is the interval with the largest frequency in a grouped display. It generally does not tell you an exact most-common individual value.

Can a graph prove that one thing caused another?

A graph alone generally cannot establish causation. The way data were gathered, possible other influences and relevant evidence must be considered.

Are all 37 examples required for Secondary 2 G1?

No. The article includes G1-aligned graph reading and broader critical interpretation. The school’s current curriculum and actual question instructions take priority.

How will a tutor know my child has improved?

Have the learner read a different graph, calculate with the right denominator and rewrite a new claim accurately without copying a model explanation. Retest after a delay.

Official sources and accuracy note

Numerical check: calculate the difference, ratio or percentage with its correct denominator.
Source check: state the sample size, time window and categories actually shown.
Representation check: name whether the figure shows an individual value, an interval, a count, a rate or a summary.
Claim check: distinguish ‘the data show’ from ‘the model suggests’ or ‘we would need more evidence’.

A five-day family data-literacy practice plan

Day 1: read axes, units and tick spacing from two graphs in the child’s current textbook.
Day 2: compute a percentage from a fictional survey count with a clearly stated sample size.
Day 3: compare mean and median in a five-value dataset containing one unusually large value.
Day 4: make a dot plot or equal-width frequency histogram and identify what exact information each display preserves.
Day 5: rewrite one overconfident chart claim in a precise sentence that names the measured group and limitation.

How a small-group G1 Mathematics lesson can teach this

First 10 minutes: retrieve fractions, percentages and one straightforward axis-reading item.
Next 20 minutes: compare accurate and misleading displays of the same fictional values.
Next 25 minutes: individually diagnose a scale, denominator or summary-statistic error.
Next 15 minutes: translate a numerical table into a graph and discuss what information grouping removes.
Next 10 minutes: write accurate interpretations of unfamiliar data without help.
Final 10 minutes: correct one overstated conclusion and plan a later independent retest.

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Yes. An axis that begins far above zero can emphasise small differences, and unclear units can encourage mistaken comparisons. Inspect the numerical labels and context before interpreting appearance.

Is median always better than mean?

No. They answer different questions. The mean uses every value, while the median describes the central position. Select the summary appropriate to the dataset and purpose.

What is a modal class?

It is the interval with the largest frequency in a grouped display. It generally does not tell you an exact most-common individual value.

Can a graph prove that one thing caused another?

A graph alone generally cannot establish causation. The way data were gathered, possible other influences and relevant evidence must be considered.

Are all 37 examples required for Secondary 2 G1?

No. The article includes G1-aligned graph reading and broader critical interpretation. The school’s current curriculum and actual question instructions take priority.

How will a tutor know my child has improved?

Have the learner read a different graph, calculate with the right denominator and rewrite a new claim accurately without copying a model explanation. Retest after a delay.

Official sources and accuracy note

One-minute diagnostic
The five-question method
37 original misleading-data examples
How a pupil can rewrite a misleading conclusion
A practical family exercise
Three-pupil data lesson
Frequently asked questions
Official sources

A one-minute data-reading diagnostic

Show two hypothetical bars labelled 92 and 96, drawn with a vertical axis that starts at 90. Ask: ‘Is 96 three times 92?’ The answer is no; 96 is only four higher, around 4.35% higher relative to 92. Next show a pie-sector angle of 90°. The sector is 25%, not 90%. Finally tell the pupil ‘six of ten respondents chose A’ and ask whether that establishes the preference of all residents. It does not. These three tasks test scale, proportional interpretation and the limits of a sample.

A tutor should ask the learner to point to the evidence for each answer. Can the child read the actual axis numbers instead of judging height? Can the child identify the whole circle before dividing? Can the child name the group actually surveyed? If the answer is uncertain, focus teaching on that one inference rather than distributing a long statistics worksheet.

How to move from a picture to a defensible claim

Start by reading the axes and legend

The y-axis may start at zero or may be truncated. The x-axis may show time, a measurement or a category. Some charts use a second axis for a different quantity. A pupil should say what one tick represents and the units on both axes before taking any difference or gradient. This habit catches many calculations that appear correct but answer a different question.

Ask which whole is being described

A rate of 20% may refer to twenty of one hundred observations, four of twenty or something else. When comparing groups of different sizes, raw counts can point in the opposite direction from their relative shares. For example, fifteen of fifty is 30%, less than ten of twenty at 50%, despite fifteen being the larger count. Always identify the denominator before interpreting the headline.

Know what an average keeps and loses

A mean uses every value and is sensitive to extreme observations. A median is the central value after sorting and can be less sensitive to one extreme. A mode tells which value appears most often. None is automatically ‘the best average’. The purpose of the summary and the dataset decide which is informative. For a grouped histogram, individual observations may no longer be recoverable at all.

Separate observation, model and conclusion

A fictional chart may show an upward association. It can describe the recorded observations. A straight-line model may make a conditional prediction. Neither fact alone proves a causal story or a guaranteed future result. The pupil should learn phrases such as ‘among the surveyed respondents’, ‘for the dates shown’, and ‘if the model remains valid’ instead of claiming more than the data support.

37 original graph and data-claim clinics

The numbers, surveys and examples below are fictional learning situations and not statistics about real Bukit Timah schools, families or businesses. Each short clinic starts with a tempting claim and replaces it with a more accurate mathematical statement. Study a few in each category and then try identifying the claim’s limitation without looking at the response.

1. An axis begins at 90 instead of zero

Claim or chart: A bar graph displays results 92 and 96, but the y-axis begins at 90. A viewer says the second is three times as large.

What the data actually support: The actual ratio is 96/92≈1.043, so the second value is about 4.35% larger, not three times as large. A truncated axis can exaggerate appearance.

First interpretation mistake: Compare the printed values before judging the relative size of the bars.

Independent check: Name the baseline and calculate 96−92=4 units; do not rely on apparent bar height alone.

2. One square means five units

Claim or chart: The vertical axis marks 0,5,10,15 with evenly spaced ticks. A pupil reads the third tick as 3.

What the data actually support: The third labelled tick represents ten units when counting 0 as the first labelled tick. Read the scale labels.

First interpretation mistake: Count value intervals, not simply grid squares.

Independent check: State how many units one square or one major interval actually represents.

3. A line’s upward movement is mistaken for total increase

Claim or chart: A graph value rises from 18 to 24. A learner reports a rise of 24 units.

What the data actually support: The increase is final minus initial, 24−18=6 units. The value at the end and the change are different quantities.

First interpretation mistake: Underline ‘how much did it rise?’ before calculating.

Independent check: Reconstruct 18+6=24.

4. Two graph axes use different units

Claim or chart: A line plot shows time in hours horizontally and distance in kilometres vertically; a learner calls gradient 30 km.

What the data actually support: Gradient is distance change divided by time change and has units km/h, not just kilometres.

First interpretation mistake: Check both axis labels before calculating rise/run.

Independent check: If the line rises 60 km over two hours, gradient is 30 km/h.

5. Two y-axis scales cannot be compared as one

Claim or chart: A graphic shows one dataset against a left y-axis and another against a differently scaled right y-axis.

What the data actually support: Do not infer that equal visual heights mean equal numerical values. Read the appropriate axis for each series.

First interpretation mistake: Track the units, minima and tick intervals of both axes.

Independent check: Re-express both series in comparable units before making a numerical comparison.

6. A proportion lacks its denominator

Claim or chart: A note says ‘twelve pupils preferred option A’ without stating how many were surveyed.

What the data actually support: Twelve is a count, not enough to calculate a share or compare prevalence fairly.

First interpretation mistake: Ask for the total number who answered the question.

Independent check: If twelve of twenty chose A, that is 60%; if twelve of one hundred, it is 12%.

7. Percent and percentage points differ

Claim or chart: A rate rises from 20% to 25%. A pupil calls it a 5% relative increase.

What the data actually support: The change is 5 percentage points. Relative to the old 20%, the increase is 5/20=25%.

First interpretation mistake: Name the reference base used for each statement.

Independent check: Check that 20% multiplied by 1.25 gives 25%.

8. The mean is pushed upward by one extreme value

Claim or chart: Fictional data are 2,3,4,5,46. Someone says the typical value must be 12.

What the data actually support: Mean is (2+3+4+5+46)/5=12, while median is four. The higher final observation influences the mean strongly.

First interpretation mistake: Explain the question’s meaning of typical before choosing a summary.

Independent check: Compare mean and median explicitly, without declaring one universally superior.

9. Median is not the middle of the written order

Claim or chart: The values 8,2,5,9,4 are displayed. A learner chooses 5 only because it appears in the written middle.

What the data actually support: The sorted list is 2,4,5,8,9, so median is five; the result happens to be correct but the method is not justified.

First interpretation mistake: Sort before locating the central observation.

Independent check: Change the written order to 8,9,2,4,5; the median must remain five.

10. A set can contain two modal values

Claim or chart: The observations are 1,1,2,2,3. A pupil insists there must be one unique mode.

What the data actually support: Both one and two occur twice, more often than three. The set has two modes.

First interpretation mistake: Count frequency instead of selecting a maximum numerical value.

Independent check: A mode is about repetition, not arithmetic size.

11. Range does not describe all spread

Claim or chart: Two datasets are 1,5,5,5,9 and 1,1,5,9,9. A pupil says they are identical because their ranges match.

What the data actually support: Both have range eight, but their distributions are different. A matching range is one shared feature, not full identity.

First interpretation mistake: Compare frequencies or dot diagrams as well as min and max.

Independent check: Identify where observations cluster in each list.

12. A pie-chart angle represents a share

Claim or chart: A sector in a pie chart spans 90 degrees. What percent of the whole is that?

What the data actually support: A full turn is 360°. The share is 90/360=1/4=25%.

First interpretation mistake: Do not read 90 degrees as 90%.

Independent check: Check that four such sectors fill a circle.

13. A pie chart must represent one whole

Claim or chart: Pie-sector shares are labelled 35%, 25%, 20%, and 30%. Are these a complete, non-overlapping partition?

What the data actually support: They total 110%, so the labels cannot represent one complete partition as written.

First interpretation mistake: Check the sum before trusting the graphic.

Independent check: Ask whether categories overlap or whether an entry is incorrect.

14. A pictogram key multiplies the symbol count

Claim or chart: A pictogram key says one full icon equals five participants. A row has three full icons.

What the data actually support: The row represents 15 participants, not three.

First interpretation mistake: Read the key before counting symbols.

Independent check: If half an icon means 2.5 people, the representation may be inappropriate for a whole-person dataset without an explanation.

15. Dots count observations, not distinct labels

Claim or chart: A dot diagram has two dots at 2 and three dots at 4. Someone reports sample size two.

What the data actually support: The list is 2,2,4,4,4, containing five observations.

First interpretation mistake: Each dot is one observation; axis values name categories of numerical position.

Independent check: Count the dots rather than the number of distinct x-axis labels.

16. A histogram bin is not an exact individual value

Claim or chart: The tallest equal-width histogram class is 10 to under 15. A learner says the exact mode is 12.5.

What the data actually support: We know the modal class is 10 to under 15, but not the precise individual values. The midpoint is not a proven exact observation.

First interpretation mistake: Avoid claiming grouped data reveal information already hidden by grouping.

Independent check: Several raw datasets with different modes can generate the same class frequency.

17. A class boundary can be double-counted

Claim or chart: The bins are 5 to under 10 and 10 to under 15. Where does a measured ten belong?

What the data actually support: Ten is in the second bin only; the first excludes ten.

First interpretation mistake: Write interval boundaries explicitly before sorting observations.

Independent check: No datum should occupy two bins in a mutually exclusive frequency table.

18. Category bars are not a numeric histogram

Claim or chart: A vertical bar chart labels categories ‘red’, ‘blue’, and ‘green’. A pupil calls it a continuous histogram.

What the data actually support: It is a categorical bar chart, because the x-axis values are categories rather than adjoining numerical intervals.

First interpretation mistake: Choose graph type according to the variable, not whether rectangles point upward.

Independent check: For a numeric histogram, define nonoverlapping class ranges.

19. A missing category may mean zero or missing data

Claim or chart: A chart lists responses A,B,D but omits C. Can we assume C had zero responses?

What the data actually support: Not without clarification. The category might have zero frequency or may have been excluded in data preparation.

First interpretation mistake: Check the source table or a chart note.

Independent check: Missing information should not silently become a zero count.

20. Two sample sizes make raw counts hard to compare

Claim or chart: Class A has ten pupils selecting an option among twenty; Class B has fifteen among fifty. Which has higher share?

What the data actually support: A has 10/20=50%; B has 15/50=30%. B’s raw count is larger, but A’s share is higher.

First interpretation mistake: Compare denominators before describing prevalence.

Independent check: A fair comparison uses the same definition and base.

21. A small sample is not a guarantee about everyone

Claim or chart: A fictional survey asks ten people in one queue; six choose option A. A headline says ‘60% of all residents prefer A’.

What the data actually support: The 60% describes only those ten respondents. Generalisation to all residents is not justified without appropriate sampling evidence.

First interpretation mistake: Name the observed group and its selection method.

Independent check: Do not publish a population claim when only a small convenience sample was measured.

22. A change in total can change a share even when counts rise

Claim or chart: A club grows from 40 to 100 members. Participants in an activity increase from 20 to 30. Did the activity’s participation rate rise?

What the data actually support: No: it fell from 20/40=50% to 30/100=30%, even though the participant count rose.

First interpretation mistake: Separate absolute counts from proportions.

Independent check: Compute both numerators relative to their own period totals.

23. A price graph can hide a fixed starting charge

Claim or chart: A fictional delivery graph follows C=5+2d. Someone says cost is directly proportional to distance because the graph is straight.

What the data actually support: Not directly proportional: at d=0, cost is $5, not zero. The relationship is linear with a fixed fee.

First interpretation mistake: Check the y-intercept and the ratio C/d.

Independent check: At 1 km cost $7; at 2 km cost $9, not $14.

24. A line can have a zero gradient at a nonzero height

Claim or chart: A graph shows y=6 for all x. A pupil says the y-value is zero because the gradient is zero.

What the data actually support: The gradient is zero but the line stays at height y=6.

First interpretation mistake: Distinguish height from change per x-unit.

Independent check: The intercept is six even though the line is horizontal.

25. An extrapolation is a model assumption

Claim or chart: A line fits measurements between zero and five hours. A headline predicts the same rule must continue for one hundred hours.

What the data actually support: The extrapolated value can be calculated, but the observation range does not prove the underlying relationship remains valid that long.

First interpretation mistake: Use words such as ‘the model predicts if the rule continues’.

Independent check: Check whether physical limits, caps or changed conditions could invalidate the trend.

26. Correlation does not by itself establish cause

Claim or chart: A fictional data graphic shows umbrella sales and wet weather occurring together. Does umbrella purchasing cause rain?

What the data actually support: No. A correlation does not, by itself, establish that one measured variable causes the other.

First interpretation mistake: Look for other influences and study design before interpreting causality.

Independent check: Describe the association without inventing a causal direction.

27. One time series can be seasonally cyclical

Claim or chart: A fictional park attendance graph peaks each school holiday. Can one holiday peak prove a permanently increasing trend?

What the data actually support: No. A repeating seasonal pattern and a long-term trend are different ideas. Compare equivalent periods over enough time.

First interpretation mistake: Do not treat a selected high week as proof of permanent growth.

Independent check: Read the time span and note comparable weeks before making a claim.

28. An average can be weighted by group size

Claim or chart: Group A has two values each averaging 10; Group B has eight values averaging 20. Can the combined mean be 15?

What the data actually support: No. Combined total is 2×10+8×20=180 over ten values, giving 18.

First interpretation mistake: A simple average of 10 and 20 gives 15 but ignores unequal group sizes.

Independent check: Check the combined mean lies closer to twenty because Group B is larger.

29. A probability estimate and a guarantee are different

Claim or chart: A fair coin has P(heads)=1/2. Will exactly five heads occur in ten tosses?

What the data actually support: Not guaranteed. Five heads is one possible count; random variation means other counts may occur.

First interpretation mistake: Distinguish an expected average proportion from a certain observed tally.

Independent check: A fair model does not remember earlier tosses in independent trials.

30. A graph can use averages with different reference groups

Claim or chart: Two schools show average score increases, but one reports a class mean and the other a whole-year mean. Can they be compared directly?

What the data actually support: Not without knowing whether the underlying groups, topic difficulty and measurement rules are comparable.

First interpretation mistake: Ask what was averaged and which population produced each mean.

Independent check: Make a like-for-like comparison before describing an advantage.

31. The chart title can omit a time window

Claim or chart: A bar graph says ‘Daily visitors’ but provides no date or time period. Can the result support ‘visitors this week’?

What the data actually support: Not without information about which day or days were counted.

First interpretation mistake: Check source date and the definition of daily.

Independent check: Do not silently change the period from a day to a week.

32. Stacked bars contain parts as well as totals

Claim or chart: A stacked bar has sections of 5, 8 and 7. A pupil reads only the tallest segment as its total.

What the data actually support: The complete bar total is 5+8+7=20. The largest individual section is eight.

First interpretation mistake: Know whether the question asks for one category or the combined stack.

Independent check: Count all included segments and inspect the legend.

33. A repeated data point is not a mistake by default

Claim or chart: A dot diagram shows the value four three times. A pupil deletes two dots as duplicates.

What the data actually support: Repeated observations can be genuine and must be retained unless a documented data-cleaning reason says otherwise.

First interpretation mistake: A dot diagram shows frequency through repetition.

Independent check: Deleting valid duplicates changes mean, mode and sample size.

34. A numerical graph can be technically correct but unsuitable

Claim or chart: A very detailed line graph is used to compare favourite school subjects, a categorical variable.

What the data actually support: A bar chart is generally clearer for distinct categories. A connected line can imply an artificial continuity or order.

First interpretation mistake: Match display to the structure of the variable.

Independent check: Check whether adjacent x-axis categories represent meaningful numerical progression.

35. Two percentages need one stated baseline to compare

Claim or chart: A fictional price goes from $50 to $60 then back to $50. Someone says each move is 20%.

What the data actually support: The rise is (60−50)/50=20%, but the fall is (60−50)/60≈16.67%. They use different starting values.

First interpretation mistake: Treating an equal dollar change as an equal relative percentage change.

Independent check: State the starting value of each comparison.

36. A rounded number does not reveal the original exactly

Claim or chart: A report says the mean is 7.3 rounded to one decimal place. Does that prove the unrounded mean equals exactly 7.3?

What the data actually support: No. Under conventional rounding it could represent values from about 7.25 up to but not including 7.35, subject to tie rules.

First interpretation mistake: Treat a rounded display as exact hidden data.

Independent check: State the reported precision when comparing measurements.

37. A graph’s units can change a comparison

Claim or chart: One chart uses metres and another centimetres. Both display 50 as a value. Are they equal lengths?

What the data actually support: No. 50 m equals 5000 cm, while 50 cm is 0.5 m. A common unit is required.

First interpretation mistake: Bare numerals do not determine quantities without units.

Independent check: Convert before calculating differences or ratios.

Teach pupils to rewrite a misleading claim accurately

A learner should not stop at ‘this chart is bad’. Ask for a specific rewrite. ‘Bar B is three times higher’ might become ‘Bar B’s reported value is four units above Bar A; the vertical axis starts at ninety.’ ‘Most residents prefer A’ might become ‘Six of the ten people surveyed chose A; the sample does not establish the preferences of all residents.’ A mathematically careful rewrite reports what is known while openly stating what is uncertain.

Use a simple check line below each answer

Numerical check: calculate the difference, ratio or percentage with its correct denominator.
Source check: state the sample size, time window and categories actually shown.
Representation check: name whether the figure shows an individual value, an interval, a count, a rate or a summary.
Claim check: distinguish ‘the data show’ from ‘the model suggests’ or ‘we would need more evidence’.

A five-day family data-literacy practice plan

Day 1: read axes, units and tick spacing from two graphs in the child’s current textbook.
Day 2: compute a percentage from a fictional survey count with a clearly stated sample size.
Day 3: compare mean and median in a five-value dataset containing one unusually large value.
Day 4: make a dot plot or equal-width frequency histogram and identify what exact information each display preserves.
Day 5: rewrite one overconfident chart claim in a precise sentence that names the measured group and limitation.

How a small-group G1 Mathematics lesson can teach this

First 10 minutes: retrieve fractions, percentages and one straightforward axis-reading item.
Next 20 minutes: compare accurate and misleading displays of the same fictional values.
Next 25 minutes: individually diagnose a scale, denominator or summary-statistic error.
Next 15 minutes: translate a numerical table into a graph and discuss what information grouping removes.
Next 10 minutes: write accurate interpretations of unfamiliar data without help.
Final 10 minutes: correct one overstated conclusion and plan a later independent retest.

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Yes. An axis that begins far above zero can emphasise small differences, and unclear units can encourage mistaken comparisons. Inspect the numerical labels and context before interpreting appearance.

Is median always better than mean?

No. They answer different questions. The mean uses every value, while the median describes the central position. Select the summary appropriate to the dataset and purpose.

What is a modal class?

It is the interval with the largest frequency in a grouped display. It generally does not tell you an exact most-common individual value.

Can a graph prove that one thing caused another?

A graph alone generally cannot establish causation. The way data were gathered, possible other influences and relevant evidence must be considered.

Are all 37 examples required for Secondary 2 G1?

No. The article includes G1-aligned graph reading and broader critical interpretation. The school’s current curriculum and actual question instructions take priority.

How will a tutor know my child has improved?

Have the learner read a different graph, calculate with the right denominator and rewrite a new claim accurately without copying a model explanation. Retest after a delay.

Official sources and accuracy note

One-minute diagnostic
The five-question method
37 original misleading-data examples
How a pupil can rewrite a misleading conclusion
A practical family exercise
Three-pupil data lesson
Frequently asked questions
Official sources

A one-minute data-reading diagnostic

How to move from a picture to a defensible claim

Start by reading the axes and legend

Ask which whole is being described

Know what an average keeps and loses

Separate observation, model and conclusion

37 original graph and data-claim clinics

1. An axis begins at 90 instead of zero

Claim or chart: A bar graph displays results 92 and 96, but the y-axis begins at 90. A viewer says the second is three times as large.

What the data actually support: The actual ratio is 96/92≈1.043, so the second value is about 4.35% larger, not three times as large. A truncated axis can exaggerate appearance.

First interpretation mistake: Compare the printed values before judging the relative size of the bars.

Independent check: Name the baseline and calculate 96−92=4 units; do not rely on apparent bar height alone.

2. One square means five units

Claim or chart: The vertical axis marks 0,5,10,15 with evenly spaced ticks. A pupil reads the third tick as 3.

What the data actually support: The third labelled tick represents ten units when counting 0 as the first labelled tick. Read the scale labels.

First interpretation mistake: Count value intervals, not simply grid squares.

Independent check: State how many units one square or one major interval actually represents.

3. A line’s upward movement is mistaken for total increase

Claim or chart: A graph value rises from 18 to 24. A learner reports a rise of 24 units.

What the data actually support: The increase is final minus initial, 24−18=6 units. The value at the end and the change are different quantities.

First interpretation mistake: Underline ‘how much did it rise?’ before calculating.

Independent check: Reconstruct 18+6=24.

4. Two graph axes use different units

Claim or chart: A line plot shows time in hours horizontally and distance in kilometres vertically; a learner calls gradient 30 km.

What the data actually support: Gradient is distance change divided by time change and has units km/h, not just kilometres.

First interpretation mistake: Check both axis labels before calculating rise/run.

Independent check: If the line rises 60 km over two hours, gradient is 30 km/h.

5. Two y-axis scales cannot be compared as one

Claim or chart: A graphic shows one dataset against a left y-axis and another against a differently scaled right y-axis.

What the data actually support: Do not infer that equal visual heights mean equal numerical values. Read the appropriate axis for each series.

First interpretation mistake: Track the units, minima and tick intervals of both axes.

Independent check: Re-express both series in comparable units before making a numerical comparison.

6. A proportion lacks its denominator

Claim or chart: A note says ‘twelve pupils preferred option A’ without stating how many were surveyed.

What the data actually support: Twelve is a count, not enough to calculate a share or compare prevalence fairly.

First interpretation mistake: Ask for the total number who answered the question.

Independent check: If twelve of twenty chose A, that is 60%; if twelve of one hundred, it is 12%.

7. Percent and percentage points differ

Claim or chart: A rate rises from 20% to 25%. A pupil calls it a 5% relative increase.

What the data actually support: The change is 5 percentage points. Relative to the old 20%, the increase is 5/20=25%.

First interpretation mistake: Name the reference base used for each statement.

Independent check: Check that 20% multiplied by 1.25 gives 25%.

8. The mean is pushed upward by one extreme value

Claim or chart: Fictional data are 2,3,4,5,46. Someone says the typical value must be 12.

What the data actually support: Mean is (2+3+4+5+46)/5=12, while median is four. The higher final observation influences the mean strongly.

First interpretation mistake: Explain the question’s meaning of typical before choosing a summary.

Independent check: Compare mean and median explicitly, without declaring one universally superior.

9. Median is not the middle of the written order

Claim or chart: The values 8,2,5,9,4 are displayed. A learner chooses 5 only because it appears in the written middle.

What the data actually support: The sorted list is 2,4,5,8,9, so median is five; the result happens to be correct but the method is not justified.

First interpretation mistake: Sort before locating the central observation.

Independent check: Change the written order to 8,9,2,4,5; the median must remain five.

10. A set can contain two modal values

Claim or chart: The observations are 1,1,2,2,3. A pupil insists there must be one unique mode.

What the data actually support: Both one and two occur twice, more often than three. The set has two modes.

First interpretation mistake: Count frequency instead of selecting a maximum numerical value.

Independent check: A mode is about repetition, not arithmetic size.

11. Range does not describe all spread

Claim or chart: Two datasets are 1,5,5,5,9 and 1,1,5,9,9. A pupil says they are identical because their ranges match.

What the data actually support: Both have range eight, but their distributions are different. A matching range is one shared feature, not full identity.

First interpretation mistake: Compare frequencies or dot diagrams as well as min and max.

Independent check: Identify where observations cluster in each list.

12. A pie-chart angle represents a share

Claim or chart: A sector in a pie chart spans 90 degrees. What percent of the whole is that?

What the data actually support: A full turn is 360°. The share is 90/360=1/4=25%.

First interpretation mistake: Do not read 90 degrees as 90%.

Independent check: Check that four such sectors fill a circle.

13. A pie chart must represent one whole

Claim or chart: Pie-sector shares are labelled 35%, 25%, 20%, and 30%. Are these a complete, non-overlapping partition?

What the data actually support: They total 110%, so the labels cannot represent one complete partition as written.

First interpretation mistake: Check the sum before trusting the graphic.

Independent check: Ask whether categories overlap or whether an entry is incorrect.

14. A pictogram key multiplies the symbol count

Claim or chart: A pictogram key says one full icon equals five participants. A row has three full icons.

What the data actually support: The row represents 15 participants, not three.

First interpretation mistake: Read the key before counting symbols.

Independent check: If half an icon means 2.5 people, the representation may be inappropriate for a whole-person dataset without an explanation.

15. Dots count observations, not distinct labels

Claim or chart: A dot diagram has two dots at 2 and three dots at 4. Someone reports sample size two.

What the data actually support: The list is 2,2,4,4,4, containing five observations.

First interpretation mistake: Each dot is one observation; axis values name categories of numerical position.

Independent check: Count the dots rather than the number of distinct x-axis labels.

16. A histogram bin is not an exact individual value

Claim or chart: The tallest equal-width histogram class is 10 to under 15. A learner says the exact mode is 12.5.

What the data actually support: We know the modal class is 10 to under 15, but not the precise individual values. The midpoint is not a proven exact observation.

First interpretation mistake: Avoid claiming grouped data reveal information already hidden by grouping.

Independent check: Several raw datasets with different modes can generate the same class frequency.

17. A class boundary can be double-counted

Claim or chart: The bins are 5 to under 10 and 10 to under 15. Where does a measured ten belong?

What the data actually support: Ten is in the second bin only; the first excludes ten.

First interpretation mistake: Write interval boundaries explicitly before sorting observations.

Independent check: No datum should occupy two bins in a mutually exclusive frequency table.

18. Category bars are not a numeric histogram

Claim or chart: A vertical bar chart labels categories ‘red’, ‘blue’, and ‘green’. A pupil calls it a continuous histogram.

What the data actually support: It is a categorical bar chart, because the x-axis values are categories rather than adjoining numerical intervals.

First interpretation mistake: Choose graph type according to the variable, not whether rectangles point upward.

Independent check: For a numeric histogram, define nonoverlapping class ranges.

19. A missing category may mean zero or missing data

Claim or chart: A chart lists responses A,B,D but omits C. Can we assume C had zero responses?

What the data actually support: Not without clarification. The category might have zero frequency or may have been excluded in data preparation.

First interpretation mistake: Check the source table or a chart note.

Independent check: Missing information should not silently become a zero count.

20. Two sample sizes make raw counts hard to compare

Claim or chart: Class A has ten pupils selecting an option among twenty; Class B has fifteen among fifty. Which has higher share?

What the data actually support: A has 10/20=50%; B has 15/50=30%. B’s raw count is larger, but A’s share is higher.

First interpretation mistake: Compare denominators before describing prevalence.

Independent check: A fair comparison uses the same definition and base.

21. A small sample is not a guarantee about everyone

Claim or chart: A fictional survey asks ten people in one queue; six choose option A. A headline says ‘60% of all residents prefer A’.

What the data actually support: The 60% describes only those ten respondents. Generalisation to all residents is not justified without appropriate sampling evidence.

First interpretation mistake: Name the observed group and its selection method.

Independent check: Do not publish a population claim when only a small convenience sample was measured.

22. A change in total can change a share even when counts rise

Claim or chart: A club grows from 40 to 100 members. Participants in an activity increase from 20 to 30. Did the activity’s participation rate rise?

What the data actually support: No: it fell from 20/40=50% to 30/100=30%, even though the participant count rose.

First interpretation mistake: Separate absolute counts from proportions.

Independent check: Compute both numerators relative to their own period totals.

23. A price graph can hide a fixed starting charge

Claim or chart: A fictional delivery graph follows C=5+2d. Someone says cost is directly proportional to distance because the graph is straight.

What the data actually support: Not directly proportional: at d=0, cost is $5, not zero. The relationship is linear with a fixed fee.

First interpretation mistake: Check the y-intercept and the ratio C/d.

Independent check: At 1 km cost $7; at 2 km cost $9, not $14.

24. A line can have a zero gradient at a nonzero height

Claim or chart: A graph shows y=6 for all x. A pupil says the y-value is zero because the gradient is zero.

What the data actually support: The gradient is zero but the line stays at height y=6.

First interpretation mistake: Distinguish height from change per x-unit.

Independent check: The intercept is six even though the line is horizontal.

25. An extrapolation is a model assumption

Claim or chart: A line fits measurements between zero and five hours. A headline predicts the same rule must continue for one hundred hours.

What the data actually support: The extrapolated value can be calculated, but the observation range does not prove the underlying relationship remains valid that long.

First interpretation mistake: Use words such as ‘the model predicts if the rule continues’.

Independent check: Check whether physical limits, caps or changed conditions could invalidate the trend.

26. Correlation does not by itself establish cause

Claim or chart: A fictional data graphic shows umbrella sales and wet weather occurring together. Does umbrella purchasing cause rain?

What the data actually support: No. A correlation does not, by itself, establish that one measured variable causes the other.

First interpretation mistake: Look for other influences and study design before interpreting causality.

Independent check: Describe the association without inventing a causal direction.

27. One time series can be seasonally cyclical

Claim or chart: A fictional park attendance graph peaks each school holiday. Can one holiday peak prove a permanently increasing trend?

What the data actually support: No. A repeating seasonal pattern and a long-term trend are different ideas. Compare equivalent periods over enough time.

First interpretation mistake: Do not treat a selected high week as proof of permanent growth.

Independent check: Read the time span and note comparable weeks before making a claim.

28. An average can be weighted by group size

Claim or chart: Group A has two values each averaging 10; Group B has eight values averaging 20. Can the combined mean be 15?

What the data actually support: No. Combined total is 2×10+8×20=180 over ten values, giving 18.

First interpretation mistake: A simple average of 10 and 20 gives 15 but ignores unequal group sizes.

Independent check: Check the combined mean lies closer to twenty because Group B is larger.

29. A probability estimate and a guarantee are different

Claim or chart: A fair coin has P(heads)=1/2. Will exactly five heads occur in ten tosses?

What the data actually support: Not guaranteed. Five heads is one possible count; random variation means other counts may occur.

First interpretation mistake: Distinguish an expected average proportion from a certain observed tally.

Independent check: A fair model does not remember earlier tosses in independent trials.

30. A graph can use averages with different reference groups

Claim or chart: Two schools show average score increases, but one reports a class mean and the other a whole-year mean. Can they be compared directly?

What the data actually support: Not without knowing whether the underlying groups, topic difficulty and measurement rules are comparable.

First interpretation mistake: Ask what was averaged and which population produced each mean.

Independent check: Make a like-for-like comparison before describing an advantage.

31. The chart title can omit a time window

Claim or chart: A bar graph says ‘Daily visitors’ but provides no date or time period. Can the result support ‘visitors this week’?

What the data actually support: Not without information about which day or days were counted.

First interpretation mistake: Check source date and the definition of daily.

Independent check: Do not silently change the period from a day to a week.

32. Stacked bars contain parts as well as totals

Claim or chart: A stacked bar has sections of 5, 8 and 7. A pupil reads only the tallest segment as its total.

What the data actually support: The complete bar total is 5+8+7=20. The largest individual section is eight.

First interpretation mistake: Know whether the question asks for one category or the combined stack.

Independent check: Count all included segments and inspect the legend.

33. A repeated data point is not a mistake by default

Claim or chart: A dot diagram shows the value four three times. A pupil deletes two dots as duplicates.

What the data actually support: Repeated observations can be genuine and must be retained unless a documented data-cleaning reason says otherwise.

First interpretation mistake: A dot diagram shows frequency through repetition.

Independent check: Deleting valid duplicates changes mean, mode and sample size.

34. A numerical graph can be technically correct but unsuitable

Claim or chart: A very detailed line graph is used to compare favourite school subjects, a categorical variable.

What the data actually support: A bar chart is generally clearer for distinct categories. A connected line can imply an artificial continuity or order.

First interpretation mistake: Match display to the structure of the variable.

Independent check: Check whether adjacent x-axis categories represent meaningful numerical progression.

35. Two percentages need one stated baseline to compare

Claim or chart: A fictional price goes from $50 to $60 then back to $50. Someone says each move is 20%.

What the data actually support: The rise is (60−50)/50=20%, but the fall is (60−50)/60≈16.67%. They use different starting values.

First interpretation mistake: Treating an equal dollar change as an equal relative percentage change.

Independent check: State the starting value of each comparison.

36. A rounded number does not reveal the original exactly

Claim or chart: A report says the mean is 7.3 rounded to one decimal place. Does that prove the unrounded mean equals exactly 7.3?

What the data actually support: No. Under conventional rounding it could represent values from about 7.25 up to but not including 7.35, subject to tie rules.

First interpretation mistake: Treat a rounded display as exact hidden data.

Independent check: State the reported precision when comparing measurements.

37. A graph’s units can change a comparison

Claim or chart: One chart uses metres and another centimetres. Both display 50 as a value. Are they equal lengths?

What the data actually support: No. 50 m equals 5000 cm, while 50 cm is 0.5 m. A common unit is required.

First interpretation mistake: Bare numerals do not determine quantities without units.

Independent check: Convert before calculating differences or ratios.

Teach pupils to rewrite a misleading claim accurately

Use a simple check line below each answer

Numerical check: calculate the difference, ratio or percentage with its correct denominator.
Source check: state the sample size, time window and categories actually shown.
Representation check: name whether the figure shows an individual value, an interval, a count, a rate or a summary.
Claim check: distinguish ‘the data show’ from ‘the model suggests’ or ‘we would need more evidence’.

A five-day family data-literacy practice plan

Day 1: read axes, units and tick spacing from two graphs in the child’s current textbook.
Day 2: compute a percentage from a fictional survey count with a clearly stated sample size.
Day 3: compare mean and median in a five-value dataset containing one unusually large value.
Day 4: make a dot plot or equal-width frequency histogram and identify what exact information each display preserves.
Day 5: rewrite one overconfident chart claim in a precise sentence that names the measured group and limitation.

How a small-group G1 Mathematics lesson can teach this

First 10 minutes: retrieve fractions, percentages and one straightforward axis-reading item.
Next 20 minutes: compare accurate and misleading displays of the same fictional values.
Next 25 minutes: individually diagnose a scale, denominator or summary-statistic error.
Next 15 minutes: translate a numerical table into a graph and discuss what information grouping removes.
Next 10 minutes: write accurate interpretations of unfamiliar data without help.
Final 10 minutes: correct one overstated conclusion and plan a later independent retest.

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Yes. An axis that begins far above zero can emphasise small differences, and unclear units can encourage mistaken comparisons. Inspect the numerical labels and context before interpreting appearance.

Is median always better than mean?

No. They answer different questions. The mean uses every value, while the median describes the central position. Select the summary appropriate to the dataset and purpose.

What is a modal class?

It is the interval with the largest frequency in a grouped display. It generally does not tell you an exact most-common individual value.

Can a graph prove that one thing caused another?

A graph alone generally cannot establish causation. The way data were gathered, possible other influences and relevant evidence must be considered.

Are all 37 examples required for Secondary 2 G1?

No. The article includes G1-aligned graph reading and broader critical interpretation. The school’s current curriculum and actual question instructions take priority.

How will a tutor know my child has improved?

Have the learner read a different graph, calculate with the right denominator and rewrite a new claim accurately without copying a model explanation. Retest after a delay.

Official sources and accuracy note

Title: what exactly was measured, where and over which period?
Axes: what are the units, origin, tick intervals and any secondary scale?
Data: are the figures individual values, categories, grouped intervals, counts or percentages?
Base: what sample size or original quantity does a percentage describe?
Claim: does the conclusion go further than the evidence, especially about cause, future outcomes or an entire population?

This guide supports families seeking a Secondary 2 G1 Mathematics tutor in Bukit Timah. It builds on the MOE G1 data-handling strand and related number, ratio and graph knowledge. It includes some general critical-data-literacy extensions; these are not a claim that every concept is explicitly examined in a given Secondary 2 G1 school assessment. The school’s actual topic list remains the practical reference.

Find the data claim you need to check

One-minute diagnostic
The five-question method
37 original misleading-data examples
How a pupil can rewrite a misleading conclusion
A practical family exercise
Three-pupil data lesson
Frequently asked questions
Official sources

A one-minute data-reading diagnostic

How to move from a picture to a defensible claim

Start by reading the axes and legend

Ask which whole is being described

Know what an average keeps and loses

Separate observation, model and conclusion

37 original graph and data-claim clinics

1. An axis begins at 90 instead of zero

Claim or chart: A bar graph displays results 92 and 96, but the y-axis begins at 90. A viewer says the second is three times as large.

What the data actually support: The actual ratio is 96/92≈1.043, so the second value is about 4.35% larger, not three times as large. A truncated axis can exaggerate appearance.

First interpretation mistake: Compare the printed values before judging the relative size of the bars.

Independent check: Name the baseline and calculate 96−92=4 units; do not rely on apparent bar height alone.

2. One square means five units

Claim or chart: The vertical axis marks 0,5,10,15 with evenly spaced ticks. A pupil reads the third tick as 3.

What the data actually support: The third labelled tick represents ten units when counting 0 as the first labelled tick. Read the scale labels.

First interpretation mistake: Count value intervals, not simply grid squares.

Independent check: State how many units one square or one major interval actually represents.

3. A line’s upward movement is mistaken for total increase

Claim or chart: A graph value rises from 18 to 24. A learner reports a rise of 24 units.

What the data actually support: The increase is final minus initial, 24−18=6 units. The value at the end and the change are different quantities.

First interpretation mistake: Underline ‘how much did it rise?’ before calculating.

Independent check: Reconstruct 18+6=24.

4. Two graph axes use different units

Claim or chart: A line plot shows time in hours horizontally and distance in kilometres vertically; a learner calls gradient 30 km.

What the data actually support: Gradient is distance change divided by time change and has units km/h, not just kilometres.

First interpretation mistake: Check both axis labels before calculating rise/run.

Independent check: If the line rises 60 km over two hours, gradient is 30 km/h.

5. Two y-axis scales cannot be compared as one

Claim or chart: A graphic shows one dataset against a left y-axis and another against a differently scaled right y-axis.

What the data actually support: Do not infer that equal visual heights mean equal numerical values. Read the appropriate axis for each series.

First interpretation mistake: Track the units, minima and tick intervals of both axes.

Independent check: Re-express both series in comparable units before making a numerical comparison.

6. A proportion lacks its denominator

Claim or chart: A note says ‘twelve pupils preferred option A’ without stating how many were surveyed.

What the data actually support: Twelve is a count, not enough to calculate a share or compare prevalence fairly.

First interpretation mistake: Ask for the total number who answered the question.

Independent check: If twelve of twenty chose A, that is 60%; if twelve of one hundred, it is 12%.

7. Percent and percentage points differ

Claim or chart: A rate rises from 20% to 25%. A pupil calls it a 5% relative increase.

What the data actually support: The change is 5 percentage points. Relative to the old 20%, the increase is 5/20=25%.

First interpretation mistake: Name the reference base used for each statement.

Independent check: Check that 20% multiplied by 1.25 gives 25%.

8. The mean is pushed upward by one extreme value

Claim or chart: Fictional data are 2,3,4,5,46. Someone says the typical value must be 12.

What the data actually support: Mean is (2+3+4+5+46)/5=12, while median is four. The higher final observation influences the mean strongly.

First interpretation mistake: Explain the question’s meaning of typical before choosing a summary.

Independent check: Compare mean and median explicitly, without declaring one universally superior.

9. Median is not the middle of the written order

Claim or chart: The values 8,2,5,9,4 are displayed. A learner chooses 5 only because it appears in the written middle.

What the data actually support: The sorted list is 2,4,5,8,9, so median is five; the result happens to be correct but the method is not justified.

First interpretation mistake: Sort before locating the central observation.

Independent check: Change the written order to 8,9,2,4,5; the median must remain five.

10. A set can contain two modal values

Claim or chart: The observations are 1,1,2,2,3. A pupil insists there must be one unique mode.

What the data actually support: Both one and two occur twice, more often than three. The set has two modes.

First interpretation mistake: Count frequency instead of selecting a maximum numerical value.

Independent check: A mode is about repetition, not arithmetic size.

11. Range does not describe all spread

Claim or chart: Two datasets are 1,5,5,5,9 and 1,1,5,9,9. A pupil says they are identical because their ranges match.

What the data actually support: Both have range eight, but their distributions are different. A matching range is one shared feature, not full identity.

First interpretation mistake: Compare frequencies or dot diagrams as well as min and max.

Independent check: Identify where observations cluster in each list.

12. A pie-chart angle represents a share

Claim or chart: A sector in a pie chart spans 90 degrees. What percent of the whole is that?

What the data actually support: A full turn is 360°. The share is 90/360=1/4=25%.

First interpretation mistake: Do not read 90 degrees as 90%.

Independent check: Check that four such sectors fill a circle.

13. A pie chart must represent one whole

Claim or chart: Pie-sector shares are labelled 35%, 25%, 20%, and 30%. Are these a complete, non-overlapping partition?

What the data actually support: They total 110%, so the labels cannot represent one complete partition as written.

First interpretation mistake: Check the sum before trusting the graphic.

Independent check: Ask whether categories overlap or whether an entry is incorrect.

14. A pictogram key multiplies the symbol count

Claim or chart: A pictogram key says one full icon equals five participants. A row has three full icons.

What the data actually support: The row represents 15 participants, not three.

First interpretation mistake: Read the key before counting symbols.

Independent check: If half an icon means 2.5 people, the representation may be inappropriate for a whole-person dataset without an explanation.

15. Dots count observations, not distinct labels

Claim or chart: A dot diagram has two dots at 2 and three dots at 4. Someone reports sample size two.

What the data actually support: The list is 2,2,4,4,4, containing five observations.

First interpretation mistake: Each dot is one observation; axis values name categories of numerical position.

Independent check: Count the dots rather than the number of distinct x-axis labels.

16. A histogram bin is not an exact individual value

Claim or chart: The tallest equal-width histogram class is 10 to under 15. A learner says the exact mode is 12.5.

What the data actually support: We know the modal class is 10 to under 15, but not the precise individual values. The midpoint is not a proven exact observation.

First interpretation mistake: Avoid claiming grouped data reveal information already hidden by grouping.

Independent check: Several raw datasets with different modes can generate the same class frequency.

17. A class boundary can be double-counted

Claim or chart: The bins are 5 to under 10 and 10 to under 15. Where does a measured ten belong?

What the data actually support: Ten is in the second bin only; the first excludes ten.

First interpretation mistake: Write interval boundaries explicitly before sorting observations.

Independent check: No datum should occupy two bins in a mutually exclusive frequency table.

18. Category bars are not a numeric histogram

Claim or chart: A vertical bar chart labels categories ‘red’, ‘blue’, and ‘green’. A pupil calls it a continuous histogram.

What the data actually support: It is a categorical bar chart, because the x-axis values are categories rather than adjoining numerical intervals.

First interpretation mistake: Choose graph type according to the variable, not whether rectangles point upward.

Independent check: For a numeric histogram, define nonoverlapping class ranges.

19. A missing category may mean zero or missing data

Claim or chart: A chart lists responses A,B,D but omits C. Can we assume C had zero responses?

What the data actually support: Not without clarification. The category might have zero frequency or may have been excluded in data preparation.

First interpretation mistake: Check the source table or a chart note.

Independent check: Missing information should not silently become a zero count.

20. Two sample sizes make raw counts hard to compare

Claim or chart: Class A has ten pupils selecting an option among twenty; Class B has fifteen among fifty. Which has higher share?

What the data actually support: A has 10/20=50%; B has 15/50=30%. B’s raw count is larger, but A’s share is higher.

First interpretation mistake: Compare denominators before describing prevalence.

Independent check: A fair comparison uses the same definition and base.

21. A small sample is not a guarantee about everyone

Claim or chart: A fictional survey asks ten people in one queue; six choose option A. A headline says ‘60% of all residents prefer A’.

What the data actually support: The 60% describes only those ten respondents. Generalisation to all residents is not justified without appropriate sampling evidence.

First interpretation mistake: Name the observed group and its selection method.

Independent check: Do not publish a population claim when only a small convenience sample was measured.

22. A change in total can change a share even when counts rise

Claim or chart: A club grows from 40 to 100 members. Participants in an activity increase from 20 to 30. Did the activity’s participation rate rise?

What the data actually support: No: it fell from 20/40=50% to 30/100=30%, even though the participant count rose.

First interpretation mistake: Separate absolute counts from proportions.

Independent check: Compute both numerators relative to their own period totals.

23. A price graph can hide a fixed starting charge

Claim or chart: A fictional delivery graph follows C=5+2d. Someone says cost is directly proportional to distance because the graph is straight.

What the data actually support: Not directly proportional: at d=0, cost is $5, not zero. The relationship is linear with a fixed fee.

First interpretation mistake: Check the y-intercept and the ratio C/d.

Independent check: At 1 km cost $7; at 2 km cost $9, not $14.

24. A line can have a zero gradient at a nonzero height

Claim or chart: A graph shows y=6 for all x. A pupil says the y-value is zero because the gradient is zero.

What the data actually support: The gradient is zero but the line stays at height y=6.

First interpretation mistake: Distinguish height from change per x-unit.

Independent check: The intercept is six even though the line is horizontal.

25. An extrapolation is a model assumption

Claim or chart: A line fits measurements between zero and five hours. A headline predicts the same rule must continue for one hundred hours.

What the data actually support: The extrapolated value can be calculated, but the observation range does not prove the underlying relationship remains valid that long.

First interpretation mistake: Use words such as ‘the model predicts if the rule continues’.

Independent check: Check whether physical limits, caps or changed conditions could invalidate the trend.

26. Correlation does not by itself establish cause

Claim or chart: A fictional data graphic shows umbrella sales and wet weather occurring together. Does umbrella purchasing cause rain?

What the data actually support: No. A correlation does not, by itself, establish that one measured variable causes the other.

First interpretation mistake: Look for other influences and study design before interpreting causality.

Independent check: Describe the association without inventing a causal direction.

27. One time series can be seasonally cyclical

Claim or chart: A fictional park attendance graph peaks each school holiday. Can one holiday peak prove a permanently increasing trend?

What the data actually support: No. A repeating seasonal pattern and a long-term trend are different ideas. Compare equivalent periods over enough time.

First interpretation mistake: Do not treat a selected high week as proof of permanent growth.

Independent check: Read the time span and note comparable weeks before making a claim.

28. An average can be weighted by group size

Claim or chart: Group A has two values each averaging 10; Group B has eight values averaging 20. Can the combined mean be 15?

What the data actually support: No. Combined total is 2×10+8×20=180 over ten values, giving 18.

First interpretation mistake: A simple average of 10 and 20 gives 15 but ignores unequal group sizes.

Independent check: Check the combined mean lies closer to twenty because Group B is larger.

29. A probability estimate and a guarantee are different

Claim or chart: A fair coin has P(heads)=1/2. Will exactly five heads occur in ten tosses?

What the data actually support: Not guaranteed. Five heads is one possible count; random variation means other counts may occur.

First interpretation mistake: Distinguish an expected average proportion from a certain observed tally.

Independent check: A fair model does not remember earlier tosses in independent trials.

30. A graph can use averages with different reference groups

Claim or chart: Two schools show average score increases, but one reports a class mean and the other a whole-year mean. Can they be compared directly?

What the data actually support: Not without knowing whether the underlying groups, topic difficulty and measurement rules are comparable.

First interpretation mistake: Ask what was averaged and which population produced each mean.

Independent check: Make a like-for-like comparison before describing an advantage.

31. The chart title can omit a time window

Claim or chart: A bar graph says ‘Daily visitors’ but provides no date or time period. Can the result support ‘visitors this week’?

What the data actually support: Not without information about which day or days were counted.

First interpretation mistake: Check source date and the definition of daily.

Independent check: Do not silently change the period from a day to a week.

32. Stacked bars contain parts as well as totals

Claim or chart: A stacked bar has sections of 5, 8 and 7. A pupil reads only the tallest segment as its total.

What the data actually support: The complete bar total is 5+8+7=20. The largest individual section is eight.

First interpretation mistake: Know whether the question asks for one category or the combined stack.

Independent check: Count all included segments and inspect the legend.

33. A repeated data point is not a mistake by default

Claim or chart: A dot diagram shows the value four three times. A pupil deletes two dots as duplicates.

What the data actually support: Repeated observations can be genuine and must be retained unless a documented data-cleaning reason says otherwise.

First interpretation mistake: A dot diagram shows frequency through repetition.

Independent check: Deleting valid duplicates changes mean, mode and sample size.

34. A numerical graph can be technically correct but unsuitable

Claim or chart: A very detailed line graph is used to compare favourite school subjects, a categorical variable.

What the data actually support: A bar chart is generally clearer for distinct categories. A connected line can imply an artificial continuity or order.

First interpretation mistake: Match display to the structure of the variable.

Independent check: Check whether adjacent x-axis categories represent meaningful numerical progression.

35. Two percentages need one stated baseline to compare

Claim or chart: A fictional price goes from $50 to $60 then back to $50. Someone says each move is 20%.

What the data actually support: The rise is (60−50)/50=20%, but the fall is (60−50)/60≈16.67%. They use different starting values.

First interpretation mistake: Treating an equal dollar change as an equal relative percentage change.

Independent check: State the starting value of each comparison.

36. A rounded number does not reveal the original exactly

Claim or chart: A report says the mean is 7.3 rounded to one decimal place. Does that prove the unrounded mean equals exactly 7.3?

What the data actually support: No. Under conventional rounding it could represent values from about 7.25 up to but not including 7.35, subject to tie rules.

First interpretation mistake: Treat a rounded display as exact hidden data.

Independent check: State the reported precision when comparing measurements.

37. A graph’s units can change a comparison

Claim or chart: One chart uses metres and another centimetres. Both display 50 as a value. Are they equal lengths?

What the data actually support: No. 50 m equals 5000 cm, while 50 cm is 0.5 m. A common unit is required.

First interpretation mistake: Bare numerals do not determine quantities without units.

Independent check: Convert before calculating differences or ratios.

Teach pupils to rewrite a misleading claim accurately

Use a simple check line below each answer

Numerical check: calculate the difference, ratio or percentage with its correct denominator.
Source check: state the sample size, time window and categories actually shown.
Representation check: name whether the figure shows an individual value, an interval, a count, a rate or a summary.
Claim check: distinguish ‘the data show’ from ‘the model suggests’ or ‘we would need more evidence’.

A five-day family data-literacy practice plan

Day 1: read axes, units and tick spacing from two graphs in the child’s current textbook.
Day 2: compute a percentage from a fictional survey count with a clearly stated sample size.
Day 3: compare mean and median in a five-value dataset containing one unusually large value.
Day 4: make a dot plot or equal-width frequency histogram and identify what exact information each display preserves.
Day 5: rewrite one overconfident chart claim in a precise sentence that names the measured group and limitation.

How a small-group G1 Mathematics lesson can teach this

First 10 minutes: retrieve fractions, percentages and one straightforward axis-reading item.
Next 20 minutes: compare accurate and misleading displays of the same fictional values.
Next 25 minutes: individually diagnose a scale, denominator or summary-statistic error.
Next 15 minutes: translate a numerical table into a graph and discuss what information grouping removes.
Next 10 minutes: write accurate interpretations of unfamiliar data without help.
Final 10 minutes: correct one overstated conclusion and plan a later independent retest.

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Yes. An axis that begins far above zero can emphasise small differences, and unclear units can encourage mistaken comparisons. Inspect the numerical labels and context before interpreting appearance.

Is median always better than mean?

No. They answer different questions. The mean uses every value, while the median describes the central position. Select the summary appropriate to the dataset and purpose.

What is a modal class?

It is the interval with the largest frequency in a grouped display. It generally does not tell you an exact most-common individual value.

Can a graph prove that one thing caused another?

A graph alone generally cannot establish causation. The way data were gathered, possible other influences and relevant evidence must be considered.

Are all 37 examples required for Secondary 2 G1?

No. The article includes G1-aligned graph reading and broader critical interpretation. The school’s current curriculum and actual question instructions take priority.

How will a tutor know my child has improved?

Have the learner read a different graph, calculate with the right denominator and rewrite a new claim accurately without copying a model explanation. Retest after a delay.

Official sources and accuracy note

Title: what exactly was measured, where and over which period?
Axes: what are the units, origin, tick intervals and any secondary scale?
Data: are the figures individual values, categories, grouped intervals, counts or percentages?
Base: what sample size or original quantity does a percentage describe?
Claim: does the conclusion go further than the evidence, especially about cause, future outcomes or an entire population?

Find the data claim you need to check

One-minute diagnostic
The five-question method
37 original misleading-data examples
How a pupil can rewrite a misleading conclusion
A practical family exercise
Three-pupil data lesson
Frequently asked questions
Official sources

A one-minute data-reading diagnostic

How to move from a picture to a defensible claim

Start by reading the axes and legend

Ask which whole is being described

Know what an average keeps and loses

Separate observation, model and conclusion

37 original graph and data-claim clinics

1. An axis begins at 90 instead of zero

Claim or chart: A bar graph displays results 92 and 96, but the y-axis begins at 90. A viewer says the second is three times as large.

What the data actually support: The actual ratio is 96/92≈1.043, so the second value is about 4.35% larger, not three times as large. A truncated axis can exaggerate appearance.

First interpretation mistake: Compare the printed values before judging the relative size of the bars.

Independent check: Name the baseline and calculate 96−92=4 units; do not rely on apparent bar height alone.

2. One square means five units

Claim or chart: The vertical axis marks 0,5,10,15 with evenly spaced ticks. A pupil reads the third tick as 3.

What the data actually support: The third labelled tick represents ten units when counting 0 as the first labelled tick. Read the scale labels.

First interpretation mistake: Count value intervals, not simply grid squares.

Independent check: State how many units one square or one major interval actually represents.

3. A line’s upward movement is mistaken for total increase

Claim or chart: A graph value rises from 18 to 24. A learner reports a rise of 24 units.

What the data actually support: The increase is final minus initial, 24−18=6 units. The value at the end and the change are different quantities.

First interpretation mistake: Underline ‘how much did it rise?’ before calculating.

Independent check: Reconstruct 18+6=24.

4. Two graph axes use different units

Claim or chart: A line plot shows time in hours horizontally and distance in kilometres vertically; a learner calls gradient 30 km.

What the data actually support: Gradient is distance change divided by time change and has units km/h, not just kilometres.

First interpretation mistake: Check both axis labels before calculating rise/run.

Independent check: If the line rises 60 km over two hours, gradient is 30 km/h.

5. Two y-axis scales cannot be compared as one

Claim or chart: A graphic shows one dataset against a left y-axis and another against a differently scaled right y-axis.

What the data actually support: Do not infer that equal visual heights mean equal numerical values. Read the appropriate axis for each series.

First interpretation mistake: Track the units, minima and tick intervals of both axes.

Independent check: Re-express both series in comparable units before making a numerical comparison.

6. A proportion lacks its denominator

Claim or chart: A note says ‘twelve pupils preferred option A’ without stating how many were surveyed.

What the data actually support: Twelve is a count, not enough to calculate a share or compare prevalence fairly.

First interpretation mistake: Ask for the total number who answered the question.

Independent check: If twelve of twenty chose A, that is 60%; if twelve of one hundred, it is 12%.

7. Percent and percentage points differ

Claim or chart: A rate rises from 20% to 25%. A pupil calls it a 5% relative increase.

What the data actually support: The change is 5 percentage points. Relative to the old 20%, the increase is 5/20=25%.

First interpretation mistake: Name the reference base used for each statement.

Independent check: Check that 20% multiplied by 1.25 gives 25%.

8. The mean is pushed upward by one extreme value

Claim or chart: Fictional data are 2,3,4,5,46. Someone says the typical value must be 12.

What the data actually support: Mean is (2+3+4+5+46)/5=12, while median is four. The higher final observation influences the mean strongly.

First interpretation mistake: Explain the question’s meaning of typical before choosing a summary.

Independent check: Compare mean and median explicitly, without declaring one universally superior.

9. Median is not the middle of the written order

Claim or chart: The values 8,2,5,9,4 are displayed. A learner chooses 5 only because it appears in the written middle.

What the data actually support: The sorted list is 2,4,5,8,9, so median is five; the result happens to be correct but the method is not justified.

First interpretation mistake: Sort before locating the central observation.

Independent check: Change the written order to 8,9,2,4,5; the median must remain five.

10. A set can contain two modal values

Claim or chart: The observations are 1,1,2,2,3. A pupil insists there must be one unique mode.

What the data actually support: Both one and two occur twice, more often than three. The set has two modes.

First interpretation mistake: Count frequency instead of selecting a maximum numerical value.

Independent check: A mode is about repetition, not arithmetic size.

11. Range does not describe all spread

Claim or chart: Two datasets are 1,5,5,5,9 and 1,1,5,9,9. A pupil says they are identical because their ranges match.

What the data actually support: Both have range eight, but their distributions are different. A matching range is one shared feature, not full identity.

First interpretation mistake: Compare frequencies or dot diagrams as well as min and max.

Independent check: Identify where observations cluster in each list.

12. A pie-chart angle represents a share

Claim or chart: A sector in a pie chart spans 90 degrees. What percent of the whole is that?

What the data actually support: A full turn is 360°. The share is 90/360=1/4=25%.

First interpretation mistake: Do not read 90 degrees as 90%.

Independent check: Check that four such sectors fill a circle.

13. A pie chart must represent one whole

Claim or chart: Pie-sector shares are labelled 35%, 25%, 20%, and 30%. Are these a complete, non-overlapping partition?

What the data actually support: They total 110%, so the labels cannot represent one complete partition as written.

First interpretation mistake: Check the sum before trusting the graphic.

Independent check: Ask whether categories overlap or whether an entry is incorrect.

14. A pictogram key multiplies the symbol count

Claim or chart: A pictogram key says one full icon equals five participants. A row has three full icons.

What the data actually support: The row represents 15 participants, not three.

First interpretation mistake: Read the key before counting symbols.

Independent check: If half an icon means 2.5 people, the representation may be inappropriate for a whole-person dataset without an explanation.

15. Dots count observations, not distinct labels

Claim or chart: A dot diagram has two dots at 2 and three dots at 4. Someone reports sample size two.

What the data actually support: The list is 2,2,4,4,4, containing five observations.

First interpretation mistake: Each dot is one observation; axis values name categories of numerical position.

Independent check: Count the dots rather than the number of distinct x-axis labels.

16. A histogram bin is not an exact individual value

Claim or chart: The tallest equal-width histogram class is 10 to under 15. A learner says the exact mode is 12.5.

What the data actually support: We know the modal class is 10 to under 15, but not the precise individual values. The midpoint is not a proven exact observation.

First interpretation mistake: Avoid claiming grouped data reveal information already hidden by grouping.

Independent check: Several raw datasets with different modes can generate the same class frequency.

17. A class boundary can be double-counted

Claim or chart: The bins are 5 to under 10 and 10 to under 15. Where does a measured ten belong?

What the data actually support: Ten is in the second bin only; the first excludes ten.

First interpretation mistake: Write interval boundaries explicitly before sorting observations.

Independent check: No datum should occupy two bins in a mutually exclusive frequency table.

18. Category bars are not a numeric histogram

Claim or chart: A vertical bar chart labels categories ‘red’, ‘blue’, and ‘green’. A pupil calls it a continuous histogram.

What the data actually support: It is a categorical bar chart, because the x-axis values are categories rather than adjoining numerical intervals.

First interpretation mistake: Choose graph type according to the variable, not whether rectangles point upward.

Independent check: For a numeric histogram, define nonoverlapping class ranges.

19. A missing category may mean zero or missing data

Claim or chart: A chart lists responses A,B,D but omits C. Can we assume C had zero responses?

What the data actually support: Not without clarification. The category might have zero frequency or may have been excluded in data preparation.

First interpretation mistake: Check the source table or a chart note.

Independent check: Missing information should not silently become a zero count.

20. Two sample sizes make raw counts hard to compare

Claim or chart: Class A has ten pupils selecting an option among twenty; Class B has fifteen among fifty. Which has higher share?

What the data actually support: A has 10/20=50%; B has 15/50=30%. B’s raw count is larger, but A’s share is higher.

First interpretation mistake: Compare denominators before describing prevalence.

Independent check: A fair comparison uses the same definition and base.

21. A small sample is not a guarantee about everyone

Claim or chart: A fictional survey asks ten people in one queue; six choose option A. A headline says ‘60% of all residents prefer A’.

What the data actually support: The 60% describes only those ten respondents. Generalisation to all residents is not justified without appropriate sampling evidence.

First interpretation mistake: Name the observed group and its selection method.

Independent check: Do not publish a population claim when only a small convenience sample was measured.

22. A change in total can change a share even when counts rise

Claim or chart: A club grows from 40 to 100 members. Participants in an activity increase from 20 to 30. Did the activity’s participation rate rise?

What the data actually support: No: it fell from 20/40=50% to 30/100=30%, even though the participant count rose.

First interpretation mistake: Separate absolute counts from proportions.

Independent check: Compute both numerators relative to their own period totals.

23. A price graph can hide a fixed starting charge

Claim or chart: A fictional delivery graph follows C=5+2d. Someone says cost is directly proportional to distance because the graph is straight.

What the data actually support: Not directly proportional: at d=0, cost is $5, not zero. The relationship is linear with a fixed fee.

First interpretation mistake: Check the y-intercept and the ratio C/d.

Independent check: At 1 km cost $7; at 2 km cost $9, not $14.

24. A line can have a zero gradient at a nonzero height

Claim or chart: A graph shows y=6 for all x. A pupil says the y-value is zero because the gradient is zero.

What the data actually support: The gradient is zero but the line stays at height y=6.

First interpretation mistake: Distinguish height from change per x-unit.

Independent check: The intercept is six even though the line is horizontal.

25. An extrapolation is a model assumption

Claim or chart: A line fits measurements between zero and five hours. A headline predicts the same rule must continue for one hundred hours.

What the data actually support: The extrapolated value can be calculated, but the observation range does not prove the underlying relationship remains valid that long.

First interpretation mistake: Use words such as ‘the model predicts if the rule continues’.

Independent check: Check whether physical limits, caps or changed conditions could invalidate the trend.

26. Correlation does not by itself establish cause

Claim or chart: A fictional data graphic shows umbrella sales and wet weather occurring together. Does umbrella purchasing cause rain?

What the data actually support: No. A correlation does not, by itself, establish that one measured variable causes the other.

First interpretation mistake: Look for other influences and study design before interpreting causality.

Independent check: Describe the association without inventing a causal direction.

27. One time series can be seasonally cyclical

Claim or chart: A fictional park attendance graph peaks each school holiday. Can one holiday peak prove a permanently increasing trend?

What the data actually support: No. A repeating seasonal pattern and a long-term trend are different ideas. Compare equivalent periods over enough time.

First interpretation mistake: Do not treat a selected high week as proof of permanent growth.

Independent check: Read the time span and note comparable weeks before making a claim.

28. An average can be weighted by group size

Claim or chart: Group A has two values each averaging 10; Group B has eight values averaging 20. Can the combined mean be 15?

What the data actually support: No. Combined total is 2×10+8×20=180 over ten values, giving 18.

First interpretation mistake: A simple average of 10 and 20 gives 15 but ignores unequal group sizes.

Independent check: Check the combined mean lies closer to twenty because Group B is larger.

29. A probability estimate and a guarantee are different

Claim or chart: A fair coin has P(heads)=1/2. Will exactly five heads occur in ten tosses?

What the data actually support: Not guaranteed. Five heads is one possible count; random variation means other counts may occur.

First interpretation mistake: Distinguish an expected average proportion from a certain observed tally.

Independent check: A fair model does not remember earlier tosses in independent trials.

30. A graph can use averages with different reference groups

Claim or chart: Two schools show average score increases, but one reports a class mean and the other a whole-year mean. Can they be compared directly?

What the data actually support: Not without knowing whether the underlying groups, topic difficulty and measurement rules are comparable.

First interpretation mistake: Ask what was averaged and which population produced each mean.

Independent check: Make a like-for-like comparison before describing an advantage.

31. The chart title can omit a time window

Claim or chart: A bar graph says ‘Daily visitors’ but provides no date or time period. Can the result support ‘visitors this week’?

What the data actually support: Not without information about which day or days were counted.

First interpretation mistake: Check source date and the definition of daily.

Independent check: Do not silently change the period from a day to a week.

32. Stacked bars contain parts as well as totals

Claim or chart: A stacked bar has sections of 5, 8 and 7. A pupil reads only the tallest segment as its total.

What the data actually support: The complete bar total is 5+8+7=20. The largest individual section is eight.

First interpretation mistake: Know whether the question asks for one category or the combined stack.

Independent check: Count all included segments and inspect the legend.

33. A repeated data point is not a mistake by default

Claim or chart: A dot diagram shows the value four three times. A pupil deletes two dots as duplicates.

What the data actually support: Repeated observations can be genuine and must be retained unless a documented data-cleaning reason says otherwise.

First interpretation mistake: A dot diagram shows frequency through repetition.

Independent check: Deleting valid duplicates changes mean, mode and sample size.

34. A numerical graph can be technically correct but unsuitable

Claim or chart: A very detailed line graph is used to compare favourite school subjects, a categorical variable.

What the data actually support: A bar chart is generally clearer for distinct categories. A connected line can imply an artificial continuity or order.

First interpretation mistake: Match display to the structure of the variable.

Independent check: Check whether adjacent x-axis categories represent meaningful numerical progression.

35. Two percentages need one stated baseline to compare

Claim or chart: A fictional price goes from $50 to $60 then back to $50. Someone says each move is 20%.

What the data actually support: The rise is (60−50)/50=20%, but the fall is (60−50)/60≈16.67%. They use different starting values.

First interpretation mistake: Treating an equal dollar change as an equal relative percentage change.

Independent check: State the starting value of each comparison.

36. A rounded number does not reveal the original exactly

Claim or chart: A report says the mean is 7.3 rounded to one decimal place. Does that prove the unrounded mean equals exactly 7.3?

What the data actually support: No. Under conventional rounding it could represent values from about 7.25 up to but not including 7.35, subject to tie rules.

First interpretation mistake: Treat a rounded display as exact hidden data.

Independent check: State the reported precision when comparing measurements.

37. A graph’s units can change a comparison

Claim or chart: One chart uses metres and another centimetres. Both display 50 as a value. Are they equal lengths?

What the data actually support: No. 50 m equals 5000 cm, while 50 cm is 0.5 m. A common unit is required.

First interpretation mistake: Bare numerals do not determine quantities without units.

Independent check: Convert before calculating differences or ratios.

Teach pupils to rewrite a misleading claim accurately

Use a simple check line below each answer

Numerical check: calculate the difference, ratio or percentage with its correct denominator.
Source check: state the sample size, time window and categories actually shown.
Representation check: name whether the figure shows an individual value, an interval, a count, a rate or a summary.
Claim check: distinguish ‘the data show’ from ‘the model suggests’ or ‘we would need more evidence’.

A five-day family data-literacy practice plan

Day 1: read axes, units and tick spacing from two graphs in the child’s current textbook.
Day 2: compute a percentage from a fictional survey count with a clearly stated sample size.
Day 3: compare mean and median in a five-value dataset containing one unusually large value.
Day 4: make a dot plot or equal-width frequency histogram and identify what exact information each display preserves.
Day 5: rewrite one overconfident chart claim in a precise sentence that names the measured group and limitation.

How a small-group G1 Mathematics lesson can teach this

First 10 minutes: retrieve fractions, percentages and one straightforward axis-reading item.
Next 20 minutes: compare accurate and misleading displays of the same fictional values.
Next 25 minutes: individually diagnose a scale, denominator or summary-statistic error.
Next 15 minutes: translate a numerical table into a graph and discuss what information grouping removes.
Next 10 minutes: write accurate interpretations of unfamiliar data without help.
Final 10 minutes: correct one overstated conclusion and plan a later independent retest.

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Yes. An axis that begins far above zero can emphasise small differences, and unclear units can encourage mistaken comparisons. Inspect the numerical labels and context before interpreting appearance.

Is median always better than mean?

No. They answer different questions. The mean uses every value, while the median describes the central position. Select the summary appropriate to the dataset and purpose.

What is a modal class?

It is the interval with the largest frequency in a grouped display. It generally does not tell you an exact most-common individual value.

Can a graph prove that one thing caused another?

A graph alone generally cannot establish causation. The way data were gathered, possible other influences and relevant evidence must be considered.

Are all 37 examples required for Secondary 2 G1?

No. The article includes G1-aligned graph reading and broader critical interpretation. The school’s current curriculum and actual question instructions take priority.

How will a tutor know my child has improved?

Have the learner read a different graph, calculate with the right denominator and rewrite a new claim accurately without copying a model explanation. Retest after a delay.

Official sources and accuracy note

Did you know? A graph can show a true number and still leave a very misleading impression. If one bar begins at zero while another graphic begins at ninety, their apparent sizes cannot be interpreted in the same way. A Secondary 2 G1 Mathematics student who learns to inspect the axis, the denominator and the sample behind a picture can become a more thoughtful reader of both school questions and everyday claims.

Quick answer: five checks before trusting a graph

Title: what exactly was measured, where and over which period?
Axes: what are the units, origin, tick intervals and any secondary scale?
Data: are the figures individual values, categories, grouped intervals, counts or percentages?
Base: what sample size or original quantity does a percentage describe?
Claim: does the conclusion go further than the evidence, especially about cause, future outcomes or an entire population?

Find the data claim you need to check

One-minute diagnostic
The five-question method
37 original misleading-data examples
How a pupil can rewrite a misleading conclusion
A practical family exercise
Three-pupil data lesson
Frequently asked questions
Official sources

A one-minute data-reading diagnostic

How to move from a picture to a defensible claim

Start by reading the axes and legend

Ask which whole is being described

Know what an average keeps and loses

Separate observation, model and conclusion

37 original graph and data-claim clinics

1. An axis begins at 90 instead of zero

Claim or chart: A bar graph displays results 92 and 96, but the y-axis begins at 90. A viewer says the second is three times as large.

What the data actually support: The actual ratio is 96/92≈1.043, so the second value is about 4.35% larger, not three times as large. A truncated axis can exaggerate appearance.

First interpretation mistake: Compare the printed values before judging the relative size of the bars.

Independent check: Name the baseline and calculate 96−92=4 units; do not rely on apparent bar height alone.

2. One square means five units

Claim or chart: The vertical axis marks 0,5,10,15 with evenly spaced ticks. A pupil reads the third tick as 3.

What the data actually support: The third labelled tick represents ten units when counting 0 as the first labelled tick. Read the scale labels.

First interpretation mistake: Count value intervals, not simply grid squares.

Independent check: State how many units one square or one major interval actually represents.

3. A line’s upward movement is mistaken for total increase

Claim or chart: A graph value rises from 18 to 24. A learner reports a rise of 24 units.

What the data actually support: The increase is final minus initial, 24−18=6 units. The value at the end and the change are different quantities.

First interpretation mistake: Underline ‘how much did it rise?’ before calculating.

Independent check: Reconstruct 18+6=24.

4. Two graph axes use different units

Claim or chart: A line plot shows time in hours horizontally and distance in kilometres vertically; a learner calls gradient 30 km.

What the data actually support: Gradient is distance change divided by time change and has units km/h, not just kilometres.

First interpretation mistake: Check both axis labels before calculating rise/run.

Independent check: If the line rises 60 km over two hours, gradient is 30 km/h.

5. Two y-axis scales cannot be compared as one

Claim or chart: A graphic shows one dataset against a left y-axis and another against a differently scaled right y-axis.

What the data actually support: Do not infer that equal visual heights mean equal numerical values. Read the appropriate axis for each series.

First interpretation mistake: Track the units, minima and tick intervals of both axes.

Independent check: Re-express both series in comparable units before making a numerical comparison.

6. A proportion lacks its denominator

Claim or chart: A note says ‘twelve pupils preferred option A’ without stating how many were surveyed.

What the data actually support: Twelve is a count, not enough to calculate a share or compare prevalence fairly.

First interpretation mistake: Ask for the total number who answered the question.

Independent check: If twelve of twenty chose A, that is 60%; if twelve of one hundred, it is 12%.

7. Percent and percentage points differ

Claim or chart: A rate rises from 20% to 25%. A pupil calls it a 5% relative increase.

What the data actually support: The change is 5 percentage points. Relative to the old 20%, the increase is 5/20=25%.

First interpretation mistake: Name the reference base used for each statement.

Independent check: Check that 20% multiplied by 1.25 gives 25%.

8. The mean is pushed upward by one extreme value

Claim or chart: Fictional data are 2,3,4,5,46. Someone says the typical value must be 12.

What the data actually support: Mean is (2+3+4+5+46)/5=12, while median is four. The higher final observation influences the mean strongly.

First interpretation mistake: Explain the question’s meaning of typical before choosing a summary.

Independent check: Compare mean and median explicitly, without declaring one universally superior.

9. Median is not the middle of the written order

Claim or chart: The values 8,2,5,9,4 are displayed. A learner chooses 5 only because it appears in the written middle.

What the data actually support: The sorted list is 2,4,5,8,9, so median is five; the result happens to be correct but the method is not justified.

First interpretation mistake: Sort before locating the central observation.

Independent check: Change the written order to 8,9,2,4,5; the median must remain five.

10. A set can contain two modal values

Claim or chart: The observations are 1,1,2,2,3. A pupil insists there must be one unique mode.

What the data actually support: Both one and two occur twice, more often than three. The set has two modes.

First interpretation mistake: Count frequency instead of selecting a maximum numerical value.

Independent check: A mode is about repetition, not arithmetic size.

11. Range does not describe all spread

Claim or chart: Two datasets are 1,5,5,5,9 and 1,1,5,9,9. A pupil says they are identical because their ranges match.

What the data actually support: Both have range eight, but their distributions are different. A matching range is one shared feature, not full identity.

First interpretation mistake: Compare frequencies or dot diagrams as well as min and max.

Independent check: Identify where observations cluster in each list.

12. A pie-chart angle represents a share

Claim or chart: A sector in a pie chart spans 90 degrees. What percent of the whole is that?

What the data actually support: A full turn is 360°. The share is 90/360=1/4=25%.

First interpretation mistake: Do not read 90 degrees as 90%.

Independent check: Check that four such sectors fill a circle.

13. A pie chart must represent one whole

Claim or chart: Pie-sector shares are labelled 35%, 25%, 20%, and 30%. Are these a complete, non-overlapping partition?

What the data actually support: They total 110%, so the labels cannot represent one complete partition as written.

First interpretation mistake: Check the sum before trusting the graphic.

Independent check: Ask whether categories overlap or whether an entry is incorrect.

14. A pictogram key multiplies the symbol count

Claim or chart: A pictogram key says one full icon equals five participants. A row has three full icons.

What the data actually support: The row represents 15 participants, not three.

First interpretation mistake: Read the key before counting symbols.

Independent check: If half an icon means 2.5 people, the representation may be inappropriate for a whole-person dataset without an explanation.

15. Dots count observations, not distinct labels

Claim or chart: A dot diagram has two dots at 2 and three dots at 4. Someone reports sample size two.

What the data actually support: The list is 2,2,4,4,4, containing five observations.

First interpretation mistake: Each dot is one observation; axis values name categories of numerical position.

Independent check: Count the dots rather than the number of distinct x-axis labels.

16. A histogram bin is not an exact individual value

Claim or chart: The tallest equal-width histogram class is 10 to under 15. A learner says the exact mode is 12.5.

What the data actually support: We know the modal class is 10 to under 15, but not the precise individual values. The midpoint is not a proven exact observation.

First interpretation mistake: Avoid claiming grouped data reveal information already hidden by grouping.

Independent check: Several raw datasets with different modes can generate the same class frequency.

17. A class boundary can be double-counted

Claim or chart: The bins are 5 to under 10 and 10 to under 15. Where does a measured ten belong?

What the data actually support: Ten is in the second bin only; the first excludes ten.

First interpretation mistake: Write interval boundaries explicitly before sorting observations.

Independent check: No datum should occupy two bins in a mutually exclusive frequency table.

18. Category bars are not a numeric histogram

Claim or chart: A vertical bar chart labels categories ‘red’, ‘blue’, and ‘green’. A pupil calls it a continuous histogram.

What the data actually support: It is a categorical bar chart, because the x-axis values are categories rather than adjoining numerical intervals.

First interpretation mistake: Choose graph type according to the variable, not whether rectangles point upward.

Independent check: For a numeric histogram, define nonoverlapping class ranges.

19. A missing category may mean zero or missing data

Claim or chart: A chart lists responses A,B,D but omits C. Can we assume C had zero responses?

What the data actually support: Not without clarification. The category might have zero frequency or may have been excluded in data preparation.

First interpretation mistake: Check the source table or a chart note.

Independent check: Missing information should not silently become a zero count.

20. Two sample sizes make raw counts hard to compare

Claim or chart: Class A has ten pupils selecting an option among twenty; Class B has fifteen among fifty. Which has higher share?

What the data actually support: A has 10/20=50%; B has 15/50=30%. B’s raw count is larger, but A’s share is higher.

First interpretation mistake: Compare denominators before describing prevalence.

Independent check: A fair comparison uses the same definition and base.

21. A small sample is not a guarantee about everyone

Claim or chart: A fictional survey asks ten people in one queue; six choose option A. A headline says ‘60% of all residents prefer A’.

What the data actually support: The 60% describes only those ten respondents. Generalisation to all residents is not justified without appropriate sampling evidence.

First interpretation mistake: Name the observed group and its selection method.

Independent check: Do not publish a population claim when only a small convenience sample was measured.

22. A change in total can change a share even when counts rise

Claim or chart: A club grows from 40 to 100 members. Participants in an activity increase from 20 to 30. Did the activity’s participation rate rise?

What the data actually support: No: it fell from 20/40=50% to 30/100=30%, even though the participant count rose.

First interpretation mistake: Separate absolute counts from proportions.

Independent check: Compute both numerators relative to their own period totals.

23. A price graph can hide a fixed starting charge

Claim or chart: A fictional delivery graph follows C=5+2d. Someone says cost is directly proportional to distance because the graph is straight.

What the data actually support: Not directly proportional: at d=0, cost is $5, not zero. The relationship is linear with a fixed fee.

First interpretation mistake: Check the y-intercept and the ratio C/d.

Independent check: At 1 km cost $7; at 2 km cost $9, not $14.

24. A line can have a zero gradient at a nonzero height

Claim or chart: A graph shows y=6 for all x. A pupil says the y-value is zero because the gradient is zero.

What the data actually support: The gradient is zero but the line stays at height y=6.

First interpretation mistake: Distinguish height from change per x-unit.

Independent check: The intercept is six even though the line is horizontal.

25. An extrapolation is a model assumption

Claim or chart: A line fits measurements between zero and five hours. A headline predicts the same rule must continue for one hundred hours.

What the data actually support: The extrapolated value can be calculated, but the observation range does not prove the underlying relationship remains valid that long.

First interpretation mistake: Use words such as ‘the model predicts if the rule continues’.

Independent check: Check whether physical limits, caps or changed conditions could invalidate the trend.

26. Correlation does not by itself establish cause

Claim or chart: A fictional data graphic shows umbrella sales and wet weather occurring together. Does umbrella purchasing cause rain?

What the data actually support: No. A correlation does not, by itself, establish that one measured variable causes the other.

First interpretation mistake: Look for other influences and study design before interpreting causality.

Independent check: Describe the association without inventing a causal direction.

27. One time series can be seasonally cyclical

Claim or chart: A fictional park attendance graph peaks each school holiday. Can one holiday peak prove a permanently increasing trend?

What the data actually support: No. A repeating seasonal pattern and a long-term trend are different ideas. Compare equivalent periods over enough time.

First interpretation mistake: Do not treat a selected high week as proof of permanent growth.

Independent check: Read the time span and note comparable weeks before making a claim.

28. An average can be weighted by group size

Claim or chart: Group A has two values each averaging 10; Group B has eight values averaging 20. Can the combined mean be 15?

What the data actually support: No. Combined total is 2×10+8×20=180 over ten values, giving 18.

First interpretation mistake: A simple average of 10 and 20 gives 15 but ignores unequal group sizes.

Independent check: Check the combined mean lies closer to twenty because Group B is larger.

29. A probability estimate and a guarantee are different

Claim or chart: A fair coin has P(heads)=1/2. Will exactly five heads occur in ten tosses?

What the data actually support: Not guaranteed. Five heads is one possible count; random variation means other counts may occur.

First interpretation mistake: Distinguish an expected average proportion from a certain observed tally.

Independent check: A fair model does not remember earlier tosses in independent trials.

30. A graph can use averages with different reference groups

Claim or chart: Two schools show average score increases, but one reports a class mean and the other a whole-year mean. Can they be compared directly?

What the data actually support: Not without knowing whether the underlying groups, topic difficulty and measurement rules are comparable.

First interpretation mistake: Ask what was averaged and which population produced each mean.

Independent check: Make a like-for-like comparison before describing an advantage.

31. The chart title can omit a time window

Claim or chart: A bar graph says ‘Daily visitors’ but provides no date or time period. Can the result support ‘visitors this week’?

What the data actually support: Not without information about which day or days were counted.

First interpretation mistake: Check source date and the definition of daily.

Independent check: Do not silently change the period from a day to a week.

32. Stacked bars contain parts as well as totals

Claim or chart: A stacked bar has sections of 5, 8 and 7. A pupil reads only the tallest segment as its total.

What the data actually support: The complete bar total is 5+8+7=20. The largest individual section is eight.

First interpretation mistake: Know whether the question asks for one category or the combined stack.

Independent check: Count all included segments and inspect the legend.

33. A repeated data point is not a mistake by default

Claim or chart: A dot diagram shows the value four three times. A pupil deletes two dots as duplicates.

What the data actually support: Repeated observations can be genuine and must be retained unless a documented data-cleaning reason says otherwise.

First interpretation mistake: A dot diagram shows frequency through repetition.

Independent check: Deleting valid duplicates changes mean, mode and sample size.

34. A numerical graph can be technically correct but unsuitable

Claim or chart: A very detailed line graph is used to compare favourite school subjects, a categorical variable.

What the data actually support: A bar chart is generally clearer for distinct categories. A connected line can imply an artificial continuity or order.

First interpretation mistake: Match display to the structure of the variable.

Independent check: Check whether adjacent x-axis categories represent meaningful numerical progression.

35. Two percentages need one stated baseline to compare

Claim or chart: A fictional price goes from $50 to $60 then back to $50. Someone says each move is 20%.

What the data actually support: The rise is (60−50)/50=20%, but the fall is (60−50)/60≈16.67%. They use different starting values.

First interpretation mistake: Treating an equal dollar change as an equal relative percentage change.

Independent check: State the starting value of each comparison.

36. A rounded number does not reveal the original exactly

Claim or chart: A report says the mean is 7.3 rounded to one decimal place. Does that prove the unrounded mean equals exactly 7.3?

What the data actually support: No. Under conventional rounding it could represent values from about 7.25 up to but not including 7.35, subject to tie rules.

First interpretation mistake: Treat a rounded display as exact hidden data.

Independent check: State the reported precision when comparing measurements.

37. A graph’s units can change a comparison

Claim or chart: One chart uses metres and another centimetres. Both display 50 as a value. Are they equal lengths?

What the data actually support: No. 50 m equals 5000 cm, while 50 cm is 0.5 m. A common unit is required.

First interpretation mistake: Bare numerals do not determine quantities without units.

Independent check: Convert before calculating differences or ratios.

Teach pupils to rewrite a misleading claim accurately

Use a simple check line below each answer

Numerical check: calculate the difference, ratio or percentage with its correct denominator.
Source check: state the sample size, time window and categories actually shown.
Representation check: name whether the figure shows an individual value, an interval, a count, a rate or a summary.
Claim check: distinguish ‘the data show’ from ‘the model suggests’ or ‘we would need more evidence’.

A five-day family data-literacy practice plan

Day 1: read axes, units and tick spacing from two graphs in the child’s current textbook.
Day 2: compute a percentage from a fictional survey count with a clearly stated sample size.
Day 3: compare mean and median in a five-value dataset containing one unusually large value.
Day 4: make a dot plot or equal-width frequency histogram and identify what exact information each display preserves.
Day 5: rewrite one overconfident chart claim in a precise sentence that names the measured group and limitation.

How a small-group G1 Mathematics lesson can teach this

First 10 minutes: retrieve fractions, percentages and one straightforward axis-reading item.
Next 20 minutes: compare accurate and misleading displays of the same fictional values.
Next 25 minutes: individually diagnose a scale, denominator or summary-statistic error.
Next 15 minutes: translate a numerical table into a graph and discuss what information grouping removes.
Next 10 minutes: write accurate interpretations of unfamiliar data without help.
Final 10 minutes: correct one overstated conclusion and plan a later independent retest.

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Yes. An axis that begins far above zero can emphasise small differences, and unclear units can encourage mistaken comparisons. Inspect the numerical labels and context before interpreting appearance.

Is median always better than mean?

No. They answer different questions. The mean uses every value, while the median describes the central position. Select the summary appropriate to the dataset and purpose.

What is a modal class?

It is the interval with the largest frequency in a grouped display. It generally does not tell you an exact most-common individual value.

Can a graph prove that one thing caused another?

A graph alone generally cannot establish causation. The way data were gathered, possible other influences and relevant evidence must be considered.

Are all 37 examples required for Secondary 2 G1?

No. The article includes G1-aligned graph reading and broader critical interpretation. The school’s current curriculum and actual question instructions take priority.

How will a tutor know my child has improved?

Have the learner read a different graph, calculate with the right denominator and rewrite a new claim accurately without copying a model explanation. Retest after a delay.

Secondary 2 G1 Mathematics Tutor Bukit Timah | Spot Misleading Graphs and Data Claims

Official sources and accuracy note

Official sources and accuracy note

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Is median always better than mean?

What is a modal class?

Can a graph prove that one thing caused another?

Are all 37 examples required for Secondary 2 G1?

How will a tutor know my child has improved?

Related Secondary 2 G1 Mathematics guides

Official sources and accuracy note

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Is median always better than mean?

What is a modal class?

Can a graph prove that one thing caused another?

Are all 37 examples required for Secondary 2 G1?

How will a tutor know my child has improved?

Related Secondary 2 G1 Mathematics guides

Official sources and accuracy note

How a small-group G1 Mathematics lesson can teach this

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Is median always better than mean?

What is a modal class?

Can a graph prove that one thing caused another?

Are all 37 examples required for Secondary 2 G1?

How will a tutor know my child has improved?

Related Secondary 2 G1 Mathematics guides

Official sources and accuracy note

How a small-group G1 Mathematics lesson can teach this

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Is median always better than mean?

What is a modal class?

Can a graph prove that one thing caused another?

Are all 37 examples required for Secondary 2 G1?

How will a tutor know my child has improved?

Related Secondary 2 G1 Mathematics guides

Official sources and accuracy note

A five-day family data-literacy practice plan

How a small-group G1 Mathematics lesson can teach this

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Is median always better than mean?

What is a modal class?

Can a graph prove that one thing caused another?

Are all 37 examples required for Secondary 2 G1?

How will a tutor know my child has improved?

Related Secondary 2 G1 Mathematics guides

Official sources and accuracy note

A five-day family data-literacy practice plan

How a small-group G1 Mathematics lesson can teach this

G1 subject level and published SEC syllabus

Frequently asked questions

Can a graph be numerically correct but visually misleading?

Is median always better than mean?

What is a modal class?

Can a graph prove that one thing caused another?

Are all 37 examples required for Secondary 2 G1?

How will a tutor know my child has improved?

Related Secondary 2 G1 Mathematics guides

Official sources and accuracy note

A one-minute data-reading diagnostic

How to move from a picture to a defensible claim

Start by reading the axes and legend

Ask which whole is being described

Know what an average keeps and loses

Separate observation, model and conclusion

37 original graph and data-claim clinics

1. An axis begins at 90 instead of zero

2. One square means five units

3. A line’s upward movement is mistaken for total increase

4. Two graph axes use different units