How Statistics, Percentages and Averages Can Strengthen—or Distort—an Argument

Numbers can make an argument feel objective.

But numbers do not interpret themselves.

A numerical claim becomes meaningful only when students understand what was counted, compared, averaged, omitted and used as the baseline.

Quick Read

Percentages need denominators and baselines. Averages need the right measure and distribution. Growth figures need starting values. Numerical comparisons need matched groups and time periods. Strong readers ask what the number actually represents before accepting the conclusion built on it.

Number → denominator → baseline → comparison → context → conclusion.

One-Sentence Answer

Statistics strengthen an argument when the numerical measure matches the claim, and distort it when scale, baseline, grouping or context changes the apparent meaning.

A Percentage Without a Denominator Is Incomplete

Suppose a headline says:

Complaints increased by 100%.

That sounds dramatic. But 100% growth could mean complaints rose from 2 to 4, or from 2,000 to 4,000.

The percentage is mathematically correct in both cases, but the scale differs enormously.

The Developmental Route

Stage	Numerical reading is becoming
Lower Primary	Reading simple counts, comparisons and tables accurately
Middle Primary	Understanding fractions, percentages and simple averages in context
Upper Primary	Checking denominators, baselines and whether numerical evidence answers the question
Secondary	Evaluating relative versus absolute change, average selection, sample size, distribution, framing and whether numerical claims justify broader conclusions

Relative Change Can Sound Larger Than Absolute Change

Suppose participation rises from 1% to 2%.

Relative increase: 100%.
Absolute increase: 1 percentage point.

Both are correct. They answer different questions.

Strong readers ask which framing is more informative for the issue being discussed.

Percent and Percentage Points Are Different

If support rises from 40% to 50%, the increase is:

10 percentage points;
25% relative to the original 40%.

Confusing these measures can exaggerate or understate change.

Averages Can Hide Distribution

An average score of 70 can arise from very different groups.

Everyone scores close to 70.
Half score 40 and half score 100.
Most score 65 but a few very high scores raise the mean.

The same mean can conceal very different distributions.

Mean, Median and Mode Answer Different Questions

Students should know that “average” can mean different measures.

Mean: total divided by number of values.
Median: middle value when ordered.
Mode: most frequent value.

If a few extremely high incomes pull the mean upward, the median may better represent a typical person.

The correct measure depends on the claim.

The Baseline Controls the Story

“Scores improved by 15%” tells us little without the original level.

A 15% increase from 20 marks and a 15% increase from 80 marks are different in absolute terms and may have different practical significance.

Whenever a number describes change, ask: change from what?

Time Windows Can Change Apparent Trends

A writer can make growth look dramatic by choosing a low starting year, or make a decline look mild by selecting a short period.

Students should check whether the chosen time window reflects the larger trend.

Group Size Matters

If 80% of a group agrees, the interpretation changes depending on whether the group contains 5 people or 5,000.

A percentage hides sample size unless the reader looks for it.

This connects to How Students Judge Whether Evidence Is Relevant and Sufficient.

Representativeness Matters More Than Size Alone

A very large sample can still mislead if it systematically excludes important groups.

For example, an online poll about public transport may overrepresent people who follow a particular social-media account.

Students should therefore ask both:

How many people or cases were measured?
Who was actually included?

Correlation Is Not Automatically Causation

Numerical association can strengthen a causal hypothesis without proving it.

If students who sleep more also score higher, several explanations remain possible:

sleep may improve performance;
better-organised students may both sleep more and study more effectively;
another factor may affect both variables.

See How Students Track Cause and Effect Across a Text.

Large Numbers Can Still Be Irrelevant

A writer may use a precise statistic that does not answer the actual claim.

If the argument concerns whether a programme improves learning, the number of students who downloaded the app may show reach, not educational effectiveness.

Precision does not create relevance.

Numbers Can Be Technically True but Selectively Framed

A report may say:

Nine out of ten participants improved.

Questions remain:

Improved by how much?
Compared with what?
How many participants began?
Did anyone drop out?
How was improvement measured?

A true statement can still leave out information essential to interpretation.

Graphs and Axes Can Change Perception

A graph whose vertical axis begins at 95 instead of 0 may make a small difference look enormous.

That does not automatically make the graph dishonest; truncated axes can help reveal small variation. But the reader should know how the scale affects visual impression.

Source Credibility Still Matters

Students should ask who collected the data, how it was measured and whether the source has an incentive to select favourable statistics.

See How Students Judge Source Credibility, Reliability and Bias.

Numerical Claims Still Depend on Definitions

If a report says “90% of students succeeded”, the reader needs to know what counted as success.

A pass? A grade improvement? Completion? Satisfaction?

Numbers become interpretable only after the category being counted is defined.

Statistical Language Can Manipulate Certainty

Words such as dramatically, significantly, nearly all and only add evaluation around numbers.

Students should separate the measured result from the language used to frame it.

This connects to connotation and modality.

A Worked Example

Claim: “Our new reading programme doubled student participation.”

Suppose participation rose from 4% to 8%.

The doubling claim is correct. But a careful reader should also notice:

the absolute increase is 4 percentage points;
92% still did not participate under this measure;
the word participation needs definition;
the comparison period and group size matter;
the increase does not by itself prove the programme caused the change.

A more informative sentence might be:

Participation increased from 4% to 8%, doubling relative to the previous term, although overall participation remained low.

The number is preserved while the reader receives the scale.

Statistics in Argumentative Writing

Students should not insert statistics simply because numbers look authoritative.

A useful numerical example should answer the claim and be explained.

Statistic → interpretation → relationship to claim → limitation.

Statistics in Comprehension

When a passage includes a number, students should ask why that particular number was chosen.

Does it demonstrate scale, comparison, trend, rarity or urgency? Does the writer provide context or rely on the number’s apparent authority?

A Diagnostic Map

Accepts a percentage without asking “of what?”: recover the denominator.
Confuses percent and percentage points: calculate both explicitly.
Uses mean as “typical” automatically: inspect distribution and median.
Sees relative growth but ignores baseline: state the starting value.
Uses large sample as proof of representativeness: ask who was included.
Infers causation from correlation: identify alternative explanations.
Uses precise but irrelevant numbers: reconnect statistic to the claim.
Reads graphs by shape only: inspect axes and scales.
Accepts “success rate” without definition: identify the operational criterion.

How We Teach Numerical Claims

We give students two headlines built from the same data and ask which one creates the stronger impression.

Then students reconstruct the underlying numbers and write a third version that preserves both accuracy and context.

This makes numerical framing visible without teaching students to distrust statistics automatically.

Why a 3-Pax English Class Helps

One student can read the headline, another can inspect the denominator and baseline, and the third can judge whether the conclusion matches the data.

The roles rotate, making numerical literacy part of language reasoning rather than a separate calculation exercise.

What Parents Can Look For

When your child reads a dramatic percentage, ask:

What were the actual numbers before and after?

That single question often reveals whether the percentage is genuinely large or merely sounds large.

Frequently Asked Questions

Are statistics objective?

Measurements can be objective while choices about what to measure, how to group data and how to present results still shape interpretation.

Is a larger percentage always more important?

No. Importance depends on the baseline, denominator, absolute scale and practical consequences.

Why is this part of English?

Informational and persuasive texts increasingly combine language with numerical evidence. Students need to interpret how numbers support, frame or overstate claims.

The Larger Idea: Numbers Need Language to Become Arguments

A statistic does not arrive with its own interpretation.

Someone chooses the comparison, the baseline, the category and the words surrounding the number.

Strong readers respect numerical evidence enough to ask exactly what it measures before allowing it to carry a conclusion.

Continue through the English Tuition Sengkang learning system.

Small Group Tutorials

How Statistics, Percentages and Averages Can Strengthen—or Distort—an Argument | English Tuition Sengkang