ACARA v9 CONTENT DESCRIPTION “identify statistically informed arguments presented in traditional and digital media; discuss and critique methods, data representations and conclusions”
Builds on: Comparing Data Sets (AC9M6ST01). Reading mode, range and shape from data leads to judging the claims others build from data — and spotting when a display or sample is being used to mislead.
Statistics can mislead
Numbers feel objective, but the way they are gathered and shown can shape, or distort, the story they tell. A graph, a headline, or a survey can all be technically accurate yet still leave a false impression. Critiquing a statistical claim means looking past the surface to ask how the data was collected, how it is being displayed, and whether the conclusion really follows. The same skills you used to read a data set fairly are now turned outward, to test the claims that fill news, advertising and social media.
The truncated axis
Where a graph's axis starts can make a small difference look enormous.
Starting the axis at 100 stretches a tiny gap into a dramatic-looking difference — a common way to mislead.
How a graph can deceive
The most common trick is the axis. A bar chart whose vertical scale starts at ninety rather than zero turns a tiny difference into a towering one, because the eye reads bar heights, not the numbers behind them. Other displays leave the scale off entirely, so no real size can be judged, or stretch one dimension to make a quantity look larger. None of this is lying with numbers; it is misleading with pictures. Checking where an axis starts, and whether a scale is even shown, is the first defence against a deceptive graph.
Spot the problem
A statistical claim can mislead through its graph, its sample, or its missing detail.
What is the problem with this? Pick A, B or C.
The sample behind the claim
Every statistic about a group comes from a sample, and the sample decides how much the claim can be trusted. A survey of three people cannot speak for a whole school: a few unusual answers would swing the result entirely. A sample also has to be representative, a fair cross-section of the group, not just one class or one club. When a claim is made, asking how many were measured and who they were often matters more than the headline figure itself, because a small or lopsided sample can produce almost any result by chance.
How big is the sample?
A conclusion is only as trustworthy as the sample it came from. Small samples mislead.
A sample of 3 is too small — a few unusual answers could distort the whole result.
The words around the numbers
Even with sound data, the words chosen to report it can spin the meaning. The plain fact that sales rose from fifty to fifty-five becomes "sales soar to record high"; sixty out of a hundred becomes "almost everyone". The figures are unchanged, but the language inflates them. A fair report describes what the numbers actually show, in proportion, while a spun one reaches for drama. Separating the data from the words wrapped around it lets you judge a claim by its evidence rather than its tone.
Same data, different headline
The very same numbers can be reported fairly or spun to grab attention.
The same fact, 60 of 100 liked it, dressed up as a dramatic headline — the words spin the data.
Questions a critic asks
Critiquing statistics well comes down to a few good questions, asked every time. Who was measured, and how many? Is the sample representative? Does the graph have a proper scale starting at a sensible point? Does the conclusion match the size of the data, or stretch beyond it? These questions do not require advanced mathematics, only the habit of pausing before accepting a claim. Asking them turns a passive reader, who takes a headline at face value, into an active one who tests whether the evidence really supports it.
Questions to ask
A good critic of statistics asks who was measured, how, and what is left out.
Which question best tests the claim? Pick A, B or C.
Does the conclusion follow?
The final test of a claim is whether its conclusion truly follows from its data. A result from one class does not justify a statement about a whole school; a single month of rising sales does not prove they will rise forever. A fair conclusion stays within what the data supports, matching the strength of the claim to the strength of the evidence. Overreaching, drawing a big conclusion from thin data, is one of the most common faults in statistical arguments, and catching it is the heart of thinking critically about numbers.
Does the conclusion follow?
A conclusion is only fair if the data actually supports it, without overreaching.
Does the conclusion fairly follow from the data? Decide fair or unfair.
Becoming a careful reader
The world is full of statistical claims, and the goal of this unit is not suspicion of all numbers but careful reading of them. Knowing how a graph can deceive, why sample size and representativeness matter, how words can spin figures, and when a conclusion overreaches gives you the tools to weigh evidence for yourself. These habits carry straight into later study and everyday life, where being able to question a claim, fairly and on the evidence, is one of the most useful things statistics can teach.
Quick self-check
1. A bar chart starts its vertical axis at 90 instead of 0. What is the effect?
2. A claim about a whole school is based on asking just 3 students. The main problem is...
3. A headline reads "sales soar to record high" when sales rose from 50 to 55. This headline is...
4. When you see a statistical claim in the media, a good first question is...
5. A survey of one football club concludes what the whole town thinks. This is unfair because...