Graph
Axis scale? Broken/truncated? Nonlinear? Units?
Learning goals: what you will learn
- explain the central idea: Evaluate statistical claims in media by examining evidence, sampling, representation, bias, inference and ethical issues.
- choose and apply an appropriate method without relying on keyword matching
- check results using units, substitution, estimation, a second representation or contextual reasonableness
- justify a conclusion and communicate limitations where the context requires them
Prerequisite knowledge
Revise percentages, averages, data displays, sampling ideas and basic interpretation of graphs and tables.
Concept teaching
Critique statistics by separating the claim from the evidence. Check sampling, variable definitions, missing context, axis choices, rates versus counts, uncertainty, causal language and ethical implications.
Sample
Who was included/excluded? Size? Selection method? Representative?
Claim
Does the evidence support association, prediction or causation?
Ethics
Privacy, fairness, data sovereignty, stigmatising framing and transparent uncertainty.
Worked examples
1. Truncated axis
Bars 98 and 100 look dramatically different if axis starts at 97. The numbers differ by only about 2%; report scale before interpreting visual impact.
2. Non-representative sample
A school polls only students in an elite sports program about weekly exercise. The sample is unlikely to represent all students.
3. Self-selection bias
An online voluntary poll may overrepresent people with strong opinions; a large sample does not automatically remove selection bias.
4. Counts versus rates
100 cases in a city of 10,000 is 1%; 200 cases in a city of 100,000 is 0.2%. Raw count is higher in second city but population rate is lower.
5. Association not causation
If screen time and low sleep are associated, the report cannot claim screen time caused low sleep without addressing design/confounding factors.
6. Public-health prediction
If side-effect rate in a suitable sample is 2%, estimating 200 cases among 10,000 assumes the sample rate transfers to that population; state that assumption and uncertainty.
7. Misaligned evidence
A report claims “students learn better” but presents only attendance data. Attendance may be relevant but does not directly measure learning outcome.
8. Algorithmic bias
If training data underrepresents one group, an AI model may have lower accuracy for that group. Overall accuracy can hide subgroup performance differences.
9. Ethical rate framing
Presenting infection rates by a small community without context can stigmatise. Use appropriate denominators, uncertainty and careful language.
10. Indigenous data sovereignty
When analysing reports about First Nations peoples, critique who controls collection, interpretation, access and use of data—not only the arithmetic. Indigenous data sovereignty is a substantive ethical/statistical consideration.
Common misconceptions and corrections
- Assuming a large sample is automatically representative.
- Equating association with causation.
- Ignoring denominator/population size.
- Criticising a truncated axis as always wrong; it can be useful if clearly labelled and interpreted honestly.
- Treating ethics as separate from statistical validity/reporting.
Guided practice
A headline says 'Study app users score 25% higher'. List the minimum information needed before accepting the claim: sample, comparison group, baseline, measure, graph/summary and possible confounders. Then rewrite the claim cautiously.
Teacher check: require a written method choice and one verification step before revealing the worked solution.
Independent practice
- Why can an axis starting at 95 exaggerate a 98-to-100 difference?
- Identify bias in surveying only gym members about exercise.
- Can a 50,000-person voluntary web poll still be biased? Why?
- Compare 40/1000 with 100/5000 as percentages.
- What extra evidence is needed before an association can be called causal?
- State an assumption in projecting a sample infection rate to a population.
- Why is attendance data insufficient to prove learning improvement?
- How can overall AI accuracy hide group bias?
- Name one ethical issue in publishing small-community health statistics.
- What does Indigenous data sovereignty add to a statistical critique?
Check answers and explanations
- The shortened scale makes a small absolute difference occupy a large visual height.
- Selection bias/non-representative sample.
- Yes; self-selection can remain severe regardless of sample size.
- 4% versus 2%.
- A design that addresses confounding and supports causal inference, not scatter alone.
- The sample is sufficiently representative/relevant to the target population.
- It measures attendance, not learning achievement directly.
- High performance for majority groups can dominate the aggregate metric.
- Privacy, stigma, small numbers or misleading rates.
- It asks who has authority/control over data and its interpretation/use.
Reasoning and problem-solving task
Find or invent a plausible media claim such as 'students using App X improve marks by 40%'. Write a statistical audit covering sample selection, comparison group, measurement, absolute versus relative change, causation, missing information and ethical considerations.
Important questions and answers
- Why can an axis starting at 95 exaggerate a 98-to-100 difference?
- The shortened scale makes a small absolute difference occupy a large visual height.
- Identify bias in surveying only gym members about exercise.
- Selection bias/non-representative sample.
- Can a 50,000-person voluntary web poll still be biased? Why?
- Yes; self-selection can remain severe regardless of sample size.
Assessment-style questions
A media report claims a treatment 'cuts risk by 50%' because risk fell from 2 in 1000 to 1 in 1000. Explain relative and absolute change, identify two questions about study design and write a more informative headline. [6 marks]
Marking focus: method selection, mathematically correct working, interpretation and justification.
Review hints
- ask who is missing
- check the denominator
- inspect axes and units
- separate association from causal language
Exit ticket: mastery check
- I can explain the concept without copying a formula sheet.
- I can solve a routine example and check the result.
- I can choose a method in an unfamiliar problem.
- I can explain a common error and correct it.
- I can connect the answer back to the context, including units or limitations.
Teacher and parent guidance
For teachers
Use real headlines but grade the reasoning structure, not students' opinions about the topic. Require evidence-specific language such as 'selection bias because…'.
For parents and carers
When you see a statistic in the news, ask together: 'Compared with what?', 'Out of how many?', and 'Who was included?'. Those three questions expose many weak claims.
Curriculum alignment
Australian Curriculum: AC9M10ST01 — Year 10 Statistics. Evaluate statistical claims in media by examining evidence, sampling, representation, bias, inference and ethical issues.
Victoria: VC2M10ST04 — Level 10 Statistics
NSW: Stage 5 Core — Data classification, visualisation and analysis; Path — Data analysis and statistical enquiry
Alignment explanation: The explicit teaching and worked examples address the Australian Curriculum concept directly. The Victorian mapping follows the current Version 2.0 descriptor structure; where Victoria combines or extends content, that difference is stated rather than hidden. NSW uses a Stage 5 Core–Paths structure, so this page maps to the relevant content group(s) and Working mathematically processes instead of inventing a Year 10 one-to-one code.
| Lesson component | Australian Curriculum | Victoria | NSW |
|---|---|---|---|
| Explicit concept teaching and worked examples | AC9M10ST01 | VC2M10ST04 — Level 10 Statistics | Stage 5 Core — Data classification, visualisation and analysis; Path — Data analysis and statistical enquiry |
| Guided and independent practice | Builds fluency and application for AC9M10ST01 | Practises the mapped Level 10/10A knowledge as applicable | Practises the mapped Stage 5 Core/Path content |
| Reasoning and assessment tasks | Applies reasoning/problem solving in the descriptor context | Supports Victorian reasoning and modelling expectations | Embeds Working mathematically: reasoning, problem solving and communication |
Australian Curriculum elaborations
AC9M10ST01: analyse claims, inferences and conclusions of statistical reports in the media, including ethical considerations and potential sources of bias.
- E1: identify misleading representations/irrelevant or non-representative data and biased use. Examples 1,7.
- E2: investigate source/sample size and representativeness. Examples 2–3.
- E3: investigate population rates and ethical presentation of infection/cases per population. Examples 4,9.
- E4: use secondary data for health predictions and discuss ethics, validity and sample size. Example 6.
- E5: recognise bias in machine/deep learning and effects on fairness/accuracy/ethics. Example 8.
- E6: use Indigenous data sovereignty to critique/evaluate “Closing the Gap” reporting. Example 10.
Practice and teaching resources
Official curriculum references
- Australian Curriculum Version 9 — ACARA
- Victorian Curriculum Mathematics Version 2.0 — VCAA
- NSW Mathematics K–10 Syllabus (2022) — NSW Curriculum/NESA
Official wording is paraphrased on SkillrHub; use the linked curriculum sites as the source of record.
Other curriculum comparisons retained
Closest US CCSS HSS-IC.B.6 and HSS-ID standards; UK GCSE statistics/data interpretation; NSW Stage 5 and Victorian Level 10 statistical investigation/critique; comparable Canadian/NZ statistical literacy.
🎥 Optional Video Lesson
The SkillrHub lesson remains the primary learning resource. This optional video reinforces the explanation; you can complete the lesson and practice without watching.
Before you watch:
- Pause after each worked example.
- Try the examples yourself.
- Return to the SkillrHub lesson before continuing.
Recommended: How statistics can be misleading - Mark Liddell
TED-Ed — How grouping data can reverse an apparent statistical relationship.
As you watch: Why might an overall percentage tell a different story from the subgroup percentages?
Load video player Loads YouTube in this lesson. See the video notice below.
Try it: Find a statistical claim and list the groups, sample sizes and missing context you would check before accepting it.
Video unavailable, inaccurate or unsuitable for this year? Report a video problem to SkillrHub by email. You can continue with the written lesson and practice resources.
About these videos
Videos are curated from trusted independent educational creators and played through YouTube. Rights remain with their respective owners. Inclusion does not imply that a creator or YouTube endorses SkillrHub.
YouTube’s terms and privacy policy apply to its player. Advertising, recommendations and external links may appear, and videos may change or become unavailable. SkillrHub’s written lessons and practice resources remain available separately.
To report a content, suitability or rights concern, email skillrhublearning@gmail.com with the lesson code and video link. Please do not include personal student information.
Curriculum equivalents: Victoria, NSW and international
Curriculum equivalents for Analyse claims, inferences and conclusions of statistical reports in the...
Mapped skill: analyse claims, inferences and conclusions of statistical reports in the media, including ethical considerations and identification of potential sources of bias
These references identify matching or closely related learning. Curriculum sequence, terminology and depth vary, so teachers should use the mapped skill and lesson difficulty to confirm suitability.
| Region | Curriculum framework | Closest level or code |
|---|---|---|
| Australia | Australian Curriculum v9.0 | AC9M10ST01 · Year 10 |
| Victoria | Victorian Curriculum F–10 Version 2.0 — Mathematics | VC2M10ST04 · Level 10 |
| New South Wales | NSW Mathematics K–10 Syllabus (2022) | MA5-DAT-P-01 · Stage 5 |
| United States (USA) | Common Core State Standards for Mathematics | Grades 9–10 band |
| Canada (Ontario) | Ontario Curriculum — Mathematics | Grade 10 |
| United Kingdom (England) | National Curriculum in England — Mathematics | Year 11, Key Stage 4 |
| India | NCERT / CBSE — Mathematics | Class 10 |
Australian Curriculum v9.0 is the canonical source for this SkillrHub lesson. Victoria and NSW entries name the closest published state codes or outcomes; international entries are planning references rather than claims of identical curricula.
Help improve SkillrHub
Questions or feedback?
Ask about this lesson, suggest an improvement or report an error. Facebook opens only when you choose an option below.
Topic reference: AC9M10ST01 — Statistical Reports, Bias and Misleading Claims — AC9M10ST01
Privacy: Please don’t share personal student or school information. Younger students should ask a parent, guardian or teacher to post on their behalf.