Put the claim on the board
A student looks at some results and announces, “Most students at our school want a lunchtime drawing club.” Before the class decides whether that sounds right, add this heading: Made-up class survey.
In this scenario, all 30 students in one Year 6 class were invited to answer, but only 12 chose to respond. Display the results and two possible captions. That’s enough for a short discussion about whether a sentence matches the people behind the data. It also gives students practice interpreting results and communicating a finding in context, as described in the Mathematics Hub’s Year 6 statistical investigation guidance.
Count the answers you actually have
Show the made-up survey question: “Would you like a lunchtime drawing club? Choose one: Yes, No, Not sure.” Explain that a respondent is simply someone who answered the question.
Made-up results: 12 respondents from one Year 6 class
| Answer | Number of respondents |
|---|---|
| Yes | 8 |
| No | 3 |
| Not sure | 1 |
| Total | 12 |
Check the total together: 8 + 3 + 1 = 12. Although 30 students were invited, only 12 answered, so 30 − 12 = 18 students didn’t supply an answer. There’s no need to introduce percentages; the counts tell students everything they need for this comparison.
Choose the caption that fits
Ask students to underline the group named in each caption before choosing between them.
- Caption A: “Most students at our school want a lunchtime drawing club.”
- Caption B: “Most of the 12 students who answered in this class said Yes to a lunchtime drawing club.”
Caption B fits the made-up results. Half of 12 is 6, and 8 is greater than 6, so a majority of the respondents said Yes. Just as importantly, the caption names those respondents rather than a larger group.
Caption A stretches the claim from 12 volunteers in one class to the whole school. The Australian Bureau of Statistics explains that volunteer responses aren’t automatically representative of a wider population. The results don’t support caption A, but they don’t prove it false either. We simply don’t know what the whole school thinks.
Leave missing answers unknown
Someone may suggest, “The other 18 must have said No.” Return to the made-up results. Those 18 students haven’t told us what they think. The number of recorded No answers is still three.
Being invited isn’t the same as answering. The ABS describes an answer that wasn’t obtained as non-response. Even when everybody is invited, missing responses and other collection errors can affect what the results show.
Ask whose views are still missing
Now ask, “Who else would we need to hear from before making a claim about the school?” Help students distinguish between the 18 non-respondents in this class and students in other classes and year levels.
Students might suggest asking across different year levels, keeping the question consistent and recording how many people were invited as well as how many answered. Deciding who will supply the data and whether a question will collect relevant information are useful parts of planning a class survey.
More volunteers won’t automatically make the responses representative. On the other hand, a survey doesn’t always need to ask everybody: a suitably designed sample can support a wider conclusion. For this discussion, the practical point is simple—the collection plan and the final claim need to describe the same group.
Continue the data discussion
If you’d like further activities about displays, claims and investigation planning, the Year 6 Data and Statistics Unit listing includes “Choosing Displays and Checking Claims — two student pages and two answer pages” and “Data Investigation Planner — two planning pages”. The made-up drawing-club survey in this article is a separate teacher-led activity.
Repair the claim
Finish by asking students to rewrite caption A and name one missing piece of information. One clear repair is: “Eight of the 12 students who answered in this class said Yes to a lunchtime drawing club. We still need to find out what other students think.”
Students might identify the views of the 18 class non-respondents or students in other year levels as missing. Use their answers to guide the next discussion, then leave them with a question they can reuse whenever they meet a survey claim: “Does our sentence describe the people our answers actually represent?”