You have made line plots of measurements. Statistics starts one step earlier, with the question you ask.
Key idea
A statistical question expects answers that vary, so you collect many pieces of data to answer it. "How many hours do the students in my class sleep?" is statistical. "How many hours did I sleep last night?" is not.
Why it works
If every answer would be the same, one look settles the question; there is nothing to summarize. Statistics is about data that vary: it asks what is typical, how spread out the answers are, and what shape they make. A question that expects variety is what makes those ideas useful.
| Question | Statistical? | Why |
|---|---|---|
| How old am I? | No | one answer |
| How old are the people at the library today? | Yes | many ages that vary |
| How tall is our school’s flagpole? | No | one measurement |
| How tall are the sunflowers in the garden? | Yes | heights vary from plant to plant |
Worked example: Is it statistical?
Is "How many pets do the families on our street have?" a statistical question?
- It asks every family, and different families have different numbers of pets.
- The answers will vary.
Answer: Yes.
Worked example: Make it statistical
Change "How far is my walk to school?" into a statistical question.
- The question has one answer because it is about one person.
- Ask about a group instead, where the answers will differ.
Answer: For example: "How far do the students in my class walk to school?"
Common mistake
Calling a question statistical because it has a number in its answer. "How old am I?" is answered with a number, but only one number. Quick check: imagine collecting the answers. Would you get a list of different values? If not, it is not statistical.