Skill: Formulate statistical questions

A statistical question is a question whose answers are expected to vary from person to person or item to item. To formulate one:

1. Choose a clear population or sample.
   Decide who or what you will collect data from.

2. Choose the characteristic to measure.
   This could be a measurement, a count, or a category.

3. Expect different answers.
   If everyone must give the same answer, or there is only one known answer, the question is not statistical.

Worked example

Suppose a teacher wants to learn about homework habits.

Step 1: Choose the population.
The population could be all Grade 7 students at the school.

Step 2: Choose what to measure.
The characteristic is the number of minutes each student spends on homework on a typical school night.

Step 3: Check whether answers will vary.
Students will probably spend different amounts of time. For example, their answers might be $20$ minutes, $35$ minutes, or $60$ minutes.

Step 4: Write the statistical question.

How many minutes do Grade 7 students at our school typically spend on homework on a school night?

This is a statistical question because:

• It names a clear population: Grade 7 students at the school.
• It names what will be measured: minutes spent on homework.
• It expects variability because students may give different answers.

Compare it with this question:

How many minutes did Maya spend on homework last night?

That is not a statistical question because it asks about one person and one specific event, so it has one answer rather than a collection of varying data.