Writing unbiased survey questions

Name:

Date:

Class:

Quick Questions

Shade in the A or B circle for each question.

1. Which survey question is unbiased?

A How would you rate the school canteen's food?

B How amazing is the school canteen's delicious food?

2. Which question asks about ONE thing only?

A How easy is the app to navigate and how good does it look?

B How easy is the app to navigate?

3. A survey asks: 'Don't you agree that homework should be banned?' What type of problem is this?

A A neutral question with balanced wording

B A leading question that pushes people to agree

4. Which set of answer choices covers every honest response about how often students use a library?

A Never / Rarely / Sometimes / Often / Daily

B Often / Very often / Almost every day

5. Which rating scale is better for measuring satisfaction with a new game?

A 1 = Very unsatisfied to 5 = Very satisfied, with a neutral middle

B 1 = Satisfied, 2 = Very satisfied, 3 = Extremely satisfied

Match the Pairs

Draw a line connecting each item on the left to its match on the right.

Left column		Right column
1Don't you agree our canteen food is awful?		AThe wrong group is surveyed. Only a small, unrepresentative group is asked, which biases the results.
2How happy are you with the canteen prices and food quality?		BThe scale is skewed positive and has no negative or neutral option, so some honest responses are missing.
3A balanced rating scale gives choices at both ends and a middle, so every honest response fits.		C1 Very hard, 2 Hard, 3 OK, 4 Easy, 5 Very easy
4Choosing who to survey: ask people who will actually use the solution, not just friends.		DA double-barrelled question: it asks about two things at once, so one answer can't be accurate for both.
5Rate our app: Excellent / Great / Good		ESurvey: 40 Year 9 and 10 students who catch the school bus
6Survey only the chess club about the new sports timetable		FA leading question: its wording pushes the person toward a negative answer instead of staying neutral.

Definitions

unbiased survey question

Fill in the Blanks

Complete each line of code by filling in the missing parts.

Q: Don't you agree our app is amazing? Type: _______ question

Q: Is the app fast and easy to use? Fix: ask about ___ thing per question

Rate: The app is easy to use 1 = Strongly disagree 5 = Strongly _____

Q: How do you travel to school? Options: Bus, Train, Walk, Car, _____

Q: Who should we ask about the new app? A: a random ______ of all our users

Spot the Error

Shade in the letter of the line that contains an error.

Question 1

A Survey: School canteen

B Q1: Don't you love the food?

C 1 Poor 2 OK 3 Good 4 Great

Question 2

A Survey: School app

B Q1: How often do you use it?

C Q2: Is the app quick and fun?

Question 3

A Q1: How hard was PE today?

B 1 Easy 2 Easier 3 Easiest

C Tick one number.

Question 4

A Goal: Improve the bus service

B Users: students who ride it

C Sample: mixed year levels

D Ask: 50 students who walk

Question 5

A Q1: How amazing was the play?

B 1 Very poor 3 OK 5 Very good

C Q2: Did you go on Friday?

D Answer: Yes / No

Answer Key

Writing unbiased survey questions

Section A: Quick Questions

A — How would you rate the school canteen's food?

B — How easy is the app to navigate?

B — A leading question that pushes people to agree

A — Never / Rarely / Sometimes / Often / Daily

A — 1 = Very unsatisfied to 5 = Very satisfied, with a neutral middle

Section B: Match the Pairs

Don't you agree our canteen food is awful? → A leading question: its wording pushes the person toward a negative answer instead of staying neutral.

How happy are you with the canteen prices and food quality? → A double-barrelled question: it asks about two things at once, so one answer can't be accurate for both.

A balanced rating scale gives choices at both ends and a middle, so every honest response fits. → 1 Very hard, 2 Hard, 3 OK, 4 Easy, 5 Very easy

Choosing who to survey: ask people who will actually use the solution, not just friends. → Survey: 40 Year 9 and 10 students who catch the school bus

Rate our app: Excellent / Great / Good → The scale is skewed positive and has no negative or neutral option, so some honest responses are missing.

Survey only the chess club about the new sports timetable → The wrong group is surveyed. Only a small, unrepresentative group is asked, which biases the results.

Section C: Definitions

unbiased survey question — A survey question that is neutral, asks about one thing only, and gives answer choices that cover every honest response, so it does not push people toward a particular answer.

Section D: Fill in the Blanks

Q: Don't you agree our app is amazing?
Type: leading question

Q: Is the app fast and easy to use?
Fix: ask about one thing per question

Rate: The app is easy to use
1 = Strongly disagree
5 = Strongly agree

Q: How do you travel to school?
Options: Bus, Train, Walk, Car, Other

Q: Who should we ask about the new app?
A: a random sample of all our users

Section E: Spot the Error

B — This is a leading question. It pushes people to say they love the food. It should be neutral, like 'How do you rate the food?'.

C — This is double-barrelled. It asks about speed and fun in one question, so one answer can't cover both. Ask about one thing only, such as 'Is the app quick to load?'.

B — This rating scale only has 'easy' options. A student who found PE hard has no honest choice. The scale needs to run from very easy to very hard with a middle option.

D — The goal is about bus riders, so asking students who walk gives the wrong sample. You should survey the people who actually use the bus.

A — The word 'amazing' is a loaded word that pushes people toward a positive answer. It should be neutral: 'How was the school play?'.