A questionnaire turns opinions into a number
A satisfaction questionnaire is a short list of statements that a person rates, so that how they feel becomes a number you can compare. It matters because times and errors show what happened, but not whether people were happy.
A ten-item usability questionnaire gives one score from 0 to 100 for each person.
By the end of this lesson
Note
I can score a ten-item questionnaire and say what its mean and spread show.
Key words
| Word | What it means |
|---|---|
| Likert scale | an answer from strongly disagree (1) to strongly agree (5) |
| Item | one statement on the questionnaire |
| Reversed item | a negative statement, so a high answer is bad |
| Mean | the total of the scores divided by how many there are |
| Spread | how far apart the scores are |
The cells students fill in are left blank in this preview.
Answers are numbers from 1 to 5
On a Likert scale a person rates a statement from 1 (strongly disagree) to 5 (strongly agree). Each statement is an item.
The ten items of a usability questionnaire alternate. The odd ones are positive, such as "I found it easy to use". The even ones are negative, such as "I found it more complicated than it needed to be".
A negative item is reversed before you add
A high answer to a positive item is good. A high answer to a negative item is bad. You cannot add them as they are.
So each item is turned into points from 0 to 4. A positive item gives the answer minus 1. A negative, or reversed, item gives 5 minus the answer. Then 4 always means the best.
Worked example: scoring one person's ten answers
| Item | Kind | Answer | Rule | Points |
|---|---|---|---|---|
| 1 | Positive | 4 | 4 - 1 | 3 |
| 2 | Negative | 1 | 5 - 1 | 4 |
| 3 | Positive | 5 | 5 - 1 | 4 |
| 4 | Negative | 2 | 5 - 2 | 3 |
| 5 | Positive | 4 | 4 - 1 | 3 |
| 6 | Negative | 2 | 5 - 2 | 3 |
| 7 | Positive | 3 | 3 - 1 | 2 |
| 8 | Negative | 2 | 5 - 2 | 3 |
| 9 | Positive | 4 | 4 - 1 | 3 |
| 10 | Negative | 1 | 5 - 1 | 4 |
| Total | 32, times 2.5 is 80 |
The cells students fill in are left blank in this preview.
Score the twelve people in the lab
docker exec jupyter python3 steps/sus.pyShown here, not run. Create a free account to press Run and type it into your own lab’s terminal.
What the lab prints for the twelve people
| Person | Score | Person | Score |
|---|---|---|---|
| P1 | 80.0 | P7 | 42.5 |
| P2 | 62.5 | P8 | 32.5 |
| P3 | 40.0 | P9 | 62.5 |
| P4 | 50.0 | P10 | 37.5 |
| P5 | 47.5 | P11 | 80.0 |
| P6 | 77.5 | P12 | 47.5 |
The cells students fill in are left blank in this preview.
Mean and spread say more than one score
The script ends with the line "mean score: 55.0". That is the total of the twelve scores divided by twelve.
The spread says how far apart the scores are. Here they run from 32.5 to 80.0, with a standard deviation of about 17. People disagreed a lot, so the mean alone hides that some people liked the shop.
A score near 68 reads as average
A score has no meaning alone. It needs something to compare with. For this ten-item questionnaire, scores around 68 are usually read as average across many products.
A mean of 55.0 is well under that, so people found this shop harder than most. Compare the same product before and after a change, and one product against another.
A common mistake
Warning
Many people think a score of 55 means 55 per cent of people were satisfied. It does not. The score runs from 0 to 100, but it is not a percentage of anything. Read it against the usual average of about 68, and against the same product earlier.
Going further
Write ten statements about a page you know, alternating positive and negative. Ask five classmates to rate them from 1 to 5.
Score each person with your function and find the mean and the spread. Do the negative items change anyone's score a lot?