Ready-made answer sets
Six common scales, lowest point first. Copy puts one answer per line.
Agreement
"The checkout was easy to use."
- Strongly disagree
- Disagree
- Neither agree nor disagree
- Agree
- Strongly agree
Frequency
"How often do you use the mobile app?"
- Never
- Rarely
- Sometimes
- Often
- Always
Satisfaction
"How satisfied are you with our support?"
- Very dissatisfied
- Dissatisfied
- Neither satisfied nor dissatisfied
- Satisfied
- Very satisfied
Importance
"How important is price when you choose a plan?"
- Not at all important
- Slightly important
- Moderately important
- Very important
- Extremely important
Likelihood
"How likely are you to renew next year?"
- Very unlikely
- Unlikely
- Neither likely nor unlikely
- Likely
- Very likely
Quality
"How would you rate the onboarding call?"
- Very poor
- Poor
- Fair
- Good
- Excellent
What a Likert scale is
Each Likert question (an "item") is a statement or question with a fixed set of ordered, labeled answers, from one extreme to the other. The psychologist Rensis Likert introduced it in 1932 to measure attitudes, and it is still the most common question format in surveys.
Good items are about one thing, use plain words and give every point a label. "The checkout was fast and easy" is two questions; split it into "fast" and "easy".
5-point or 7-point?
Use 5 points for most surveys, and always on phones. Every label is easy to tell apart, and people answer faster.
Use 7 points when you need finer detail, for example to see small shifts between two survey rounds, and your audience is used to rating things carefully.
Whichever you choose, keep it the same across the survey and across rounds. A result on a 5-point scale can't be compared directly with one on a 7-point scale.
How to analyze the answers
Likert answers are ordered, but the gaps between them aren't equal: the step from "Disagree" to "Neutral" isn't the same size as "Agree" to "Strongly agree". So don't average blindly. Report these instead:
- Top-2-box: the share who picked the two most positive answers. "58% agree" is easy to read and easy to track over time.
- The median: the middle answer when everyone's answers are lined up.
- The full spread: a stacked bar shows whether people agree, split or don't care.
Try it with your own counts:
Agree (top-2-box)
58%
of 100 people picked Agree or Strongly agree. The median answer is Agree.
- Disagree
- 20.0%
- Neutral
- 22.0%
- Agree
- 58.0%
Common mistakes
- Unbalanced answers. "Poor, Good, Very good, Excellent" has three positive points and one negative, which pushes results up. Keep the same number on each side.
- Double-barreled items. "Support was quick and helpful" can't be answered if it was quick but not helpful.
- Leading wording. "How much did you enjoy the new design?" assumes they did. Ask "How would you rate the new design?"
- Labels only at the ends. Unlabeled middle points mean different things to different people. Label every point.
- Flipping the order between questions. If positive is on the right in one question, keep it there in all of them.
Questions about Likert scales
- Should a Likert scale have a neutral middle option?
- Usually yes. Without one, people who truly have no opinion are pushed to pick a side or skip the question. Drop it only when you need people to lean one way and you know they have an opinion.
- Can I average Likert answers?
- You can, but the average hides the shape. A mean of 3.0 can come from everyone answering 3, or from half answering 1 and half answering 5. Report the share who agree (top-2-box) and the median, and show the full spread.
- Is a 1 to 10 rating a Likert scale?
- Not strictly. A Likert item has a label on every point, like 'Agree' or 'Disagree'. A 1 to 10 scale with labels only at the ends is a numeric rating scale.