What Is a Likert Scale?
A Likert scale is a survey question that asks people to place themselves on a range of ordered answers, most often five, such as from "strongly disagree" to "strongly agree". It turns an opinion into a number you can count and compare.
A Likert scale is the workhorse of survey research. It is how most surveys measure agreement, satisfaction, likelihood and importance. It is one of about a dozen types of survey questions, and the one you will use most.
The scale is named after Rensis Likert, the psychologist who published it in 1932. Strictly, a Likert scale is a set of several statements whose answers are added up. In everyday use the term covers any single question with ordered answer labels, and that is how this article uses it.
Below are scales you can copy as they are, followed by how to choose between them and how to report the results.
Key point: pick the scale that matches what you are measuring. Do not force everything into "agree or disagree".
5-Point Likert Scale Examples
Five points is the standard. It is quick to answer, works on a phone, and gives enough spread to see differences.
What you measure | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|
Agreement | Strongly disagree | Disagree | Neither agree nor disagree | Agree | Strongly agree |
Satisfaction | Very dissatisfied | Dissatisfied | Neither satisfied nor dissatisfied | Satisfied | Very satisfied |
Frequency | Never | Rarely | Sometimes | Often | Always |
Likelihood | Very unlikely | Unlikely | Neither likely nor unlikely | Likely | Very likely |
Importance | Not at all important | Slightly important | Moderately important | Very important | Extremely important |
Quality | Very poor | Poor | Fair | Good | Excellent |
Ease | Very difficult | Difficult | Neither easy nor difficult | Easy | Very easy |
Familiarity | Not at all familiar | Slightly familiar | Moderately familiar | Very familiar | Extremely familiar |
Value for money | Very poor value | Poor value | Fair value | Good value | Excellent value |
Relevance | Not at all relevant | Slightly relevant | Moderately relevant | Very relevant | Extremely relevant |
Example Questions for Each Scale
Agreement: "The checkout process was easy to complete."
Satisfaction: "How satisfied are you with our customer service?"
Frequency: "How often do you use the mobile app?"
Likelihood: "How likely are you to buy from us again in the next 3 months?"
Importance: "How important is free delivery when you choose where to shop?"
Quality: "How would you rate the quality of the product you received?"
Ease: "How easy was it to find what you were looking for?"
Familiarity: "How familiar are you with our brand?"
Value for money: "How would you rate our value for money?"
Relevance: "How relevant is this product to your needs?"
7-Point Likert Scale Examples
Seven points give respondents room for a mild opinion. Use them when your audience has a real view on the topic and small differences matter, for example when you compare two versions of a product.
What you measure | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
|---|---|---|---|---|---|---|---|
Agreement | Strongly disagree | Disagree | Somewhat disagree | Neither agree nor disagree | Somewhat agree | Agree | Strongly agree |
Satisfaction | Completely dissatisfied | Mostly dissatisfied | Somewhat dissatisfied | Neither | Somewhat satisfied | Mostly satisfied | Completely satisfied |
Likelihood | Extremely unlikely | Unlikely | Somewhat unlikely | Neutral | Somewhat likely | Likely | Extremely likely |
5-Point or 7-Point: Which Should You Use?
Use five points unless you have a reason not to.
Choose 5 points for general audiences, mobile surveys, and long questionnaires. Fewer options mean faster answers and less fatigue.
Choose 7 points for engaged audiences and for studies where you need to detect small differences.
Avoid 10 or 11 points for labelled scales. People cannot tell a 6 from a 7 when each needs a word. A 0 to 10 scale works for recommendation questions because only the two ends are labelled.
Whatever you choose, keep it the same across the whole survey and across every wave. You cannot compare a 5-point result with a 7-point one.
Should You Keep the Neutral Midpoint?
Keep it in most cases. Some people have no opinion, and a scale with no middle forces them to invent one.
Remove the midpoint only when you need a decision and every respondent is able to give one, for example when existing customers rate a feature they use daily. A four-point scale then reads: strongly disagree, disagree, agree, strongly agree.
A related choice is whether to add "Don't know" or "Not applicable". Add it whenever some respondents may not have experienced what you ask about. It keeps guesses out of your data.
How to Report Likert Scale Results
There are two common ways to turn the answers into one number.
Top-2-box score. The share of respondents who chose the two most positive answers. It is easy to explain and is the standard in market research.
Average score. Each answer counts from 1 to 5 and you take the mean. It uses all the data, but it hides how the answers are spread.
Worked Example
You ask 200 customers how satisfied they are, on a 5-point scale.
Answer | Responses |
|---|---|
Very dissatisfied (1) | 12 |
Dissatisfied (2) | 22 |
Neither (3) | 36 |
Satisfied (4) | 78 |
Very satisfied (5) | 52 |
Top-2-box: 78 + 52 = 130, and 130 ÷ 200 = 65%.
Average: (12×1 + 22×2 + 36×3 + 78×4 + 52×5) ÷ 200 = 3.68.
Report both, along with the full spread. An average of 3.68 could come from many people in the middle or from a split between fans and critics, and those call for different actions.
Then check how far the number can be trusted. With 200 responses, a score of 65% has a margin of error of about 6.6 points at 95% confidence. The margin of error calculator gives the figure for your own sample.
When the question is about satisfaction on a 1 to 5 scale, the top-2-box score has a name: CSAT. The CSAT calculator works it out from your counts.
To see whether the score differs between groups, such as new and long-standing customers, break it down with a cross tabulation.
Five Mistakes to Avoid
1. Two ideas in one statement. "The app is fast and easy to use" cannot be answered by someone who finds it fast but confusing. Ask one thing per question.
2. An unbalanced scale. "Poor, Good, Very good, Excellent" has three positive answers and one negative. The results will lean positive whatever people think. Use the same number of options on each side.
3. Agree or disagree for everything. People tend to agree with statements, whatever they say. Where you can, ask directly. "How satisfied are you with delivery?" is better than "I am satisfied with delivery."
4. Switching direction. If 5 means "very satisfied" in one question and "very dissatisfied" in the next, some respondents will not notice. Keep the positive end in the same place.
5. Long grids. Twelve statements in one grid invite people to click the same column all the way down. Keep grids to five or six rows, and put an attention check inside the grid to catch those who do.
Use These Scales in a Survey
In SegmentOS the Likert scale question type takes your own labels, so any scale on this page can be pasted in. For questions to use them with, see these survey question examples. The ready-made templates use 1 to 5 rating scales with a label at each end, and each one includes an attention check that catches respondents who answer without reading.
Key Takeaways
Definition: a Likert scale asks people to choose from a range of ordered answers, usually five.
Match the scale to the question: agreement, satisfaction, frequency, likelihood and importance each have their own labels.
Five points is the default: use seven only when small differences matter.
Keep the midpoint unless every respondent can give a clear view.
Report the top-2-box score and the spread, and check the margin of error before you compare two results.



