Tickbox

All articles

Likert scale: definition, examples and how to read the results

A Likert scale is a group of related statements that people rate on an ordered list of labels, usually from ‘Strongly disagree’ to ‘Strongly agree’. Each label has a number, and you add or average the numbers to get one score for an attitude. This guide gives labels for 7 scale types and 28 example questions for marketing teams. It also explains when to use an odd or even number of points, and how to read the results.

What is a Likert scale?

The Likert scale is named after the American social psychologist Rensis Likert, who described the method in 1932 in ‘A Technique for the Measurement of Attitudes’. Source: Wikipedia: Likert scale.

People use the name for two different things. A Likert item is one statement with its answer labels. A Likert scale, in the strict sense, is the group of items that you combine to measure one attitude. In everyday use, people also say ‘Likert scale’ for one item. Source: Wikipedia: Likert scale. Questions that use the same format with other labels, such as satisfaction or frequency, are often called Likert-type questions. This guide includes them, because surveys for marketing teams use them often.

Example of a Likert item: The pricing page told me what I would pay. Strongly disagree / Disagree / Neither agree nor disagree / Agree / Strongly agree

Likert scale, rating scale or Net Promoter Score?

A Likert item asks for a reaction on a set of labeled points. A rating scale gives one thing a score, such as 1 to 5. Net Promoter Score is one fixed question with a scale from 0 to 10 and its own calculation, and you can get the score with our NPS calculator. If you are still choosing a question type, read 10 types of survey questions, with examples.

Example labels for 7 types of Likert scale

Choose the type that matches what you want to measure. Some scales run between two opposites and have a neutral point in the middle, such as agreement, satisfaction and likelihood. To turn a satisfaction scale into a score, use our CSAT calculator. Others run from nothing to a lot, such as importance and familiarity. Our labels are based on a list of response anchors by Wade Vagias (2006). We changed some wording, and we made the 7-point importance scale into a 5-point scale. Source: Wade Vagias: Likert-Type Scale Response Anchors (2006), copy at Marquette University.

Scale typeWhat it measures5-point labels
Agreement scaleAttitudes and opinionsStrongly disagree / Disagree / Neither agree nor disagree / Agree / Strongly agree
Satisfaction scaleA reaction to an experienceVery dissatisfied / Dissatisfied / Neither satisfied nor dissatisfied / Satisfied / Very satisfied
Frequency scaleHow often a behavior happensNever / Rarely / Sometimes / Often / Always
Importance scaleWhat matters in a decisionNot at all important / Slightly important / Moderately important / Very important / Extremely important
Likelihood scaleWhat people intend to doVery unlikely / Unlikely / Neither likely nor unlikely / Likely / Very likely
Quality scaleA judgment about one specific thingPoor / Fair / Good / Very good / Excellent
Familiarity scaleAwareness of a brand or a subjectNot at all familiar / Slightly familiar / Somewhat familiar / Moderately familiar / Extremely familiar

The same scale with 3 to 9 points

Here is the agreement scale at each common length. The other types can grow or shrink in the same way. The notes in the last column are our own advice. The research is in the sections below.

PointsLabelsWhen to use it
3Disagree / Neither agree nor disagree / AgreeA very quick check, when you need only the direction of opinion.
4Strongly disagree / Disagree / Agree / Strongly agreeA short scale with no middle point.
5Strongly disagree / Disagree / Neither agree nor disagree / Agree / Strongly agreeShort forms, and forms that most people answer on a phone.
6Strongly disagree / Disagree / Somewhat disagree / Somewhat agree / Agree / Strongly agreeMore detail with no middle point.
7Strongly disagree / Disagree / Somewhat disagree / Neither agree nor disagree / Somewhat agree / Agree / Strongly agreeMore detail, when the results must separate small differences.
9Strongly disagree / Disagree / Somewhat disagree / Slightly disagree / Neither agree nor disagree / Slightly agree / Somewhat agree / Agree / Strongly agreeRarely needed. It is hard to find labels that people can tell apart.

28 example Likert scale questions for marketing teams

There are 4 examples for each scale type. Use each one with the labels of its type, and change the words to match your product. These questions are illustrations, not survey data.

Agreement scale

Labels: Strongly disagree / Disagree / Neither agree nor disagree / Agree / Strongly agree

  1. The homepage made it clear what the product does.
  2. Our emails are relevant to my work.
  3. I trust what this brand says about its product.
  4. Our case studies describe companies like mine.

Satisfaction scale

Labels: Very dissatisfied / Dissatisfied / Neither satisfied nor dissatisfied / Satisfied / Very satisfied

  1. How satisfied are you with the sign-up process?
  2. How satisfied are you with the content of our newsletter?
  3. How satisfied are you with the first week of using the product?
  4. How satisfied are you with the speed of replies from our customer support?

Frequency scale

Labels: Never / Rarely / Sometimes / Often / Always

  1. How often do you read our newsletter when it arrives?
  2. How often do you read online reviews before you buy software?
  3. How often do you share our articles with a colleague?
  4. How often do you click on sponsored search results?

Importance scale

Labels: Not at all important / Slightly important / Moderately important / Very important / Extremely important

  1. How important is price when you choose a tool like ours?
  2. How important are customer reviews when you compare vendors?
  3. How important are integrations with your current tools?
  4. How important is a free trial when you decide to buy?

Likelihood scale

Labels: Very unlikely / Unlikely / Neither likely nor unlikely / Likely / Very likely

  1. How likely are you to renew your subscription next year?
  2. How likely are you to attend our next webinar?
  3. After seeing this ad, how likely are you to visit our website?
  4. How likely are you to try the new plan in the next 3 months?

Quality scale

Labels: Poor / Fair / Good / Very good / Excellent

  1. How would you rate the quality of our product demo?
  2. How would you rate the content of the webinar?
  3. How would you rate the help articles on our website?
  4. How would you rate the speakers at our event?

Familiarity scale

Labels: Not at all familiar / Slightly familiar / Somewhat familiar / Moderately familiar / Extremely familiar

  1. How familiar are you with our brand?
  2. How familiar are you with the products that we sell?
  3. How familiar are you with our latest campaign?
  4. How familiar are you with marketing automation tools?

Odd or even: should your scale have a middle point?

An odd number of points (3, 5 or 7) includes a middle option, such as ‘Neither agree nor disagree’. An even number of points (4 or 6) has no middle option, so each respondent must lean to one side.

Each choice has a cost. In their review of questionnaire design, Jon Krosnick and Stanley Presser explain that a middle option can give people who do not want to think about the question an easy answer. But if you remove it, people who really are neutral must choose a side, and the measurement becomes less accurate. They report a study where adding middle points improved reliability and validity, and they conclude that this suggests that middle points are desirable. Source: Krosnick and Presser: Question and Questionnaire Design, Handbook of Survey Research (2010).

The research does not give one rule for all surveys. A 2017 review by Chyung and colleagues concludes that a middle point is appropriate in some situations and not in others. Source: Chyung and colleagues: Evidence-Based Survey Design: The Use of a Midpoint on the Likert Scale, Performance Improvement (2017).

Our advice

How many points should a Likert scale have?

In a study by Carolyn Preston and Andrew Colman, 149 respondents rated a store or restaurant on scales with 2 to 11 points, and on a 101-point scale. The scales with 2, 3 and 4 points performed relatively poorly on several measures of reliability, validity and discriminating power. The measures improved up to about 7 points. Test-retest reliability tended to decrease above 10 points. Respondents preferred the 10-point scale, closely followed by the 7-point and 9-point scales. This is one study on one subject. Source: Preston and Colman: Optimal number of response categories in rating scales, Acta Psychologica (2000).

Krosnick and Presser conclude that 7-point scales are probably optimal in many cases. They also note that most of this research used visual formats, so it is less certain for telephone surveys. Source: Krosnick and Presser: Question and Questionnaire Design, Handbook of Survey Research (2010).

So the research gives more support to 7 points than to 5 points. Use 7 points when detail is important. Use 5 points when the form is short or when most people answer on a phone, because 5 labels take less space. This is a practical trade-off, and not a research result. When you repeat a survey, keep the same scale so that you can compare the results.

Put a word on each point, not only on the two ends. Krosnick and Presser report that reliability is higher when all points have word labels, and that labels which divide the scale into approximately equal parts improve reliability and validity. Source: Krosnick and Presser: Question and Questionnaire Design, Handbook of Survey Research (2010).

How to write better Likert statements

How to read Likert scale results

Start with the share of answers for each label. Then add a summary number that fits your decision. In this example, two statements each have 200 answers. The numbers are an illustration, not survey data. To get these numbers and a chart from your own answers, use our Likert scale calculator.

AnswerStatement AStatement B
Strongly disagree10 (5%)40 (20%)
Disagree30 (15%)20 (10%)
Neither agree nor disagree40 (20%)10 (5%)
Agree80 (40%)50 (25%)
Strongly agree40 (20%)80 (40%)
Top 2 box60%65%
Bottom 2 box20%30%
MedianAgree (4)Agree (4)
Mean3.553.55

Both statements have the same median and the same mean, but they do not tell the same story. Statement A shows broad, moderate agreement. Statement B divides people: 40% strongly agree, and 30% disagree. If you report only the mean, you lose this information.

Six steps to read the results

  1. Give each label a number, for example 1 for ‘Strongly disagree’ and 5 for ‘Strongly agree’. If a statement is negative, reverse its numbers before you combine it with other statements.
  2. Show the share for each label. Gail Sullivan and Anthony Artino write that a frequency distribution of responses will likely be more helpful than a mean to describe this data. Source: Sullivan and Artino: Analyzing and Interpreting Data From Likert-Type Scales, Journal of Graduate Medical Education (2013).
  3. Add the top 2 box score. This is the percentage of people who chose one of the two highest labels: 60% for Statement A. Source: SurveyMonkey: How to use a Top 2 Box score in your survey analysis. To score an agreement scale about effort, such as ‘The company made it easy to solve my problem’, use our customer effort score calculator. Look at the bottom 2 box score too, because it shows the people who are not happy. Do not compare these scores between scales with different lengths.
  4. Be careful with the mean. Likert answers are ordinal: the order is known, but the distance between two labels cannot be measured. For this reason, experts have argued for the median. Sullivan and Artino also say that parametric tests can be used on Likert responses, with an adequate sample size and data that is approximately normal, and they refer to a review by Geoff Norman. Source: Sullivan and Artino: Analyzing and Interpreting Data From Likert-Type Scales, Journal of Graduate Medical Education (2013). Our advice is to use a mean only to follow one question over time, and to show it with the shares for each label.
  5. Combine statements only if they measure one idea. You can score a Likert scale as the sum or average of its items. Source: Wikipedia: Likert scale. For example, 4 trust statements with 1 to 5 points each give a total from 4 to 20. Do not add a satisfaction question to a frequency question.
  6. Check whether a difference is large enough to mean anything. With 200 answers from a random sample, the margin of error on a percentage is about 7 percentage points at 95% confidence. So the gap between 60% and 65% in this example can be only chance. You can check your own numbers with our margin of error calculator.

A Likert scale shows how many people feel a certain way, but it does not show why. Put an open-ended question after the statements that are important to you, and read the answers with the scores.

Likert scales in several languages

To combine the results from several languages, each label must keep its meaning and its strength. The European Social Survey translation guidelines were written for large international surveys, but a small team can use the same main ideas. Source: European Social Survey: ESS Round 11 Translation Guidelines (2022).

  1. Keep the order of strength. In an example from the guidelines, the translation of ‘quite interested’ must be weaker than the translation of ‘very interested’.
  2. Keep the strength of the end points. The translation of ‘extremely’ does not have to be literal, but it must not be a word that means ‘very’, because ‘very’ is less intense.
  3. Expect some limits. The guidelines say that some languages may have no word that works at both ends of a scale, so the two ends cannot always match. Write down each decision of this type.
  4. Be careful when you compare countries. A 26-country study by Anne-Wil Harzing found major differences in response styles, such as a tendency to choose average or extreme answers. A gap in scores can come from response style and not from opinion, so compare changes over time in one language first. Source: Anne-Wil Harzing: What if fully agree doesn’t mean the same thing across cultures?.

Common questions about Likert scales

Is a Likert scale quantitative or qualitative?

It gives quantitative data, because you count closed-ended answers. The numbers show an order, and not exact distances. For qualitative data, add an open-ended question.

Can I use one Likert item alone?

Yes, for a quick measure of one thing. Read it as one item: show the shares for each label, and do not call it a measure of a broad attitude.

How do you pronounce ‘Likert’?

Rensis Likert is reported to have said his name as ‘LICK-ert’, but many people say ‘LIKE-ert’. Source: Wikipedia: Likert scale.

Where Tickbox fits

Tickbox is a form builder with voice and video. Forms receive text, audio and video responses, and AI analyzes the responses so they are easy to read. AI analysis is on all plans, including Free. Forms support multiple languages. You can use this guide with any form tool, or start from a free form template.

Read what Tickbox is

Sources for the facts in this article

Each external fact above links to its source, and the list below shows what each source supports. We checked these sources on 4 November 2026. For the Chyung study, we checked the statement against the abstract only. For the Vagias list, we used a copy on a university website. The definitions, example questions and advice are our own, except where a source link is given. The example questions and the numbers in the results table are illustrations, not survey data.