What is a Likert scale and how should you use it in surveys?

Picture this: You went to see your favorite artist in concert. They were great, but the venue was very hot and had frequent sound issues. Afterward, you get an email from the production company with a single question: “Did you enjoy the show? (Yes, No).” 

You don’t know how to answer. On one hand, you’re happy about having seen your favorite artist live, but overall, it wasn’t a good experience. 

This is a great example of when using a Likert scale comes in handy. It’s one of the simplest, yet most effective tools for capturing people’s opinions, behaviors, or attitudes when answers are more nuanced than a binary yes or no. 

In this article, we’ll get into what a Likert scale is, its benefits and limitations, examples from real surveys, and tips for designing effective scale-based questions. Let’s get started.

What is a Likert scale?

A Likert scale is a type of closed-ended question that allows respondents to answer by choosing an option from a rating scale. These questions usually contain four to eleven answer options. 

For example: 

• Very unlikely
• Unlikely
• Neutral
• Likely
• Very likely 

Here is an example of Likert scales you have likely seen before :

Who created the Likert scale?

The Likert scale was created by Rensis Likert, an American socialist, economist and psychologist, in 1932. He first introduced this concept in an article called: “А Technique for the Measurement of Attitudes.” 

He continued expanding on this idea in 1934 in an article published by the Journal of Social Psychology titled “A Simple and Reliable Method of Scoring the Thurstone Attitude Scales.”

According to Likert, the biggest problem with research was that people can have an infinite number of attitudes, but these can be grouped together in a range. Therefore, he created the Likert scale. 

Why is the Likert scale controversial?

The Likert scale is controversial because the data can be compromised by assumptions, biases and misuses, especially during analysis. Reasons include:

Assuming equal intervals

Some people see Likert scale responses as interval data when they’re actually ordinal. This means assuming the gap between “Agree” and “Strongly agree” is equal to the one between “Neutral” and “Disagree”. 

In reality, people’s interpretations of these points can vary significantly. So, instead of viewing this scale as intervals, consider each response as a unique ordinal value. 

Falling into biases

Likert scales can contribute to central tendency and acquiescence biases. The first one happens when people avoid extreme response categories and gravitate toward the middle (e.g., choosing “Neutral” or “Somewhat agree”). 

The acquiescence bias, on the other hand, happens when people agree with statements regardless of their content. These factors make Likert scales controversial since biases can mask true opinions, skew data and create the illusion of consensus or positivity where none exists.

Cultural and language interpretations

If you’re surveying a global audience, consider explaining the meaning of each answer beforehand, as the perception of certain words (e.g., “neutral”) can vary across cultures or languages. This makes cross-cultural or multilingual comparisons tricky, and relying on a Likert scale without proper explanation could skew your data.

The pros and cons of Likert scales

Likert scales are a great technique to understand slight differences in people’s sentiments, but aren’t the right choice for all cases. Let’s explore the benefits and drawbacks of using Likert scales in questions.

Benefits of using Likert scale questions

Using Likert scale questions in your surveys is beneficial because these are:

Easy to understand and complete

According to Pew Research, respondents give more honest answers when they easily understand the question. 

Likert scale survey questions require little cognitive load for people to choose an option. They also offer a clear structure, help reduce survey fatigue and limit acquiescence bias.

Ideal for focused surveys

Likert scales allow you to measure one concept at a time by asking diverse questions with scale-based answers. 

For instance, if you want to measure how satisfied your users are with a new feature, you can ask different questions using Likert scales to gauge their true opinion on this particular topic. 

Versatile

Present these types of questions in multiple formats, such as site intercepts, pop-ups, long-form surveys or email-based questionnaires. Respondents can simply pick an option and share their opinion from wherever they’re interacting with the survey. You can also adopt different scales depending on your research goals (more on this below!)

Non-polarizing

Unlike yes-or-no questions, Likert scales allow respondents to express degrees of opinion, making feedback more detailed and less forced. For example, if you want to know if respondents care about sustainability, a yes or no question doesn’t give you enough information. A person can say yes because they only buy cruelty-free brands, but they only drink mineral water. Likert scales give color to these types of nuanced answers.

Good for trend and sentiment analysis

Scale-based answers are easier to quantify and analyze, especially over time. For instance, you can compare last quarter’s customer satisfaction vs. this one without hassle.

Limitations of using Likert scale questions

While Likert scales are easy to complete and analyze, they come with some limitations. We’ll discuss these below.

Lack of “why”

Scale-based questions alone don’t reveal what’s driving a sentiment. Pair them with open-ended follow-ups to get context.

Subjectivity and bias

People interpret scale points differently depending on culture and background, which introduces bias. That said, Likert type questions can reduce other types of survey bias, such as:

• Binary thinking: Forcing respondents into overly simplistic yes/no answers.
• Cognitive overload: Asking too much from respondents, tiring them out, and getting random answers.
• Interviewer influence: Swaying the respondents through the interviewer’s tone, body language, or wording.

💡Pro-tip: For more on how to avoid misleading responses and overclaiming in surveys, check out this guide on five common survey mistakes.

Neutral answers can skew results

 Some respondents overuse the midpoint to avoid taking a stance. To balance this, try rephrasing the same question in different ways and include open-ended follow-ups.

Survey fatigue and patterns

In long surveys, people may fall into response patterns without reading carefully, e.g., choosing “Strongly agree” repeatedly. Mix up the wording or order of questions to keep them engaged.

Not for multi-topic questions

Likert scales work best for single-focus questions. Avoid using them to assess multiple ideas at once, or the data will be muddled.

Types of Likert scales

Now that you’re familiar with the pros and cons of Likert scales, let’s explore the reasons why these questions are so versatile: their length and format. Here are different scales you can choose from based on your needs:

4-point Likert scale 

4-point Likert scale questions present the respondent with four different options to choose from on a point scale. These questions tend to skip the middle option so respondents are compelled to give a positive or negative response. For example:

How satisfied are you with our customer service?

• Not at all satisfied
• Dissatisfied
• Satisfied
• Very satisfied

However, forced choices (i.e., skipping the middle option) could lead to false responses, especially from those who genuinely feel neutral.

5-point Likert scale 

A 5-point Likert scale includes two positive, two negative, and one neutral option. Unlike the previous example, it allows people to feel neutral about a question. Here’s an example:

How satisfied are you with the onboarding program?

• Not at all satisfied
• Dissatisfied
• Neither satisfied nor dissatisfied
• Satisfied
• Very satisfied

A 5-point Likert scale enables people to choose from a balanced range of responses. This helps capture more accurate attitudes and reduce respondent frustration. However, it can contribute to central tendency bias (i.e., preference for neutral response options) as people may overuse the middle option.

6-point Likert scale

A 6-point Likert scale gives respondents even more options to choose from, but like in the four-point scale, it doesn’t give a neutral option. An example of a 6-point Likert scale is: 

Rate your agreement with this statement: The customer service at Attest is quick to respond.

• Fully disagree
• Disagree
• Somewhat disagree
• Somewhat agree
• Agree
• Fully agree

Similar to the 4-point one, but a 6-point Likert scale gives more chances for people to express how they truly feel. However, it may force respondents to choose a side when they genuinely feel neutral about a question.

 7-point Likert scale 

Like with the 5-point Likert scale, this offers seven options for respondents to choose from, with the middle one being neutral. Here’s an example: 

How easy was it to navigate the new feature? 

• Very difficult
• Difficult
• Somewhat difficult
• Neither difficult nor easy
• Somewhat easy
• Easy
• Very easy

A 7-point Likert scale provides greater nuance than 5- or 6-point versions, allowing respondents to choose an option that more accurately reflects their views. This can lead to more precise data and improved scale reliability. However, the added options can increase cognitive load.

10-point Likert scale 

The 10-point Likert scale gives respondents 10 different options to choose from, increasing the precision of each answer. In this case, the scale is mostly based on numbers and not descriptors.  

Here’s an example:

How likely are you to recommend our product to others?

• 1 = Not at all likely 
• 2
• 3
• 4
• 5 
• 6
• 7
• 8
• 9
• 10 = Extremely likely

A 10-point Likert scale offers a highly granular view of respondents’ attitudes, often preferred in quantitative research or finer statistical analysis. 

On the downside, the increased range can overwhelm participants, making it harder to choose a precise option. Also, without clear anchors or labels for each point, people may interpret numbers differently, which can compromise the consistency and validity of the data.

Frequency Likert scale 

This is a regular Likert scale, but instead of using qualifiers, it uses time measures. 

Frequency scales are useful for capturing how often a behavior or event occurs, offering broader context than agreement scales. For example: 

How often do you go to the gym?

• Never
• A couple of times per year
• Once a month
• Every week
• Every day

These questions help quantify habits or usage patterns, making them ideal for behavioral research. However, vague or unevenly spaced options can be interpreted inconsistently across survey respondents, causing the data to lack reliability or comparability.

💡Remember: Likert scale questions are ordinal, not nominal. This means the answers help classify people in response categories based on a meaningful order, for example: Satisfied customers. Whereas nominal data groups people based on categories that don’t have an inherent order such as gender or race.

When to use each type of scale (plus Likert scale examples)

Different types of Likert scales can be used for different goals. For example, to address how strongly someone feels about something (intensity), how often they attempt to do it (frequency), or how likely they are to do it (likelihood). Here are examples for you to get inspired when building your next survey.

Net Promoter Score (NPS) 

When assessing your customer loyalty, use an 11-point Likert scale ranging from 0-10. Use it to measure how likely it is for a customer to recommend you to others.

ℹ️ Example: “How likely are you to recommend our product to a friend?”

➡️ Answers: 0 (Not at all likely) – 10 (Extremely likely)

[H3] Customer satisfaction score (CSAT)

Use a five or seven-point Likert scale to measure customer satisfaction with your general product, a particular feature or the performance of your customer support team.

ℹ️ Example: “How satisfied were you with your recent support experience?”

➡️ Answers: Not at all satisfied, dissatisfied, neither satisfied nor dissatisfied, satisfied, very satisfied.

[H3] Employee engagement survey 

Capture how your employees feel about the company by asking them to rate different factors that add up to overall engagement. Use a 7-point Likert scale to gain more in-depth insights. 

ℹ️ Example: “I feel valued by my team.” Or, “I feel supported by the company.”

➡️ Answers: Extremely disagree, disagree, somewhat disagree, neither agree nor disagree, somewhat agree, agree, extremely agree.

Consumer profiling 

Ask questions to understand your customer habits or behaviors for segmentation purposes. You can then use the data to tailor offers to their desires. Use a frequency response scale in this case. 

ℹ️ Example: “How often do you shop online in a typical week?”

➡️ Answer: Never, rarely, sometimes, often, always.

Brand sentiment survey

Gauge your customers’ opinion towards your brand. Since you want to identify whether they have a positive or negative opinion, use a 6-point Likert scale with no neutral option. 

ℹ️ Example: “This brand aligns with my values.”

➡️ Answers: Extremely disagree, disagree, somewhat disagree, somewhat agree, agree, extremely agree.

Designing effective Likert-scale questions

Here are some best practices to follow to write a customer survey and get actionable insights.

1. Determine the scale 

Likert scales are flexible and there’s not a single right choice. However, there are some things to keep in mind before choosing your response scale: 

• Pick a scale that you can apply to most of your survey questions. Having a consistent scale makes it easier for respondents to answer without making mistakes.
• Consider that less answer options are easier to analyze and also easier for your respondents to answer (less survey fatigue). However, these also have less nuance. So, consider the length of your scale depending on the amount of detail you need from each question.
• Using larger scales with more points gives you more granularity, but a higher cognitive load. So, stick to these when you have short surveys.
• Determine if you want to include a neutral midpoint. Odd scales allow respondents to neither agree nor disagree with your statement, but even point scales help avoid fence-sitting and force people to take a stance. We recommend removing the neutral option in opinion-based questions. However, the choice is ultimately yours. 

2. Write specific questions

As with any survey question, focus is key. Asking about two things at once is confusing and unhelpful.

For example, if you want feedback on your support team’s speed and politeness, don’t ask: “Rate your agreement with this statement: The customer support team was fast and polite.” What if they were fast but rude?

Instead, stick to one idea per question and use clear, neutral wording to avoid leading your respondents. For example:

❌ “Was the onboarding process quick and helpful?”

✔️ “Was the onboarding process easy to follow?”

3. Use clear and balanced response options

When creating your scales, ensure they use simple language and are symmetric to avoid skewing the data. Here are some recommendations:

• Start with extremes like “Not at all” or “Extremely” to anchor responses and build a full, clear range.
• Borrow from others and use the language you see in popular surveys. Move away from using jargon or novel scales. If your customers are used to certain ratings or qualifiers, stick to what they know to avoid confusion. 
• If your scale has a midpoint, make sure it reflects true neutrality (e.g., “Somewhat”, “Neither agree nor disagree”, “Neutral”).
• Avoid vague or overlapping terms like “a little” vs. “somewhat”, or “sometimes” vs. “a few times.” These can confuse respondents, and remember: Confused respondents give unreliable answers.
• Keep both ends of the scale symmetrical so each side carries equal weight. For example: “Completely agree,” ⇄ “Completely disagree,” or “Extremely satisfied” ⇄ “Extremely dissatisfied.”

4. Choose between bipolar and unipolar scales

Bipolar scales measure a person’s attitude or opinion across two opposing ends, such as good vs. bad. For example: “Strongly disagree” to “Strongly agree.” They capture both direction and intensity of opinion.

Unipolar scales, on the other hand, measure the intensity of a single attribute without an opposite end. For example: “Not at all helpful” to “Extremely helpful.” These show degrees of an attribute, like interest or satisfaction.

Usually, the question signals which type to use, but here’s a quick decision guide:

• Use bipolar scales when you need to measure direction and intensity (e.g., how positive or negative someone feels).
• Use unipolar scales when you’re measuring intensity only (e.g., how helpful or clear something was).

5. Ask questions rather than making statements

To write better survey questions and avoid getting biased answers, frame items as questions, not declarative statements. This invites people to think about how they truly feel about your question, rather than quickly agree or disagree with a statement. 

For example:

❌ “The customer support was helpful.”
✔️ “How helpful was the customer support?”

Analyzing Likert-scale data

People filled out your Likert scale questionnaire; it’s now time to analyze the responses. Before you jump in, take some time to refresh your research goals. This will guide the type of insights you look for in your data and give you a blueprint on how to analyze survey results.

Then, interpret and present Likert scale ordinal data by following these tips: 

Avoid using the averages

Circling back to why Likert scales are controversial, it comes down to how people do data analysis. 

As we mentioned above, likert responses are ordinal, not nominal or interval. Meaning that the options have a meaningful order, but the distance between points isn’t necessarily equal. 

Therefore, using the average (mean) to analyze Likert scale survey question data isn’t reliable because it assumes equal spacing between responses. Instead, focus on the mode (most common response) or median (middle response) to understand the overall trend.

For example, let’s say this is your data set: 

Question 1	Extremely dissatisfied	Dissatisfied	Satisfied	Extremely satisfied
Number of answers	5	12	10	15

This makes:

• Mean: 2.83 (skewed)
• Mode: Extremely satisfied
• Median: Satisfied  

Calculate the percentage of each response

Whenever you’re analyzing data, count how many respondents chose each answer option, and convert that into percentages. For example, using the same previous data set, we can say:

• Extremely dissatisfied: 5 [(5/42)*100)] = 11,9%
• Dissatisfied: 12 [(12/42)*100)] = 28,57%
• Satisfied: 10 [(10/42)*100)] = 23,8%
• Extremely satisfied: 15 [(15/42)*100)] = 35,71%

This makes it easier to spot patterns, like how many people agreed vs. disagreed. Comparing those percentages helps teams spot trends or gaps, such as users who are satisfied but not strongly satisfied. From there, you can dig deeper to understand what would move those users toward a more positive experience.

Group similar responses

If you’re looking to gain broader insights about your respondents’ feelings, try combining scale point answers into bigger categories. For example, you can group “somewhat satisfied,” “satisfied” and “very satisfied” together to show total positive sentiment. 

This kind of grouping gives a high-level view that’s much easier for you to act on, especially in an employee feedback or customer satisfaction survey. Sometimes, understanding that 40% of customers are unhappy is more powerful than analyzing each group individually.

Use charts to visualize the data

We recommend using bar or pie charts to show the data distribution. Bar charts are great for comparing categories side-by-side. For instance, if you want to see customer satisfaction across attributes, such as: Satisfaction with onboarding, customer support, new features or the overall product. 

This example below shows public opinion on three climate-related policies, broken down by level of support or opposition: 

Source

Pie charts, on the other hand, help you view proportions at a glance and make quick, broader comparisons. For example, gauging what percentage of your total users are satisfied with your product: 

Source

The goal of using charts is to make the data clear and accessible, especially for presentations, stakeholder updates or reports. Also, data visualization simplifies decision-making. 

In fact, according to a study by the International Journal of Research Publication and Reviews (IJRPR), “Visual representation transforms numerical data into comprehensible and meaningful insights, leveraging the human brain’s ability to process visuals faster than text.”

Be aware of response bias

Remind respondents throughout the survey that their answers are anonymous and that there are no right or wrong responses. This helps reduce social desirability bias, where people respond in ways they think are more socially acceptable.

By encouraging honesty, you’re more likely to avoid inflated neutral or positive scores, especially on sensitive topics. Still, even with those reminders, be mindful of bias when interpreting results that lean heavily toward the middle or positive end of the scale. 

To get a fuller picture, complement your survey findings with additional research, such as industry benchmarks, behavioral data or one-on-one customer interviews. You can also add open-ended questions in your survey to identify the “why” behind rating-based answers.

Ready to create your own Likert-scale questions?

Likert scales are powerful tools for gathering quantitative data and understanding customer attitudes. However, like all other research methods, these need to be used and analyzed thoughtfully, considering potential biases that may have affected the data.

Be mindful of how you design your scales. Choose between unipolar or bipolar formats and analyze the results using appropriate statistical methods (remember this is an ordinal scale and answers represent an order!) Avoid relying on averages and instead focus on patterns, medians, and modes to uncover meaningful insights. 

When done right, Likert scale data can guide smarter, more empathetic decisions. Need help building your Likert scale questionnaires? Explore and get inspired by Attest’s survey template library.