Unlike simple yes or no questions, understanding customer sentiment towards your brand, product or service is often more complex. You have to account for attitudes, opinions and perceptions, all of which can influence how much a customer likes (or dislikes) what you offer.
Under these circumstances, you need a more comprehensive and representative way to measure customer sentiment — and one of the best ways to do so is with a likert scale.
In this guide, we’ll cover everything you need to know about likert scales, from what a likert scale is to how it works, and how you can use it to elevate your market research.
Likert scale defined
A likert scale, also referred to as a rating system, is a measurement method used in research studies to evaluate attitudes, opinions and/or perceptions.
For each question or statement, subjects choose from a range of possible responses. The responses, for example, typically include:
- Strongly agree
- Strongly disagree
In studies where responses are coded numerically, ‘Strongly agree’ would be defined as 1 or 5, respectively increasing or decreasing for each response, e.g. in the above example, 5, 4, 3, 2 and 1.
Some likert scales use a seven-point scale with 1 being ‘Strongly Agree’ and 7 being ‘Strongly disagree’ (or reversed). In the middle, a neutral statement like ‘neither agree nor disagree’.
But as well as judging positive and negative statements, likert scales can judge frequency, quality, or feelings of importance towards a specific variable. For example, you could use a likert scale to understand how customers view certain product features, or what product upgrades they would most like to see next based on a range of options.
With this in mind it can be a very versatile scale to use when gathering quantitative data. The granularity it provides over simple yes or no responses means that you can uncover degrees of opinion that give you a more accurate and representative understanding of the feedback you receive.
Here are a few examples of how likert scales look:
What are the benefits of using likert scales?
Likert scales have several benefits, especially if you want to align data to a specific scale (as is often done in quantitative surveys). Here are just a few of those benefits:
1. They’re easy to understand
Likert scales are super easy to understand and complete as responders simply need to rank their preference based on the point scale you choose.
For example, depending on whether they strongly agree or strongly disagree with a statement, they just mark or select their response. This is sometimes referred to as a symmetric agree disagree scale.
Likert scales are also easy to analyze based on the responses given as they can be collated numerically and filtered based on responses to questions.
2. Ideal for single topic surveys
Likert scales are ideal for surveys that deal with a single topic as the data obtained can be easily analyzed to judge sentiment or feelings towards particular things.
For example, NPS surveys often use likert scales to judge the sentiment towards customer service. Rather than ranging from strongly agree to strongly disagree, you would instead use ‘highly satisfied’ to ‘highly dissatisfied’. This enables you to get a good grasp on how customers feel about your service.
At the same time, you can use likert scales to quickly assess how customers feel about specific parts of your customer service, product or brand, and then follow-up with a more detailed study.
3. Likert scale questionnaires are super versatile
Whenever you want to evaluate preferences, sentiment, perspectives, behaviors or opinions, likert scale questionnaires can provide you with the information you need.
You can implement them as part of a standard questionnaire to provide a bit of variety, or you can implement site intercepts on specific pages to get quick feedback. For example, you could have a likert scale questionnaire pop up after a webinar concludes. This way, you get immediate feedback on content and ideas on what to improve.
4. They don’t force specific responses
Rather than extreme response categories, e.g. giving respondents only two options when discussing potentially polarizing topics, likert scales provide a degree of flexibility. That said, when it comes to difficult topics, respondents may feel that they have to answer a certain way to avoid being seen as ‘extreme’. It’s therefore important to remind them that their survey responses are anonymous.
5. They’re great for sentiment analysis
A likert scale is perhaps most effective when you’re trying to assess sentiments towards your business, brand, product or service.
Likert scale responses can be used to judge sentiment, along with the reasons for the sentiment. For example, you could collate the data in a statistical analysis platform and filter the responses to see what percentage of customers are satisfied with your business, brand, product or service, versus those that are not.
You could go a step further and break the percentages down, e.g. those who are highly satisfied versus those who are just satisfied. How can you convert those customers into true evangelists? Again, another area of opportunity for organizations to capitalize on.
However, it’s important you only use a likert scale questionnaire when asking about a singular topic, otherwise you risk confusing your respondents and damaging the legitimacy of your study.
6. They keep respondents happy
One of the pitfalls with conventional survey design is that in a bid to design surveys quickly, researchers use overly broad questions that are limited to yes or no answers. These sorts of questions can frustrate respondents (as they give them no real way to provide context or accurate answers), leading to them rushing through surveys, affecting the quality of your data.
What are the limitations of likert scales?
While likert scaling is a highly effective way to measure opinions and sentiments, they do have some limitations depending how you create them. Here are a few:
1. Response choices limit real understanding
While likert scales can help you determine sentiment, they aren’t as effective at helping you understand why people feel a certain way. There’s also no interpretation of the sentiment between each choice, whether positive or negative.
For example, a respondent might ‘slightly agree’ with a statement, but the question is: why? What made them feel that way and what influences their responses? This kind of granularity can only be achieved with qualitative methods.
With this in mind, to increase the accuracy of your survey data, it’s worth running any likert-based questionnaires in conjunction with qualitative research methods.
2. Respondents might focus on one side of the sentiment
Depending how the questions are written, respondents might focus on one side of the scale. For example, if they feel that their answers might somehow affect their reputation, lifestyle or portray them in a negative way, they’ll pick positive responses.
Also, depending on the topic, respondents may be less likely to take extreme sides of the likert scale, instead agreeing, disagreeing or even remaining neutral.
3. Previous questions can influence responses
With any quantitative survey, respondents can get into a ‘rhythm’ of answering questions. The result of this is that they start to respond a certain way (this can be exacerbated further by poor questioning, the length of the survey and/or flicking between themes frequently).
When should I use a likert scale question?
A likert scale question is best used when you want to assess responses based on variables, e.g. sentiment, satisfaction, quality, importance, likelihood. Think of them when you want to assess how people perceive your products, or what employees think about your new offices.
For example, you might ask a respondent: “How would you rate the quality of our products?”, and provide a response scale of:
- Very poor
As mentioned above, this gives respondents a range and a degree to which they agree or disagree, rather than a simple yes or no answer which is often insufficient. Ultimately, you’ll want to use likert scales when you need to measure sentiment about something in more detail.
How to write likert scale survey questions
When writing likert scale questions, to ensure you get accurate responses and that responders understand what’s being asked of them, there are several things to consider:
1. Keep them simple
The best way to get accurate results is to ask simple, specific questions. Make it crystal clear what you’re asking respondents to judge, whether it’s their preference, opinion or otherwise.
For example, asking them: “How satisfied are you with our service?” and providing a standard scale, from very satisfied to very dissatisfied, provides no room for confusion.
2. Make sure they’re consistent
Respondents should fully understand the likert scale they’re being expected to record answers against, this means your answers on either side of the scale should be consistent.
For example, if you say “completely agree” at one extreme, the other extreme should be “completely disagree”.
3. Use appropriate scaling — unipolar scales and bipolar scales
Any likert scale will use either a unipolar scale or bipolar scale.
A bipolar scale should be used when you want respondents to answer with either an extremely positive or negative opinion. Sometimes, an even-point scale is used, where the middle option of “neither agree nor disagree” or “neutral” is unavailable. This is sometimes referred to as a “forced choice” method.
A unipolar scale works in the same way, but it starts from zero at one end, while an extreme is at the other. For example, if you ask how appealing your product is, your unipolar responses would go from “not appealing at all” to “extremely appealing”.
You should also aim to keep your scales odd because scales with an odd number of values ensure there’s a midpoint. Keep your scales limited to 5 or 7 points.
4. Don’t make statements, ask questions
Creating an effective likert scale means asking questions, not statements. This way, you avoid bias and present respondents with the opportunity to answer.
The reason for this is that people tend to automatically agree with positive or established statements, or unconsciously respond in a positive way (acquiescence bias). Overall, this can damage the validity of your study.
Also, asking questions rather than making statements encourages less biased responses because respondents have to think about their answers.
For example, asking “How satisfied are you with the quality of this service?” provides respondents with a chance to answer truthfully.
5. Switch your scale points
Switching scale points prevents respondents from falling into a rhythm and giving biased responses.
For example, if your point scale starts at 1, ‘completely agree’ and ends with 5, ‘completely disagree’, then you switch these around for a few questions so 1 is completely disagree and 5 becomes completely agree. This keeps respondents on their toes and engaged with the survey.
Here are a few examples of likert scale survey question types:
How to analyze likert scale data
Unlike many survey types, you can’t use the ‘mean’ as a measure of tendency because the mean response to a likert survey has no meaning. In other words, understanding the average of those who strongly agree or disagree will tell you nothing of relevance.
Instead, when analyzing your likert scale data, the right approach is to measure the most frequent response to understand the overall sentiment of respondents.
For example, 87% ‘strongly agree’ that you offer a good service. You can also compare the percentages for each response to see where respondents ultimately fall. This is incredibly useful for when you want to nurture customers — perhaps there’s something you can do for those who answered ‘agree’ rather than ‘strongly agree’ when you asked about the quality of your service.
So, how do you present your results? Well, the easiest way to present likert scale survey results is to use a simple bar or pie chart showing the distribution of response types or answer options.
You could also visualize your responses using a diverging stacked bar chart:
Go beyond standard likert scales with Qualtrics
Understanding engagement or sentiment towards your products or services is an essential part of improving your business. And with Qualtrics CoreXM — you can go even further.
Designed to empower everyone to gather experience insights and take action, Qualtrics CoreXM is an all-in-one solution to carry out customer, product, marketing and brand research, and then implement effective strategies.
From customer satisfaction surveys and event feedback to product concept testing and simple polls, create and deploy the research projects you need to enhance every aspect of your business.
Listen to everyone wherever they are providing feedback — whether directly in surveys and chatbot windows or indirectly via online reviews. Capture experience data across more than 125 resources and use that data for more targeted research and highly personalized experiences.
Use advanced analytics to interpret the feedback data and then automatically alert the right people to tell them what actions to take. All in real-time, with no legwork required. It’s time to go from measuring to acting and start closing experience gaps across your business.
Want to see how Qualtrics CoreXM can supercharge your research?