Are you happy? Are you satisfied? These questions are probably some of the hardest to answer in life and are relevant within market research too. How do we objectively measure the level of satisfaction of a person on a particular subject?
The Likert scale was invented in the 1930s to solve this problem!
What is a Likert scale?
The Likert scale was named after organisational psychologist Rensis Likert in 1930. The scale objectively quantifies feelings and attitudes as well as levels of consensus on certain topics or objects. Some researchers call this scale a satisfaction scale since it measures levels of satisfaction in particular.
How is it defined?
A Likert scale is defined as an ordered scale in which respondents choose an option that best matches their opinion.
It is used, for example in market studies, when seeking to obtain a valuation value, to quantify intangible or abstract concepts. In some ways, it is similar to a multiple-choice question (MCQ), except that it limits responses to a single set of logical values (for example, from not at all satisfied to very satisfied). Additionally, when a Likert scale is used, respondents cannot answer yes or no but must make specific choices based on "agree" or "disagree" with a specific question.
Here is an example of the application of the Likert scale taken from the latest FMCG survey by Appinio.
Why using a Likert?
The popularity of this scale is that it is one of the most reliable ways to measure opinions, perceptions and behaviours. No other research tool can objectively measure or quantify these variables. Researchers often use this psychometric scale to understand opinions and views about a brand, product or target market, employee satisfaction and more, making it one of the most widely used scales in surveys and an integral part of market research. For example, if you recently implemented a work-from-home policy at your office and you would like to know how smooth the transition was for employees, you can easily create an HR questionnaire using the Likert scale to assess the level of satisfaction with the new directive and the way it has been managed.
Other attractive features about this tool are:
- Likert questions are easy to understand
- Likert scales offer raw answers before focusing on further analysis of the results - Likert scales offer more precise data or feedback rather than a question that can be answered with "Yes" or by "No"
- Responses from a Likert scale contribute to a faster understanding of the problem, which allows for deeper insight into the underlying reasons for a specific problem
- Ease of creation: creating Likert-type questions is simple and poses no problem
- Ease of understanding for survey respondents: there is no need to give additional guidance or explain each question in detail
- Ease of use: Respondents should only select one option for each question. They don't have to answer an open question or a question that's too complex, which can lead to inaccurate data
- Ease of analysis of responses: Likert scales allow the collection of quantitative data, which means it is easier to understand and analyze the results
- A very comprehensive approach: Likert scales allow you to collect a variety of opinions on topics that would usually be difficult to measure, such as feelings and perceptions of customers
- Easy to design and apply to surveys
- Provides a range of opinions to respondents
When can a Likert scale be used?
You cannot use Likert scales to assess attributes such as a person's age or gender, for example. However, it is the best instrument to assess individual attitudes and to gather more precise information on a given subject.
The graph below illustrates the variables that can be measured using this scale:
Taking into consideration the variables that the scale can measure, different types of companies use it for internal and external affairs. Here are some examples of topics that you might understand better using the Likert scale questions:
- Customer reactions to a new product
- Staff satisfaction
- Customer satisfaction
- Feedback on events
- Target markets
- Public health assessments
- Needs assessments (linked to guidelines)
How to create a Likert scale
Usually, the prompt provided for a Likert scale survey is not asked in the form of a question, but of a declarative statement that measures a certain attitude. The statement is then followed by a series of logically distributed response categories. (eg. Very dissatisfied to very satisfied). The last step is to assign a numeric value to each value in the series.
The graphic below summarizes the steps required to create your own Likert scale:
What types of Likert scales can you create for your own research?
A popular variation of the Likert scale is the “Likert-type scale”. This variant has similar characteristics to the traditional scale, such as the ordered set of response categories and a balanced number of positive and negative options.
One difference is that Likert-type scales can have labels for each answer option or only for the final categories, otherwise known as anchor categories.
A second difference is that Likert-type scales do not use the traditional spectrum of “Strongly disagree” to “Strongly agree” responses. They use other categories of ordered responses, such as:
Never, Sometimes, Often, Always, to assess a frequency
Mild, Moderate, Severe, to assess the intensity
Not at all, A little, A lot, to estimate the quantity
How to decide on the number of possible answers in the scale?
Usually, Likert scales have five, seven or even nine answer options. More answer choices allow for more precise results.
The number of responses given to express a person's attitude leads us to differentiate between the two types of Likert scales: odd and even scales.
Even Likert scale
The even scale consists of removing the intermediate option "Neither agree nor disagree". This method is sometimes called "forced choice" because the neutral option is not available. Even scales are generally scales with four or eight options.
This method is not bad and has its advantages, but it does not adapt to every situation. It is therefore important to analyze your precise situation before removing the neutral choice.
The table below compares the pros and cons of using a pair Likert scale.
Odd Likert scale
Unlike even scales, researchers use the odd Likert scale to give respondents the opportunity to respond in a neutral way. This naturally means that all odd scales have an intermediate value which is often labeled "neither disagree nor agree", "neutral" or "undecided." Usually, odd scales are 5, 7, or 9 point scales.
The dilemma of odd Likert scales is the risk that respondents will have different interpretations of the midpoint, even if it is labelled correctly.
The midpoint can be interpreted as: "I don't know", "I am not sure", "I don't care", "I have no opinion", "Neither", "I am neutral", "I equally agree and disagree", "I am undecided" or I do not want to answer".
This wide range of options means only one thing: confusion. This can lead to measurement errors and bias. Having a midpoint can be seen as an easy option to take when a respondent is unsure at the time of responding, so it is questionable whether this is a true neutral option.
Respondents may be less discriminating and not take the time to weigh the merits of each response category.
On the other hand, when clearly explained, midpoints allow participants who truly have no opinion on the question or who are just indifferent to answer honestly.
The table below compares the pros and cons of using a Likert scale:
What is the best type of Likert scale to use?
Rating scales have been the subject of a long-standing debate. Advocates of even-numbered scales argue that it forces respondents to choose an opinion: either they like something or they don't. However, the downside of this type of scale is its subjective nature. On the other hand, odd rating scales are defended from a mathematical point of view. Odd scales also suffer from response-style bias, a type of bias that respondents show regardless of the content as a systematic tendency to answer a series of questions on a certain basis rather than what they actually think (in the case of the Likert scale).
To avoid the problem of bias, in either case, you should keep in mind that a well-structured survey always starts with a clear objective and is never meant to elicit answers or force respondents to form an opinion; in many cases, an opinion may simply be neutral.
So here's what to keep in mind when deciding whether to choose an odd or even scale:
Remember the main objective of the survey
The subject of the survey: If the subject is not very controversial, you can use an even scale without a neutral point.
If you know your respondents: are they generally frustrated with the lack of a midpoint? If so, it is strongly recommended to use a neutral response option, in case this type of option is not offered, the results may not be fully valid.
What are the advantages and disadvantages of the Likert scale?
Like any type of scale in research, the Likert scale has positive and negative aspects that every researcher should keep in mind before creating a survey.
The benefits of using a Likert scale are numerous and justify its popularity. Here's a look at those benefits:
Other advantages of Likert scales include:
Choice of quantifiable answers
Ease of response for participants
Very easy to code answers
Speed and efficiency
Inexpensive method for data collection
On the other hand, the downsides to creating a survey with a Likert scale are as follows:
Such a scale is one-dimensional and only gives 5-7 options for choices: the space between each choice may not necessarily be equal.
This does not allow the real attitudes of the people questioned to be measured.
There is always the risk that people's answers are influenced by previous questions, or that they focus heavily on one aspect of the answer (agree/disagree).
People may also avoid choosing the "extreme" options on the scale, due to the negative implications associated with "extremists", although an extreme choice would be most accurate.
What to keep in mind when creating Likert scales
Use Wide Ladders: As a general rule, Likert and other scales recommend using as wide a ladder as possible. Responses can always be broken down into concise groups, if necessary, for analysis.
Decide on a scale: the items must always have at least two extreme positions and the choice of an intermediate response option which serves as a graduation between the extremes.
Be specific to avoid confusion. The questions should be clear and easy to understand. The more precise it is, the more valuable the data will be because the respondents will have answered in a completely honest and precise manner.
For example, rather than asking “Do you like our products?”, it is better to ask more specific questions such as “Are you satisfied with the quality of our products?” or "Do you think our products are good value for money?"
Don't offer too many options. If you offer too many options on your Likert scale (we recommend using fewer than 7), respondents are more likely to choose an option at random, resulting in inaccurate data.
Cover all the bases: Your Likert scale should cover the full range of answers, as well as a midpoint. If your answers only range from "Extremely satisfied" to "Fairly satisfied", for example, the respondents who are not satisfied will not know which answer to choose and thus will not be able to answer honestly; the results will then be skewed.
Use words rather than numbers: Numbered scales, such as 1 to 5, can be confusing and lead to inaccurate data. Respondents may not know which end of the scale is positive and which end is negative. It is best to indicate your scale options using words.
Always indicate your midpoint
Provide balanced response options
Make sure the labels are clear and concise
In conclusion, Likert scales may be the best alternatives to measure people's attitudes and satisfaction levels in record time and at a very low cost. However, as with most types of scales in research, one should be aware of the disadvantages of using them and minimize them by wording them clearly and using a wide range of options.
Are you ready to create your next survey using Likert scales? Use our free dashboard and sign up here:
Learn more about Appinio
Appinio is a market research platform that enables companies around the world to obtain thousands of opinions from consumers belonging to specific target groups within minutes. For the first time, anyone can validate decisions and ideas in real time, with results representative of consumers. Appinio provides insight into over 90 markets with its fastest market research.