Likert Scale: Definition, Examples, and Complete Guide

The Likert scale is a psychometric measurement method that allows respondents to express their degree of agreement, satisfaction, or opinion on a graduated scale, typically ranging from 4 to 7 points. Invented in 1932 by American psychologist Rensis Likert, this survey technique has become the global standard for measuring attitudes and opinions in market research, satisfaction surveys, and academic research.

The three most commonly used Likert scale formats are the 5-point scale (from “Strongly Disagree” to “Strongly Agree”), the 7-point satisfaction scale, and the even-numbered 4 or 6-point scale that forces a choice by eliminating the neutral option. Unlike open-ended questions where analysis is subjective and time-consuming, the Likert scale enables precise quantification of responses and rigorous statistical analysis while remaining simple for participants to understand.

>> Create your Online Survey with Drag’n Survey, click here

History of the Likert Scale

It bears the name of its inventor, psychologist Rensis Likert. Born in the United States in 1903 and deceased in 1981, this American psychologist was a professor at the University of Michigan. He gained recognition in management circles with his book “Participative Management.” This book presents his research findings where he sought to understand why, within an identical organizational context, measuring relationships between supervisors and subordinates sometimes leads to completely different results.

Presentation of the Likert Scale

It measures judgment and contains between four and seven response choices. These response options measure the degree of agreement or disagreement. The wording of the proposed responses is variable. It can incorporate notions of satisfaction, agreement, or numerical ratings.

The number of options can be even or odd. When it’s even, the response is called “forced choice.” Participants are not given the option to choose a neutral position (neither satisfied nor dissatisfied). This is also called the “neither-nor” condition.

Scale Type	Number of Points	Advantages	Disadvantages	Recommended Use Case
Odd-numbered scale	5 or 7 points	Offers a neutral option, reduces frustration for undecided respondents	Risk of “refuge response” with overuse of the neutral option	Customer satisfaction surveys, general opinion studies
Even-numbered scale	4 or 6 points	Forces taking a position, avoids evasive responses	May frustrate genuinely neutral respondents, risk of abandonment	Performance evaluations, product acceptability tests
Numerical scale	0 to 10	Familiar to respondents, allows for more nuance	Ambiguity about the exact meaning of each number	Net Promoter Score (NPS), general ratings
Visual scale (slider)	Continuous or discrete	Dynamic and engaging, reduces survey fatigue	Less precise on mobile, requires suitable software	Long surveys, UX/UI studies

Best Practices for the Likert Scale

This type of question is very useful in a survey. It allows respondents to simply check the response that corresponds to them. However, this ease of use doesn’t exclude certain risks for result analysis.

Using the Neutral Option

Using a neutral option can be used by respondents as a “refuge response.” Indeed, it offers the possibility of not freely expressing one’s opinion. Respondents use it out of fatigue or simply because they don’t know what to answer.

The risk for the person in charge of the study is not knowing how to interpret the results. They may be skewed and not reflect reality. It is therefore recommended to use the “neither-nor” condition when it’s essential for understanding the question. Otherwise, it’s preferable to increase the number of response choices so that respondents can find the category that corresponds to them.

Using Numerical Graduation

When using numerical graduation, being as precise as possible is recommended. Using numbers can create confusion among respondents. It’s sometimes difficult to identify which part is positive or negative.

Offering Responses of the Same Nature

It’s preferable that the proposed response options be of the same nature. For example, one would go from “Extremely Satisfied” to “Extremely Dissatisfied.” This is called unidirectional graduation or unipolar graduation. If they are not of the same nature, it can create confusion for the respondent.

Different Display Modes for the Likert Scale

The Likert scale can be presented in different forms. For example, multiple-choice questions can be used as shown in the image above. When a survey mainly consists of measurement levels, there’s a risk of creating respondent fatigue. To make the survey more dynamic, it’s possible to use sliders and rating bars when using online survey software.

Slider-Type Question

The slider offers a display mode that differs from multiple-choice questions. Varying formats helps maintain and capture the attention of survey participants. With participation being more dynamic, this reduces the risk of abandonment during participation by respondents.

Rating Bar Question

A rating bar is a more dynamic measurement scale. It optimizes the responsiveness of participants in an online survey. It appears in the form of checkboxes or stars.

Rating Bar with Checkboxes

With online survey software, it’s possible to modify the color of the checkboxes. This further emphasizes the negative or positive character of the proposed responses.

Rating Bar with Stars

The rating bar with stars corresponds exactly to the question with checkboxes. It’s simply a diversification of the display mode that helps maintain and capture the audience’s attention.

Creating an Online Survey with Likert Scale Questions

Online survey software has become simple and highly efficient tools. Today, creating a survey is within everyone’s reach.

That’s why, by creating a free account on an online solution, you’ll have access to numerous features. You can create a study to obtain information necessary for decision-making and implementing action plans. Users have access to several types of questions, particularly multiple-choice questions and rating bars.

Once the online survey is created, it’s possible to distribute it through different channels (emailing, web link, QR Code, paper format, and Iframe Code). Once distributed, results update in real time.

Solutions available online allow you to analyze results directly on the software site and extract them in Excel and PDF formats.

Which Likert Scale to Choose for Your Survey?

Choosing the optimal Likert scale format depends primarily on your objective and target audience. For most customer satisfaction surveys and corporate opinion studies, the 5-point scale offers the best compromise between cognitive simplicity for respondents and richness of collected data. Academic studies or research projects requiring fine granularity will benefit from a 7-point scale. If your goal is to force a clear position and avoid evasive responses, favor an even-numbered scale with 4 or 6 points, particularly effective for performance evaluations. In all cases, ensure you formulate clear and unambiguous statements, perfectly balance your scale, and remain consistent throughout your survey. The statistical analysis of your results will be more reliable when your initial design has been rigorous.

Frequently Asked Questions

What is the difference between a 5-point and 7-point Likert scale?

The 5-point Likert scale offers five response levels (for example: Strongly Disagree, Disagree, Neutral, Agree, Strongly Agree), which represents a good balance between simplicity and nuance for the respondent. The 7-point scale adds two additional intermediate levels, allowing finer measurement of opinions and better discrimination between respondents. Research shows that the 5-point scale is sufficient for most customer satisfaction surveys, while the 7-point scale is preferable in academic studies or when seeking to detect subtle attitude variations. In practice, beyond 7 points, respondents’ ability to distinguish nuances decreases significantly without adding value to the analysis.

Should you use an even or odd scale to avoid bias?

The choice between even and odd scales depends on your research objective, not a question of bias. An odd scale includes a neutral midpoint (such as “Neither Agree nor Disagree”), which respects the reality that some respondents don’t have a strong opinion on a subject. An even scale eliminates this neutral option and forces participants to position themselves on one side or the other, called “forced choice.” Use an odd scale when neutrality is a legitimate and informative response for your study, for example in general satisfaction surveys. Opt for an even scale when you need to clearly identify positive or negative trends, particularly in performance evaluations or acceptability tests where indifference is not a relevant option. The important thing is to be consistent throughout your survey and never alternate between the two types.

How to statistically analyze Likert scale results?

Likert scale data is considered ordinal, meaning it indicates an order (agree > neutral > disagree) but the intervals between points are not necessarily equal. For descriptive analysis, you can calculate frequencies, percentages, and the median of your responses, and present results in bar charts or distribution diagrams. For more advanced analyses, although debated by statisticians, it’s common to treat Likert scales as interval data when you have at least 5 points, which allows using mean and standard deviation. You can then apply parametric tests like the t-test or ANOVA to compare groups. If you remain rigorous about the ordinal character, favor non-parametric tests like the Mann-Whitney test or Kruskal-Wallis test. To analyze multiple Likert items measuring the same concept, calculate a composite score (sum or mean of items) after verifying internal consistency with Cronbach’s alpha coefficient.

What are common errors to avoid with a Likert scale?

The first major error is formulating ambiguous or double-barreled statements, such as “The product is fast and economical,” which doesn’t allow knowing which aspect the respondent’s agreement concerns. Also avoid double negative formulations (“I am not dissatisfied”) which create confusion and interpretation errors. Another frequent error is unbalancing the scale by offering more positive than negative options, or vice versa, which biases responses. Never mix different types of scales in the same survey (satisfaction, frequency, agreement) without clear transition, as this increases the error rate. Technically, the most critical error is using imprecise labels for scale points, such as a numerical scale 1-5 without indicating what the extremes mean. Finally, watch for order effects: randomize your items when possible to avoid position bias, and never include all statements in the same direction (all positive or all negative), as respondents tend to agree automatically.

More about surveys:
5 tips to avoid creating confusing questions in a survey, click her
Everything you need to know about satisfaction surveys, click here

Read the article:
Polish – Skala Likerta, click here
German – Likert-Skala, click here
French – échelle de Likert, click here
Portuguese – escala Likert, click here