What Is a Likert Scale and How to Use It?
The Likert scale is one of the most widely used measurement tools in survey research. Every time you've marked somewhere between 'strongly agree' and 'strongly disagree,' you've encountered a Likert scale. In this article, we explain what it is, how many points it should have, how to analyze it, and the most common mistakes.
What is a Likert scale?
A Likert scale is a response format used to measure the degree of an attitude, opinion, or perception. A statement is presented and respondents indicate their level of agreement on an ordered set of options.
The classic example has five options: strongly disagree, disagree, neither agree nor disagree, agree, strongly agree.
The fundamental logic is offering a degree rather than a sharp yes/no distinction. Because attitudes are rarely black or white. The Likert scale attempts to capture these shades.
When to use a Likert scale
A Likert scale works for any measurement involving degree. It's suitable for measuring satisfaction, intent, perception, importance, and frequency.
It's the right tool when you want to measure how satisfied a customer is, how important a feature is considered, or how persuasive an ad is.
Conversely, it's not suitable for categorical or single-choice information. Questions like 'which brand do you use' don't involve degree. Use the scale only where you're truly measuring intensity or degree.
How many scale points to use?
This is the most debated question in Likert scale design, and there's no single correct answer.
5-point scale
Easiest for respondents. Options clearly differentiate, speeds up the survey, and provides sufficient distinction for most consumer research.
7-point scale
Allows finer distinctions. Preferred when you want to capture small differences between attitudes. The cost is being slightly more demanding for respondents.
10+ point scale
Respondents can't meaningfully differentiate between options, reducing data quality. Exception: standardized metrics like NPS that use 0-10.
Thinking more points always means better is a common misconception.
The neutral midpoint question
Another important design decision is whether to include a neutral option in the middle.
Odd-numbered scale (5, 7)
Places a neutral point in the middle. Gives genuinely undecided respondents an honest answer option. But sometimes serves as an 'escape hatch' for those unwilling to think.
Even-numbered scale (4, 6)
Removes the neutral point and forces respondents to lean one way. Preferred for topics where you want a clear position.
The decision depends on the nature of what you're measuring. If indecision itself is valuable information, keep the neutral point.
How to analyze Likert data
A Likert scale produces 'ordinal' data—options follow an order but the distance between them may not be equal.
Option-level distribution
Reporting each option's percentage separately is the safest approach. Makes no assumptions and shows the full distribution.
Mean calculation
Assigning numerical values and calculating averages is practical and facilitates comparison. But technically controversial as it assumes equal intervals.
Top-two box
Looking at the total percentage of the two most positive options. Often provides a more robust and understandable metric than averages.
Common mistakes
The most frequent mistakes in Likert scale design and how to avoid them:
Building an unbalanced scale
Positive and negative option counts should be equal. A scale skewing positive artificially inflates results.
Writing a leading statement
Even if the scale is neutral, a leading statement corrupts results. The statement must be at least as neutral as the scale.
Ignoring straight-lining
With consecutive Likert questions, some respondents give the same answer to all. This behavior needs detection and filtering.
Stacking too many Likert questions
Pages of identical format tire respondents. Spreading them evenly throughout the survey produces better results.
Reporting only the mean
Giving only the average hides the distribution. The same mean can come from very different distributions. Present with distribution information.
How to use Likert scales on Sorbunu
On Sorbunu, scale questions are one of the ready-made question types—you don't have to design Likert scales from scratch.
When you set the number of scale points and write your statement, a balanced and standard structure is automatically built. This prevents common design mistakes like unbalanced scales from the start.
Results are presented both as option-level distributions and summary metrics. Quality issues like straight-lining are filtered by background control layers.
A well-prepared Likert scale combined with proper question design and adequate sample produces reliable attitude measurement.
Frequently Asked Questions
5-point and 7-point are most common. 5-point suffices for most consumer research. 7-point provides finer distinctions. With 10+ points, respondents struggle to differentiate options.
Depends on the situation. Odd-numbered scales include a neutral midpoint for genuinely undecided respondents. Even-numbered scales force respondents to lean one way. Keep the neutral point if indecision is meaningful information; remove it if you want clear positions.
Common but technically controversial. Likert is ordinal data and distances between options may not be equal. Means are practical but can hide distributions. Reporting option-level distributions or top-two box percentages is also recommended.
A method of looking at the total percentage of the two most positive scale options. Often provides a more understandable and comparable indicator than averages, frequently used in satisfaction and intent measurements.
Likert is a general attitude measurement format, typically 5 or 7 points. NPS is a specific metric based on a single 0-10 recommendation question. Likert has broad application while NPS focuses solely on measuring loyalty and recommendation tendency.
No fixed limit, but stacking too many leads to respondent fatigue and straight-lining. Spreading them evenly throughout the survey and adding variety with different question types preserves response quality.
Related Solutions
Customer Satisfaction Research
Measure satisfaction with balanced scales and track over time.
Brand Awareness & Perception Research
Measure brand perception reliably with scale questions.
Product & Concept Testing
Evaluate attitudes toward products and concepts on a degree basis.
Build Scale Questions Easily
Create balanced Likert scales with ready templates on Sorbunu, reach your target audience, and go to field in minutes.