In short
A Likert scale is a rating scale that measures attitudes or opinions by asking people how much they agree or disagree with a statement — typically on a 5-point scale from "strongly disagree" to "strongly agree." It's the most widely used scale in surveys because it turns a subjective opinion into a number you can compare and track.
What is a Likert scale?
A Likert scale (pronounced LICK-ert, after psychologist Rensis Likert who created it in 1932) is a psychometric rating scale used to measure attitudes, opinions, and perceptions. Instead of a yes/no answer, respondents rate their level of agreement with a statement across a range of options — most commonly five: strongly disagree, disagree, neutral, agree, strongly agree. Each option is assigned a number (1–5), which lets you average, compare, and trend the responses.
The key idea: a Likert scale measures intensity of feeling, not just direction. "Agree" and "strongly agree" both point the same way, but they're meaningfully different — and the scale captures that.
How a Likert scale works
A Likert item has two parts: a statement and a symmetric agree–disagree scale. The statement should be a single, clear proposition; the scale offers balanced options on either side of a neutral midpoint. For example:
| Statement: "The staff resolved my issue quickly." |
|---|
| 1 — Strongly disagree |
| 2 — Disagree |
| 3 — Neither agree nor disagree |
| 4 — Agree |
| 5 — Strongly agree |
A true Likert scale is several of these items combined to measure one underlying concept; a single item is technically a "Likert-type" item. In everyday use, most people call any agree–disagree question a Likert scale.
Likert scale examples
Likert statements work for almost any attitude you want to measure. Some examples:
- Customer service: "It was easy to get the help I needed."
- Product: "This product meets my needs."
- Employee experience: "I feel my ideas are valued at work."
- Healthcare: "I was treated with dignity and respect."
- Events: "The session was a good use of my time."
Response scales aren't limited to agreement — Likert-type scales also measure frequency (never → always), quality (poor → excellent), likelihood (very unlikely → very likely), and satisfaction (very dissatisfied → very satisfied).
How many points: 4, 5, or 7?
The number of points changes how much nuance you capture and whether you allow a neutral answer.
| Scale | Neutral option? | Best for |
|---|---|---|
| 4-point | No (forced choice) | Pushing respondents off the fence when you need a lean |
| 5-point | Yes | The default — balanced, familiar, easy to answer |
| 7-point | Yes | More nuance and sensitivity for research or benchmarking |
| 10/11-point | Varies | Fine-grained scoring (e.g. NPS uses 0–10) |
Odd vs even: odd-numbered scales (5, 7) include a neutral midpoint; even-numbered scales (4, 6) force a positive or negative lean. Use even scales when a neutral response isn't useful; use odd scales when a genuine "neither" is a valid answer.
Unipolar vs bipolar: a bipolar scale runs from a negative through a neutral to a positive (disagree → agree). A unipolar scale runs from none to a lot of one attribute (not at all satisfied → extremely satisfied). Match the type to what you're actually measuring.
How to create a Likert scale
- Define the one thing you're measuring. Each statement should map to a single, clear concept — satisfaction, ease, trust, likelihood to return.
- Write balanced statements. Avoid double-barrelled ("fast and friendly"), leading, or negative wording that's hard to answer.
- Choose your points. 5-point is the safe default; go 7 for research depth, 4 to force a lean.
- Label every point, not just the ends — labelled points are answered more consistently.
- Keep the scale direction consistent across the survey so respondents don't have to re-orient.
How to analyze Likert scale data
This is where Likert scales get debated. Because the "distance" between points isn't guaranteed to be equal (is "strongly agree" exactly one unit better than "agree"?), Likert data is technically ordinal. That affects which statistics are strictly correct:
- Best practice for single items: report the median and mode, and show the distribution (how many chose each point) — often as a stacked bar (a "diverging" chart).
- Common in practice: many teams treat 5+ point scales as interval and report the mean (average score). This is widely accepted for tracking trends, even if purists disagree.
- Top-box scoring: report the percentage choosing the top one or two options (e.g. "% who agree or strongly agree") — simple, intuitive, and great for dashboards.
Is a Likert scale ordinal or interval?
Strictly, a Likert scale is ordinal: the options have a clear order, but the gaps between them aren't guaranteed to be equal. In practice, scales with five or more points are frequently analyzed as if they were interval (allowing means and standard deviations), which is a pragmatic and common convention. If your analysis needs to be statistically defensible, treat single items as ordinal and use non-parametric methods.
Is Likert scale data qualitative or quantitative?
It's quantitative — each response is a number you can count and compare — but it measures a qualitative concept (an attitude or opinion). That dual nature is exactly why Likert scales are so useful: they put a number on a feeling. Pair them with an open comment and you get the score and the reason.
Advantages and limitations
Why they work: Likert scales are quick to answer, easy to understand, capture intensity, and produce data you can quantify and trend. Watch out for: acquiescence bias (people lean toward agreeing), central-tendency bias (avoiding the extremes), and the ordinal-data debate above. Clear, balanced statements and a sensible number of points minimize these.
Likert scales vs smiley and star ratings
A Likert scale asks people to read a statement and agree or disagree. A smiley face rating scale asks them to tap the face that matches how they feel — no reading required. For in-the-moment, high-traffic feedback (a kiosk at an exit, a QR code on a receipt), smiley scales get far higher response rates; for detailed post-experience surveys, Likert scales capture more nuance. Many programs use both: a smiley tap to capture volume, a short Likert follow-up for depth. See how we deploy them on the smiley face feedback page.
Where Likert scales fit in your feedback program
Likert questions are a core building block of survey design, alongside customer satisfaction metrics like CSAT and NPS. Used well — short, balanced, captured close to the experience — they turn how people feel into a number you can act on.
Questions
Frequently asked
What is a Likert scale?
A Likert scale is a rating scale that measures attitudes by asking how much someone agrees or disagrees with a statement, usually on a 5-point scale from 'strongly disagree' to 'strongly agree.' Each point is assigned a number so responses can be averaged, compared, and tracked.
Is a Likert scale ordinal or interval?
Strictly, it's ordinal — the options have a clear order but the gaps between them aren't guaranteed to be equal. In practice, scales with five or more points are often analyzed as interval (using means), which is a common and pragmatic convention.
Is Likert scale data qualitative or quantitative?
It's quantitative — each answer is a number you can count and compare — but it measures a qualitative concept like an attitude or opinion. That's why Likert scales are useful: they put a number on a feeling.
How do you analyze Likert scale data?
For single items, report the median, mode, and the distribution of responses (a stacked/diverging bar chart), or use 'top-box' scoring (the percentage choosing the top one or two options). Many teams also report the mean for 5+ point scales to track trends over time.
How do you pronounce Likert?
It's pronounced 'LICK-ert,' after Rensis Likert, the psychologist who developed the scale in 1932 — not 'LY-kert.'
Turn survey scores into action
Book a free assessment and we'll help you design feedback questions that people actually answer — and that you can act on.