Likert Scale

A Likert scale is a survey format that measures attitudes by asking people to place themselves on an ordered range of response options, typically running from strong disagreement to strong agreement. A statement is presented — the checkout process was easy, say — and respondents pick one option from a symmetric set such as strongly disagree, disagree, neutral, agree, strongly agree. The options are ordered and evenly labeled around a neutral middle, so the answer captures both the direction and the strength of an opinion, not just a yes or no. Named after the psychologist Rensis Likert, who introduced it in the 1930s, the format has become the standard way to quantify subjective attitudes, satisfaction, and perceptions in surveys across marketing, research, and social science.

The scale usually offers five or seven points, though other lengths are used. An odd number includes a neutral midpoint, letting people sit on the fence; an even number forces a lean one way or the other by removing the middle. Strictly, a single such item is a Likert item, and a Likert scale is the combination of several related items whose responses are summed or averaged to measure an underlying attitude more reliably than any one question could — though in everyday use the term names both. The appeal is that it turns fuzzy opinions into ordered, comparable data: you can see how agreement is distributed, compare groups, and track sentiment over time, all from a simple, familiar question format that respondents understand instantly.

The subtlety of Likert scales lies in the kind of data they produce. The responses are ordinal: the options have a clear order (agree is more than neutral), but the gaps between them are not guaranteed to be equal — the distance from agree to strongly agree may not equal the distance from neutral to agree. This matters for analysis. Treating Likert responses as if they were evenly spaced numbers and taking a plain average is common and often useful, but it assumes equal intervals the scale does not strictly provide, so purists prefer methods for ordinal data, like reporting the distribution or the median. The practical answer usually depends on how the results will be used and how much precision the decision demands.

A Likert scale is easy to confuse with related rating formats. A semantic differential asks respondents to place themselves between two opposite adjectives (for example, cheap to expensive) rather than to agree or disagree with a statement. A simple numeric rating (rate this one to ten) or a star rating captures intensity without the ordered agree-disagree labels. A single Likert item differs from a full Likert scale of several combined items, which is more reliable. And a Likert scale is not a Net Promoter Score, which asks a specific likelihood-to-recommend question and sorts respondents into set groups. Each format suits different questions, so choosing a Likert scale means specifically wanting ordered agreement or intensity on labeled statements.

Designing Likert scales well starts with clear, single-idea statements. Each item should express one unambiguous claim, so a respondent is not stuck agreeing with half of it — avoid double-barreled statements that bundle two ideas, and avoid loaded or leading wording that pushes an answer. Keep the response options balanced and symmetric around the midpoint, label them clearly, and decide deliberately whether to include a neutral middle: an odd scale allows genuine neutrality, an even one forces a direction. For a reliable measure of an attitude, use several related items rather than relying on one, and word some in reverse to catch respondents who answer on autopilot. Consistency in scale length and labeling across a survey keeps responses comparable.

The failures are double-barreled or leading statements that make answers meaningless, unbalanced options that bias responses, and analyzing ordinal Likert data as if the intervals were exactly equal without acknowledging the assumption. Response biases lurk too: acquiescence bias (a tendency to agree), central-tendency bias (clustering on the safe middle), and social-desirability bias (answering to look good) all distort results, and thoughtful design — reverse-worded items, careful wording, sometimes an even scale — helps blunt them. The discipline is to write clean single-idea statements, offer balanced ordered options, combine items for reliability, and analyze the ordinal data honestly, so a Likert scale delivers what it promises: a clear, comparable read on where people stand.

The Likert scale — an ordered agree-to-disagree survey format measuring attitudes, usually on five or seven points — was introduced by psychologist Rensis Likert in the 1930s.

Etymology: source.

Search interest for this term over the last five years:

View interest-over-time on Google Trends →

Areas of marketing where likert scale is a core concern: