Likert scale

A Likert scale is an ordinal rating scale used to measure attitudes or opinions, typically offering five or seven response options running from one extreme, such as "strongly disagree," to the opposite extreme, "strongly agree," often with a neutral midpoint. It is named after psychologist Rensis Likert, who developed the method in the 1930s.

Strictly, a "Likert scale" is the sum or average of several related statements, a multi-item scale, while a single question in this format is more precisely a Likert-type item; everyday use often blurs the distinction. Because the response categories are ordered but not necessarily evenly spaced, the data is technically ordinal, which is why some statisticians argue means should not be calculated on individual items and recommend medians instead, though treating summed multi-item scales as approximately interval is common practice.

Likert-type items are a building block in questionnaire design work, and how they perform can shift with survey mode effects. They also underpin familiar metrics like the Net Promoter Score and customer satisfaction score, which convert Likert-style responses into a single summary number. A common pitfall is comparing average scores across groups as if the scale had a fixed, universal meaning, when acquiescence bias, cultural response tendencies, and anchor-label wording can all shift how the same attitude gets reported.