The Likert Scale( frequently known as agree-disagree scale) was first published by physiologist Rensis Likert in 1932. The technique presents respondents with a series of attitude dimensions ( a battery), for each of which they are asked whether, and how strongly, they agree or disagree, using one of the numbers of the positions on a five-point scale.

With face-to-face interviewer-administered scale batteries, the responses may be shown on a card whilst the interviewer reads out each of the statements in turn. With telephone interviewing, the respondent may sometimes be asked to remember what the response categories are, but preferably be asked to write them down.

The technique is easy for administrators in self-completion questionnaires, either paper or electronic, and may often be given to respondents as a self-completion section in an interviewer-administrator survey.
Responses using a Likert scale can be given scores for each statement, usually from 1 to 5, negative to positive, or -2 to +2. As these are interval data, means and standard deviations can be calculated for each statement.

The full application of the Likert scale is then to sum the scores from each respondent to provide an overall attitudinal score for each individual. Likert’s intention was that the statement would represent different aspects of the same attitude. The overall score, though, is rarely calculated in commercial research (Albaum,1997) where the statement usually covers a range of attitudes. The responses to individual statements are of more interest in determining the specific aspects of attitude that drive behaviour and choice in a market, or summations are made over small groups of items. The data will tend to use in factor analysis, in order to identify the groups of attitudinal dimensions, data are then often used in various forms of cluster or segmentation analyses, in order to segment data into groups of respondents with similar attitudes.

There are four interrelated issues that questionnaires writers must be aware of when using Likert scales :
Order effect
Acquiescence
Central tendency
Pattern answering