Likert scales (also known as "summated rating scales") are popular and easy to use. They are simpler to construct than a Thurstone scale and generally yield more reliable results. Like a Thurstone scale, respondents are asked to indicate the strength of their agreement or disagreement to a series of about 20 statements, but then each category of response is allocated a score and a measure of the respondent's overall attitude is given by their total score. A Likert scale therefore takes account of every response, unlike a Thurstone scale where the researcher focuses only on those statements with which the respondent agrees. Responses to individual statements can also be analysed to gain a fuller understanding of respondents' views.
An example of a Likert scale is shown below.
Extreme statements are not usually included in a Likert scale because nearly everyone in the target population is likely to respond to them in the same way, and therefore they are of little value in gauging subtleties of attitudes. However, because it is not an interval scale, no conclusions can be drawn about the distances between different positions on the scale. Neutral statements do not tend to work well on a Likert scale because respondents' scores tend to cluster around the middle point