Likert Scale

A Likert scale measures attitudes by asking respondents to rate agreement with a series of statements, typically on five or seven points from strongly disagree to strongly agree. Strictly, a Likert item is a single statement and a Likert scale is the summed or averaged set of items measuring one construct.

That distinction drives a long-running debate about analysis. A single item produces ordinal data, best summarised with medians and analysed with non-parametric tests. A multi-item scale averaged across items is widely treated as interval in practice, which is what permits means, standard deviations, regression and SEM.

Practical guidance: use at least three or four items per construct so reliability can be assessed, include reverse-worded items to catch inattentive responding (remembering to recode them before analysis), and report Cronbach's alpha or composite reliability for each construct rather than for the questionnaire as a whole.