International Encyclopedia of Communication Online

Extract

In communication research scales are used to assess the intensity or the strength of personal variables like traits, states, →  attitudes , feelings, and so on (→  emotion ). A rating is based on a statement expressing a perception, an attribution, or an attitude toward something that is presented to the subject. He or she is asked to indicate the degree of his or her agreement with this statement by marking a corresponding label from a list provided by the scale. Bipolar Likert scales ( Likert 1932 ) extend from negative to positive numbers (e.g., −2 to +2), whereas unipolar Likert scales start from non-negative numbers like 0 or 1 (e.g., +1 to +5). Likert scales are most frequently constructed as five-point scales, but scales with seven or even nine points are also common. The labels assigned to points on a Likert scale (like “strongly disagree,” “somewhat disagree,” “neither agree nor disagree,” “somewhat agree,” “strongly agree”) must be chosen with great care in order to hold up equal semantic distances between the labels over the entire scale. As it cannot be ensured that subjects perceive intervals between adjacent points on the scale as equidistant, responses collected from subjects should preferably be cautiously treated as ordinal (and not as interval) data (→  Measurement Theory ; Scales and Indices ; Rating Methods ). A Guttman scale ( Guttman 1950 ) is constructed ... log in or subscribe to read full text