Abstract
This paper describes experiments relating to the choice of words and phrases for use in questionnaires and public opinion polls. It is concerned particularly with expressions of degree or intensity of belief or opinion, and of temporal frequency.
Sets of phrases were presented to groups of subjects in random order, in serial order, and in context and the subjects were asked to place each item on a nine-point scale. From the results, the mean scale position and the ambiguity of each item were calculated; these data were combined with an index of the length of each item (in words per syllable) and the frequency of its occurrence in the language generally to provide criteria for the choice of suitable word-scales. A number of recommended sets of phrases are given, together with their appropriate scale ratings, positions and ranges.
Get full access to this article
View all access options for this article.