Answers are grounded in this paper and cite their sources.
We examined the effects of three variables on rating scale response: (a) polarization of opinion regarding scale content (high vs. low), (b) intensity of item wording (strong vs. weak), and (c) psychological width of the rating scale (wide vs. narrow). Subjects were randomly assigned to one of the four combinations of item wording and scale width. All subjects responded to both the high- and low-content polarization attitude scales. Analyses showed that item means, standard deviations, and skew differed as a function of content polarization, item wording, and rating scale width. The results suggest that, to minimize rating scale design effects, high-polarization scale content should be balanced with weak item wording, and low-polarization content should be balanced with strong item wording regardless of the rating scale width. We also found that variability was maximized by strongly worded, highly polarized items but only through the overuse of scale endpoints.
No takes yet. Share an insight, caveat, or question.
Lam et al. (1994) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: