Answers are grounded in this paper and cite their sources.
Each of six forms of a 10-item teacher evaluation rating scale, having 2, 3, 4, 5, 6, or 7 response categories per form, was administered to over 100 students at a Pacific Coast college. Means of item responses increased linearly and item variances increased curvilinearly with number of response categories. Internal consistency reliability coefficients (alpha) of total scores did not change systematically with increases in the number of response categories.
No takes yet. Share an insight, caveat, or question.
Lewis R. Aiken (1983) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: