Key points are not available for this paper at this time.
The question as to whether the format of a scale influences results has been examined infrequently and with conflicting answers. Two Likert-type formats, one with all choice points defined and the other with only end-points defined, were administered to 121 subjects. Each subject completed half of the items in the defined and the other half in the end-defined condition. Results were not significantly different between forms, nor did subjects indicate a format preference. Although the end-defined items exhibited greater variability than did the every-point defined items, the results suggest that minor Likert-type format changes do not critically affect outcomes.
Dixon et al. (Thu,) studied this question.