Response option configuration of online administered Likert scales

Likert scales are used to make relative and absolute judgments about measures of attitude. Despite its ubiquitous use, only few studies have investigated the effects of altering the configurations of the response options. The purpose of this experiment was to explore the effects of response option orientation and directionality in Likert scales administered online on internal consistency reliability, factorial validity, measured score, completion time, and mouse click counts. A positively stemmed 15-item 5-point Likert scale was administered to 1693 adults. Six independent groups were constructed (2 x 3 design). The response options were configured (a) vertically or horizontally and (b) unidirectional-ascending (strongly disagree to strongly agree), unidirectional-descending (strongly agree to strongly disagree), or bidirectional (mixed). Results indicated that vertically unidirectional response options should be used when absolute judgments are being made using online-administered Likert scales. When relative judgments are being made, horizontally unidirectional response options should be used.