Descending Beats Ascending: Effects of Order on the Likert Scale on Consumer Ratings

In marketing surveys, the question order bias and response order bias are known from the viewpoint of order, while the response scale bias and central tendency bias are known from the viewpoint of the Likert scale, and a large amount of literature exists to support this. However, there is no evidence of bias due to the ascending or descending order of the Likert scale. This study filled this gap. An online survey of loyalty at Japanese universities was collected from 338 individuals. The result showed that since respondents prefer the first option, the descending order with the highest score on the upper side obtained a higher score. When determining the loyalty indices on the Likert scale in measuring marketing effectiveness, the true effectiveness cannot be known unless the order of options is consistent. In addition, since less serious respondents prefer higher options, the ascending order should be adopted to improve the quality of answers with good scores. This makes it possible to improve the accuracy of identifying customers with high loyalty. Thus, ascending options in the loyalty index question eliminate spurious loyalty due to insincere answers. Research design that understands the mechanisms of bias and limits uncertainty is essential in both business and academic research. This study provides useful suggestions for effect measurement, which is indispensable in digital marketing.