Considerable research has examined the effect of response option order in ordinal bipolar questions such as satisfaction questions. However, no research we know of has examined the effect of the order of presentation of concepts in the question stem or whether stem order moderates response option order. In this article, we experimentally test the main and interaction effects of both stem and response option order for items in self-administered surveys on response distributions and answer changes in eight satisfied/dissatisfied questions. We find consistent evidence that response option order impacts answers. We also find that the order of “satisfied” or “dissatisfied” in the question stem impacts response distributions for four of our eight items but does not moderate the effect of response option order. We discuss the implications of our findings for questionnaire design and secondary data analyses.
ASJC Scopus subject areas