A Cognitive Model of Response-Order Effects in Survey Measurement

Survey researchers have long been aware that the order in which response alternatives are presented to respondents may profoundly affect the obtained results (cf. S. L. Payne, 1951). However, the exact nature of the impact of response order is not well understood. Theoretically, primacy effects, that is, higher endorsements of items presented early in the list, as well as recency effects, that is, higher endorsements of items presented late in the list, may be obtained. Moreover, the conditions under which either of these effects may emerge are not well specified, and the area is characterized by a large number of apparently inconsistent findings.