View article

Response-order effects can bias evaluations. Measurement error is introduced by order effects, and this, in turn, can affect associations between variables. This study attempts to identify types of questions in mail questionnaires which are susceptible to order effects and illustrates the consequences of response-order effects on associations. Data from a Cooperative Extension Service evaluation showed order effects for multiple response and attributive questions but not for single response ones. Aspects of question complexity and social desirability may contribute to order effects. In at least one instance, response-order effects were found to influence associations between items and thus, the substantive conclusions.