Suchergebnisse

The sequence of items on a questionnaire or interview schedule may influence the nature of response; such order effects have both positive & negative aspects. A method is proposed for testing & controlling these order effects, while preserving the positive aspects of question sequencing. The method is based on the theory of randomization & the availability of highly flexible computer text-editing algorithms. An illustrative application of the procedure is discussed, & its impact on response rates, the coding of data, the development of personalized research materials, & cost are reviewed. AA.