Psychological sources of response effects in self-administered and telephone surveys

Abstract

The impact of mode of data collection (self-administered questionnaire vs. telephone interview) on the emergence of response effects and the accuracy of recall from memory was explored in a cross-cultural experiment, conducted in the U.S. and the Federal Republic of Germany. As predicted on the basis of psychological considerations, question order effects were obtained under telephone interview conditions but not under self-administered conditions, where question order is eliminated by the opportunity to browse back and forth through the questionnaire. On the other hand, the impact of the content of related questions was more pronounced under self-administered than under telephone interview conditions, independent of the order in which they were presented. This reflects respondents' differential opportunity to elaborate on related questions under both administration modes, as well as the necessity to rely on the content of presumably related questions in determining the meaning of ambiguous questions under self-administered conditions. Finally, respondents' recall of the date of public events was more accurate under self-administered than under telephone interview conditions, reflecting the beneficial effect of having sufficient time to work on the recall task.