Abstract:

This paper introduces a new design for the study of Internet vs. other data collection modalities. With this so-called crossover design it is possible, for the first time, to separate modality from population effects. In this design, respondents in the experimental conditions are confronted with both survey modalities, CATI and Web. Results show that the mode of survey administration does hardly influence the outcome of a survey. A number of guidelines are given, by means of which a market researcher can conduct a multi-modal study in a reliable and practical manner.