Alternative question forms for attitude scale questions in telephone interviews

The absence of visual aids in telephone interviews has led investigators to seek methods for asking questions which are commonly accompanied by aids in personal interviews. This study experimentally compared two approaches to asking seven-point scale attitude questions in a national (N = 4, 300) telephone health survey. A random half-sample received the attitude questions as single-step numerical selection tasks. The questions were administered to the other half-sample in two stages, the first asking for a general verbal statement of attitude and the second asking for a more detailed specification. The two question forms produced comparable results in univariate distributions. Items administered with the one-step form showed slightly higher intercorrelations. This study represents one step in the process of empirically testing data collection procedures which have been deemed necessary in the rich folklore of survey research. Overall, the results show more comparability than might be anticipated, given the multiple differences between the forms.