Question format and response style behavior in attitude research

This study focuses on the impact of number of response categories on response behavior. The study consists of four measurements:

• In January 2008, all respondents were presented with twelve statements. The statements were identical for each respondent, but the response scale could differ per respondent. Each respondent was randomly assigned to one of nine groups.
• In February 2008, a second measurement was taken among the same respondents. They were presented the same statements as in January, but with a different response scale. The type of response scale depended on the group to which the respondents had been assigned in January 2008. Panel members that had not participated in January were able to join; they were assigned to one of the groups at random.
• A third and fourth measurement were taken in the LISS panel in January and February 2009. These measurements also concerned twelve statements, but these were different to the ones presented in 2008. The statements were identical for all respondents, but the response scale could differ per respondent. In February 2009, the same statements (with a different response scale) were fielded as in January 2009.

The January measurements were taken following a questionnaire on religion and ethnicity. The February measurements were taken following a questionnaire on leisure time activities (in 2008, the statements were preceded by a questionnaire on leisure time activities and Internet use).