Abstract: | Research on the use of no opinion-filters suggeststhat respondents are the less likely to offer a substantiveresponse the more strongly the filter question is worded. Aseries of experiments is reported that demonstrates that filterquestions influence respondents' perception of their task: themore strongly the filter question is worded, the more respondentsassume that they will have to answer difficult questions, andthat they may not have the required knowledge. Accordingly,filter questions discourage respondents from offering globalopinions that they may hold. In line with this assumption, allrespondents who reported not having an opinion in response toa filter question, subsequently provided substantive responseson a global opinion questionpresumably because the globalquestion asked was less demanding than expected on the basisof the filter. Analyses of these substantive responses indicatedthat respondents who initially reported not having an opiniondiffered from respondents who reported having one. Methodologicalimplications of these findings for the use of filter questionsand for research on the nature of floating arediscussed. |