Question Wording as Discourse Indicators

Survey Questions as Indicators of Social Discourse

Public opinion and public discourse are often viewed as two parallel systems interacting with each other to provide interpretation and meaning to relevant events and policies. Discourse analysis reveals the evolving culture of issues and events in terms of their interpretive frames. The analysis of public opinion provides a glimpse into how individuals adopt such frames and which perspectives prevail on the aggregate. Frequency distributions of public opinion polls provide the latter. Public discourse packages of important public issues can be captured extremely well by content analysis of questions in public opinion polls through systematic examination of the language of survey questions.

It is commonly accepted that question-answer situations must be understood in the social context in which they take place. This contextual approach applies to the interview setting as well as to the content and form of the questions themselves, which must be seen as segments of the ongoing social discourse. In fact, survey questions and the way they frame issues are increasingly viewed as part of the elite discourse and political debate, which effectively shape popular discourse and thinking. Pollsters who design surveys are in the same position as public officials, editors, journalists, and newsmakers of all sorts: All choose how public issues are to be framed, and their choices have consequences.

Indeed there are good reasons to think of survey questions as useful indicators of social discourse on a specifie issue. First, to be valid and reliable and to engage respondents, thus not adding to nonresponse, [Page 667]survey questions must be formulated in a way that carries widely shared meaning. Thus, pollsters often use words that respondents would use themselves in the same context, or they define key concepts in the questions in terms familiar to the respondents. Either way, the choice of wording reflects the ongoing political debate and, eventually, public discourse. Second, public opinion surveys and polls tend to focus on topical issues and current events and are likely to share similar choice criteria used by editors to decide what is news. Just like news items, they tend to concentrate on salient events and to reflect critical discourse moments. Third, when pollsters change question wording—often a hard decision for pollsters to make—they usually intend to either adapt to meaningful social changes or to write a better question. But writing better questions often means simply fine-tuning the questions so as to fit into a frame of reference perceived, by the pollster, as dominant or most relevant. Sometimes changes in questions are made to increase respondents' capability or willingness to respond. These changes do not occur arbitrarily but often indicate pollsters' assumptions about the salience of an issue, the public's level of knowledge, and the prevailing social norms affecting willingness to respond. In short, the reasons for introducing changes in question form and wording may, in fact, reflect developments in public discourse.

...