Topic Saliency

The saliency of a topic—that is, its importance or relevance to potential respondents—can affect response patterns to surveys. There are several explanations as to how and why topic saliency affects response rates. First, if a potential respondent believes the topic to be important, he or she may be more likely to rationalize incurring the costs of responding to the survey. Second, responding to a survey topic that is of personal interest may have intrinsic rewards, such as providing an opportunity to exhibit one's knowledge or share one's opinion. Third, responding to a survey about a salient topic may be motivated by perceived direct benefits. Survey participation may be viewed as an opportunity to advance one's own needs, interests, or agenda. All of these explanations may apply to explain why a single respondent or respondents are more apt to complete a survey about a salient topic.

Research suggests that people are more likely to respond to a survey if the topic is of interest to them. For example, teachers are more likely than non-teachers to respond to, and cooperate with, a survey about education and schools; senior citizens are more likely than younger adults to respond to a survey about Medicare and health.

In addition to its impact on survey participation, topic saliency is important for attitude formation and response retrieval. Theories about the representation of attitudes in memory suggest that attitudes reported by the respondent as being more important or as being more strongly held are also more stable over time. Attitudes about salient topics require less cognitive effort to recall, resulting in attitude responses that are more stable over time, more stable when presented [Page 892]with counterarguments, and more consistent with other attitudes and considerations.

There is great debate about whether attitudes are saved in memory and retrieved when the situation or the survey question arises (the online model of attitude formation), or whether attitudes are continually constructed from multiple considerations that are sampled each time they are needed (the memory model). The online model implies a continuum ranging from nonattitudes to “true” attitudes, where attitude stability and consistency are partly determined by topic saliency. The memory model implies that attitude formation is a stochastic process subject to some variability where the respondent samples considerations off the “top-of-the-head” when asked a survey question. In this model, more salient topics result in a response distribution for each individual that is more tightly clustered and therefore also more stable.

Whether the attitude is “true” or constructed on the spot, attitudes about salient topics are often considered to be more resistant to differences in questionnaire form. However, the evidence is mixed and depends highly on how topic saliency is measured. Self-reports of attitude importance, certainty, or strength are more resistant to certain questionnaire design features. When salience is measured by interest in politics or by the cognitive accessibility of those attitudes, evidence is tenuous on whether topic salience is related to questionnaire form or other survey response patterns.

Topic saliency has implications for survey operations, countering survey nonresponse, questionnaire design, and analysis. People are more likely to cooperate with, and respond more quickly to, a survey if the survey topic is of interest to them, suggesting a need to compensate for the potential bias this effect may cause by using a multiple contact strategy. Questionnaire designers and analysts need to consider the implications of their question form for their response distribution and thus for the interpretation of the results.

...