Question context and the impact of question order

The issue

A key concern often raised in the literature is that preceding questions may affect how respondents interpret the meaning of an item and/or the type of information that is temporarily accessible to respondents when constructing their answers - effects often collectively referred to as priming. The precise influence of priming effects is not always simple to predict or easy to detect, however, and there may be individual differences in the extent to which question context exerts an influence. For example, asking about marital status immediately before a life satisfaction question may not exert a strong influence on respondents whose marital status has remained stable for a number of years, but may be more salient for those recently married, divorced or widowed - potentially evoking positive feelings in some individuals, but more negative ones in others.

Context effects may therefore exert an influence on both the mean level of a measure, when summed across respondents, and/or the distribution of data, when context effects impact differently among different subgroups within a sample. This can in turn threaten both the comparability of data from different surveys with differing contexts, as well as the comparability of data from different groups of respondents on the same survey. Context effects may also inflate or suppress relationships between variables by making certain information and/or mood states more accessible to respondents. For example, if a set of prior questions about health status prompts respondents to draw more heavily on health-related information when answering a subsequent life satisfaction question, this could lead to a higher correlation between health status and life satisfaction than one might find in other surveys where the questions are arranged differently.

< Prev   CONTENTS   Source   Next >