Key messages on question context and the impact of question order
Available evidence suggests that question context - and particularly question order - can influence subjective well-being reports, and that in some cases these effects can be large. Given that subjective well-being questions may be included in a range of different surveys covering different subject matters, identifying and applying the best approach for minimising survey context effects should be a priority. Producing measures that are robust to context effects is important for ensuring data comparability - across surveys and across countries.
Locating subjective well-being questions as early on in the survey as possible should limit interference from other items. However, the novel nature of subjective well-being questions may come as a shock to respondents if presented right at the beginning of a household survey, before interviewers have had the opportunity to build some degree of rapport. It is most important to avoid placing subjective well-being questions immediately after questions that are likely to elicit a strong emotional response. Use of introductory text and transition questions may also help to reduce context effects. These and other practical recommendations are discussed further in Chapter 3.
Finally, it is unlikely that subjective well-being measures are uniquely susceptible to context effects, and the impact of prior subjective well-being questions on responses to subsequent questions on other topics (such as self-reported health status, or subjective poverty) remains an important area for further study.12 Until more is known about this impact, it seems strongly advisable to maintain some distance between these items in the survey through the use of more neutral transition questions and introductory text that clearly distinguishes between question topics.