Key messages on question order within a subjective well-being question module
Although overall effect sizes appear to be small in most cases, the presence of order effects within groups of subjective well-being questions has some clear implications for survey design. First, it seems advisable to ask the most general evaluative questions first, followed by domain-specific evaluative questions as necessary. If evaluative subjective well-being is measured by single-item scales, using only one of these measures should reduce redundancy and any potential for respondent confusion or fatigue. This means that a choice must be made between, for example, the Cantril Ladder, a life satisfaction question and an overall happiness question, rather than including them all in one survey. Where domain-specific measures are to be included, covering a wide range of domains should reduce the likelihood of respondents focusing on any one particular domain (such as marital or relationship satisfaction).
The approach of running from the general to the specific suggests that surveys should move from global evaluative measures to eudaimonic questions and then to more specific affect measures - although further work is needed to explore how responses to each of these questions may interact and concerning the buffering text and/or question instructions that might be best used to break up the question module. In the case of affect measures, although in theory randomised presentation of positive and negative affect items is likely to represent the optimal solution, in practice this could heighten the risk of respondent or interviewer error, particularly in less technologically-advanced survey settings. Thus, the best way to ensure comparability of results may be to use a fixed item order across all surveys. This requires further investigation.