The effects of question order and response-choice on self-rated health status in the English Longitudinal Study of Ageing (ELSA).

BACKGROUND

One of the most ubiquitous global health measures is a single self-rated health item. This item may be sensitive to its position in questionnaires and to response-choice wording. The aims of this paper were to investigate the effects of question order and response choice on self-reported health status.

METHOD

A secondary analysis of wave 1 of the English Longitudinal Study of Ageing (ELSA). Participants were a nationally representative sample of people aged 50 years and over living at home. Over 11 000 respondents were interviewed face-to-face in their homes, and were randomly assigned to one of two versions of a self-rated health item.

RESULTS

The health status item asked after, rather than before, a module of health questions, resulted in more optimal health assessments, although the effect size was small. The version of the health status item with "excellent", rather than "very good" as the first response category resulted in more optimal health assessments, although it had a smaller ceiling effect.

CONCLUSIONS

There was support for the insertion of the health status question at the beginning of health questionnaires, as it may be influenced by questions about health and disease if placed at the end, although the effect size was small. Evidence for the version of the item with "excellent", rather than "very good", as the first response choice was more mixed as, although optimism bias appeared higher, the ceiling effects were lower. The smaller ceiling effects for the "excellent" version has important implications for the ability to detect improvements in follow-up studies.