A study of non-response in a questionnaire survey of parents' views of paediatric care

The aim of this study was to examine whether there were differences in quality ratings between respondents and non‐respondents to a questionnaire concerning parents’ views of paediatric care. This study also examined whether quality ratings differed when questionnaire respondents were asked to reassess certain aspects of hospital care in a follow‐up questionnaire. A total of 1094 questionnaires were distributed to parents at the hospital. Three weeks later, follow‐up questionnaires were sent home to a random sample of 140 parents who had visited the hospital during the 2 weeks when the main questionnaire was distributed. Forty‐six per cent of respondents to the follow‐up questionnaire had never received the original questionnaire in hospital, while only seven individuals in the follow‐up sample were active non‐respondents. Analysis of variance revealed that respondents to the follow‐up questionnaire who had never received the main questionnaire did not differ significantly from respondents to the main questionnaire in their ratings of key quality domains. There were no statistically significant differences in quality ratings between parents who responded to both questionnaires and parents who responded to the main questionnaire. For hospital management, it is important to be able to trust questionnaire results. Non‐response bias can jeopardize the validity of questionnaire results, which is why studies of non‐response are important. The current study pinpoints a number of difficulties that need to be considered when conducting investigations of non‐response.