Reformatting examples for survey questions with highest non-response rates
| Question | Non-response rate (%) | Details of reformatting |
| B12 (four parts) | 44 | On a scale of 1–10, how satisfied are you with how you receive information about your baby on the neonatal unit? This question contains four subquestions (verbal information, telephone information, written information and overall). There was a large proportion of non-respondents on analysis for telephone (34%), written (44%) and overall (8.1%). As these performed well during cognitive testing, this was likely due to the questions’ format. We have added a letter prefix to each subquestion (a, b, c and d) so it is more obvious all questions need to be answered. We have added the response option N/A to all subquestions. |
| B14 open | 73 | If there is anything else you would like to tell us about how you receive information about your baby on the neonatal unit, then please do so here. In view of the high proportion of non-respondents and the fact that another open question at the end allows parents to give similar feedback, we removed this question. |
| E1 open | 61 | If there is anything else you would like to tell us about the way you are given updates about your baby on the neonatal unit, then please do so here. After removing question B14, this remains the survey’s only generic open question. We reworded it to ‘If there is anything else you would like to tell us about your experience of care on the neonatal unit then please do so here.’ |
N/A, not applicable.