An exploration of demographic bias in a questionnaire survey of hearing-impaired children: Implications for comparisons of children with and without cochlear implants

SummaryObjectivesResponders to questionnaire surveys, who are self-selecting, are generally accepted to be unrepresentative of the total available population, at least in demographic terms. Since demographic and other variables are known to be predictive of outcome, it is important to understand the extent of that unrepresentativeness when using survey data to report comparisons of outcome. This paper aims (i) to evaluate the extent to which a sample of hearing-impaired children surveyed by postal questionnaire was representative of the population of hearing-impaired children in the United Kingdom (UK), and (ii) to identify demographic differences between children with and without cochlear implants.MethodsData from a previously reported total ascertainment of hearing-impaired children in the UK which identified 17,160 with permanent bilateral hearing impairment >40 dB HL were compared with data collected by postal questionnaire for a sample of 3224 children, including 527 with cochlear implants.ResultsThe sampled children were similar to the ascertained population in gender, age at onset of hearing impairment, and number of additional disabilities, but came from more recent birth cohorts and from more affluent families. Compared with profoundly impaired non-implanted children, implanted children had greater degrees of hearing loss, fewer additional disabilities, a later age of onset, were younger, came from more affluent families, were more likely to use spoken language at home, and to be taught using spoken language only.ConclusionsComparisons of outcomes and generalisation of results require adjustment for relevant variables to avoid confounding estimates of the effectiveness of interventions including cochlear implantation.