The usual suspects: why don’t they participate?
Some groups are unlikely candidates for survey participation. However, even groups deemed to be very difficult may turn out to be not so difficult in the end. People with learning disabilities – and their parents and carers – were much keener to answer questions about their own situation than expected (Stoop & Harrison, 2012a). Ethnic minorities may be hard to reach but, once reached by someone who speaks their language, may participate just as often as the majority population (Feskens, Hox, Lensvelt-Mulders, & Schmeets, 2007). Busy people may have little time to spare but are used to doing lots of things quickly and are thus generally not underrepresented in burdensome time use studies (Van Ingen, Stoop, & Breedveld, 2009).
When trying to identify hard-to-reach respondents, it rapidly becomes clear that being hard to reach may be related to sociodemographic and socioeconomic characteristics of individuals under study, but also depends greatly on the survey design. Survey mode, in particular, has been identified as an important factor in reaching different groups in the population. Geographically isolated groups may be easy to reach by telephone; mobile-only households may have no problem with a web-based survey; linguistic minorities may be quite happy to complete a questionnaire translated into their language; and the illiterate may enjoy a conversation with a face-to-face interviewer. Mixed-mode surveys aim to overcome the fact that different groups of people may be more or less easy to reach and more or less willing to respond in a particular mode (de Leeuw, 2008; de Leeuw, Dillman, & Hox, 2008; Dillman & Messer, 2010). More generally, securing participation by the hard to reach or hard to survey will depend on the design and topic of the survey and the efforts one is willing to make.