ABSTRACT
In our experience of using questionnaires we have found that the provision of short and clear instructions to the respondent, at relevant places in the questionnaire, can generate more useful and reliable data. If zero is a meaningful answer at some point in your questionnaire, make this clear to those completing. For example, you may be interested to know, as part of a health evaluation programme, how many times your respondents have visited a doctor in the last six months. Without an instruction to the contrary respondents might skip this question if they have not visited a doctor in the preceding six months, and you will lose valuable data as a result. At the analysis stage this will be viewed as a nonresponse and will have an impact on any claims you make about the data.
