7.2. Appropriate designs for different sampling methods

A questionnaire for use by a trained interviewer conducting face-to-face or telephone interviews can be much more elaborately constructed than one used for general distribution (say at a conference) or for a postal or email-based survey or a web-based survey.

In the former situation, the time and cost of obtaining the interviews at all is so large that the additional expenditure for training specialist interviewers and possibly providing them with aids such as laptop computers with purpose-designed questionnaire software is often considered worthwhile. Many large scale government-funded surveys seeking national statistical data are conducted in this way. Questionnaires may then have complicated question routing for various alternative pathways through the questionnaire, as indicated by answers to key initial questions. Sophisticated questionnaire software will have these paths encoded within it. In other cases the interviewer will be familiar with the routes to take depending on the responses given. This is the more common situation in surveys with visiting inspectors of the extension service, or in general in projects which require advanced diagnosis of disease and/or sampling of colonies.

In the second situation, where respondents to the survey have full control of the progress of the response and interpretation of the questions, it is vital that shorter and simpler questionnaires are used, with clear instructions and clear questions, to avoid low response rates and inaccurate responses. This is almost always the situation relevant to surveys of beekeepers.