7.1. Completeness of the questionnaire
In the framing of survey questions, the aims and objectives of the desired analysis, and the methods to be used in the analysis, should be borne in mind, in order to ask all of the questions needed to enable collection of the appropriate data. For example, for modelling the odds of colony loss through CDS, questions should be asked relating to any suspected risk factors as well as collecting data on numbers of colonies lost and what number of losses are attributed to each of a list of possible causes.
In a survey concerned with honey bee colony losses, migratory habits need to be stated as clearly as possible by the respondents in order to avoid misunderstanding about the place(s) where losses were/were not recorded, as well as possible causes of losses, since migratory habit could be one such cause or a contributing factor in honey bee loss.
It is also important to collect and record auxiliary information for statistical purposes such as weighting and multivariate analysis.