Business Research Methods: Questionnaire Design
Let parsimony, relevance and elegance guide you when designing a questionnaire for quantitative research.
If it is decided that a quantitative approach will be relevant to answering the research questions set, then it is likely that a questionnaire of some sort will be required. In fact, a questionnaire is just one of the various types of research instrument that can be used but it is probably the easiest to imagine and the most commonly used. How, then, should it be designed?
The principles to bear in mind include parsimony, relevance and elegance. The first, parsimony, means reducing the number of questions to be asked (or observations to be made) to the least possible. Research, as I have described elsewhere, is a labour-intensive process and, especially in western or developed countries where it is necessary to use qualified personnel, that makes it very expensive. Reducing the unit cost of questionnaires will have a significant impact on the overall cost of the project. In addition, the longer that questionnaires take to complete, the lower the quality of the data. This is because people, respondents, tend (unless they are very keenly engaged with a product or service) to become bored with subjects and will start to answer quickly and thoughtlessly just to get it all over with.
Relevance relates directly to the research questions. There are again efficiency and cost implications in this, since asking questions that are not directly relevant to answering the research question confuses the situation and may lead to raised expectations from the client: do not attempt to answer questions you are not being paid to answer – this is a business and not a charity. So, irrespective of how interesting an additional question may be, cut it out if at all possible.
Third, elegance, relates to the wording of a question and the way it may be answered. This concerns the way in which accurate information is sought, bearing in mind that people will have incentives not to answer certain questions accurately. In particular, try to avoid asking people about personal details that may be sensitive. The obvious example of this is in terms of income. There are all kinds of reasons why people will not wish to reveal their salary and, in any case, in contemporary society the level of an individual’s salary (if any) may be only tangential to overall income. In such cases, find a proxy that may do the job for you: some attempts to do this include finding out what car a person drives, what kind of watch is being worn, location of house and so forth. The more sophisticated and less obvious the approach, the more likely it is that people will allow their guard to slip and to give not just an honest answer but a useful one.