Open-ended questions
Open-ended questions provide no indication of possible answers and rarely define any parameters to restrict the respondent. These are essentially descriptive questions that require a more detailed and personal response. Very often I see them being described as the qualitative element in a questionnaire. This is something of an over- statement; they are descriptive but rarely are they truly qualitative. The depth of narrative needed to label something as qualitative data is almost impossible to harvest from a questionnaire. The descriptive data that open-ended questions may produce can add detail to the closed question and can often bring a totally new perspective to an issue, one even the researcher had not considered, but please do not rely on them as a source of qualitative data. If your purpose is to harvest both quantifiable data and individual detail then I would urge you to employ more than one data collection technique in your research design. That is not to say open-ended questions are unsuccessful, they can be, but it’s what they are successful at that is the issue. Relying on respondents to fill out vast empty spaces on a questionnaire is not something I would ever recommend except in the case of a Delphi questionnaire, which is something quite different aimed at a very particular audience. So, enough about what open-ended questions can’t do, what cam they do?