I do not quite understand your question: you create the survey according to your research question(s). And you identify your sample according to your research. For example, we are preparing a research on museum marketing and part of our sample is people who never visit a museum..