Select a specific group of people as the focus of your research and limit your survey to them. Avoid establishing an initial screen that is too limited to encompass enough individuals to sustain your research. For instance, if you are studying the patient opinion of the rehabilitation center at a local hospital, you can establish an initial screen by selecting only people who are patients at the hospital to take your survey. This screen will limit your research to only individuals who may have access to the rehabilitation center.
Include a second screen within your survey by asking survey-takers to comment on the focus of your research. Ask them questions about how they have interacted with your focus, such as the number of times they interacted with your focus and the amount of time they spend interacting with your focus. As an example, you can ask your survey takers how many times they visited the rehabilitation center in the hospital and what activities they participated in while they were there.
Separate your surveys into categories, with one stack for surveys of individuals who never interacted with your focus and another stack for individuals who have interacted with it. Store the unusable surveys in a secure location. Avoid disposing of these surveys, in case you decide to use them for other research purposes or as a comparison against your usable surveys. For instance, you may decide later to compare the patient's opinion of the rehabilitation center to other patients' opinions of the hospital in general.
Review the remaining surveys, focusing specifically on their relation to your research. Analyze the survey results, looking for any similarities in people's responses. Alternately, you can further separate these surveys based on additionally screening questions in your survey, such as separating them by the number of times they interacted with the focus of your study. Use this technique to determine the differences between people who interacted heavily with your focus and those who only interacted with it rarely. As an example, you may separate your surveys by people who used the rehabilitation center a few times and those who used it many times over a long period.