ANALYZING AN EXISTING SURVEY

Without a doubt, the best way to get survey data is to design and carry out a survey focused on precisely the research questions you want to study. Realistically, though, you often have to settle for "re-using" a survey that somebody else has carried out. Using data from a survey that was not designed for your study is often called secondary analysis to distinguish it from the primary analysis that was the purpose of the original survey.

Secondary analysis lets you do research that you could not otherwise do all on your own. But you must keep in mind that the data were not collected specifically for your purposes. The survey questions may not have measured exactly what you wanted them to, but you are stuck with them nonetheless. Remember to interpret them as they were asked, not as you wish they had been asked.

When you plan to use existing data, you do not have to worry about the thousands of details that go into conducting a survey. Instead, you have to make sure that the survey was carried out properly in the first place. Was it conducted by a reputable organization? Were the questions well phrased? Was the sample well chosen ? Were the forms carefully processed ? Most important, have you formulated research questions that you can reasonably hope to answer with the existing data?