Designing an effective survey requires as much art as skill. The more practice you get, the more you will come to appreciate the nuances of survey design. The process of designing a survey involves numerous iterations, with input from multiple members of a research team or sometimes from an advisory panel. Draft surveys are usually pre-tested and revised before they are administered. In the end, every little detail potentially matters and may contribute to measurement error, non-response bias, or both. Issues to think about: Questions Rumor has it that students at FAU are biking less than they used to. This might be happening for any number of reasons. Hypotheses include: Boca Raton has grown so students are living farther from campus (or outside of Boca) Students are biking less as children and are less comfortable biking when they get to Boca Students are working harder than ever and don’t have the time to bike Biking is not seen as particularly cool any more It’s not possible to go back in time to do a longitudinal study to see how these factors have changed, or even to see if biking levels have in fact changed. But it is possible to do a cross sectional study that determines the relative importance of these factors in explaining current levels of biking among students. Provide the following: Who do you want to survey? Who is your population? How will you access them? Write a question to assess the dependent variable. Open ended, closed ended, scale? Should this question be quantitative (measureable)? Write one question for each of the hypotheses listed above. Open ended, closed ended, scale? Should this question be quantitative (measureable)? Are there any other variables that might be influencing the level of biking? What else might you ask to find out? How would you pretest your survey? How would you administer your survey – mailed, group, phone, in-person interview, or electronic? What are the pros and cons of this approach? ...