Chapter specific application exercises will help you think about research design in practice or have you explore a relevant resource.
Exercise 1: Comparing Survey Methods
Many survey methods have advantages and disadvantages in common. To help you think about survey methods more critically, identify a unique advantage and disadvantage, not shared with any other methods, for each of the following survey methods: mail, face-to-face interview, email, web interface, landline phone, smartphone.
Exercise 2: Survey Wording and Formatting
Are you in favor of ____________? Yes / No
Fill in the blank in the above question with a topical or campus issue. Now take this basic question and rewrite it in open-ended, multiple-choice, semantic differential, and Likert-type formats so that you are capturing more subtle responses to the question. You may need several questions in each format. Check your wording carefully for any leading or double-barreled questions. Having written the questions, which, if any, would you now eliminate, and why? Remember that there may be good reasons for keeping questions that appear to duplicate each other. What additional questions will you need to capture information about your respondents as well as their knowledge of the issue and their possible actions toward it?
Exercise 3: Mobile Technologies
This chapter suggested that successful use of mobile technologies for surveys means that researchers must first successfully contact respondents and then present survey questions that will motivate respondents to begin the survey and continue with it until all questions have been answered honestly by all respondents. Assume that you need to survey your campus community on a contentious issue. How would you contact respondents in the mobile community? What incentives, if any, might entice them to begin—and complete—your survey? To minimize the break-off rate for cell phone, smartphone, and tablet users, how many questions do you think could be asked? What format(s) should they have?
Exercise 4: Balancing Respondent and Researcher Interests in an Age of Gender Fluidity
Traditionally, survey researchers provided two response options with respect to gender—male or female. Facebook now recognizes over 50 options. Suppose you are designing a survey and wish to maximize respondents’ opportunities to provide their gender identity. You have several options:
- A list of each option along with a “check one” instruction.
- A list of each option along with a “check as many as apply” instruction.
- A text box with a “please enter your gender identity” instruction.
- A reduced number of options designed to meet your research interests.
- The traditional male/female option.
- Don’t ask the question; gender is theoretically unimportant to your research.
Consider that you need to make the user experience relevant and comfortable for respondents, especially perhaps with how they would want to describe themselves. Consider also that you need to be able to record, store and analyze all your survey data. If you do not intend to analyze 50 different response options why provide them? Consider also sample size. If a survey sample of 100 people had 50 gender categories equally represented, you would have only two people in each category.
How does your thinking about this issue change if designing your survey specifically for (a) the web (b) smartphones?
For further explanation of the Facebook categories see, for example,