Click on the following links. Please note these will open in a new window.

Group Projects

1. Have each group develop a ten-question closed-ended questionnaire for a survey about student stress (or some other topic).   Each member of each group should keep a copy of the ten questions.  Once the groups are ready, pair each individual with a person from a different group.  Have them conduct cognitive interviews with their partner, in which they explain their responses.  When interviews are complete, have each group reform and rewrite their questionnaire, given the feedback from the cognitive interviews.
    
2. Select a topic of interest for the class or allow smaller groups to select their own topic.  (An easy topic of interest is satisfaction in relationships; there are ample data available in scientific and popular print!) Instruct each group to collect all the survey data they can find on this topic from academic and popular sources. Have them collect not only the findings, but also the surveys themselves. Identify the most popular survey questions (excluding demographic measures) as well as the most unusual questions or measurements. Given all this information, have the students design what they believe to be the best survey possible. Remind them that surveys should not only be exhaustive, but also as brief and easily understood as possible.
    
3. Reach out to colleagues in your department, asking them to give over 10 minutes of class time for students from your class to survey other undergraduate students. If you can get enough agreement/appeal from other professors, have students craft an undergraduate social work survey that addresses topics such as enjoyment of classes, rigor, life outside the classroom, and professional plans. Have students collect and code data from as many classes as possible. Discuss the results in class. Finally, prepare a one page summary to be distributed to all the classes and students who participated.
          
4. Have a “worst question” competition.  Ask students to devise three questions for a survey by a political campaign or to develop information useful in marketing a product.  Choose the general focus in advance.  Each student should write questions that violate the guidelines for writing questions in the textbook.  Students should try to phrase their questions not only so that they result in misleading answers that are favorable to the political campaign or product, but also so that they would appear to many people to be perfectly reasonable questions.  Distribute a list of all the questions to the whole class and have them discuss what’s wrong with each question.  Have students vote on the worst question.  You may conclude by having students rewrite questions so they conform to guidelines outlined in the chapter.
    

Mini Projects

1.  Survey Research in Existing Literature 
As surveys are among the most popular forms of research published in the social sciences, this exercise allows you to praise or critique surveys for their quality. 

1. Locate five research articles in peer reviewed social work journals that present survey data.
2. Note the following information for each survey:  sampling design, manner of administration, response rate, and number of surveys completed.  This information should be easily identifiable in the methods section of the article.
3. Note the key findings of each article, as stated in the abstract (or in the absence of the abstract, in the findings section of the article). 
4. For each article, discuss the generalizability of findings based on the survey.  In particular, consider any potential problems for generalizability based on response rates.

2.  Open- or Closed-Ended Questions 
Any decision about research design involves trade-offs and compromises between such goals as achieving measurement reliability and validity, generalizability, and cost-effectiveness.  This project focuses on one such trade-off:  the decision to use open- or closed-ended questions.

1. Begin by formulating a research question that is appropriate for a survey of college students.  Identify at least one hypothesis and several variables that are involved in the question.
2. Draft 12 closed-ended questions that measure these variables.  Some questions may be part of an index; some may be alternative measures of the same variable.  For a shorter project, select your questions (and the research question) from the General Social Survey or another instrument to which you have access.
3. Circulate your questions for classmates to critique.  Consider carefully their comments and revise your questions as warranted.
4. Now write three to four open-ended questions that you think will measure the same variables as the closed-ended questions.  Circulate your questions to classmates for critique and revise as warranted.
5. Prepare the closed-ended questions as a short questionnaire and the open-ended questions as a short interview schedule.
6. Administer the closed-ended version of the survey to five other students.  It will be easiest to do this one student at a time rather than in a group.  Ask each student to answer all the survey questions while you wait.  Collect their survey sheet.
7. Now interview each student using the interview schedule.  Explain that you will be going back over the same issues, but with somewhat different questions.  Record their responses verbatim. 
8. Ask your respondents whether they provided different information in response to the interview than in the questionnaire.  If they perceive any such difference, ask them to explain.  Then thank your respondents.
9. Compare responses to the open-ended and closed-ended questions.  Prepare a summary sheet that indicates the differences you found for each respondent.  Write a short report on the differences and similarities.  What changes would you make in the closed-ended questions based on the open-ended responses?  Has your confidence in closed-ended questions been increased or decreased?  Explain.

3.  A Short Survey Two Ways 
Many people are used to receiving tons of junk emails as well as junk snail mail. The goal of this exercise is to see how likely individuals are to respond to two different survey mediums, while more broadly learning first hand the largest weakness of surveys as a research tool.

1. Select a current issue in the news of interest to social workers that can be assessed relatively easily. For the purposes of this exercise, avoid topics that are very emotionally charged.
2. Pose one or more hypotheses about possible influences on attitudes about this issue or concern.
3. Write about ten questions that measure the variables in your hypothesis/es.
4. Organize your questions into a logical flow and pretest them on classmates. 
5. Develop a brief introductory statement that covers the focus of your survey, its auspices (a class project at your college or university), and its length (hopefully, only a few minutes).  Your statement should include the request that you would like to speak to any adult in the household.  Try this out on a few friends and make whatever revisions seem warranted.
6. Select one or more communities in which you will conduct your mail survey as well as the emails of releant individuals of a similar magnitude. Devise a scheme to sample randomly the appropriate range of emails and addresses (using a random numbers table).  Select 25 random emails and addresses.
7. Describe your experiences in a short paper.  Report on the number of emails and mailed surveys answered and responses to your survey questions.  What have you learned about email interviewing? What have you learned about snail mail interviewing? Was there a difference in the respondent rates? Which survey questions seem to need improvement?