Week 9.
Surveys & questionnaires.
Lecture notes
Uses of surveys:
Core topic areas for survey research:
Survey format issues:
General issues in survey design:
|
 |
Lecture notes are here.
Readings
Chapter 13, plus readings on decreasing value of political polls, biased polling, and whether people lie on sex surveys (click "hands counting" image). The web site I mention in lecture as an example of a biased survey is at: http://edworkforce.house.gov/.
Click the image for the article describing a survey study of smoking among gay & bisexual men (used in lecture).Discussion group Assignment
Word version of Week 9 Assignment.
Design a survey
In the text book (see Box 13.1) and in class we reviewed four steps to survey design first outlined by Campbell & Stanley (1953). Think of a survey topic - people's attitudes toward some topic, knowledge about some issue, or some important behavior(s) - and using Campbell's four steps, briefly describe how you would design a survey.
1. What general theme or question are you interested in.
- What is the "phenomenon" you will investigate?
- Why is it important? What theory underlies your question or approach?
- Hint: keep your overall question specific and straightforward. You will need to operationally define it using a (relatively) small number of survey items.
2. Operationally define your phenomenon: Are you assessing knowledge, attitudes or beliefs, or actual behaviors (or some combination?
- Why are you assessing the term(s) you chose?
- How would the study differ if you operationalized your phenomenon differently, e.g., by assessing self-reported behaviors rather than attitudes, or knowledge instead of behavior?
- Can participants validly report the knowledge, attitudes or behaviors you are interested in?
3. What sampling approach will you use? (Sampling notes here or in the text).
- Will you use a probability or a non-probability sample?
- What form of sample will you use within each of these categories, and what are the implications for your results of choosing one or another?
4. Write some questions.
- Your items can be open- or closed-ended, Likert-type rating scales, checkboxes, or simple multiple-choice questions.
- Why did you choose the type of question you did, and how will that choice affect your results?
