Brief Summary of Chapter 8

In Chapter 8, we discuss the most misused of all research methods--the survey method.

Many people use surveys when they shouldn't-- survey methods are not useful for every problem. If you want to know whether a treatment causes an effect, you need to do an experiment, rather than a survey. If people are going to lie, then you should use a correlational method other than the survey (see Chapter 7 to learn about those methods).

Many people don't spend enough time planning their survey. Consequently, they may

  1. conduct a survey without having hypotheses.
    As a result, at the end of the study, they have a great deal of data, but no meaningful information. Before conducting your survey, we advocate not only having hypotheses, but also knowing what pattern of results will support (and what pattern won't support) those hypotheses.
  2. have questions that participants don't understand or interpret differently than the researcher intended.
    We advocate editing questions based on our checklist and then pilot testing your questions.
  3. be overwhelmed about how to make sense of participants' responses.
    We advocate using fixed-response questions (filling out rating scales, answering multiple-choice questions, etc.) rather than open-ended questions (such as essay questions)
  4. be unable to generalize their results.
    We suggest that you should spend the time to identify your population and then, if possible, take a random sample from that population.

Back to Chapter 8 Menu