Survey Research

1
Chapter 9

• Survey Research

2
Topics Appropriate for Survey Research

• Exploratory, explanatory or descriptive
• Usually individuals (as opposed to institutions,
  groups, etc.) are the units of analysis
• Need to think through if a survey is a good way
  to collect the information (i.e., you wouldnt do
  a mail survey to find illegal aliens)

3
Guidelines for asking survey questions

• Open-ended questions (respondents write in their
  own answer in their own words)
• Close-ended questions (respondents choose from a
  list of answersALWAYS include a not applicable
  or other category)
• Make items clear
• Ask about a single topic with a single question

4
Guidelines for asking survey questions (cont.)

• Ask questions the respondent can answer
  Examples
• Few people would be able to answer what the
  future value of their social security account is
  in terms of an annuity adjusted for inflation to
  2001 dollars
• Infants cannot answer a survey themselves
• Illiterate people could not be expected to answer
  a 34 page written survey

5
Guidelines for asking survey questions (cont.)

• Respondents must be willing to answer
• Dont expect drug users to identify their drug
  source on a written survey
• Illegal aliens are unlikely to tell you where
  they are employed.
• Questions should be relevant
• Dont ask someone in MT about how they feel about
  recent legislation affecting Texans

6
Guidelines for asking survey questions (cont.)

• Short items are best
• Just spit it out and ask it directly
• Try to keep all in a positive tense
• Less confusion about what youre asking
• Avoid biased items
• Like the PETA questionnaire we looked at in class

7
Questionnaire Construction

• Looks count
• uncluttered, consistent, professional looking
• Contingency Questions
• If yes then go to question 9 (for example, if
  you are a pet owner then go to the questions
  regarding pets, otherwise skip the pet questions)

8
Question Construction (cont.)

• Matrix questions
• Put like things together, so that people get in
  the swing of answering them (be especially
  careful here not to start doing some in positive
  and some in negative tenses it will goof-up
  your answers)

9
Question Construction (cont.)

• Ordering items is important it will affect the
  answers
• ALWAYS pretest your questionnaire
• Write clear instructions on how to fill out your
  questionnaire and how to send it back to you.

10
Self-Administered Questionnaires

• Mailed questionnaires
• There are recommendations (i.e., the Dillman
  method) for how to mail, follow-up, etc.
• Monitor your returns
• Response rates (50, 60, 70 from the book 33
  more of a reality for university studies) (Have
  to worry about response bias)

11
Interview Surveys

• The role of the survey interviewer is important
• Consistency of asking questions from observation
  to observation
• Recording exactly
• Probing for responses consistently
• Bias from interviewer to interviewer

12
Telephone Surveys

• Need to make sure telephones is an appropriate
  way to reach your audience
• Length of survey is always a concern (expense
  wise and how much time people are willing to talk
  on the phone to you)

13
Survey Research Pros and Cons

• Weak on validity
• Hard to get to the real content youre after
  not as accurate as observing directly or living
  among a population
• Strong on reliability
• Same survey goes out to everyone consistency of
  how things are asked is high