Survey Research at WCU

1
Survey Research at WCU

• Hal Herzog

2
The truth about the subject pool

• Little incentive to give you good data.
• Usually young and inexperienced.
• Often have a haphazard approach to surveys.
• Compromise the validity and reliability of your
  results.
• Source of unwanted variation in data
  (experimental error)
• Reduce the effect size of your results

3
The truth about experimenters

• Often unconsciously give the message that the
  research is not important.
• Show up late.
• Dont show up at all.
• Treat subjects unprofessionally.
• Undergraduate assistants are not trained.
• Sloppy signage on rooms.
• Surveys are often poorly designed.

4
My experience with surveys

• Galvin and Herzog (1998) animal rights
  demonstrators
• Handed out 750 surveys
• Got 231 back (33 return rate)
• Scott Plous (1997) psychologists
• Sent surveys to 5,000 APA members
• Got 3,982 back (80.4 return rate)
• Plous and Herzog (2001) Animal care committees
• Sent 556 surveys to university faculty members
• Got 494 back (87 return rate)

5
Tailored Design MethodDon Dillman

• Mail and Internet Surveys The Tailored Design
  Method
• Based on social exchange theory (Thibaut and
  Kelly 1950)
• It works increases typical return rates from
  25 to 75.
• TDM principles apply to getting better data from
  WCU Subject Pool
• Basic information in Chapters 1 3.

6
Some Principles of Treatment of SubjectsReward
and Reciprocity

• Show positive regard
• introduce yourself
• ask their name
• Say Thank You in advance.
• Give tangible rewards in advance.
• Certificate of participation?
• Make the questions interesting.
• Sponsorship by legitimate authority.
• Make task appear important.
• Ask for advice.

7
Some Principles of Survey Design

• Booklet format
• Cover have a picture on it.
• Introductory letter.
• Order of questions.
• Pay attention to details e.g. font size, bold,
  italics, etc.
• Provide a navigational path.
• Use back page for open-ended question.

8
Some Principles for Writing Questions

• Use complete sentences.
• Choose simple over specialized words.
• Use as few words as possible to pose the
  question.
• Use mutually exclusive response categories.
• Separate undecided from neutral on Likert
  scale items.
• Included an open-ended question at the end.

9
Other tips.

• Consent form should be user-friendly.
• Question/answer format?
• Pilot studies are critical.
• Assess question clarity.
• These details make a difference!