Writing Good Questions

1
Writing Good Questions
2
A QUESTIONNAIRE IS ONLY AS GOOD AS THE
QUESTIONS IT ASKS
3
(No Transcript)
4
(No Transcript)
5
Plan What to Measure

• What information is required?
• Who are appropriate target respondents?
• What data collection method will be used to
  survey respondents?

6
Preface Open-ended versus Close-ended Questions
7
Close-ended Questions
Advantages Disadvantages

• Communication skills of respondent less critical
• Speedy response
• Easier to answer
• Data quickly coded entered
• Easier to analyze
• Less-skilled or no interviewer needed

• Cant obtain in-depth response
• Poor at providing new insights
• Harder to write
• Answer may not fully reflect respondents
  attitude
• Categories hint at right answers

8

9
Open-ended Questions
Advantages Disadvantages

• Wide range of responses
• Encourages response
• Good for probing

• Biased by respondent articulateness
• Interviewer bias
• Hard to record answers
• Coding inconsistency
• Hard/costly to code
• Reduced cross-study comparability
• Tabulating complexity
• Costly

10
(No Transcript)
11
Guidelines for Writing Good Questions
12
(No Transcript)
13
Be Clear and Precise
14

15
(No Transcript)
16
(No Transcript)
17
Responses Should Be Mutually Exclusive and
Exhaustive
18
(No Transcript)
19
Use Natural and Familiar Language
20
Use Natural and Familiar Language

• Grinder
• Hoagie
• Hero
• Submarine
• Poor Boy

21
Avoid Leading Questions
22
Avoid Leading Questions
Leading version 1. Do you believe that private
citizens have the right to own firearms to defend
themselves, their families, and property from
violent criminal attack? Yes No Undecided Impr
oved version 2. Do you believe that a ban on
the private ownership of firearms would be
significantly reduce the number of murders and
robberies in your community? Yes No Undecided
23
Avoid Double-barreled Questions
24
Avoid Double-barreled Questions
25
Ask One Question at a Time
26
State Alternatives Explicitly
27
Questions Should Yield Reliable and Valid Answers

• Relevance
• Memory
• Omission
• Telescoping
• Creation

28
Provide Appropriate Time Referents
29

30

31

32
Example of Order Bias
33
Sources of Order Effects in Self-administered
Questionnaires
34
Other Guidelines
35

36

37

38

39
Procedures for Softening Impact of Potentially
Objectionable Questions
40
Asking Sensitive Question

• Respondent randomly selects which of two
  questionsone innocuous and one sensitiveto
  answer
• Questions Last digit of SS odd?
• Sensitive question?
• n1000
• If 300 yes answers, then 50 of 500 (10)
  answered yes to sensitive question

41
Poor Questions
42
(No Transcript)
43
(No Transcript)
44
(No Transcript)