Chapter 7 : Questionnaire design
Questionnaire design
A questionnaire is only as good as the questions it
asks
The major decisions in questionnaire design in 5 steps
1. What Should Be Asked?
- Questionnaire relevance
– Each
question asked should have a reason behind asking it (they must be relevant to
your research objectives)
– Each
question asked should stem from your conceptual diagram
- Questionnaire accuracy
– Question
wording and sequence affect accuracy
– When
questions are not lengthy, difficult to answer, or personal opinion threatening, there is a greater chance of
unbiased responses
2.
How the question should be phrased
Open-Ended Questions
·
Complete
freedom of response (eg. short answer questions)
·
More
appropriate for exploratory research
·
Difficult
to code but not impossible
- Close-Ended Questions
· Fixed
alternative responses (eg. Multiple choice questions)
· More
appropriate for descriptive research
· Easy
to code for quantitative analysis
· But
little can be done to correct a close-ended question that does not provide
enough alternatives
Open-end
question: http://www.mediacollege.com/journalism/interviews/open-ended-questions.html
Closed end question
- Simple-Dichotomy Questions
– Two
alternatives: yes/no questions
Have you
travelled overseas in the last 6 months?
Yes
No
- Determinant-Choice Questions
- Multiple
choices (include Likert scales)
How many times have you travelled overseas in
the last 6 months?
·
None
· 1 - 3 times
· 4 – 5 times
·
6 or more times
3.
Checklist
Questions
Please
check which of the following sources of information you have used to gather trave information
·
Newspapers
·
Internet
·
Travel agent
·
Family and friends
4. Better
alternative to Checklist Questions
How often have
you used the following sources of information to gather travel information?
Never All the
time
Newspapers 1 2
3 4 5
6 7
Internet 1 2
3 4 5
6 7
Travel
agent 1 2
3 4 5
6 7
Family and
friends 1 2
3 4 5
6 7
Questionnaire Wording
v No hard and fast rules
v Only guidelines
Ø Avoid Complexity: use simple, conversational language
(avoid terms like “brand image” etc.)
Ø Avoid leading and loaded questions
Ø Avoid ambiguity: be as specific as possible
Ø Avoid double-barreled items
Ø Avoid making assumptions
Ø Avoid burdensome questions
a.
Avoid
Complex quest
Vitamino Plus
should be banned from Australia since it contains two compounds Benzoic
Altissimo and Citric Zen amine which are widely seen as the main cause for the
disease cardiac dyslectic and also the a chief contributor for conjunctiva
symbiosis. Do you agree?
Agree 1 2
3 4 5
6 7 Disagree
Complex, too
long, confusing, and leading
b.
Avoid
Leading question
1. Generally, it is perceived that Target provides better
service quality compared to K-Mart. Do
you agree?
Agree 1 2 3
4 5 6 7 Disagree
Leading because
of the “bandwagon” effect
2. Due to the current water crisis from extended
droughts, do you agree that the authorities should ban sprinklers in Western Australia?
Agree 1
2 3 4
5 6 7
Disagree
c.
Avoided
Loaded (full) Quest
Due
to the current water crisis from extended droughts in Western Australia, 90% of
the farming industries are out of jobs with a staggering loss of 50 million dollars
for 2001. Do you agree that the authorities should ban sprinklers?
Agree 1
2 3 4
5 6 7
Disagree
Loaded question which suggests a socially desirable
response
and emotional
How did you rank in your studies in your primary school?
Top 25%
Second 25%
Third
25%
Last 25%
Social desirability may influence the response
d.
Avoid ambiguous
1.
Ambiguous
question
Which media do you depend most for
getting the latest information?
Newspapers, TV , Radio
2.
Better
question
How often do you depend on each of the
following media for getting the latest information on News?
Never All the
time
Newspapers 1 2 3
4 5 6
7
TV 1 2 3
4 5 6
7
Radio 1 2 3
4 5 6
7
e. Avoid
mention of example
Do you agree
that leafy vegetables should be in your daily diet?(we cannot use example in
the question that are designed.)
Agree 1
2 3 4
5 6 7
Disagree
f.
Avoid
Making Assumptions
g.
Avoid
burdensome (troublesome) question
Burdensome
Questions: tax the respondent’s memory
a. A long list of burdensome questions may influence
accuracy by increasing mental fatigue(tiredness)
b. Open-ended questions are more burdensome than
close-ended questions in most situations
c. Aided recall in advertising research may reduce the
burden on the respondent’s memory
Wrong
example VS good example
Wrong example
How much do you depend on print media sources such as
magazines, newspapers and flyers to get information about products that you are
likely to buy?
Very
Little 1
2 3 4
5 6 7 A
Lot
Good example
How much do
you depend on the following sources of information when looking to buy a new
computer?
Very A
Lot
Little
Newspapers 1 2 3
4 5 6
7
Flyers 1 2 3
4 5 6 7
Magazines 1 2 3
4 5 6
7
Question sequence
a.
Start
with broad easy questions
–
This
creates interest and cooperation
b.
Demographic
and personal questions placed last
–
This
will reduce embarrassment and be less threatening
c.
Order
Bias
–
Created
by question order
–
Created
by order of determinant choices
d.
Funnel(guide)
Technique
–
Reduces
bias by first asking general broad questions then more specific questions
e.
Filter
Questions
–
Minimize
chance of asking questions that are not applicable to the respondent
Questionnaire layout
Questionnaire layout should make it look:
- Uncomplicated: uncluttered
- Easy to complete: clear instructions in bold
- Not too time consuming: use multiple-grid format (ie. tables in MS
Word)
- Clear: do not break questions over page
- Like an important survey: choose a good title (eg. “Internet
Banking Survey 2004”)
- Professional: if 4 pages or less, consider printing double sided
on A3 paper then folding to A4
Pretesting is important
1. Read the questionnaire yourself at least 10 times
- Then pretest with 10 – 20 respondents
3. Check for:
a.
Grammar
- Spelling
- Flow of questions
- Question wording
- Completeness of determinant choices
- Time taken to complete
No comments:
Post a Comment