TataKelola dan KamSiber Kecerdasan Buatan v022.pdf
Week6 7a- developing a questionnaire
1.
2. A form containing a set of questions asked to
respondents, and designed as a way of gathering
information for a survey.
It serves four basic purposes :
to collect the appropriate data
to make data comparable and amenable to analysis
to minimize bias in formulating and asking question
to make questions engaging and varied
It can be a modified instrument or, an intact instrument
that someone else has already implemented or, an
original document
3.
4. Decide what information you need to
collect for your survey
Clarification of purpose and objectives helps to
gain a focus on specific information required
from the research.
A good questionnaire should meet the research
questions and research purpose / objectives.
Each question in the questionnaire should thus
be directly linked to the research question(s)
and the purpose /objectives.
5. Define the target population – who do
you wish to target?
Provide a clear, concise and unambiguous
formulation of who the target population is,
i.e. who the intended study is aimed at.
A clear description of the target population
allows the researcher to ask relevant
background questions and to formulate the
questions in such a way that it is understood
by the respondents.
6. Formulate the questions in order to answer
the research question(s)
List ALL the information needed to be able to
answer all the research question(s)
Scrutinise the existing body of knowledge, e.g.
available questionnaires or articles regarding similar
studies in order to select questions that are relevant
for your purposes
Adapt and refine the available questions
Formulate your own questions where necessary
Decide on the appropriate level of measurement for
each question
Determine the appropriate statistical techniques to
be used.
7. Organise the questions
Questions should be organised in logical order
e.g. background questions in the same section;
opinion-related questions dealing with the same
topic in a separate section, etc.
It is often a good idea to start with a question
that is non- threatening, yet interesting to the
respondent.
Provide Clear and Easy- to – Read Instruction to
the questions.
8. Consult experts
Once you have a first draft of the
questionnaire:
Consult experts in the specific field of study
to ascertain their opinion whether you have
addressed all relevant issues and formulated
the questions in an understandable and
unambiguous way
Consult an expert in questionnaire design to
assist with the formulation of the questions
and the response format.
9. Do a pilot study among a small group
of respondents similar to those in the
target population
A pilot study is imperative in any study since
it
Enables the researcher to identify and rectify
problems prior to the survey being
conducted.
Provides an indication of the response rate
that can be expected.
10. Adhere to ethical standards
The rights of respondents as human beings should be
respected at all times (Cohen, Manion & Morrison 2004). This
implies that:
The decision to take part in a survey (by completing the
questionnaire) remains the choice of the respondent .
A respondent should be given the option to withdraw from the
study at any time
A respondent cannot be coerced into providing information,
especially not information that may be perceived as sensitive or
incriminating
Respondents should be given the assurance that their responses
will remain anonymous and that the information they provide will
be treated as confidential at all times
11.
12. There are two categories of questions that can be included in
questionnaires: open-ended and closed-ended.
OPEN CLOSED
Elicit “rich” qualitative data
Require The respondents to give
opinions, ideas , suggestions or
comments and allow respondents to
provide lengthy responses.
Encourage thought and freedom of
expression
May discourage responses from less
literate respondents
Take longer to answer and may put
some people off
Are more difficult to analyze –
responses can be misinterpreted.
Elicit quantitative data
Require the respondents to provide
opinion, ideas or comments based on
the option given.
Can encourage ‘mindless’ replies
Are easy for all literacy levels to
respond to
Are quick to answer and may improve
your response rate
Are easy to ‘code’ and analyze
13. Examples of open-ended questions:
1. Which cellular phone service provider do you use?
__________________________________________
__________________________________________
2. Why have you chosen the service provider?
___________________________________________
___________________________________________
15. Scale Questions
The most commonly used scale question is the Likert-scale
questions. Likert-scale questions may use quantifiers such as
Strongly Agree, Agree, Disagree, Strongly Disagree.
Examples of Likert-scale questions
Please indicate the level of agreement with each statement by ticking (✔)
the appropriate bracket.
Strongly Agree Disagree Strongly
Agree Disagree
1. I use my hand phone everyday. ( ) ( ) ( ) ( )
2. I change my hand phone every ( ) ( ) ( ) ( )
year.
3. I can not survive without my ( ) ( ) ( ) ( )
hand phone.
http://www.gifted.uconn.edu/siegle/research/instrument%20reliability%20and%20vali
dity/likert.html
16. Listing / Choice Questions
Listing or choice questions provide items or choices for the
respondents to choose.
Example of Listing /choice question
For the question below, you may tick (✔) as many boxes as you
think appropriate.
1. Where do you get information about a new hand phone brand?
Friend ( )
TV/Radio ( )
Newspaper ( )
Magazine ( )
The Internet ( )
Others, please specify __________________
17. Yes/ No question
Yes / No questions require the respondents to
tick the appropriate box either Yes or No.
Example of yes/no question
Please tick (✔) the appropriate box.
YES NO
1. Do you receive a loan? ☐ ☐
18. Ranking questions
Ranking questions provide degree of preference or
choice.
Example of ranking question
1. Please rank the following criteria you use in buying
hand phone.
( 1 – most preferred to 5 – least preferred )
__________ Price
__________ Coverage
__________ Brand Name
__________ Size
__________ Function
19. Category questions
Category questions ask respondents to identify the
response provided in different categories , for example
, age range, monthly income, etc.
Example of a category question
Please tick (✔) in the appropriate space.
1. How much do you spend to pay the phone bill each month?
_________ below RM 50
_________ RM 51 – RM 100
_________ RM 101 – RM 200
_________ more than RM 201
20.
21. Question Wording
The wording of every question is of the utmost
importance. A ‘good’ question should satisfy EACH of
the following criteria:
Avoid Complexity
Avoid leading questions
Avoid loaded questions that contain words which may bias
the responses
Avoid Ambiguity
Avoid jargon and colloquialisms
Avoid double-barrelled questions
Avoid “ combination” questions
Avoid double negatives
Minimise bias
22. Avoid Complexity
- The way people are asked to show their responses should be
simple
Poor item : Are you aged between 20 and 30? Delete which ever
does not apply.
Yes No
Better item : Are you aged between 20 and 30 inclusive? Tick
appropriate answer.
Yes No
- Ask them to tick what does apply, rather than delete what does
not. Ticking or circling is more definite than underlining. You can ask
people to put a (x) against the alternative which applies, but this
may cause problems if people think of X as indication “wrong”. A
tick is less confusing.
23. Avoid leading questions
- A leading question is a question that suggests or implies
certain answers.
e.g. “Wouldn’t you say that...”, “Isn’t it fair to say...”
Poor item: Do you agree with most people that capital
punishment should be restored?
Yes No
Better item: Do you believe that for some crimes capital
punishment should be restored, should not
be restored or do you have no opinion ?
Should be restored ; Should not be restored ; No
opinion
24. Avoid loaded questions that contain words
which may bias the responses
- A loaded question is a question that suggests social-
desirability answers or is emotionally charged.
Poor item : Do you agree that racist organizations such as the
ABC should be banned ?
This is a poor item because it labels the target organization as
racist ; the respondent might not have considered it as racist
without this suggestion.
25. Avoid Ambiguity: Be as Specific as Possible
- Avoid words like “regularly”, “often”, “locally” , “occasionally” , “usually”, “many”, “good”,
“ frequently”, “fair” , and “poor”.
– Each of these words has many meanings. For one person , frequently reading a magazine
may be six or seven issues; for another it could be twice a year.
- Some adjectives have high variability and others have low variability.
High variable : e.g. a clear mandate, most, numerous,
(should be avoided) a substantial majority, a minority of, a large proportion of,
a significant number of, many, a considerable number of,
and several.
- Other adjectives produce less variability and generally have more shared
meaning.
e.g. lots, almost all, virtually all, nearly all, a majority of, a consensus of, a small
number of,
not very many of, almost none, hardly any, a couple, and a few.
Poor Item: Do you frequently consult your doctor?
Better item: How many times have you consulted your doctor in the last
six months?
None, 1 or 2 times ; 3-5 times ; more than 5 times
26. Avoid jargon and colloquialisms
- ensure your language caters for all levels of literacy.
Remember for some respondents English may be their
second language.
Poor item : Do you worry that you may have
halitosis?
Better item: Do you worry that you may have
bad breath?
27. Avoid double-barrelled questions
- A question may induce bias because it covers two
issues at once.
Poor item : “Do you enjoy playing badminton and
tennis?” or “Do you agree with the
recommendations of the Stern Review on
the economics of climate change?”
Better item: - Do you enjoy playing badminton ?
- Do you enjoy playing tennis ?
- Do you agree with the
recommendations of the Stern Review
on the economics of climate change?
28. Avoid “ combination” questions
- Do not include the word “and” in case a
respondent wants to respond “yes” to one and
“no” to the other part of the question.
Poor item : Do you use hand phone and lap top in
class ?
Better item : Do you use hand phone in class ?
Do you use lap top in class?
29. Avoid double negatives
- Double negatives in the question is likely to confuse
your respondents.
– e.g. instead of asking respondents whether they agree
with the negative statement;
“Smoking in public places should not be abolished”
use the positive
“Smoking in public places should be abolished”.
-
30. Minimise bias
- People sometimes answer questions in a way they perceive to be
socially acceptable. Make it easy for respondents to admit social
lapses by wording questions carefully.
For example, “How many times have you broken the speed limit
because you were late?”
…could be rephrased,…
“Have you ever felt under pressure to drive over the speed limit in
order to keep an appointment?”.
… Then you could ask, …
“How many times have you prioritised the appointment ?”.
31. Sources:
• Crawford, I. M. (1990) Marketing Research Centre for
Agricultural Marketing Training in Eastern and Southern
Africa. Harare Zimbabwe.
• Sudman, S. and Bradburn, N. M. (1973), Asking Questions, pp.
208 - 28.
• http://www.lse.ac.uk/library/versions/Creating%20effective%20
questionnaires%20and%20surveys.pdf
• http://www.uj.ac.za/EN/Research/Statkon/Documents/Statko
n%20Questionaire%20Design.pdf
• Rosen, L.J. (2010). The Academic Writer’s Handbook.
Longman
34. You can also ADAPT an existing
questionnaire (usually attached to
appendices of theses or dissertations).
Please make sure that the
questionnaire you choose is
QUANTITATIVE in nature.
Adaption or adaptation simply means to
reword or write using your own words,
all the questions in the questionnaire
you choose, to avoid plagiarism.