Anda di halaman 1dari 13

Using Questionnaires

 A questionnaire survey is, basically, a list


of written questions that can be completed
in one of two ways:
 Respondents complete the questionnaire personally,
writing the answers themselves, with the researcher
not present (sometimes known as a postal
questionnaire, even if it has not been posted)
 Respondents complete the questionnaire by verbally
responding to questions in the presence of a
researcher (sometimes known as a structured
interview)
Designing Your
Questionnaire 1
 Clarify Your Objectives
 Being clear about what information you want from
the survey will make the questions easier to write
 Define Your Sample
 Think carefully about the group(s) to whom the
questionnaire is given. Is the sample representative
or restricted? How big will the sample be? In what
ways will the sample introduce bias into the
responses?
 Choose your questions
 Think about whether you want closed questions, open
questions or a mixture of both
Designing Your
Questionnaire 2
 Structure your questionnaire
 Include an introduction explaining its
purpose
 Give clear instructions on how to fill it in

 Keep questions simple and unambiguous

 Be consistent in your approach (not a

mixture of tick boxes, circling answers etc.)


 Arrange questions in a logical order, putting

sensitive material towards the end


 Include a thank you at the end
Questionnaire Questions
 Two broad types of question can be
identified:

 Closed Questions

 Open Questions
Closed Questions
 A closed question is a question for
which the researcher provides a
suitable list of responses, limiting
the respondent to preset choices.

 Do you listen to hip-hop music?


a) Yes
b) No
Closed Questions 2
 Alternatively, a closed question can allow
the respondent the opportunity to choose
an answer from a predefined list:

 What kind of music do you listen to


most?:
d) Heavy Metal
e) Hip-Hop
f) Dance music
g) Garage
Closed Questions 3
 Another variation is to attempt to measure a
respondent’s attitude towards a particular
subject:

Hip-Hop music causes violence among its audience:


1)Strongly agree
2)Agree
3)Neither agree nor disagree
4)Disagree
5)Strongly disagree

This type of question is known as a Lickert scale


Closed Questions 4
 Alternatively:
Describe your opinion of hip-hop music

Violent 1 2 3 4 5 Non-violent
Sexist 1 2 3 4 5 Non-sexist
Racist 1 2 3 4 5 Non-racist

This is known as a Semantic-Differential scale


Closed Questions 5
 All of the above examples are variations
on a similar theme.
 Their defining characteristic is that they
allow the respondent very little scope in
which to develop an answer beyond the
categories selected
 As such, the primary purpose of closed
question tends to be the collection of
quantitative data
Open Questions
 In open questions, the researcher
doesn’t provide the respondent with
a set of answers from which to
choose
 Rather, the respondent is allowed to
answer questions in their own words
 This means that the primary purpose
of this type of question is to collect
qualitative data
Pilot the Questionnaire
 Before using the questionnaire in
earnest, pilot it with 4 or 5
respondents to uncover potential
problems in the questionnaire and in
the analysis

 Use the results of the pilot to adapt


and improve your questionnaire
Advantages of
Questionnaires
 Contacting large numbers of people
quickly, easily and efficiently
 Relatively quick and easy to create. Once
finished, the respondent does most of the
work
 Easy to standardise as everyone answers
the same question
 Can be used to explore sensitive material
as they can be anonymous and completed
in private
Disadvantages of
Questionnaires
 Difficult to explore complex issues and
opinions. Limited depth of response, even
with open questions.
 Difficult to ascertain the validity or truth of
respondents’ answers.
 Respondents can misunderstand questions
or deliberately invalidate the questionnaire
 A poorly chosen sample can make the
questionnaire of limited use. Self-
selecting samples can also contaminate
data