|
How to design a questionnaire
The questionnaire is a commonly used research instrument
for gathering and recording information. It is tailor made
to meet the information requirements of the market research
and therefore lies at the heart of the research process.
Before a questionnaire can be designed and used, you need
to ask yourself the following:
-
What information is required?
-
For whom is it required?
-
Which method of communication will be used?
Once
you have thought about the aim of your research and what
you intend to do with the information you receive, you can
start to put your questions together. Ask questions that
are relevant to your research. Try to resist the temptation
to ask questions out of curiosity.
What
should a questionnaire contain?
The
questionnaire should include the following:
1.
Title & Instructions
This
should be a short title that tells the respondent the nature
of the questionnaire. The title should be followed by instructions
on how to complete the questionnaire.
2.
Identify your respondents
This
usually contains the name, address of respondent, date,
time, length, place of interview and name of interviewer
in case any check back is needed. If you do ask for this
information, be sure to state your policy on confidentiality.
Bear in mind that asking respondents to identify themselves,
may have an adverse effect on the response rate and credbility
of the answers given.
3.
Put the information into appropriate groups
Identify
and group information which will help to define the individual
for the purpose of analysing responses. This include information
such as age, sex, and occupation of head of household, income
group and marital status.
4.
Ask the questions on your subject
This
forms part of the majority of the questionnaire. The early
questions should be of interest to the respondent and be
easy to answer. Questions that are either uninteresting
or rather personal should appear as late as possible in
the questionnaire. Try to vary the questioning format to
prevent the respondent from answering the questions 'automatically'.
How
should the questions be worded?
The
wording of the questions should be as such that the respondent
fully understands what is required and accurate responses
are encouraged. Here are some tips to bear in mind:
-
Avoid jargon and technical language.
-
Ensure that words have the same meaning for the respondent
as they have for the researcher.
-
Be clear and specific as possible, for example what does
frequently mean?
- Avoid
questions or words with meanings which would invite particular
responses, instead select words that are least likely
to bias response.
-
If there is more than one topic in the questionnaire,
'cushion' the change of subject with an introductory phrase.
Coding
For
quantitative data, pre-coding should be used whenever possible.
The code numbers should be printed onto the questionnaire
and the interviewer simply rings the appropriate code. Pre-coding
should help to reduce the cost of research. To code open-ended
questions, each point raised is coded with a number, which
is written on the questionnaire. Code numbers are allocated
from a coding frame, which is constructed by a research
executive.
Once
you have put your questionnaire together, check whether:
- All
questions are really needed.
-
Each question is sufficient to generate the required information.
- The
respondent will answer the question correctly.
-
The questions are in logical sequence.
- The
questionnaire is kept short, ideally one to two pages.
Incentive
You
may also wish to consider offering an incentive to complete
the questionnaire, for example, a free summary report.
Return
Make
it easy as possible for the respondent to return the questionnaire.
If the questionnaire is self-completion or sent by post,
enclose a prepaid, addressed envelope. Another means of
return is to ask the respondent to deposit the questionnaire
in a special container that is placed at a convenient location.
Covering
Letter
A covering
letter may help to persuade a respondent to complete the
questionnaire. It is usually sent with a questionnaire by
post. It should include information on the purpose of the
questionnaire, your contact details, any incentives, instructions
on how to complete the questionnaire and how long it should
take, how to return the questionnaire and your confidentiality
policy. If you are not sending your questionnaire by post,
and your questionnaire includes this information, it is
not necessary to also attach a covering letter.
Piloting
Before
committing yourself to a full-scale exercise, test out your
questionnaire on a sample selection of people from your
target market. Piloting a questionnaire on a small-scale
can help to iron out any 'bugs' so that it can be refined
before the full survey goes ahead.
Download
an example questionnaire
Suggested
next stage:Starting up: Choosing
a business structure
Useful
Sites
The
Market Research Society
British
Market Research Association
Chartered
Institute of Marketing
|