10 Step Guide to Questionnaire Design
10 Step Guide to
Questionnaire Design
Produced by RCU Limited
10 Step Guide to Questionnaire Design
Step 1
Keep it simple and make it interesting.
• Be concise when designing your questionnaire.
• Short, clearly worded questions can get you much more reliable
answers, but also make it interesting.
• Have some questions towards the beginning of the questionnaire
that really get to the heart of what you want the respondent to
tell you.
10 Step Guide to Questionnaire Design
Step 2
Keep it short.
• Keep questionnaires short and to the point.
• Don’t be tempted to try and throw in additional questions which
are not relevant to your specific research objectives.
• If your questionnaire gets too long then the respondent’s
attention and commitment will wane.
• If you have some very important questions towards the end of a
long a drawn out questionnaire you may not be able to trust the
results that they give you.
10 Step Guide to Questionnaire Design
Step 3
Use language your respondents would use.
• Always use words that your respondents would understand and
would naturally use themselves.
• If you use overly complex language in either the instructions, the
questions or the response options, you would again be unable to
trust the results that come back.
• It’s always a good idea to pre-test your survey questions with a
small group of the intended audience before you actually
administer it for real.
10 Step Guide to Questionnaire Design
Step 4
Think about the order of the questions.
• Research shows that the way we answer one set of questions is
closely determined by the questions which immediately preceded
it.
• E.g. if you ask a specific question on facilities and follow this
immediately with a question on recommendations, you will find
that the recommendations you receive will be closely correlated
with issues concerning facilities.
• Because of this it is a good idea to ask more general questions
before going on to specifics.
10 Step Guide to Questionnaire Design
Step 5
Avoid subjective terms.
• Avoid using terms in the question or response options that are
open to wide interpretation.
• Research shows that terms such as ‘frequently’ or ‘often’ are
widely differently interpreted by different respondents.
• Better to use more specific and precise terms such as ‘once a
week’ or ‘twice a week’ or you could use terms such as ‘always’ or
‘never’ to ensure interpretation is much more precise and regular.
10 Step Guide to Questionnaire Design
Step 6
Include a middle response option if it reflects
what some respondents would want to say.
• There is a bit of a myth in questionnaire design that says never
use an odd number of response options as people will
automatically get drawn to the middle option.
• In fact this isn’t true and you are more likely to distort the
responses you get if you don’t provide a middle option if that is
what the respondents actually want to tell you.
10 Step Guide to Questionnaire Design
Step 7
Avoid double negatives.
• There is a very common question structure known as the Likert
scale, that is where you are given a prompt and then asked to
either ‘strongly agree’, ‘agree’... right through to ‘strongly
disagree’.
• If your prompt itself is written with negative terminology such as
‘teachers don’t consult with students’, you have to say ‘strongly
disagree’ to actually agree with the question.
• This can clearly get very confusing and you couldn’t always trust
the responses.
10 Step Guide to Questionnaire Design
Step 8
Make scales logical.
• When attaching a scale or labels to the response options make
sure that they make sense.
• Number ratings really should run from low to high (e.g. 0 to 10).
• Research shows that people cope best when these response
options are clearly labelled (e.g. ‘good’ or ‘very good’).
• There is also contrasting evidence to show that we get the most
accurate answers with quite a wide scale or range of options.
• In practice it is much harder to provide clear labels for all the
points on a 10 point scale compared to a 5 point scale.
10 Step Guide to Questionnaire Design
Step 9
Don’t forget zero!
• One of the more useful scales that you can use is a rating out of
ten.
• If you are going to use this scale it is important to remember zero
as a response option.
• If your scale only runs from 1 to 10, you might think of the mid-point
as 5. In fact this isn’t true as the mid-point is actually 5.5 and
you may see an artificial clustering of responses around 5.
• Therefore it is better to use a 0 to 10 scale and make it clear in the
question prompt.
10 Step Guide to Questionnaire Design
Step 10
Put the personal questions at the end.
• It’s tempting when designing a questionnaire to put some easy
questions at the beginning that everyone can answer (e.g age,
gender and ethnicity).
• If the survey is anonymous and some of the issues covered in it or
are quite sensitive, this will affect the way people answer later
questions as they can worry that they may be identifiable from
their earlier responses.
• As a matter of good practice it is better to put those sorts of
questions at the end of the questionnaire, even if that feels
unnatural to do.