Questionnaire construction

Questionnaires are frequently used in quantitative marketing research and social research in
general. They are a valuable method of collecting a wide range of information from a large number
of respondents. Good questionnaire construction is critical to the success of a survey.
Inappropriate questions, incorrect ordering of questions, incorrect scaling, or bad questionnaire
format can make the survey valueless. A useful method for checking a questionnaire for problems
is to pretest it. This usually involves giving it to a small sample of respondents, then interviewing
the respondents to get their impressions and to confirm that the questions accurately captured
their opinions.
Questionnaire construction issues
The wording must be kept simple no technical or speciali!ed words. "se short sentences.
#riting style should be conversational, yet concise and accurate.
The meaning should be clear. Avoid ambiguous words and equivocal sentence structures.
Avoid double negatives. $ven single negatives should be reworded as positives.
Avoid biasing the responses. A biased question or questionnaire encourages respondents
to answer one way rather than another. Avoid %loaded& questions.
Ask one question at a time. Avoid comple' questions. If more than one question is hidden
in a survey question, the researcher will not know which one the respondent is answering.
Avoid personal or intimate questions. (ost people will not answer them.
)onsider the respondent*s frame of reference. #hat is their background, and how will
this effect their interpretation of the questions+ ,o respondents have enough
information or e'pertise to answer the question+
Ask yourself if each question is really necessary. "nneeded questions are an e'pense to
the researcher and an unwelcome imposition on the respondents. To answer this question,
you must consider the ob-ective.s/ of the research.
Ask yourself what type of data analysis techniques are available for various kinds of
questions. #ill the question provide you with the statistical analysis that you want+
#hat type of content will responses to the question yield+ #ill the question responses
provide facts, beliefs, feelings, descriptions of past behavior, or standards of action+
#hat type of scale, inde', or typology should be used+
0ow should the questions be presented on the page .or computer screen/+ 0ow much
white space+ 0ow many colours+ ,o you use pictures, charts, or other graphics+ It should
be colourful enough to gain and maintain respondent interest, but not so graphic as to
distract from the questions.
1hould questions be open2ended or should respondents* answers be limited to a fi'ed set
of responses+
#hat order should the questions be in+ Is there a %natural& grouping to the questions+
#ill previous questions bias later questions+
1hould the questions be numbered+ Generally this is a good idea.
Are possible responses mutually e'clusive+ The respondent should not find themselves in
more than one category, for e'ample in both the %married& category and the %not living
with spouse& category. )ategories should not overlap.
Is the list of possible question responses inclusive+ The respondent should not find
themselves with no category that fits their situation.
Is the questionnaire going to be administered by research staff, or will it be self2
administered by the respondents+ 1elf administered questionnaires must give clear,
detailed instructions.
Types of questions
3. Contingency questions 2 A question that is answered only if the respondent gives a
particular response to a previous question. This avoids asking questions of people that do
not apply to them .for e'ample, asking men if they have ever been pregnant/.
4. Matrix questions 2 Identical response categories are assigned to multiple questions. The
questions are placed one under the other, forming a matri' with response categories
along the top and a list of questions down the side. This is an efficient use of page space
and respondents* time.
5. Scaled questions 2 6esponses are graded on a continuum .e'ample rate the appearance
of the product on a scale from 3 to 37, with 37 being the most preferred appearance/.
$'amples of types of scales include the 8ikert scale, semantic differential scale, and
rank2order scale .1ee scale for a complete list of scaling techniques./.
9. Closed ended questions 2 6espondents* answers are limited to a fi'ed set of responses.
(ost scales are closed ended. :ther types of closed ended questions include
o ,ichotomous questions 2 The respondent answers with a %yes& or a %no&.
o (ultiple choice 2 The respondent has several option from which to choose.
;. Open ended questions 2 <o options or predefined categories are suggested. The
respondent supplies their own answer without being constrained by a fi'ed set of possible
responses. $'amples of types of open ended questions include
o )ompletely unstructured 2 =or e'ample, %#hat is your opinion of questionnaires+&
o #ord association 2 #ords are presented and the respondent mentions the first
word that comes to mind.
o 1entence completion 2 6espondents complete an incomplete sentence. =or
e'ample, %The most important consideration in my decision to buy a new house is .
. .&
o 1tory completion 2 6espondents complete an incomplete story.
o >icture completion 2 6espondents fill in an empty conversation balloon.
o Thematic apperception test 2 6espondents e'plain a picture or make up a story
about what they think is happening in the picture
Question sequence
Questions should flow logically from one to the ne't.
The researcher must ensure that the answer to a question is not influenced by previous
questions.
Questions should flow from the more general to the more specific.
Questions should flow from the least sensitive to the most sensitive.
Questions should flow from factual and behavioural questions to attitudinal and opinion
questions.
Questions should flow from unaided to aided questions
According to the three stage theory .also called the sandwich theory/, initial questions
should be screening and rapport questions. Then in the second stage you ask all the
product specific questions. In the last stage you ask demographic questions.