Chapter 4
Chapter 4
Chapter 4
Questionnaire Design 1
Learning Objectives
Questionnaire Design 2
What is a Questionnaire?
Questionnaire Design 3
Questionnaire Design Process
Questionnaire Design 4
Figure 4.1: Steps in questionnaire design process
Questionnaire Design 5
Phase I: Pre-Construction Phase
Questionnaire Design 6
Step 1: Specific Required Information
in the Light of Research Objective
Questionnaire Design 7
Step 2: An Overview of
Respondent’s Characteristics
Questionnaire Design 8
Step 3: Decision Regarding Selecting an
Appropriate Survey Technique
While constructing a questionnaire, a deep thinking process is
required to select an appropriate survey technique.
Chapter 7 provides a detailed discussion of various survey and
observation techniques and its relative advantages and
disadvantages. The questions in the questionnaire must be
constructed, sequenced, and placed according to the mode of
the survey.
Questionnaire Design 9
Phase II: Construction Phase
Phase II is the real construction phase of the questionnaire
design process.
It consists of six steps:
Decision regarding question format :structured questions versus
unstructured questions
Decision regarding question relevance and wording
Decision regarding question sequencing
Decision regarding question response choice
Decision regarding the questionnaire layout
Producing first draft of the questionnaire.
Questionnaire Design 10
Step 1: Decision Regarding Question
Format: Structured Questions Versus
Unstructured Questions
Questionnaires use two types of question formats. These are
open-ended questions and closed-ended questions.
In the case of a “closed-ended” question, the respondent has to
format the judgment to fit the response categories and when
“open-ended” questions are used, the judgment has to be
verbalized into a preliminary answer (DeLeeuw, 2001).
The closed ended question format can be further divided into
dichotomous, multiple-choice questions, and scales. The
following sections focus on open-ended questions and closed-
ended questions.
Questionnaire Design 11
Open-ended Questions
Questionnaire Design 12
Closed-ended Questions
Closed-ended questions are structured questions. The question
structure does matter (Connolly et al., 2005).
The closed-ended questions provide response alternative to the
respondents instead of giving them a free-to-express response
option.
The choice offered to the respondents can be either in the form
of a rating system or a set of response alternatives.
These response alternatives are presented to the respondents
and they select the most appropriate one.
The closed-ended questionnaires are generally cheaper, more
reliable, and faster to code, and analyse the collected data
(Swamy, 2007).
The closed-ended question format can be further divided into
dichotomous, multiple-choice questions, and scales.
Questionnaire Design 13
Dichotomous Questions
Dichotomous questions have only two response alternatives usually
presenting the two extremes “yes” or “no.” To make the alternatives
balanced, the researchers often present a third neutral alternative
“don’t know.”
Questionnaire Design 14
Multiple-Choice Questions
Questionnaire Design 15
Step 2: Decision Regarding Question
Wording
Questionnaire Design 16
Figure 4.2: Decision parameters regarding question wording
Questionnaire Design 17
Question Wordings Must Be Simple and
Easy to Understand
Questionnaire Design 18
Vague or Ambiguous Words
Must Be Avoided
Questionnaire Design 19
Double-Barrelled Questions
Must Be Avoided
Double-barrelled questions are those with wordings such as “and”
or “or.” In a double-barrelled question, a respondent may agree to
one part of the question but not to the other part.
Questionnaire Design 20
Avoid Leading and Loaded Questions
A leading question is the one which clearly reveals the researcher’s
opinion about the answer to the question.
Questionnaire Design 21
Split-Ballot Technique
A leading question generally emphasizes either the positive or
negative aspect of the question.
Sometimes, it becomes necessary to ask a question with either
the positive or negative aspect.
In this situation, a split-ballot technique is used to avoid bias
due to positive or negative aspect of the question.
This technique involves the construction of a single question in
two alternative phrases, and the question based on one phrase
is administered to half of the respondents and the question
based on the other phrase is administered to the other half of
the respondents.
Questionnaire Design 22
Split-ballot technique
Questionnaire Design 23
Avoid Using Overstated Words
The answer will always be overblown due to the first part of the
question, which generates a worry in the mind of the respondent
and results in a positive answer, which is not possible otherwise. A
more poised way of asking the same question is shown below.
Questionnaire Design 24
Implied Assumptions Must Be Avoided
Questionnaire Design 25
Respondent’s Memory Should
Not Be Overtaxed
Questionnaire Design 26
Generalization and Estimation
Must Be Avoided
Questionnaire Design 27
Respondent’s Ability to Answer
Must Be Considered
A question targeted to officers older than 55 years to assess the
importance of Internet banking is as follows:
Questionnaire Design 28
Step 3:Decision Regarding Question
Sequencing
Questionnaire Design 29
Figure 4.3: Decision parameters regarding question sequence
Identification
Difficult to
Screening Opening Transition and Logical order
answer
questions questions statements categorization of questioning
questions
questions
Questionnaire Design 30
Screening Questions
Researchers generally begin with some screening questions to
make sure that the target respondent is qualified for the
interview.
In some cases, when the researcher is very sure about the
qualification status of the respondent, he or she does not
incorporate the screening question and starts from some
“opening questions.”
Questionnaire Design 31
Opening Questions
The opening questions should be simple, encouraging, and trust
building. From the research objective point of view, these
questions may sometimes be little irrelevant but should be good
initiators.
These questions should not seek in-depth information and
should be as general as possible.
For example, a microwave company, trying to assess “shift in
consumer attitude” from traditional way of cooking, should ask a
first opening question as follows:
Questionnaire Design 32
Transition Statements
The movement from one set of questions to another set requires transition
statements.
For example, a mineral water bottle manufacturing company is encouraged with the
expanding market.
The company wants to assess the potential future market and hence conducted a
survey on non-users.
Its researchers have identified various variables to get the potential use, of which
“awareness” and “taste” are important.
It has prepared the first 11 questions with the first 5 questions based on the
“awareness” and the next 6 questions on “taste.”
After asking the first set of 5 questions, a researcher moves to the second set of 6
questions to get the potential consumer feeling for mineral water taste.
Thus, before asking the next set of 6 questions, a transition statement is required to
make the respondent familiar with the coming questions.
Hence, a transition statement “Now, I would like to understand your opinion about
the mineral water taste” will develop respondent’s connectivity for the next set of 6
questions related to “taste,” and he or she will be in a comfortable state of mind to
answer these questions.
Questionnaire Design 33
Difficult to Answer Questions
Difficult to answer, sensitive, or complicated questions should
be placed later in the questionnaire.
Placing it first will confuse the respondent and he or she will
tend to terminate the interview process.
Asking difficult questions first in a telephone interview reduces
a respondent’s interest in the interview process and he or she
tends to terminate the interview. Under telephone interview
conditions, substantively related questions affect the responses
to the target question only when asked first (Schwarz & Hippler,
1995).
Questionnaire Design 34
Identification and Categorization
Questions
Identification questions are used to generate some basic
identification information such as name, mailing address, office
phone number, personal phone number, or cell phone number.
Categorization questions are mainly used to generate
demographic information.
For example, researchers generally want to generate the
information related to age, experience, gender, and occupation
of the respondents.
Questionnaire Design 35
Logical Order of Questioning
In a questionnaire, the questions must flow in a logical
sequence. There are at least three approaches to suggest the
roadmap to place the questions in a logical sequence; they are
funnel technique, work technique, and sections technique.
Funnel technique suggests asking general questions first and
then the specific questions.
Work technique suggests that difficult-to-answer, sensitive, or
complicated questions should be placed later in the
questionnaire.
The third technique is the section technique in which questions
are placed in different sections with respect to some common
base.
Questionnaire Design 36
Step 4: Decision Regarding Question
Response Choice
Questionnaire Design 37
Step 5: Decision Regarding
Questionnaire Layout
Questionnaire Design 38
Step 6: Producing First Draft of the
Questionnaire
Questionnaire Design 39
Phase III: Post-Construction Phase
Phase III is the post-construction phase of the questionnaire
design process. It consists of four steps:
Pre-testing of the questionnaire
Revisiting the questionnaire based on the inputs obtained from
the pre-testing
Revising final draft of the questionnaire
Administering the questionnaire and obtaining responses.
Questionnaire Design 40
Step 1: Pre-Testing of the Questionnaire
Pre-testing of the questionnaire involves administering the
questionnaire to a small sample of the population to identify
and eliminate the potential problems of the questionnaire, if
any.
Testing a questionnaire can be a time-consuming process, and
this stage of questionnaire has often been overlooked in the
past in researcher’s eagerness to start collecting data (Williams,
2003).
Researchers generally use two common procedures to pre-test:
protocol analysis and debriefing.
Questionnaire Design 41
Step 1: Pre-Testing of the Questionnaire
(Cont.)
Using protocol analysis, a researcher asks the respondent to “think
aloud” while answering the question.
Debriefing is an interview conducted when a respondent has filled the
questionnaire.
After completing, the respondents are informed that the questionnaire
they have just filled was a pre-test questionnaire and are requested to
share their views about various dimensions of the questionnaire.
They are also requested to find out the problems with the
questionnaire, which they realized while filling it.
While taking a decision about the sample size for pre-testing, as a
thumb rule, a researcher should keep the nature of population diverse
in mind.
More heterogeneity in population requires relatively large size of the
sample compared with the situation when population is relatively
homogeneous.
Questionnaire Design 42
Step 2: Revisiting the Questionnaire Based
on the Inputs Obtained from Pre-Testing
Questionnaire Design 43
Step 3: Revised Final Draft of the
Questionnaire
Questionnaire Design 44
Step 4: Administration of the
Questionnaire and Obtaining Responses
Questionnaire Design 45