Questionnaire definition

► Questionnaire A structured technique for data collection

Questionnaire Design consisting of a series of questions, written or verbal, that a

respondent answers.
► A questionnaire, is a formalized set of questions for
obtaining data from respondents.
► It is a tool to generate data or we can say, it is a scale used
to measure the requisite attributes of the respondents.
► Any questionnaire has three specific objectives. First, it
Dr. Anoop Kumar Gupta must translate the information needed into a set of specific
questions that the respondents can and will answer.
MAIT ► Developing questions that respondents can and will
answer and that will yield the desired information is
1 2
difficult.

Questionnaire definition Questionnaire definition

► Second, a questionnaire must uplift, motivate ► Such an appreciation of what respondents go

and encourage the respondent to become through affects the design of how they are
involved, to cooperate, and to complete the approached, the stated purpose of the research,
task. the rewards for taking part and the whole process
► Before designing any questionnaire or any of questioning and question design.
research technique, the researcher must evaluate ► Not all respondents are the same in what they seek
‘what is the respondent going to get out of from a questionnaire or interview process.
this?’ ► Some respondents may want to see some
► The marketing researcher must have an empathy personal benefit, perhaps a tangible reward, while
with target respondents and appreciate what they others may be happy to see the social benefits.
think when approached and questioned. 3 4
 Questionnaire definition Questionnaire design process
Specify the Information Needed
► Taking care in appreciating what respondents Specify the Type of Interviewing Method
expect from the questioning process can nurture Determine the Content of Individual Questions
responses that have been well thought through,
Design the Question to Overcome the Respondent’s
are honest, and accurate. Inability and Unwillingness to Answer
► Third, a questionnaire should minimize response Decide the Question Structure
error. Response error is defined as the error that Determine the Question Wording
arises when respondents give inaccurate answers or
Arrange the Questions in Proper Order
when their answers are mis-recorded or misanalysed.
Identify the Form and Layout
► A questionnaire can be a major source of
response error. Minimizing this error is an important Reproduce the Questionnaire

objective of questionnaire design. 5

Eliminate Bugs by Pre-testing
6

1. Specify the information needed 1. Specify the information needed

► The first step in questionnaire design is to specify ► The wording and style of questions that may be
the information needed. This is also the first step in appropriate for finance directors being surveyed
the research design process. about their IT needs may not be appropriate for
► Note that, as the research project progresses, the retired persons being surveyed about their holiday
information needed becomes more and more needs.
clearly defined. ► The more diversified the respondent group, the
► It is also vital to have a clear idea of the target more difficult it is to design a single questionnaire
respondents. The characteristics of the respondent appropriate for the entire group.
group have a great influence on questionnaire
design.
7 8
 2. Specify the type of interviewing method 2. Specify the type of interviewing method

► An appreciation of how the type of interviewing method ► Mail and electronic questionnaires are self-
influences questionnaire design can be obtained by administered, so the questions must be simple,
considering how the questionnaire is administered and detailed instructions must be provided.
under each method.
► In Internet surveys and computer-assisted
► In personal interviews, respondents see the
interviewing, complex skip patterns and
questionnaire and interact face to face with the
randomization of questions to eliminate order
interviewer. Thus, lengthy, complex and varied
questions can be asked. bias can be easily accommodated.
► In telephone interviews the respondents interact with ► Questionnaires designed for personal and
the interviewer, but they do not see the questionnaire. telephone interviews should be written in a
This limits the type of questions that can be asked to conversational style.
short and simple ones. 9 10

3. Determine the content of individual 3. Determine the content of individual

questions questions

Once the information needed is specified and the type of ► It is useful to ask some neutral questions at the
interviewing method decided, the next step is to beginning of the questionnaire to establish
determine individual question content: what to include in involvement and rapport, particularly when the topic
individual questions. of the questionnaire is sensitive or controversial.
 Is the question necessary? ► Sometimes filter questions are asked to disguise the
► Every question in a questionnaire should contribute to purpose or sponsorship of the project. For example,
the information needed or serve some specific purpose. rather than limiting the questions to the brand of
If there is no satisfactory use for the data resulting from interest, questions about competing brands may be
a question, that question should be eliminated. included to disguise the sponsorship.
► In certain situations, however, questions may be asked ► Questions unrelated to the immediate problem may
that are not directly related to the needed information. sometimes be included to generate client support for
11 the project. 12
 3. Determine the content of individual 3. Determine the content of individual
questions questions

 Are several questions needed instead of one? ► Such a question is called a double-barreled
► Once we have ascertained that a question is question, because two or more questions are
necessary, we must make sure that it is sufficient to get combined into one.
the desired information. ► To obtain the required information, two distinct
► Sometimes several questions are needed to obtain the questions should be asked: ‘Do you think Coca-Cola
required information in an unambiguous manner. is a tasty soft drink?’ and ‘Do you think Coca-Cola is
Consider the question ‘Do you think Coca-Cola is a a refreshing soft drink?’
tasty and refreshing soft drink?’ A yes answer will ► Another example of multiple questions embedded in
presumably be clear, but what if the answer is no? a single question is the ‘why’ question. In the context
Does this mean that the respondent thinks that Coca- of the Formula One Racetrack study, consider the
Cola is not tasty, that it is not refreshing, or that it is question ‘Why did you buy a Renault car?’
neither tasty nor refreshing? 13 14

4. Overcoming the respondent’s inability and 4. Overcoming the respondent’s inability and
unwillingness to answer unwillingness to answer

 Is the respondent informed?  Can the respondent remember?

► Respondents are often asked about topics on which ► Many things that we might expect everyone to
they are not informed. Parents may not be informed know are remembered by only a few. Test this on
about their child’s daily purchasing or vice versa. yourself. Can you remember the brand name of the
► In situations where not all respondents are likely to be
socks you are wearing (presuming you are wearing
informed about the topic of interest, filter questions socks), what you had for lunch a week ago, or what
that measure familiarity, product use and past you were doing a month ago at noon? Further, do you
experience should be asked before questions about the know how many litres of soft drinks you consumed
topics themselves. Filter questions enable the during the last four weeks?
researcher to filter out respondents who are not ► Evidence indicates that consumers are particularly
adequately informed. poor at remembering quantities of products
15 consumed. 16
 4. Overcoming the respondent’s inability and 4. Overcoming the respondent’s inability and
unwillingness to answer unwillingness to answer

► In situations where factual data were available for ► For example, a respondent reports three trips to the
comparison, it was found that consumer reports of supermarket in the last two weeks when, in fact,
product usage exceeded actual usage by 100% or one of these trips was made 18 days ago.
more.
► Creation error takes place when a respondent
► The inability to remember leads to errors of
‘remembers’ an event that did not actually occur.
omission, telescoping and creation.
► The ability to remember an event is influenced by
► Omission is the inability to recall an event that actually
took place.
(1) the event itself, (2) the time elapsed since the
event, and (3) the presence or absence of things
► Telescoping takes place when an individual
that would aid memory.
telescopes or compresses time by remembering an
event as occurring more recently than it actually ► We tend to remember events that are important or
occurred. 17 unusual or that occur frequently. 18

4. Overcoming the respondent’s inability and 4. Overcoming the respondent’s inability and
unwillingness to answer unwillingness to answer

► People remember their wedding anniversary and ► Even if respondents are able to answer a particular
birthday. Likewise, more recent events are remembered question, they may be unwilling to do so, because too
better. much effort is required, the situation or context may
► A grocery shopper is more likely to remember what he not seem appropriate for disclosure, no legitimate
purchased on his last shopping trip than what he bought purpose or need for the information requested is
three shopping trips ago. apparent, or the information requested is sensitive.
 Is the respondent able to articulate?  Effort required of the respondents
► Respondents may be unable to articulate certain types of ► Most respondents are unwilling to devote much effort
responses. For example, if asked to describe the
to providing information. Hence, the researcher
‘atmosphere’ of a restaurant they would prefer to
should minimise the effort required of the
frequent, most respondents may be unable to phrase
their answers. 19
respondents. 20
 4. Overcoming the respondent’s inability and 4. Overcoming the respondent’s inability and
unwillingness to answer unwillingness to answer

 Context  Legitimate purpose

► Some questions may seem appropriate in certain ► Respondents are also unwilling to divulge
contexts but not in others. For example, questions information that they do not see as serving a
about personal hygiene habits may be appropriate legitimate purpose.
when asked in a survey sponsored by a health ► Why should a firm marketing breakfast cereals
organisation but not in one sponsored by a want to know their age, income and occupation?
breakfast cereal manufacturer. ► Explaining why the data are needed can make the
► Respondents are unwilling to respond to questions request for the information seem legitimate and
they consider being inappropriate for the given may increase the respondents’ willingness to
context. answer.
21 22

4. Overcoming the respondent’s inability and 4. Overcoming the respondent’s inability and
unwillingness to answer unwillingness to answer

 Sensitive information  Increasing the willingness of respondents

► Respondents may be unwilling to disclose, at least Respondents may be encouraged to provide
accurately, sensitive information because this may information which they are unwilling to give by the
cause embarrassment or threaten the respondents’ following techniques:
prestige or self-image, or be seen as too personal and ► Place sensitive topics at the end of the
an invasion of privacy. questionnaire. By then, initial mistrust has been
► If pressed for the answer, respondents may give biased overcome, rapport has been created, legitimacy of
responses, especially during personal interviews. the project has been established, and respondents
► Sensitive topics include money, personal hygiene, are more willing to give information. In this context,
family life, political and religious beliefs, and consider how sensitive classification questions such
involvement in accidents or crimes. as gender, age and income may be perceived.
23 24
 4. Overcoming the respondent’s inability and
unwillingness to answer 5. Choose question structure

► Provide response categories rather than asking for

A question may be unstructured or structured.
specific figures. Do not ask ‘What is your  Unstructured questions
household’s annual income?’ Instead, ask the ► Unstructured questions are open-ended
respondent to indicate an appropriate income questions that respondents answer in their own
category words. They are also referred to as free-
response or free-answer questions.
► The following are some examples:
 What is your occupation?
 What do you think of people who patronize secondhand
clothes shops?
25  Who is your favorite film personality? 26

5. Choose question structure 5. Choose question structure

► Open-ended questions can be good first questions on a ► Whether the interviewers record the answers verbatim
topic. They enable the respondents to express general or write down only the main points, the data depend on
attitudes and opinions that can help the researcher the skills of the interviewers. Recorders should be used
interpret their responses to structured questions. if verbatim reporting is important.
► They can also be useful as a final question in a ► Another major disadvantage of unstructured
questionnaire. questions is that the coding of responses is costly
► Respondents are free to express any views. Their and time consuming.
comments and explanations can provide the researcher ► In general, open-ended questions are useful in
with rich insights. exploratory research and as opening or closing
► A principal disadvantage is that potential for interviewer questions. They should be chosen with great care as
bias is high. their disadvantages can outweigh their advantages in a
27 large survey. 28
 5. Choose question structure 5. Choose question structure

 Structured questions ► Consider the following question:

Do you intend to buy a new car within the next six months?
► Structured questions specify the set of response
____________________________ Definitely will not buy
alternatives and the response format. A structured
____________________________ Probably will not buy
question may be multiple choice, dichotomous or a
____________________________ Undecided
scale.
____________________________ Probably will buy
► Multiple-choice questions In multiple-choice ____________________________ Definitely will buy
questions, the researcher provides a choice of ____________________________ Other (please specify)
answers and respondents are asked to select one ► Of concern in designing multiple-choice questions are the
or more of the alternatives given. number of alternatives that should be included and the order of
potential responses, known as position bias.
29 30

5. Choose question structure 5. Choose question structure

► Multiple-choice questions overcome many of the ► Scales To illustrate the difference between scales
disadvantages of open-ended questions because and other kinds of structural questions, consider
interviewer bias is reduced and these questions are the question about intentions to buy a new laptop
administered quickly. computer. One way of framing this using a scale is
as follows:
► Also, coding and processing of data are much less
Do you intend to buy a new car within the next six months?
costly and time consuming.
► In self administered questionnaires, respondent
cooperation is improved if the majority of the
questions are structured
This is only one of several scales that could be used to ask
31 this question
 6. Choose question wording 6. Choose question wording

► Question wording is the translation of the desired ► Unless the respondents and the researcher
question content and structure into words that assign exactly the same meaning to the
respondents can clearly and easily understand. question, the results will be seriously biased.
► Deciding on question wording is perhaps the ► To avoid these problems, following guidelines are
most critical and difficult task in developing a advised:
questionnaire. 1. Define the issue A question should clearly define
► If a question is worded poorly, respondents may the issue being addressed. Trainee journalists
refuse to answer it or answer it incorrectly. are cautioned to define issues in terms of who,
► The first condition, known as item non-response, what, when, where, why and way (the six Ws).
can increase the complexity of data analysis. The Consider the following question:
second condition leads to response error. 33 Which brand of shampoo do you use? 34

6. Choose question wording 6. Choose question wording

► ‘When’ is not clear; does the researcher mean last 2. Use ordinary words Ordinary words should be
time, last week, last month, last year, or ever? used in a questionnaire, and they should match
► As for ‘where’, it is implied that the shampoo is the vocabulary level of the respondents.
used at home, but this is not stated clearly. A better ► When choosing words, bear in mind the
wording for this question would be: intellectual level of the target group of
Which brand or brands of shampoo have you respondents, and how comfortable they are
personally used at home during the last month? with technical terms related to any products or
In the case of more than one brand, please list all services we are measuring.
the brands that apply. ► Most respondents do not understand technical
marketing words.
35 36
 6. Choose question wording 6. Choose question wording

3. Use unambiguous words The words used in a ► The answers to this question are fraught with response bias,
questionnaire should have a single meaning that is known because the words used to describe category labels have
to the respondents. A number of words that appear to be different meanings for different respondents. Three
unambiguous have different meanings for different people. respondents who visit a bank once a month may tick three
These include ‘usually’, ‘normally’, ‘frequently’, ‘often’, different categories: occasionally, sometimes and often. A
‘regularly’, ‘occasionally’ and ‘sometimes’ much better wording for this question would be the following:
► Consider the following question : In a typical month, how In a typical month, how often do you visit a bank?
often do you visit a bank?

37 38

6. Choose question wording 6. Choose question wording

4. Avoid leading or biasing questions A leading question is 5. Avoid implicit assumptions Questions should
one that clues the respondent to what the answer should be, as
in the following: not be worded so that the answer is dependent on
Do you think that patriotic French people should buy imported implicit assumptions about what will happen as a
cars when that would put French workers out of employment? consequence.
► Implicit assumptions are assumptions that are
not explicitly stated in the question, as in the
following example:
► This question would lead respondents to a ‘No’ answer. After all,
how could patriotic French people put French people out of work?  Are you in favour of a balanced national budget?
Therefore, this question would not help determine the  Are you in favour of a balanced national budget if it
preferences of French people for imported versus domestic cars.
would result in an increase in personal income tax?
► Do your wife still beats you? (on a lighter vein) 40
 6. Choose question wording 6. Choose question wording

► Implicit in question 1 are the consequences that will 6. Avoid generalizations and estimates Questions
arise as a result of a balanced national budget. should be specific, not general. Moreover, questions
There might be a cut in defence expenditure, an should be worded so that the respondent does not
have to make generalizations or compute estimates.
increase in personal income tax, a cut in health
Suppose that we were interested in households’ annual
spending, and so on. per capita expenditure on groceries. If we asked
► Question 2 is a better way to word this question. respondents the question
► Question 1’s failure to make its assumptions What is the annual per capita expenditure on groceries in
explicit would result in overestimating the your household?
respondents’ support for a balanced national ► they would first have to determine the annual expenditure

budget. on groceries by multiplying the monthly expenditure on

41
groceries by 12 or the weekly expenditure by 52. 42

6. Choose question wording 7. Arrange the questions in proper order

7. Use positive and negative statements Many ► The order of questions is of equal importance to the
questions, particularly those measuring attitudes wording used in the questions.
and lifestyles, are presented as statements to ► Questions communicate and set respondents in a
which respondents indicate their degree of particular frame of mind.
agreement or disagreement. ► This frame of mind is set at the start of the
► Evidence indicates that the response obtained is questioning process and can change as each
influenced by the directionality of the statements: question is posed and responded to.
whether they are stated positively or negatively ► It affects how respondents perceive individual
► In these cases, it is better to use dual statements, questions and respond to those questions.
some of which are positive and others negative.
43 44
 7. Arrange the questions in proper order 7. Arrange the questions in proper order

► As well as understanding the characteristics of ► Questions that ask respondents for their opinions
language in target respondents, questionnaire can be good opening questions, because most
designers must be aware of the logical connections people like to express their opinions.
between questions – as perceived by target
► Though classification questions seem simple to
respondents.
start a questionnaire, issues like age, gender and
The following issues help to determine the order of
income can be seen as sensitive issues.
questions.
► Opening a questionnaire with these questions
1. Opening questions The opening questions can be
crucial in gaining the confidence and cooperation of tends to make respondents concerned about the
respondents. These questions should be interesting, purpose of these questions and indeed the whole
simple and non-threatening. survey.
45 46

7. Arrange the questions in proper order 7. Arrange the questions in proper order

► However, in some instances it is necessary to 2. Type of information The type of information

qualify respondents to determine whether they are obtained in a questionnaire may be classified as:
eligible to participate in the interview. In this case i. Basic information
the qualifying questions serve as the opening ii. Classification information
questions, and they may have to be classification iii. Identification information
questions such as the age of the respondent. ► Basic information relates directly to the research
problem.
► Classification information, consisting of socio-
economic and demographic characteristics, is used
to classify the respondents, understand the results,
47 and validate the sample . 48
 7. Arrange the questions in proper order 7. Arrange the questions in proper order

► Identification information includes name, address and 3. Difficult questions Difficult questions or questions
telephone number. Identification information may be that are sensitive, embarrassing, complex or dull
obtained for a variety of purposes, including verifying should be placed late in the sequence. After rapport
that the respondents listed were actually interviewed has been established and the respondents become
and to send promised incentives or prizes. involved, they are less likely to object to these
► As a general guideline, basic information should be questions.
obtained first, followed by classification and finally 4. Effect on subsequent questions Questions asked
identification information. early in a sequence can influence the responses to
► The basic information is of greatest importance to subsequent questions. As a rule of thumb, general
the research project and should be obtained first, questions should precede specific questions. This
before risking alienation of the respondents by prevents specific questions from biasing responses to
asking a series of personal questions. 49 the general questions. 50

7. Arrange the questions in proper order 7. Arrange the questions in proper order

► Going from general to specific is called the funnel 5. Logical order Questions should be asked in a
approach. A strategy for ordering questions in a logical order. This may seem a simple rule, but as
questionnaire in which the sequence starts with the the researcher takes time to understand
general questions, which are followed by progressively respondents and how they use language, the
specific questions, to prevent specific questions from researcher should also take time to understand their
biasing general questions. logic, i.e. what ‘logical order’ means to target
► The funnel approach is particularly useful when respondents.
information has to be obtained about respondents’ general
► All questions that deal with a particular topic should
choice behavior and their evaluations of specific products.
Sometimes the inverted funnel approach may be useful. be asked before beginning a new topic. When
In this approach, questioning starts with specific questions switching topics, brief transitional phrases should be
and concludes with the general questions. used to help respondents switch their train of thought.
51 52
 7. Arrange the questions in proper order 8. Identify the form and layout

► Branching questions should be designed ► The format, spacing and positioning of questions can have a
carefully. Branching questions direct respondents significant effect on the results, particularly in self-
administered questionnaires.
to different places in the questionnaire based on
► It is good practice to divide a questionnaire into several
how they respond to the question at hand. parts. Several parts may be needed for questions pertaining
► These questions ensure that all possible to the basic information.
contingencies are covered. ► The questions in each part should be numbered, particularly
► They also help reduce interviewer and respondent when branching questions are used. Numbering of
questions also makes the coding of responses easier.
error and encourage complete responses.
► The questionnaires should preferably be pre-coded. In pre-
coding, the codes to enter in the computer can be printed
on the questionnaire.
53 54

9. Reproduce the questionnaire 9. Reproduce the questionnaire

► How a questionnaire is reproduced for administration ► Overcrowded questions with little blank space between
can influence the results. them can lead to errors in data collection and yield
► For example, if the questionnaire is reproduced on shorter and less informative replies. It gives the
poor-quality paper or is otherwise shabby in impression that the questionnaire is complex and can
appearance, the respondents will think that the project result in lower cooperation and completion rates.
is unimportant and the quality of response will be ► Although shorter questionnaires are more desirable
adversely affected. than longer ones, the reduction in size should not be
► Therefore, the questionnaire should be reproduced on obtained at the expense of crowding.
good-quality paper and have a professional ► The questionnaire should be reproduced in such a way
appearance. that it is easy to read and answer. The type should be
► The tendency to crowd questions together to make the large and clear. Reading the questionnaire should not
questionnaire look shorter should be avoided. 55 impose a strain. 56
 10. Eliminate problems by pilot-testing 10. Eliminate problems by pilot-testing
► The respondents in the pilot-test should be similar to
► Pilot-testing refers to testing the questionnaire on those who will be included in the actual survey in
a small sample of respondents to identify and terms of background characteristics, familiarity with
eliminate potential problems. the topic, and attitudes and behaviors of interest.
► Even the best questionnaire can be improved by ► Pilot-tests are best done by personal interviews, even
pilot-testing. As a general rule, a questionnaire if the actual survey is to be conducted by Internet,
should not be used in the field survey without mail or telephone, because interviewers can observe
adequate pilot-testing. respondents’ reactions and attitudes.
► A pilot-test should be extensive. All aspects of the ► After the necessary changes have been made,
questionnaire should be tested, including question another pilot-test could be conducted by Internet,
content, wording, sequence, form and layout, mail or telephone if those methods are to be used in
question difficulty, and instructions. 57 the actual survey. 58

THANK YOU

59