Module 8:

Questionnaire Design
Questionnaire-Meaning
• It is essentially a data collection instrument that has a pre-designed set of
questions following a particular structure.
• Since it includes a standard set of questions, it can be successfully used to
collect information from a large sample in a reasonably short time period.
Limitations:
• At the exploratory stage when one is still trying to identify the information
areas, variables and execution decision, it is advisable to use a more
unstructured interview.
• Secondly, when the number of respondents is small and one needs to collect
more subjective data and most of the questions to be asked are open-ended,
then a standardized questionnaire is not advisable.
Criteria for questionnaire designing
• The spelt out research objectives must be converted into clear
questions which will extract answers from the respondent.
• Sometimes it is difficult to ask direct questions, in that case disguised
questions can be asked.
• The questions should be designed to engage the respondent and
encourage a meaningful response.
• The questions should be non threatening, must encourage response
and be clear to understand.
• The question should be self-explanatory and not confusing as then
the answer one will get might not be accurate or usable for analysis.
 Types of questionnaires:
• For designing purposes the questionnaires are formed on the basis of two parameters:
The degree of construction or structure
 The degree of concealment of the research objectives
Construction or formalization refers to the degree to which the response category has
been defined.
Concealed refers to the degree to which the purpose of the study is explained or is clear
to the respondent.
1. Formalized and unconcealed questionnaire
2. Formalized and concealed questionnaire
3. Non-formalized and unconcealed questionnaire
4. Non-formalized and concealed questionnaire
1. Formalized and unconcealed questionnaire:
This is the one which is most frequently used by all management researchers.
For example, if a new brokerage firm wants to understand the investment behavior
of the population under study, they would structure the questions and answers as
follows.
 2. Formalized and concealed questionnaire:
The research studies which are trying to unravel the latent causes of behavior cannot rely on direct
questions. Thus the respondent has to be given a set of questions that can give an indication of what
are his basic values, opinions and beliefs.
For Example: A publication house that wants to launch a newspaper wants to ascertain what are the
general perceptions and current attitudes about newspapers.
Advantage
1. These are structured, one can ascertain
their impact and quantify the same
through statistical techniques.
2. Such questions increase the
subject coverage and improve the validity
of the instrument.
 3. Non-formalized and unconcealed
questionnaire
Formalized structure express attitudes in a manner that the researcher or experts think is the correct way. This,
however, might not be the way the person thinks. Thus, rather than giving them predesigned response categories, it
is better to give them unstructured questions where he has the freedom of expressing himself the way he wants to.
The advantage of the matter is that the respondent can respond in any way he/she believes is important.
For example: 1. How do you generally decide on where you are going to invest your money?
2. Give three reasons why you believe that the Commonwealth 2010 games have helped the country?
Some people might respond by stating that it has boosted tourism in the country and contributed to the country's
economy.
Some will think it will encourage more international events to be held in the country.
Some might also state that it is not a good idea and the government should instead be spending on improving the
cause of the people who are below the poverty line.
The validity of these measures is higher than the previous two. However, quantification is a little tedious and one
cannot go beyond frequency and percentages to represent the findings. The other problem is the researcher’s bias
which might lead to clubbing responses into categories which might not be homogeneous in nature.
4. Non formalized, concealed
• If the objective of the research study is to uncover socially unacceptable desires and latent or
subconscious and unconscious motivations, the investigator makes use of questions of low
structure and disguised purpose.
• For projective techniques, these kinds of questionnaires are designed.
Advantage:
The presumption behind this is that if the argument, the situation, or question, is ambiguous, it
is most likely that the revelation it would result in would be more rich and meaningful.

Limitation:
 Low structure
 The interpretation required is highly skilled
 Cost, time and effort are additional elements which might curtail the use of these techniques.
Example: Non formalized, concealed
A study conducted to measure to which segment should men’s personal care toiletries (especially
moisturizers and fairness creams) be targeted. For this investigator designed two bachelor’s
shopping list.

• Conclusion: Normal Indian adult male include beauty or cosmetic products into his normal
Questionnaire Design Procedure
1. Convert the research objectives into
information areas.
This is the first step of the design process.
• Identify the specific research questions that the study will address.
• Convert these questions into statement of objectives.
• Identify the variables under study and they should be clearly defined.
• The direction of the relation or any other assumption and makes about
the variables under study in the form of a hypothesis.
• Specify the information needed for the study.
 2. Method of administration
• Once the researcher has identified his information area; he needs to specify how the information should be
collected.
• There are variety of methods for administering this study.
• The main methods are
 Personal schedule
 Self Administered questionnaire through mail, fax, e-mail and web-based.
• There are different preconditions for using one method over the other.
• Large population can be covered by mail or fax. In case the population to be studied is computer literate, it is
possible to use e-mail or web design surveys.
• For a smaller population and more complex or sensitive issues, personal schedule is advisable.
• In computer assisted dissemination, complex skip and branching options are possible, and randomization of
questions to eliminate the other order bias can be carried out with considerable ease.
• When the researcher wants to have a higher control over the way the questions are answered, he should be
using the schedule. Sampling control is highest in schedule and lowest in web-based survey.
 3. Content of the questionnaire
• Once information needs and mode of administration has been decided, the next step is to determine
the matter to be included as questions in the measure.
• In the course of the research study, the researcher might formulate a number of questions which he
thinks address the information needs of the study.
• One needs to remember that the time of the respondent is precious and it should not be wasted.
• Unless a question is adding to the data required for reaching an answer to the formulated problem, it
should not be included.
• Some neutral questions must be asked at the beginning of the questionnaire to establish an
involvement and rapport.
• Disguised questions are asked sometimes to keep the purpose or sponsorship of the project hidden.
• One needs to ascertain whether a single question will serve the purpose or should more than one
question be asked.
4. Motivating the respondent to answer
• The question should be designed in a manner that it involves the respondent and motivates
him/her to give comprehensive information.
• Sometimes the respondent answers questions even when they do not understand or have
information about the construct being investigated.
• Thus in situations where not all the respondents are likely to be informed about the research
topic, certain qualifying or filter questions that measure the experience or knowledge must
be asked before the questions about the topics themselves. Filter questions enable the
researcher to filter out the respondents who are not adequately informed.
• Most respondents would not know what phrases to use to give an answer. If the researcher
uses a semantic differential scale, the respondent can be provided adjectives to choose
from. If the person does not know what words to use or finds the task of description too
tedious, the person will not fill in the answers.
5. Determining the type of questions.
• Open ended questions
• Openness refers to the option of responding in one's own words. They are also referred to as unstructured questions or free-response or free-
answer questions. The researcher suggest no alternatives. Thus the words, logic and structure that a person would give while filling the answers is
totally left to his discretion.
• Example: Why do you smoke cigarettes?
• Which is your favorite TV serial?
• What training programs did you last attended?
• I like Nescafe ….
• My career goal is….
• I think hybrid cars are….
• Open-ended questions can typically be used for three reasons.
• They can be used in the beginning to start the questioning process.
• These can also be used as probing or clarifying questions to understand the reason behind certain responses.
• They can be used in the end as suggestions or final opinions.
• Advantage:
• These questions have the inherent advantage of improving the validity of the construct being studied.
• They are not restrictive and the respondents are free to express any views.
• Disadvantage:
• Coding or categorizing the written responses for an open-ended question is expensive, both in terms of time as well as finances.
• Open-ended questions are also dependent upon the respondent skill to articulate well.
• They are more suited to face to face interactions rather than self administered type questionnaire where there are chances of misinterpretation or
a complete non response as well.
Determining the type of questions-contd.
• Closed ended questions
• In these questions, both the question and response formats are structured and defined.
• The respondent only needs to select the option(s) that he feels are expressive of his opinion.
• There are three kinds of formats.
• Dichotomous questions,
• multiple choice questions, and
• those that have a scaled response.
• (I) Dichotomous questions:
• These are restrictive alternatives and provide the respondent only with two answers.
• Example: Are you diabetic? Yes/No
• Have you read the new book by J K Rowling? Yes/No
• What kind of petrol do you use in your car? Normal/premium
• What kind of cola do you drink? Normal/diet
• Your working hours in the organization are fixed/flexible
• Limitation: The problem with these situations is that these are forced choices and one needs to select one of them.
• Another dilemma is whether to include a neutral response alternative. If there are only two choices, he is forced to take a stand even when he
has no opinion on either or he is uncertain about the two options. However, the problem with the neutral category is that most respondents
want to avoid taking a stand and use it as an escape, thus the researcher does not get any meaningful number for or against the issue on the
study.
• The wording of the question might result in different answers
• Advantage: Dichotomous questions are the easiest type of questions to code and analyze. They are constructed on the nominal level of
measurement and our categorical or binary in nature.
Closed ended questions-contd
• (II) Multiple choice questions
• Unlike dichotomous questions, the person is given a number of response alternatives. Here he might be
asked to choose the one that is most applicable.
• There could also be instances when multiple options are given to the respondent and he can select all those
that apply in the case. These kind of multiple choice questions are called checklists.
• Advantage:
• Multiple choice questions can effectively cancel the researchers bias that was inherent in the open-ended
questions.
• Secondly, since they have pre designed response options that require the person to pick one or all that
apply, the administration is much faster.
• Data processing with these questions is much easier as is quantification and analysis of the information
collected.
• Limitations:
• Administering them might be easier, but designing exhaust exhaustive multiple choice questions is a
challenge.
• The researcher will have to do an exploratory study to uncover possible alternatives or conduct an extensive
secondary data analysis to identify the alternatives.
 Closed ended questions-contd.
• Scales refer to the attitudinal scales.
• For example:
• The following is a question which has five sub-questions designed on the Likert scale.
• Advantage:
• They are easy to administer no matter what will be the mode.
• Coding and tabulating these questions are not difficult.
6. Criteria for question designing.
Criteria for question designing-contd
d) Avoid leading questions: Any question that provides a clue to the respondent in terms of the direction in which one wants them to answer is called a leading or
biasing question.
For example, do you think that working mother should buy ready to eat food when that might contain some chemical preservatives? Yes/No/Don't know. Instead,
Do you think that working mother should buy ready to eat food? Yes/No/Don't know.
e) Avoid loaded questions: Question that address sensitive issues are termed as loaded questions, and the response to these questions might not always be
honest as the person might want wish to admit the answer even when assured about his anonymity.
For example, have you ever cheated on your spouse?
Will you take dowry when you get married?
Do you think your boss supervisor is incompetent? Instead,
do you think most people usually cheat on their spouses?
Do you think most Indian men would take dowry when they get married?
Do you think your juniors are incompetent?
Do you think your boss supervisor is incompetent?
Do you think your neighbor is incompetent?
f) Avoid double-barreled questions:
The questions that have two separate options separated by an ‘or’ or an ‘and’ are like the following:
1. Do you think Nokia and Samsung have a wide variety of touch phones? Yes or No Instead,
A wide variety of touch phones is available for:
Nokia,
Samsung,
both.
2. In the context of training needs analysis, consider the question:
Did the training you went through make you feel more motivated and effective in your job? Instead,
Did the training we went through make you feel motivated at your job?
Did the training you went through make you more effective at your job?
7. Questionnaire structure.
The next step is to put together the questions in a sequence that is
reader/respondent friendly and generates the required data in a short and
effective manner.
a) Instructions: The questionnaires begin with standardized instructions.
These begin by greeting the respondent and then introducing the
researcher, investigator, and the affiliating body. The note then goes on
to explain the purpose of questionnaire administration.
b) Opening questions: These have to be non threatening and yet lead the
respondent to get into the right frame for answering the rest of the
questions. For example, a questionnaire on understanding the
consumers buying behavior in malls, can ask an opening question such
as: What is your opinion about shopping at a mall?
c) Study questions: After the opening questions, the bulk of the
instructions needs to be devoted to the main questions that are related
to this specific information needs of the study.
The simpler questions, which do not require a lot of thinking or response
time, should be asked first as they build tempo for answering the more
difficult, sensitive questions later on. This method of going in a sequential
manner from the general to the specific is called the funnel approach. Like a
funnel, then initial set of questions are broad and as one goes along the
questions, the answers required to become more specific as well as
restrictive. To the right is a funnel-shaped questionnaire to assess pizza
purchase behavior.
Questionnaire structure-contd.
• d) Classification information: This is information that is related to the basic
socio-economic and demographic traits of the person.
• These might include name, address, e-mail address and telephone number.
• e) Acknowledgement: The questionnaire ends by acknowledging the input
of the respondent and thanking him for his cooperation and valuable
contribution.
• f) Sequential order: The researcher must take care that there is a logical
order maintaining the questions that are asked.
• A set of questions related to a particular area of investigation must be
asked first before moving on to the next. Sometimes the set of questions
that are to be asked are dependent on the answer that a particular person
gives and there are different possibilities for each answer.
• In this case, one needs to design a separate set of questions for each
selected answer. These kinds of questions are called branching questions.
These questions are designed so that all possibilities are covered. Thus,
they require careful formulation and inclusion in the questionnaire format.
8. Physical characteristics of the questionnaire
The questionnaire is a very important document that is the first interface between the respondent and the
researcher. Thus, the appearance of the instrument is very important. Following things are important:
1. The quality of the paper on which the question is printed must be good.

2. In case the number of questions is too many and then put them together as a booklet or one can have a
double page format for the questions.

3. The font style and spacing used in the entire document should be uniform. One must ensure that every
question and its response options are printed on the same page.

4. Incase the questionnaire is long, or the researcher is economizing, one must not crowd questions together
with no line spacing to make the questionnaire seem shorter.

5. Surveys for different groups could be on different coloured paper.

6. In case there is any response instruction for an individual question, it must accompany the question.
9. Pilot testing of the questionnaire:

• Pilot testing refers to testing and administering the design instrument on a small group of people from
the population under study.
• This is to essentially cover any errors that might have still remained even after the earlier eight steps.
• Every aspect of the questionnaire has to be tested and one must record all the experiences of the
conduction, including the time taken to administer it.
• In case the question got no answers then it might be essential to rephrase the entire question. Pilot tests
should always be done in a face to face interaction.
• Here the researcher is able to observe and record responses, both verbal and non-verbal. Sometimes
the researcher might also get the questionnaire vetted by academic or industry experts for their inputs.
• The researcher must explore and see whether the information that is being collected through the
questionnaire would adequately furnish the information needs for which the instrument was designed.
10. Administering the questionnaire:
Once all the nine steps have been completed, the final instrument is ready for conduction and the questionnaire needs to be
administered according to the sampling plan.