Anda di halaman 1dari 88

Definition of Marketing Research:

Marketing research is defined as the systematic and

objective identification, collection, analysis, and
dissemination of information for the purpose of
assisting management in decision making related to the
identification and solution of problems (and
opportunities) in marketing.

1) Identification: Involves defining the marketing

research problem (or opportunity) and determining
the information that is needed to address it.

2) Collection: Data must be obtained from relevant


3) Analysis: Data are analyzed, interpreted, and

inferences are drawn.

4) Dissemination of information: The findings,

implications, and recommendations are provided in a
format that makes this information actionable and
directly useful as an input into decision making.

Classifications of marketing research.:

1) Problem identification research: The goal is to
identify existing or potential problems not apparent on
the surface. Examples include market potential, market
share, market characteristics, sales analysis, short-range
forecasting, long-range forecasting, and business trends

2) Problem solution research: The goal is to solve

specific marketing problems such as segmentation,
product, pricing promotion, and distribution research.

Steps involved in the MARKETING RESEARCH


1) Problem definition: Defining the marketing research

problem to be addressed is the most important step
because all other steps will be based on this definition.

2) Developing an approach to the problem:

Development of a broad specification of how the
problem will be addressed allows the researcher to
break the problem into salient issues and manageable

3) Research design formulation: A framework for

conducting the marketing research project that specifies
the procedures necessary for obtaining the required
information. It details the statistical methodology
needed to solve the problem and thus the data
requirements needed from data collection.

4) Fieldwork or data collection: A field force (personal

interviewing, phone, mail, or electronic surveys) gathers
project data. Although seemingly trivial in nature, to
obtain meaningful results field workers must be
accurate and thorough in data collection.

5) Data preparation and analysis: The editing, coding,

transcription, and verification of data allow researchers
to derive meaning from the data.

6) Report preparation and presentation: The findings

are communicated to the client. The report should
address the specific research questions identified in the
problem definition, describe the approach, the research
design, data collection and the data analysis procedures
adopted, and present the results and the major findings.
Customer Groups

• Consumers
• Employees
• Shareholders
MIDM • Suppliers 3
variables RESEARCH

• Economy
• Product • Technology
• Price • Competition
• Promotion • Laws & Regulations
• Distribution • Social and Cultural
• Political Factors

Assessing Providing Marketing

Information Information Decision
Needs Making

Marketing Managers

• Marketing Segment
• Target Market Selection
• Marketing Programs
• Performance & Control

The Role of Marketing Research

Nature of Marketing Research and its components :

Marketing research provides the information for

decision makers at each step of the marketing decision
process. It is the goal of marketing research to provide
relevant, accurate, reliable, valid, and current
information to management in order to facilitate

managerial decisions. Each of these characteristics can
be defined for students as:

 Information that is relevant, addresses the problem or

issue being investigated.
 Information that is accurate, correct, and precise.
 Information that is reliable, and originates from
competent, trustworthy sources.
 Information that is valid is applicable to the problem
at hand.
 Information that is current is timely and up-to-date for
both the industry and issue under consideration.

Two types of marketing research suppliers:

The two types of suppliers by their relationship to the

client, internal or external are:

a) Internal supplier—a marketing research department

located within the firm where all the marketing research
staff members are employees of the firm. Most major
corporations have their own marketing research

b) External supplier—research suppliers that are not a

part of the firm. The external supplier may offer the
entire range of marketing services including problem
definition, developing an approach, questionnaire
design, sampling, data collection, data analysis,
interpretation, and report preparation and presentation.

Services offered by a full-service marketing research


1) Syndicated services: Offered by research

organizations that provide information from a common
database to different firms that subscribe to their

2) Standardized services: Research studies conducted

for different client firms but always in the same way.

3) Customized services: Offer a wide variety of

marketing research services customized or tailor-made
to suit the specific needs of a particular client.

4) Internet services: Offered by firms that have

specialized in conducting marketing research on the

Services offered by limited-service suppliers:

1) Coding and data entry services: The supplier will
take the administered questionnaires, edit them, develop
a coding scheme, and transcribe the data onto diskettes
or magnetic tapes for input into the computer.

2) Analytical services: These services include

questionnaire design and pretesting, determining the
best means of collecting data, sampling plans, and other
aspects of the research design.

3) Data analysis services: Offer sophisticated data

analysis using multivariate techniques.

4) Branded product and services: Consist of specialized

data collection and analysis procedures developed to
address specific types of marketing research problems.

Criteria when selecting an External Supplier:

1) The firm selected should be capable of working on

the project which includes the employees of the
supplier, facilities for fieldwork, and the data
2) The firm should possess a high degree of technical

3) There should not be any personality clashes
between the client and the supplier.
4) Good communication between the client and the
supplier is essential to the success of a project.
5) The supplier should provide supervision and control
of the fieldwork and other phases of the
project and offer acceptable validation procedures.
6) The supplier should be flexible to meet the unique
needs of the client and the project.
7) The supplier should be able to complete the work on
8) The supplier should have experience in order to use
sound judgment when conducting certain
marketing research tasks.
9) The supplier should understand the role of research in
developing marketing strategies and making
marketing decisions.
10) The supplier should maintain high ethical standards.
11) The approach adopted would be influenced by the
research ideology of the supplier.
12) The supplier should have a good reputation.
13) How much the supplier is charging for conducting
the project should be a factor.
14) A location close to the client is desirable but not

Ethical considerations in marketing research:

There is the potential to abuse or misuse marketing

research by taking advantage of the respondents and the
general public, for example, by misrepresenting the
research findings in advertising. The profit motive may
occasionally cause researchers or clients to compromise
the objectivity or professionalism associated with the
marketing research process.

Marketing research has often been described as having

four stakeholders. These stakeholders are (1) the
marketing researcher, (2) the client, (3) the respondent,
and (4) the public. Ethical issues can be understood in
terms of the responsibilities these stakeholders have to
each other and to the research project. When conflict
occurs, it becomes the responsibility of the stakeholders
involved to behave honorably. Sometimes accepted
codes of conduct help guide this behavior. Often
decisions rely solely on the character of the stakeholder.

Importance of the Problem Definition Process:

A clearly defined problem serves as a guideline to the

researcher in designing and conducting research
properly. Thus, it helps the researcher in answering the
question: What is to be done? In absence of a well-
defined problem, the data collected may be worthless to
the decision maker. Stress that a clearly laid down
research problem leads to goal-directed research, which
will meet the objectives of the decision maker instead of
haphazard research, which often provides incomplete

Process of formulating the problem.

Formulating the problem is a sequential process. The

first step involves discussion with the decision-maker.
The researcher needs to understand the nature of both
the problem and the decision which management faces
in order to determine the underlying information needs.
Sometimes discussions with industry experts, analysis
of secondary data, and preliminary research are required
to identify the factors that must be considered for the
proper identification of the decision problem. The final
step is to translate the decision problem into a research

Importance of the decision maker (DM) to the


The researcher must communicate with the DM in order
to understand the nature of the problem the DM faces
and what he hopes to learn from the research. Such an
understanding will help the researcher in gathering
information relevant to the problem faced by the
decision-maker. Note that a candid and open discussion
between the researcher and DM may help in

1) The events that led to the need for making a

2) The alternative courses of action available to DM.
3) The criteria to be used in evaluating various courses
of action.
4) The information that is needed by the DM in
making the decision.

A systematic approach to working with the DM is the

problem audit. It enables the researcher to get beyond
the mere symptoms to understand the causes of the

Problem audit:
A problem audit is a comprehensive examination of a
marketing problem situation with the purpose of
understanding its origin and nature. The problem audit
involves discussions with the DM on the following

a) The events that led to the decision that action is

needed, or the history of the problem.
b) The alternative courses of action available to the
DM. The set of alternatives may be incomplete at
this stage, and qualitative research may be needed
to identify the more innovative courses of action.
c) The criteria that will be used to evaluate the
alternative courses of action. For example, new
product offerings might be evaluated on the basis of
sales, market share, profitability, return on
investment, and so forth.
d) The potential actions that are likely to be suggested
based on the research findings.
e) The information that is needed to answer the DM’s
f) The manner in which the DM will use each item of
information in making the decision.
g) The corporate culture as it relates to decision
making. In some firms, the decision making process
is dominant; in others, the personality of the DM is
more important.

Role of industry experts and secondary data in
identifying problem(s):

Note that industry experts can provide useful

information about the prevailing market conditions.
They can be especially useful in the case of industrial
marketing research where technical knowledge is
required. Regarding secondary data, it is important to
provide economical and quick information that can be
useful in understanding the problem clearly. Sometimes
focus groups are used to provide information that is
then used in refining the problem.

Role of qualitative research in the process of

developing an approach:

The purpose of qualitative research is to get a feel for

the situation rather than a conclusive result. Such
research can and should play a useful role in helping the
researcher to understand the problem more clearly.
Techniques such as focus group interviews, pilot
surveys, and in-depth interviews are often used to find
the opinion of the consumers. This helps the researcher
in refining the problem and guiding the research in the
right direction.

Factors affecting the problem definition process:

1) Past information and forecasts: Past information

and forecasts provide industry data that put the
current problem into context.
2) Resources and constraints: Resources and
constraints force the problem to be defined in an
appropriate scope.
3) Objectives: An understanding of the objectives of
the organization and decision maker allows the
researcher to pinpoint the exact desires for the
4) Buyer behavior: An understanding of the ultimate
consumer’s behavior is critical to understanding
their response to specific marketing actions.
5) Legal behavior: The legal environment may
regulate certain aspects of the marketing mix and
the research effort, thus affecting the problem
6) Economic environment: The economic environment
can affect the decisions of consumers and impact
the marketing mix.
7) Marketing and technological skills: The abilities of
the organization to develop and market products
may affect the scope of the research to be done. In

addition, technological advances offer new methods
of conducting marketing research.

Marketing Research Problem and Management

Decision Problem:

The management decision problem asks what the DM

needs to do, whereas the marketing research problem
entails determining what information is needed and how
it can be obtained in the most feasible way.

For example:

Should the price be cut Determine the buyer-
in response to a price- behavior at various
cut by a competitor? price levels.
Should the product ‘X’ Assess the probable
be introduced in the market size and share
market? for product ‘X’
What should be done to Determine the strengths
increase the relative and weaknesses of ‘Y’
market share of product vis-à-vis those of the
‘Y’? competitors.

Components of a well-defined marketing research

A well-defined marketing research problem consists of
both a broad statement and a list of specific components
of the problem. The broad statement provides
perspective on the problem and acts as a safeguard
against overlooking important aspects of the marketing
research and management decision problems. The
specific components focus on the key aspects of the
problem and provide clear guidelines on how to proceed

A broad definition does not provide guidelines for

subsequent steps in research. A narrow definition, on
the other hand, may preclude the consideration of some
courses of action. In either case, the solution reached
will not be directly related to the problem and may be of
little use to the manager.

An example -
Department Store Project: Problem Definition

MR problem -
To determine the relative strengths and weaknesses of
Shopper’s Stop, vis-à-vis other major competitors,
with respect to factors that influence store patronage.

Specifically, research should provide information on
the following questions.

1. What criteria do households use when selecting

department stores?
2. How do households evaluate Shopper’s Stop and
competing stores in terms of the choice criteria
identified in question 1?
3. Which stores are patronized when shopping for
specific product categories?
Department Store Project: Problem Definition
4. What is the market share of shopper’s Stop and its
competitors for specific product categories?
5. What is the demographic and psychological profile
of the customers of Shopper’s Stop? Does it differ
from the profile of customers of competing stores?
6. Can store patronage and preference be explained in
terms of store evaluations and customer

Research Design

Importance of a good research design:

Research design is a framework or blueprint for

conducting the marketing research project. It specifies
the precise details of the procedures necessary for
obtaining the required information. Finally, stress that it
is important to have a good research design in order to
ensure that the marketing research project is conducted
effectively and efficiently.

Exploratory and Conclusive Research Designs:

Exploratory research is used in situations where the

problem may have to be defined more precisely,
relevant courses of action identified, hypotheses
formulated, or additional insights gained before an
approach can be developed. Conclusive research would
be used to test specific hypotheses, examine specific
relationships, or make predictions.

Descriptive and Causal Research.:

Although both descriptive and causal research are

classified as conclusive research, they differ in terms of
their objectives.

Descriptive research is used to describe something,

usually market characteristics or functions.

Causal research is used to obtain evidence regarding
cause-and-effect relationships.

Distinction between exploratory, descriptive, and causal


Exploratory research is typically used to provide

structure and insight into the research problem. For
example, using focus groups to determine key factors
related to the use of your product.

Descriptive research begins with the structure already

defined and proceeds to actual data collection in order
to describe some market variable. For example,
determining the average age of purchasers of your

Causal research also proceeds from a pre-established

structure but attempts to infer causal relationships
between variables as opposed to describing variables.
For example, determining if increased advertising
spending has led to an increase in sales.

Six W’s of descriptive research:

The six W’s are used by journalists when trying to
gather facts for a story. In like manner, because
descriptive research is marked by the prior formulation
of specific hypotheses, the design requires a clear
specification of the six W’s of the research:

1) Who: Who should be considered?

2) Where: Where should the respondents be contacted
to obtain the required information?
3) When: When should the information be obtained
from the respondents?
4) What: What information should be obtained from
the respondents?
5) Why: Why are we obtaining information from the
6) Way: The way in which we are going to obtain
information from the respondents.

These questions form the basis for describing the

research to be conducted.

Descriptive research is characterized by a clear

statement of the problem, specific hypotheses, and
detailed information needs.

Causal research is appropriate to use when the purposes
are to understand which variables are the cause and
which variables are the effect, and to determine the
nature of the functional relationship between the causal
variables and the effect to be predicted.

Six components of a research design:

a) Define the information needed.

b) Design the exploratory, descriptive, and/or causal
phases of the research.
c) Specify the measurement and scaling procedures.
d) Construct and pretest a questionnaire or an
appropriate form for data collection.
e) Specify the sampling process and sample size.
f) Develop a plan of data analysis.

Primary and Secondary Data

Primary data are originated by researcher for the

specific purpose of addressing the research problem.

Secondary data is data that has already been collected

for purposes other than the problem at hand.
The data are often found internally, and also from
published materials, computerized databases, or from
syndicated services. Secondary data are characterized as
easily available and relatively inexpensive to obtain.

A comparison of Primary and Secondary Data

Collection For the problem For other
Purpose at hand problems
Collection Very involved Rapid and easy
Collection Cost High Low
Collection Time Long short

Scope of secondary data:

Secondary data can cover a broad range of factors that

affect the problem at hand. It does not always fit the
specific problem at hand, but can be useful in
developing an approach to the problem and providing a
comprehensive understanding of the problem

It is important to obtain secondary data before primary
data because secondary data, as compared to primary
data, are easily available, inexpensive, and retrieving
secondary data requires a short amount of time. In
addition, secondary data generally provide valuable
insights for collecting primary data.

The difference between internal and external secondary

data is that internal data are those available within the
organization for which the research is being conducted
although external data are those generated by sources
outside the organization.

Advantages of secondary data:

Secondary data can help you:

 Identify the problem.

 Better define the problem.
 Develop an approach to the problem.
 Formulate an appropriate research design (for
example, by identifying the key variables).
 Answer certain research questions and test some
 Interpret primary data more insightfully.

Examination of available secondary data is a
prerequisite to the collection of primary data. Proceed to
primary data only when the secondary data sources
have been exhausted or yield marginal returns.

Disadvantages of Secondary Data:

Because secondary data have been collected for

purposes other than the problem at hand, their
usefulness to the current problem may be limited in
several important ways, including relevance and

The objectives, nature, and methods used to collect the

secondary data may not be appropriate to the present

Secondary data may be lacking in accuracy, or may not

be completely current or dependable.

Internal Sources of Secondary Data:

Internal sources can supply some of the most vital data

for research. The information generated by the
corporation’s daily business operations can represent a
wealth of data useful to the researcher and should be the
starting point of a project. It offers the advantages of
being proprietary to the company and is available at a
low cost relative to outside suppliers.

For example, sales data are valuable information for any

marketing project because it shows the exact results of a
program, salesperson, or sales region. Actual costs
allow the researcher to estimate costs for a research
study or project costs for a marketing program. Detailed
information can be gathered on precise questions; for
example, the percentage of sales to industry versus
government, or sales broken out by company accounts.
With planning, sales data can be recorded in the
companies’ management information systems to allow
for optimal use by analysts.

Published External Secondary Data:

Published data is abound but the key is knowing where

to look for it. Both government and nongovernmental
published sources exist, as well as guides, directories,
and indexes to help locate the necessary information.
The difficulty is locating either the right directory or
index to guide you, or to understand the classification
system used. A good librarian is a real advantage and
can cut down the search time remarkably. You may
want to bring a copy of a directory or index to show
how complicated they can be to use.

Census data - a major source of secondary data:

The Census Bureau is the largest source of statistics.

Census data can provide important information on
demographics, manufacturers, retail trade agriculture,
transportation, and so on. The quality of census data is
very high, making it a very reliable and useful source.

Databases in marketing research:

The use of databases has increased phenomenally due to

the rocketing sales of PCs and due to the increase in the
number of vendors providing such databases. Both
online and offline databases are available consisting of
bibliographic, numeric, full-text, directory, and
specialized databases. In addition, directories of
databases exist to aid in locating the proper information.

Syndicated sources of secondary data:

Companies which collect and pool information and sell

it to a number of clients are termed syndicated sources
of data. They provide information on both the consumer
and industrial market via a variety of data gathering
techniques. Based on the unit of measurement,
syndicated services can be classified as
household/consumer-oriented or institutionally-

Household services emphasize the consumer. The most

popular form of measurement is the survey. Institutional
services emphasize the distribution chain. Information
is more difficult to obtain here because of the strategic
and proprietary nature of information in the business
realm. As a result, it is often the case that valid reports
on institutions will be more general than most marketers
would like.

Panels and surveys:

A panel is a sample of respondents that provide

specified information at regular intervals over an
extended period of time. The distinguishing feature of
purchase panels is that the respondents in the panel are
required to record specific behaviors as they occur. In
media panels, the behavior is automatically recorded by
electronic devices.

1) Panels provide longitudinal data, i.e. data obtained
from the same respondents repeatedly.

2) Panel members may provide higher quality data than

would a sample because of their willingness to serve
on the panel.


1) Most panels are not representative of the population.

2) Over a period of time maturation sets in and the panel

members must be replaced.

3) Response biases may occur as simply being on the

panel may alter behavior.

Psychographics and AIO:

Psychographics is psychological characteristics of

consumers that can be quantified. They can then be
classified as consisting of two types of variables:
lifestyle and personality.

A lifestyle may be defined as a distinctive pattern of
living that is described by the activities people engage
in, the interests they have, and the opinions they hold of
themselves and the world around them. These activities,
interests, and opinions are termed AIOs. The AIOs can
be used to segment people into groups with different
lifestyles and then the marketing effort can be geared to
meet the needs of any group(s).

Need to use multiple sources of secondary data:

No single source of secondary data can provide

information that completely fits the need of the present
problem. Moreover, each method has certain
weaknesses. One may suffer from biases, the other may
be dated, the third may not be very accurate, and so on.
Combining various secondary sources can cancel out
the weaknesses of one another and provide accurate,
comprehensive, and high-quality data.

Qualitative and Quantitative Research:

Qualitative Research is a unstructured, exploratory

research methodology based on small sample that

provides insights and understanding of the problem

Quantitative Research is a research methodology that

seeks to quantify the dats and typically, applies some
form of statistical analysis.

Qualitative Versus Quantitative Research

Objective To gain a To quantify the
qualitative data and
understanding of generalize the
the underlying results from the
reasons and sample to the
motivation population of
Sample Small number of Large number of
non representative
representative cases
Data collection Unstructured Structured
Data analysis Non statistical Statistical
Outcome Develop an initial Recommend a
understanding final course of

The primary differences between quantitative research

and qualitative research are that in quantitative research
the objective is to quantify the data and generalize the
results from the sample to the population of interest,
whereas qualitative research’s objective is to gain a
general understanding of the underlying reasons and
motivations of the group.

Other differences are that qualitative research employs a

small number of nonrepresentative cases, uses unstructured
data collection techniques, nonstatistical data analysis, and
results in an initial understanding of the problem.
Quantitative research, however, employs a large number of
representative cases in the sample, uses structured data
collection techniques, statistical data analysis, and results
in recommendations for action.

Qualitative research is called the “why” research. Its

basic purpose is to provide additional insights and
understanding of the problem at hand. This type of
research is typically based on a loosely structured
nonrepresentative sample, unstructured interviews or

observations, and a nonstatistical approach to data

Inter-relationship between qualitative research and

quantitative research:

Qualitative research is primarily used to gain an initial

understanding of underlying motivations for people’s
attitudes, preferences, and behavior. The goal is to gain
sufficient knowledge about the scope and the general
nature of the problem at hand to direct future
quantitative research. The findings from qualitative
research can then be used to generate hypotheses on key
variables that can be tested quantitatively.

Classification of Marketing Research Data

Marketing Research Data

Secondary Data Primary Data

Qualitative Data Quantitative Data

Descriptive Causal

Survey Observational Experimental
Data & Other Data Data
Categories of Qualitative Research.

Qualitative Research

Direct (Non Disguised) Indirect (Disguised)

Focus Group Depth Interview Projective Techniques

Association Completion Construction Expressive

Techniques Techniques Techniques Techniques

The three types of qualitative research can be classified
as direct or indirect. The difference is that in direct
research the purpose of the research is disclosed to the
respondent or is otherwise obvious to them, whereas the
reason for the research is not disclosed to the
respondents of indirect qualitative research. Focus
groups are an example of direct techniques whereas
word association is an illustration of indirect

Focus groups consist of 8 to 12 people interviewed by a

trained moderator to gain insights on the salient aspects
of a marketing problem by listening to a group of
people from an appropriate target market. Depth
interviews are conducted on a one-on-one basis to
uncover underlying motivations, beliefs, attitudes, and
feelings on a topic. Indirect techniques, also called
projective techniques, are used to ask the respondent to
interpret the behavior of others and thus add insight into
personally held beliefs of the respondent.

Focus group:

A focus group is an interview conducted by a trained

moderator among a small group of respondents in a
nonstructured and natural manner. The focus group is
the most popular qualitative research technique because
it is a relatively inexpensive technique that can be used
in almost any situation requiring preliminary
understanding and insights. This technique has several
advantages that include synergism, snowballing,
stimulation, security, spontaneity, serendipity,
specialization, scientific scrutiny, structure, and speed.

Planning and conducting focus groups:

1) Examine the objectives of the research project—

provides the rationale for conducting the focus
2) Specify the objectives—outlines the goals of the
study in order to guide the interview.
3) State the questions to be answered from the focus
group—a detailed set of questions to be answered.

4) Write a screening questionnaire—ensures that
participants represent an appropriate sample for the
5) Develop a Moderator’s Outline—ensures that the
moderator understands the nature of the study and
the key findings desired by the client.
6) Conduct the interview
7) Review tapes and analyze data—allows the
researcher to uncover inconsistent responses,
missed remarks, nonverbal communication and new
8) Summarize findings and plan follow-up research—
to probe further into the issues and sample
statistically significant populations.

Role and skill requirements of effective focus group


The moderator plays a vital role in the success of a

focus group. He is responsible for setting a tone in the
focus group that makes the respondents feel
comfortable enough to discuss their thoughts. He also
must establish the rules, direct the study, communicate
the objectives to the respondents, probe the respondents
for deeply held attitudes, and facilitate a free-flowing
discussion in the relevant areas. Finally, he must
summarize the group’s responses to ensure agreement
on his interpretation of their responses.

Some key qualifications of focus group moderators

include kindness, firmness, permissiveness,
involvement, the ability to project incomplete
understanding, encouragement, flexibility, and

Disadvantages of focus groups:

1) It is harder to correctly interpret the responses.

2) There is extreme dependence on the performance of
the moderator.
3) Coding and analysis is cumbersome.
4) The group is not representative of the general
5) The results are subject to researcher or client bias.

Depth interviews:

A depth interview is an unstructured and direct way of

obtaining information and is conducted on a one-on-one
basis. The respondent is probed in depth by a highly
skilled interviewer to uncover underlying motivations,
beliefs, attitudes, and feelings on a topic. The
interviewer attempts to get the subject to talk freely and
the direction of the interview is influenced heavily by
the subject’s answers. Depth interviews are preferable
to focus groups in cases where detailed probing of the
individual is required, the topic is considered
confidential or embarrassing, or detailed understanding
of complicated behavior is required.


1) Great depths of insights can be uncovered.

2) It associates the responses directly with the
3) There is no social pressure to conform to group


1) Skilled interviewers capable of conducting depth

interviews are expensive and difficult to find.
2) The lack of structure makes the results very
susceptible to the influence of the interviewer.
3) The quality and completeness of the results depends
very heavily on the skills of the interviewer.
4) The data obtained is problematic to analyze and
Depth interviews can be effectively employed in special
problem situations, such as those requiring:

1) Detailed probing of the individual.

2) Discussions on topics considered confidential,
sensitive, or embarrassing.
3) Situations where strong social norms exist and the
individual may be easily swayed by group response.
4) Detailed understanding of complicated behavior.
5) Interviews with professional people.
6) Interviews with competitors who are unlikely to
reveal the information in a group setting.
7) Situations where the product consumption
experience is sensory in nature affecting mood
states and emotions.

Projective techniques:

Projective techniques use third person associations to

indirectly ascertain the respondent’s motivations/
attitudes/ beliefs. When a personal or threatening
question is asked directly, it is assumed that the
respondent will be reluctant to answer in complete truth,
but if the same question is asked about a third party, it is

assumed a more accurate and complete answer
pertaining to the respondent will be supplied.

The four forms of projective techniques are:

1) Association techniques: An individual is presented

with a stimulus or stimuli and asked to respond with
the first thing or things that come to mind.

2) Completion techniques: The respondent is required

to complete an incomplete stimulus situation.

3) Construction techniques: The respondent is

required to construct response in the form of a
story, dialogue, or description.

4) Expressive techniques: The respondents are

presented with a verbal or visual situation and asked
to relate the feelings and attitudes of other people to
the situation.

Descriptive Research Design

Descriptive Research consist of Surveys and
Observations as the main ways in which relevant data
can be collected.

Survey Method: A structured questionnaire given to

respondents and designed to elicit specific information

This method of obtaining information is based on the

questioning of respondents. The respondents are asked
variety of questions regarding their behavior, intentions,
attitudes, awareness, motivations and demographic and
lifestyle characteristics.

Structured data collection: use of a formal

questionnaire that presents questions in a prearranged

Fixed alternative questions: Questions that require

respondents to choose from a set of predetermined

e.g. Disagree Agree

1 2 3 4 5

Shopping in department store is fun

Classification of Survey Methods

Survey Method

Telephone Mail Electronic
Interviewing Interviewing Interviewing


Mall Computer
Intercept Assisted

E-mail Internet


Observation : The recording of behavior patterns of
people, objects and events in a systematic manner to
obtain information about the phenomenon of interest.

The observer does not question or communicate with

the people being observed.

Information may be recorded as the events occur or

from records of past events.

Observational method may be structured or


Structured Observation: Observation techniques where

the researcher clearly defines the behaviors to be
observed and the methods by which they will be

e.g. Auditor performing inventory analysis in the store.

Unstructured Observation: Observation that involves a

researcher monitoring all aspects of the phenomenon
without specifying the details in advance.

e.g. observing children playing with new toys.

Measurement and Scaling

Measurement proceeds scaling in test construction.

Measurement is the assignment of numbers or other

symbols to characteristics of objects according to
certain prespecified rules. Scaling is an extension of
measurement where it involves the generation of a
continuum upon which measured objects are located.

Primary scales of measurement:

1) Nominal scale: This is used only as a labeling

scheme where numbers serve only as labels or tags for
identifying and classifying objects. The numbers in a
nominal scale do not reflect the amount of a
characteristic possessed by the objects, rather they are
used only for identification. For example, numbers on
cricket players uniforms, street names, or insurance
policy numbers.

2) Ordinal scale: This is a ranking scale in which

numbers are assigned to objects to indicate the relative
extent to which some characteristic is possessed. It is
then possible to determine whether an object has more
or less of a characteristic than some other object. For
example, rankings of tennis players for a championship,
socioeconomic status and quality rankings.

3) Interval scale: Numbers are used to rank objects

such that numerically equal distances on the scale
represent equal distances in the characteristic being
measured. Examples include time and temperature.

4) Ratio scale: This is used to identify or classify

objects, rank order the objects, and compare intervals,
differences. For example, height, age, and income.

The differences between a nominal and an ordinal scale

are that in nominal scales, the numbers serve only as
labels or tags for identifying and classifying objects
while in an ordinal scale the numbers are used as a
ranking device. Although both nominal and ordinal
scale data can be used for counting operations, ordinal
scales permit the use of statistics based on percentiles.

The advantage of a ratio scale over an interval scale is

that the origin is fixed. Hence, it is meaningful to take
ratios of scale values. Statistics such as the geometric
mean, harmonic mean, and coefficient of variation can
be applied to analyze ratio scale data. However, this
advantage is not significant because the commonly used
statistics in marketing research can be calculated on
interval data.

Two types of scales: comparative and non-comparative.

Comparative scales—a direct comparison of stimulus

objects is elicited. Thus, two brands may be compared
along a dimension such as quality.

Non-comparative scales — the respondent provides

whatever standard seems appropriate to him/her, thus,
only one object is evaluated at a time. In this case, one
brand is rated on a scale independent of other brands.

Classification of Scaling Techniques

Scaling Techniques

Comparative Scales Non comparative Scales

Continuous rating Itemized

Scale Rating Scales

Paired Rank Constant Q-Sort and Other

Comparison Order Sum Procedures

Likert Stapel
Comparative scaling techniques:

All comparative scaling techniques involve a direct

comparison of stimulus objects with one another. This
should be highlighted as each of the scales are discussed
in turn.
1. Paired comparison scaling: Here a respondent is
presented with two objects at a time and asked to
select one object in the pair according to some
criterion. The data obtained is ordinal in nature.
This is frequently used in marketing when
comparisons of products or brands are being made.
Refer Figure 3 for an example of paired comparison

2. Rank order scaling: Respondents are presented with

several objects simultaneously and asked to order or
rank them according to some criterion. This is
commonly used to measure preferences for brands
as well as the importance of attributes.

3. Constant sum scaling: Respondents are required to

allocate a constant sum of units such as points,
dollars, chits, stickers, or chips among a set of
stimulus objects with respect to some criterion.
Specific instructions are provided that if an attribute
is not at all important, it is possible to assign zero
points. If an attribute is twice as important as some
other attribute it should receive twice as many

4. Q-sort scaling: This technique uses a rank order
procedure in which objects are sorted into piles
based on similarity with respect to some criterion.
In Magnitude estimation numbers are assigned to
objects such that ratios between the assigned
numbers reflect ratios among the objects on the
specified criterion. Guttman scaling or scalogram
analysis: A procedure for determining whether a set
of objects can be ordered into an internally
consistent, one-dimensional scale.

Non-comparative scaling techniques:

Non-comparative scaling is the type of scaling, which

does not compare the object against another object or
some standard. Rather, the rater uses whatever standard
seems most appropriate to him or her.

1. Continuous rating scale: The respondents rate the

objects by placing a mark at the appropriate position on
a line that runs from one extreme of the criterion
variable to the other. The form of the continuous scale
varies considerably depending on the imagination of the
researcher. Their use in marketing has been limited
because they are not as reliable as itemized scales, the

scoring process is cumbersome, and they provide little
additional information.


Continuous Rating Scales

How would you rate Shopper’s Stop (SS) as a

department store?

Version 1

Probably the Worst

------ Probably the Best

Version 2

Probably the Worst

------ Probably the Best

0 10 20 30 40 50 60 70
80 90 100
Version 3

Very bad Neither good Very

nor bad

Probably the Worst

------ Probably the Best

0 10 20 30 40 50 60 70
80 90 100

2. Itemized rating scale: The respondents are provided

with scales having numbers and/or brief descriptions
associated with each category. The respondents are
required to select one of the specified categories that
best describes the object being rated. Several types of
itemized rating scales exist.

(i. Likert scale: The respondents are required to indicate
a degree of agreement or disagreement with each of a
series of statements related to the stimulus objects. The
Likert scale is often used in marketing. It is easy to
construct and administer, it is easy for respondents to
complete, and it is suitable for mail, telephone, and
personal surveys.

Strongly Disagree Neither Agree Strongly

disagree agree nor

1. Shoppers Stop sells high quality 1 2X

3 4 5

2. SS has poor in-store service. 1 2X 3


3. I like to shop at SS 1 2 3X

(ii) Semantic differential scale: It is a seven-point
rating scale with end points associated with bipolar
labels that have semantic meaning. Respondents are
required to rate objects on a number of itemized, seven-
point rating scales bounded at each end by one of two
bipolar adjectives. This scale is popular in marketing
and has been used in image studies, promotion strategy,
and new product development studies.

Shopper’s Stop IS:

Powerful --:--:--:--:-X-:--:--: Weak

Unreliable --:--:--:--:--:-X-:--: Reliable
Modern --:--:--:--:--:--:-X-: Old-fashioned

(iii). Stapel scale: Is a unipolar rating scale with ten

categories numbered from -5 to +5, without a neutral
point. Respondents are asked to indicate how accurately
or inaccurately each term describes the object by
selecting an appropriate numerical response category.
Though easier to construct than the semantic

differential, while giving the same results, the Stapel
scale has not been widely applied in marketing.

Shopper’s Stop

+5 +5
+4 +4
+3 +3
+2 +2X
+1 +1
-1 -1
-2 -2
-3 -3
-4X -4
-5 -5

The differences between the Stapel scale and the

Semantic differential is that with the Stapel scale there
is no need to pretest the adjectives or phrases to ensure
true bipolarity and the Stapel scale can also be

administered over the telephone. The semantic
differential is more popular than the Stapel scale.

Major decisions involved in constructing itemized

rating scales:

1) Number of scale categories to use: The number of

categories should be between five and nine; however,
there is no single, optimal number of categories that
would be applicable for all scaling situations. In the
Department Store example, the class can debate the
trade off between ability to discriminate among the
categories and ability of the respondent to answer.

2) Balanced versus unbalanced scale: A balanced

scale has an equal number of favorable and unfavorable
categories used; otherwise, the scale is unbalanced. The
scale should be balanced in order to obtain objective
data; however, if the distribution of responses is likely
to be skewed, an unbalanced scale with more categories
in the direction of skewness may be appropriate.

3) Odd or even number of categories: With an odd

number of categories, the middle scale position is
generally designated as neutral. If a neutral or
indifferent scale response is a possibility for at least
some of the respondents, an odd number of categories
should be used.

4) Forced versus non-forced scales: A forced rating

scale does not have a “no opinion” or “no knowledge”
option. Thus, the respondents without an opinion may
mark the middle scale position. If a sufficient
proportion of the respondents in fact do not have an
opinion on the topic, marking the middle position in this
manner will distort measures of central tendency and
variance. In situations where the respondents are
expected to have no opinion, the accuracy of the data
may be improved by having a nonforced rating scale
which includes a “no opinion” or “no knowledge”

5) Nature and degree of verbal description: The

strength of the adjectives used to anchor the scale has a
slight tendency to influence the distribution of the
responses. With strong anchors, respondents are less
likely to use the extreme scale categories. Have the
students reach consensus on the scale anchors. Try to
encourage multiple anchors that can be used in the

6) Physical form of the scale: There is no agreement
as to which form is the most appropriate, but scales
could be presented vertically or horizontally, categories
could be expressed in terms of boxes, discrete lines, or
units on a continuum and may or may not have numbers
assigned to them, and numerical values could be
positive or negative or both. The students should decide
which format to use for the scales. Have them justify
their reasons for the scale they choose.

Questionnaire and Form Design

Importance of a well-constructed data collection


Data collection flows from specifying the information

needed, to identifying the method of obtaining it, to
determining scaling procedures. Once this is done, a
data collection instrument can be created. This usually
is a questionnaire.

It is vital that the questionnaire be constructed so as to

(1) capture the information needed,
(2) be standardized so that data are comparable across
test sites and results can thus be generalized, and (3)

maximize the efficiency of tabulating and analyzing

Objectives of a questionnaire:

These objectives should guide researchers as they

construct a questionnaire and act as guidelines for
making judgment calls on specific aspects of
questionnaire design.

a)The questionnaire must translate the information

needed into a set of specific questions that the
respondents can and will answer.
b) A questionnaire must uplift, motivate, and encourage
the respondent to become involved in the interviewing
process, to cooperate, and to complete the interview.
c)A questionnaire should be designed to minimize
response error.

Steps involved in the questionnaire design process:

1. Specify the information needed: Translate the

information needed into a set of specific questions.
Dummy tables are particularly helpful in guiding the
design of individual questions in the questionnaire. It

is important to have a clear idea of the target

2. Type of interviewing method: In personal interviews,

lengthy, complex, but diverse types of questions can
be asked. In telephone interviews, the questions asked
are short and simple because the respondents do not
get to see the questionnaire. In mail interviews, the
questions have to be simple and detailed instructions
have to be provided because there is no interaction
with the interviewer.

3. Individual question content: For each item of

information that is needed, the researcher should ask
whether that information is being adequately obtained
by questions already formulated. Also, the data
generated by each question should be used to address
the information needed. Once it is determined that the
question is needed, it is important to make sure the
question is sufficient to obtain the desired
information. Sometimes several questions are needed
instead of one.

4. Overcoming inability to answer: There are several

factors that limit the respondent’s ability to provide
the desired information:
a)Is the respondent informed: In situations where it
might be suspected that not all respondents are
likely to be informed about the topics of interest,
filter questions measuring familiarity, product use
and past experience should be asked before asking
questions on the topics.
b)Can the respondent remember: The ability to
remember an event is influenced by the event itself,
the time elapsed since the event, and the presence or
absence of events that would assist in remembering.
c)Cans the respondent articulate: Respondents should
be provided with aids to help them articulate their
responses. For example, providing a list of
descriptions or possible answers can assist the
respondent in articulating.

5. Overcoming unwillingness to answer: The reasons for

unwillingness to respond include too much effort
being required, the situation or context may not be
seen as appropriate for disclosure, no legitimate
purpose or need is seen for the information requested,
and the information is seen as sensitive.
6. Choosing question structure: Questions can be either

structured or unstructured. Unstructured questions are

open-ended questions in which the respondents have
the freedom to reply in their own words rather than
being limited to the selection of one of the response
alternatives. These are best used as opening questions.
Structured questions prespecify the set of response
alternatives and the response format. They improve
respondent cooperation and make administering and
coding the survey easier.

In designing multiple-choice questions, the number of

response alternatives should be collectively exhaustive
in that they exhaust or include the set of all possible
alternatives. The general guideline is to list all
alternatives that may be of any importance at all and
include an alternative labeled “Other (please

In the rare cases in which none of the listed alternatives

represent the appropriate response, respondents can use
the “Other” option and record their responses in the
space provided. Also, the response alternatives should
be mutually exclusive. Overlap among alternatives
should be avoided. To control for order bias, several
forms of the questionnaire should be prepared. The
order in which the alternatives are listed should be
varied from form to form. It is desirable that each
alternative appears once in the extreme positions, once
in the middle, and once somewhere in between.
Some of the advantages of multiple-choice questions
include a reduction in interviewer bias. Such questions
are faster to administer. Also, coding and processing of
data is less costly and time consuming. In self-
administered questionnaires, respondent cooperation is
improved if the majority of the questions are structured.

Some of the disadvantages of multiple-choice questions

include the potential need for exploratory research using
open-ended questions to determine the appropriate
response alternatives. It is difficult to obtain
information on alternatives not listed. Even if an “Other
(please specify)” category is included, there is a strong
tendency for the respondents to choose from among the
listed alternatives. In addition, showing respondents the
list of possible answers has the potential of producing
biased responses. As already discussed, there is also the
potential of order bias in multiple-choice questions.

7. Choosing question wording: The translation of the

desired question content and structure into words that
can be clearly and easily understood by the
respondents. The following guidelines are used for
framing questions:

a)Define the issue
b)Use ordinary words
c)Ambiguous words should be avoided
d)Avoid leading questions
e)Avoid implicit alternatives
f) Avoid implicit assumptions
g)Avoid generalizations and estimates
h)Use positive and negative statements

8. Order of the questions: The order in which the

questions are presented is a crucial element in
questionnaire design because improper ordering of the
questions can seriously bias the responses. The
opening questions should be interesting, simple, and
nonthreatening so that it will tap the interest of the
respondents and get them involved in the interviewing
process. Questions that are difficult, sensitive,
embarrassing, complex, or dull should be placed later
in the sequence only after rapport has been

9. Form and layout: It is good practice to divide a

questionnaire into several parts, each part dealing with
a specific set of questions. Research has shown that
questions at the top of the page receive more attention
than those placed at the bottom.
Some guidelines for form and layout decisions include:

a) Questions at the top of the page receive more

attention than when placed at the bottom.
b) It is a good practice to divide a questionnaire into
several parts, each part dealing with a specific set of
c) Questions in each part should be numbered—
particularly when branching questions are used.
d) Questionnaires should preferably be precoded.
e) Questionnaires should be numbered serially.

Reproduction of the questionnaire: The following


guidelines are used:

a)The questionnaire should have a professional

b) If the questionnaire runs to several pages, it should

be made into the form of a booklet rather

than a number of sheets of paper clipped or stapled
c)The entire question should be reproduced on a
single page.
d)Vertical response category formats should be used
for individual questions.
e)The questions should not be overcrowded.
f) Directions or instructions meant for the interviewer
or the respondent for individual questions
should be placed as close to the question as
g)Color does not influence response rates to
questionnaires. However, it can be employed
advantageously in branching questions.
h)The questionnaire should be reproduced in such a
way that it is easy to read and answer.
i) Several opportunities exist for improving the
reproduction of an average questionnaire in such
a way as to obtain better quality data and
simultaneously reduce costs (e.g., by reducing
printing costs).

11.Pretesting: Pretesting refers to the testing of the

questionnaire on a small sample of respondents for
the purpose of improving the questionnaire by
identifying and eliminating potential problems. In
general, a questionnaire should not be used in the field
survey unless it has been adequately pretested. (Figure
1 and Figure 2)

The issues involved in pretesting include:

a) Extensive examination of all aspects of the
b) Similarity in pretest and survey respondent
c) Initial pretest done by a personal interview, followed
by a pretest using the interviewing mode to be
actually used if that is different.
d) Selection of actual interviewers.
e) Sample size.
f) Protocol analysis and debriefing.
g) Editing and analysis.

Observational forms versus questionnaires:

Similarities—preplanning is necessary to obtain the

information needed. Specification of the who, what,
when, where, why, and how of behavior is necessary.

Form, layout, and reproduction of forms should make it

easy for recording data and later tabulating and
analyzing data. Adequate pretesting is required.

Differences—observational forms are easier to

construct because the psychological impact of the form
of questions and the way they are asked is eliminated.


The objective of marketing research is to obtain

information about the characteristics of a population,
either using a sample or census. A representative
sample, though not the entire population, contains the
same characteristics as the population, thus,
generalizability is high and population parameters can
be inferred from the information from the sample. This
information is contained in statistics, and the inferences
that are made use statistical techniques such as
estimation procedures and hypothesis testing.

Population, census and sample:

A population is the aggregate of all the elements that

share some common set of characteristics, and that
comprise the universe for the purpose of the marketing
research problem. The population parameters are
typically numbers, such as the proportion of consumers
who are loyal to a particular brand of toothpaste.
Information about population parameters may be
obtained by taking a census or a sample.

A census involves a complete enumeration of the

elements of a population. The population parameters
can be calculated directly in a straightforward way after
the census is enumerated.

A sample, on the other hand, is a subgroup of the

population selected for participation in the study.
Sample characteristics, called statistics, are then used to
make inferences about the population parameters.

Relationship between the sample design process and

the research project:

Sampling is the fifth step in the process of formulating a

research design. The researcher has already specified
the information that is needed to address the marketing
research problem and the primary sources from which
this information would be obtained have also been
determined. Furthermore, the scaling and measurement
procedures have been specified and the questionnaire
designed. At this point sampling becomes the critical
question because the researcher must now administer
the questionnaire to the relevant group of respondents.

Sampling design process:

This process is depicted in Figure 1. The sampling
design originates from the population of interest, thus,
the population must be accurately defined in a precise
statement of who should or should not be included in
the sample. It may be very general, such as residents of
the state of Delhi, or specific, like males between 20
and 30 years of age who cannot read. To measure the
population, a sampling frame is needed. The sampling
frame consists of a list or set of directions for
identifying the elements of the target population. The
sampling unit must then be set. This is the element that
contains the elements of the population, for example, it
could be a household, an office, or the individual.

At this point, the researcher is ready to introduce

quantitative techniques into the sampling design
process. The appropriate sampling technique must be
chosen which involves making decisions between using
a Bayesian approach versus a traditional approach,
sampling with replacement versus without replacement,
and using probability versus nonprobability sampling.
When the method is determined, a particular sample
size can be calculated. This number represents the total
number of elements that must be surveyed in order for
the results to be valid. Finally, the execution of the
sampling process must be operationalized, i.e., detailed
specifications of variable definitions and interviewing
procedures. Doing this ensures that all interviewers use
the same guidelines during the interview process so that
the data are comparable.

Population sampling frame discrepancy:

Population sampling frame discrepancy is defined as a

sample frame that does not include all elements of a
population. A common example is the telephone
directory (it does not include unlisted numbers and new
people who have moved recently). It should always be
identified in a study. It can be resolved in one of the
following ways:

1) Redefine the population in terms of the sampling

frame—this is a simplistic approach that avoids the
problem that portions of the population are not included
in the sampling frame. If the telephone book is our
sampling frame, it now becomes our population.

2) Screen the respondents through qualifying

questionnaires—although this eliminates unwanted
elements from the survey, it does not address the main
issue. Screening would not address the fact that some
numbers are not included in the phone book.
3) Apply a weighting scheme to counterbalance the
sampling frame error—accurate estimates for the
weights must be made to make this an effective tool.

Factors to consider in selecting a sampling


There are three decisions to be made when selecting a

sampling technique:

1) Bayesian or traditional approach—although the

Bayesian approach is theoretically more sound, in
practice it is more challenging to implement because of
its stringent statistical requirements. Thus, most
research uses the traditional approach, in which the
entire sample is selected before the research begins.

2) With or without replacement—this is only important

if the sampling frame is small compared to the sample
size because the potential bias from counting one
element twice is proportionately greater.

3) Nonprobability or probability sampling—probability
sampling offers greater statistical inference about the
population and is generally recommended for most

Probability and nonprobability sampling


Nonprobability sampling techniques do not use chance

selection procedures. Rather, they rely on the personal
judgment of the researcher. The researcher can
arbitrarily or consciously decide the elements to be
included in the sample. Nonprobability samples may
yield good estimates of the population characteristics.
However, they do not permit an objective evaluation of
the precision of the sample results. As there is no way
of determining exactly what the chance is of selecting
any particular element for inclusion in the sample,
estimates obtained are not statistically projectable to the

In probability sampling techniques, sampling units are

selected by chance. Each sampling unit has a nonzero
chance of being selected. Furthermore, it is possible to
prespecify every potential sample of a given size that
could be drawn from the population as well as the
probability of selecting each sample. There is no
requirement that every potential sample have the same
probability of selection. However, it should be possible
to specify the probability of selecting any particular
sample of a given size. This not only requires a precise
definition of the target population, but also, generally, a
specification of the sampling frame. Given the chance
selection procedures employed, it is possible to
determine the precision of the sample estimates of the
characteristics of interest. Confidence intervals
containing the true population value with a given level
of certainty can be calculated. This permits the
researcher to make inferences or projections to the
target population from which the sample was drawn.

Probability techniques differ from nonprobability

techniques in the following ways:

a) Probability techniques select sampling units by

chance as opposed to the judgment of the researcher.

b) The number of elements to be included in the sample

set can be prespecified.

c) It is possible to specify the probability of selecting

any particular sample of a given size.
d) Estimates are statistically projectable to the

Examples of nonprobability sampling techniques:

Convenience Sampling: Exploratory research

Judgmental Sampling: Test markets selected to

determine the potential of a new product
Choice of expert witnesses in a court
Selection of department stores to test
a new merchandising display system

Quota Sampling: Mall intercept interviews

Snowball Sampling: Interviews with unclassified

populations Interviews with very
small, disbursed populations

Advantages and disadvantages of each

nonprobability sampling technique:

i) Convenience sampling:

Advantages: Least expensive, Least time consuming,
Sampling units are accessible, cooperative, and easy to

Limitations: Many potential sources of bias are present,

Not representative of any population, Not able to
generalize findings

ii) Judgmental Sampling:

Advantages: Low cost, Less time consuming than

most techniques, Convenient to use

Limitations: No direct generalizations to a

specified population, Results entirely dependent on the
judgment of the researcher

iii) Quota Sampling:

Advantages: Lower cost in selecting elements for

each quota, Greater convenience in selecting
elements for each quota, With tight controls on
interviewers and interviewing, results can be
compared to probability techniques

Limitations: No assurance that a sample is
representative, Increasing controls decreases the
ease of conducting the interview, Many sources
of bias can be present due to the selection

iv) Snowball Sampling:

Advantages: Good at locating and estimating

various characteristics that are rare in a
population, Low sample variance and costs
when used

Limitations: Difficulty in obtaining the initial

sample, Limited to studies of a particular sample

Probability Sampling Techniques-

Simple Random Sampling: The distinguishing features

of SRS are:

a) Each element in the population has a known and

equal probability of selection.

b) Each possible sample of a given size has a known
and equal probability of being the sample actually

c) A sampling frame must be compiled in which each

element has a unique identification number.

d) Random numbers determine which elements are

included in the sample.

In systematic sampling, the sample is chosen by

selecting a random starting point and then picking every
ith element in succession from the sampling frame. The
sampling interval, i, is determined by dividing the
population size N by the sample size n and rounding to
the nearest integer.

Stratified sampling is a two-step process in which the

population is partitioned or stratified into
subpopulations called strata. The strata should be
mutually exclusive and collectively exhaustive so that
every population element should be assigned to one and
only one stratum and that no population elements
should be omitted. Then, elements are selected from
each stratum by a random procedure, usually by SRS.

The criterion for the selection of stratification variables
consists of homogeneity, heterogeneity, relatedness, and
cost. The stratification variables should be selected so
that elements within a stratum are as homogeneous as
possible. However, across strata the elements should be
as heterogeneous as possible. The stratification
variables should also be closely related to the
characteristic of interest. Finally, the variables should
be selected to decrease the cost of the stratification
process, i.e., the stratification variables should be easy
to measure and apply.

Proportionate stratified sampling uses strata whose size

is proportionate to the relative size of that stratum in the
total population. However, in disproportionate stratified
sampling, the size of the sample from each stratum is
proportional to the relative size of that stratum and the
standard deviation of the distribution of the
characteristic of interest among all the elements in that
stratum. In addition, it requires that some estimate of
the relative variation within the strata be known or

In cluster sampling, first the target population is divided

into mutually exclusive and exhaustive subpopulations
called clusters. Then, a random sample of clusters is
selected based on a probability sampling technique such
as SRS. For each selected cluster, either all the elements
are included in the sample or a sample of elements is
drawn probabilistically. If all the elements in each
selected cluster are included in the sample, the
procedure is called one-stage cluster sampling. On the
other hand, if a sample of elements is drawn
probabilistically from each selected cluster, the
procedure is two-stage cluster sampling.

In cluster sampling only samples of the subpopulations

are chosen, whereas in stratified sampling all of the
subpopulations are chosen. The criteria for forming
clusters is the opposite of that for stratification. Ideally,
each cluster should be a representation of the

Examples of probability sampling techniques -

Simple Random Sampling: Not widely used

Systematic Sampling: Consumer mail interviews,
Telephone interviews, Mall
intercept interviews
Stratified Sampling: Comprehensive, cost effective

Cluster Sampling: Area sampling, Cost constrained
personal interviews

Advantages and disadvantages of each probability

sampling technique:

i) Simple Random Sampling :

Advantages: It is easily understood, Sample results

may be projected to the target population

Limitations: Difficult to construct a sample frame

which permits a simple random sample to be drawn.
Cost is high because large samples are needed. Results
show lower precision with larger standard errors
compared to other techniques. The sample may not be

ii) Systematic Sampling :

Advantages: Potential of increased

representativeness of the sample versus SRS. Random
sampling is done only once. Can be performed even
when knowledge of the composition of the sampling
frame does not exist

Limitations: Ordering of elements is critical.
Improper ordering may decrease the representativeness
of the sample

iii) Stratified Sampling :

Advantages: Increases the precision of sampling

by controlling sources of sample variation. Ensures all
important subpopulations are represented in the sample

Limitations: Disproportionate sampling requires

estimates of the population variance to compute strata
size. Extra care must be taken in creating the strata to
control for sampling variation

iv) Cluster Sampling:

Advantages: Cost effective sampling method. A

sampling frame is needed only for those clusters
selected in the sample

Limitations: Results in samples that have lower

precision. Generally difficult to form heterogeneous
clusters. Difficult to compute and interpret statistics
based on clusters

Strengths and weaknesses of basic Sampling


Convenience Least expensive, Selection bias,
Sampling least time sample not
consuming, representative,
most convenient not recommended
for descriptive or
causal research
Judgmental Low cost, Does not allow
Sampling convenient, not time generalization,
consuming subjective
Quota Sample can be Selection bias, no
Sampling controlled for assurance of
certain representativeness
Snowball Can estimate rare Time consuming
Sampling characteristics

Simple Easily understood, Difficult to
Random result projectable construct
Sampling sampling
(SRS) frame, expensive,
lower precision,
no assurance of
Systematic Can increase Can decrease
Sampling representativeness, representativeness
to implement than
SRS, sampling
frame not necessary
Stratified Includes all Difficult to select
Sampling important relevant
subpopulations, stratification
precision variables, not
feasible to stratify
on many
Cluster Easy to implement, Imprecise,
Sampling cost effective difficult to
and interpret
Choice between nonprobability and probability

 The nature of the research

 relative magnitude of nonsampling versus sampling
 variability in the population
 statistical and operational considerations
 the homogeneity of the population with respect to the
variables of interest

Choosing Nonprobability versus Probability


FACTORS Nonprobabilty Probability
Sampling Sampling
Nature of Exploratory Conclusive
Relative Nonsampling Sampling errors
magnitude of errors are larger are larger
sampling and
Variability in the Homogeneous Heterogeneous
population (low) (high)
Statistical Unfavorable Favorable
Operational Favorable Unfavorable