MARKET RESEARCH

Research Design
 Theoretical Framework and Hypothesis Development
1

OBSERVATION
Broad area of
Research Interest
Identified
4
3
THEORETICAL 5 6 7
PROBLEM
FRAMEWORK GENERATIO SCIENTIFI DATA
DEFINITION
C COLLECTION
Research N OF RESEARCH ANALYSIS, AND
Problem HYPOTHESIS DESIGN INTERPRETATION
Variables
Delineated
already
PRELIMINARY identified and
labeled DEDUCTION
DATA
Hypotheses
GATHERINGS
8 Substantiated?
Interviewing
Research Question
Literature Survey
2 answered?

No 9 10 11
Yes Report Report Decision
Writing Presentation Making
Source: Research Methods for Business, Uma Sekaran
 Research Design
Introduction
• The design of the research project is a creative plan to
obtain the necessary information in the best way
possible.
• The correct design saves time and money and will result
in valid and reliable information.
• It is expected that the market researcher should be
equipped with the technical knowledge to design
research projects that will provide a wide variety of
marketing information to key decision-makers.
• A range of required design skills are covered under this
topic.
Research Design
 Research Design
Research Design:
A research design is the operational plan for executing a
marketing research project. It describes the steps necessary to
obtain the information needed to structure or solve marketing
problems and reduce decision-making risk.
Typically, creating a research design involves the following
tasks:
• Define the information needed
• Design the exploratory, descriptive, and/or causal phases of
the research
• Specify the measurement and scaling procedures
• Construct and pretest a questionnaire or some other type of
data collection instrument
• Specify the sampling process and sample size
• Develop a data analysis plan
 Research Design
Exploratory Versus Conclusive Research
Research designs may be broadly classified as exploratory or
conclusive
• Exploratory research is also known as qualitative research
• Conclusive research is the same as quantitative research
Exploratory research is intended to develop an understanding of
the problems motivating the research. It has following characteristics:
– The information needed is only loosely defined at this stage, and the
research process is flexible and relatively unstructured.
– The sample is usually small and not statistically representative.
– The primary data are non-numerical (for example, transcripts of in-
depth interviews or focus group sessions) and are analyzed
qualitatively.
– Exploratory research findings are not conclusive but are often
essential to the construction of a valid conclusive research design.
 Research Design
Exploratory Research
An Exploratory research or study is undertaken when not
much is known about the situation at hand, or no information
is available on how similar problems or research issues have
been solved in the past.
In such cases, extensive preliminary work needs to be done to
gain familiarity with the phenomena in the situation, and
understand what is occurring.
In certain qualitative studies where data is collected through
observation or interviews are exploratory in nature. Based on
that further research work either exploratory or conclusive
are taken
 Research Design
Exploratory Research:

Exploratory research searches are for insights and

understanding about a problem

It is most useful when researchers lack the information

needed to define variables and hypotheses suitable for
statistical research.

Exploratory research methods include:

– Interviews with experts, which are usually with a small number of
individuals and very open-ended.
– Pilot surveys, which are relatively structured interviews with small
samples of respondents conducted prior to the implementation of a large-
scale quantitative investigation.
– Analysis of secondary data.
– Qualitative research, which is usually conducted via focus group
discussions or as in-person/telephone in-depth interviews.
 Research Design
Exploratory Research
Example:
Manager Marketing of a fruit juice making firm is curious to
know the choice preference of its consumers from lower
income groups. Earlier the product was target to upper middle
or higher income groups. There is a little information about this
population segments and there are difference in opinions
about the taste and development process. The manager’s
curiosity can be satisfied only by an exploratory research
study. This can be done through interviewing the people from
the lower income group, studying the economic, education,
and social conditions, upbringing, cultural values and son.
 Research Design
Conclusive Research
Studies that engage in hypothesis testing usually explain the
nature of certain relationships, or establish the differences
among groups or the independence of two or more factors in a
situation.
• The objective of conclusive research is to test hypotheses about
well-defined, quantifiable problems.
• Conclusive research is based on large, representative samples,
and the data obtained are subjected to statistical analysis.
• The findings are conclusive in that their generalizability, validity,
and reliability are sufficiently high enough to help managers
make decisions with a good understanding of risks and benefits.
• Conclusive research designs may be descriptive (providing data
that describes something) or causal (providing data that defines
cause-and-effect relationships).
 Research Design
Conclusive Research
Examples:
 Research Design
Descriptive Research: An Overview
• Descriptive research is a type of conclusive research design. The
function of descriptive research is to describe something, usually
the characteristics or attributes of a specific market.
• Descriptive research begins with the statement of testable
hypotheses. This requirement makes descriptive research less
open-ended and more narrowly focused than exploratory research.
• Large samples are needed to ensure that descriptive research
findings are representative.
• Most marketing research studies involve descriptive research.
Typical data sources include:
 Surveys -- Survey methods include telephone interviewing, in-person interviewing,
mail surveys, and electronic interviewing (e-mail, Internet).
 Observation -- Observational methods include personal observations, mechanical
observations (e.g., television "people meters"), and audits.
 Research Design
Descriptive Research: An Overview
 Research Design
Descriptive Research: Cross-Sectional Designs
• Cross-sectional studies are a common form of descriptive
marketing research. Cross-sectional research designs allow
the researchers to collect information from a given sample
one time, providing information about a specific "cross-
section" or moment of this population's history.
• Single cross-sectional designs use only one set of respondents
from the target population.
• Multiple cross-sectional designs take two or more samples of
respondents, and information from each sample is obtained
at different points in time (e.g., pre/post studies and tracking
studies use multiple cross-sectional designs).
 Research Design
Descriptive Research: Cross-Sectional Designs;
Example
 Research Design
Descriptive Research: Longitudinal Designs
• Longitudinal designs measure a sample repeatedly. Unlike a
cross-sectional design, the researcher collects data from the
same respondents over an extended period. While a cross-
sectional design captures the variables of interest at a single
moment in time, a longitudinal study provides a series of
snapshots that allows the researcher to understand how
situations and attitudes evolve over time.
• The term "panel" is often used synonymously with
"longitudinal design." A panel is a sample of respondents,
generally households, who agree to provide information at
specified intervals. Most panels are maintained by syndicated
firms, and respondents receive gifts, coupons, information, or
cash as compensation for their involvement in the research
project.
 Research Design
Descriptive Research: Longitudinal Designs:
Example:
 Research Design
Type of Investigation:
Causal:
• The study in which the researcher wants to delineate the
cause of one or more problems is called a causal study.
• The goal of causal research is to demonstrate cause-and-
effect relationships.
• Marketing managers constantly rely on assumptions
about causal relationships in their decision-making.
• If these assumptions are unwarranted, the results can be
costly.
For example, the common belief that lowering prices will
increase sales and market share does not apply to every
situation and could lead to losses of revenue if acted on
without justification.
 Research Design
Type of Investigation:
Causal
• Like descriptive research, causal research is structured by
hypothesis testing. While descriptive research assesses the
degree of association between variables, through causal
research it is possible to generalize about causes and effects.
It requires a design in which the causal or independent
variables can be controlled and systematically manipulated.
The researcher measures the effects of this manipulation on
one or more dependent variables and can then make
inferences about causality.
• Causal research designs are typically experimental and reflect
the extension of the classic scientific method to marketing
problems.
 Research Design
Type of Investigation:
Correlational:
• When researchers is interested in delineating the important
variables associated with the problem, the study is called
correlational study.
• It is mere establishment of relationship of one or more variables
with the problems.
Example:
A causal study question:
Does smoking cause cancer?
A correlational study question:
Are smoking and cancer related?
OR
Are smoking, and chewing gutka or meethi chalia associated with
cancer? If so which of these contributes most to variance in the
dependent variables?
 Research Design
Type of Investigation:
Example:
Fears of earthquake in Kashmir and adjacent areas were
instrumental (i.e. causal) in an unprecedented number of
house-owners in that region taking out an insurance policy.

Strict application of property taxes, recession, land grabbing

and the predicted law and order situation considerable
slowed down the business of real estate in Karachi.
 Research Design
Reserchers Interference in Causal and Correlational Studies:
The extent of interference by the researcher with the normal/
natural situation has a direct bearing on whether the study
undertaken is causal or correlational.
A correlational study is conducted in the natural environment
with minimum interference by the researcher.
In cause-and-effect studies the researcher tries to manipulate
the certain variables so as to study the effect of such
manipulation on the dependent variable of interest.
MINIMAL INTERFERENCE
MODERATE INTERFERENCE
EXCESSIVE INTERFERENCE
 Research Design
Reserchers Interference in Causal and Correlational Studies:
Minimal Interference:
 Research Design
Reserchers Interference in Causal and Correlational Studies:
Moderate Interference:
 Research Design
Reserchers Interference in Causal and Correlational Studies:
Excessive Interference:
 Research Design
Primary Versus Secondary Data
• Primary data are generated by a researcher in the process of
studying a specific problem. Collecting primary data can be
expensive and slow.
• Secondary data have been collected for purposes other than
the researcher's immediate need. These data can be located
quickly and usually inexpensively. The disadvantages are that
the usefulness of secondary data may be limited for the
problem at hand. The applicability, accuracy, reliability, and
validity of the secondary data must be critically assessed.
• Sources of secondary data in Pakistan??????
 Research Design
Sources Of Primary Data
• Primary data may be qualitative or quantitative. The
distinction between qualitative and quantitative data parallels
the distinction between exploratory and conclusive research
methods discussed earlier.
• Qualitative data are usually verbal and are often collected
through focus groups, interviews (sometimes called in-depth
interviews or IDIs), or projective techniques.
• Quantitative primary data are gathered through surveys,
observation, or experimentation.
 Research Design
Selecting An Appropriate Research Design
• Exploratory, descriptive, and causal research are convenient as major
classifications of research designs, but the boundaries between these
classifications are not hard and fast. A research project can require
several types of research design, depending on the problem being
studied.
The following guidelines can help:
– If the problem is vague or not well understood, exploratory research is the
best first step.
– Exploratory research should usually be followed by descriptive or causal
research.
– Some research projects do not require initial exploratory research, depending
on how well the problem can be specified. Research on a problem that is well-
understood and clearly defined might begin with descriptive or causal
research.
– Exploratory research can also follow descriptive or causal research. For
example, if the findings of descriptive or causal research are ambiguous,
exploratory research may provide insights to help clarify their meaning
 Research Design
Potential Sources Of Error
• Research designs are vulnerable to several potential sources of error.
These include errors of omission and errors of commission.
 Errors of omission include sampling coverage and non-response.
 Errors of commission are response errors (due to the researcher,
interviewer, or respondent).
• Since any of these can prevent research findings from providing a sound
basis for decisions, it's important in the design process to anticipate and
try to control as many sources of error as possible.
• The total error is the variation or gap between the researcher's
interpretation of data and the actual conditions the data represent. More
technically, in a quantitative study, total error is the difference between
the actual mean value of the variable of interest in the population and the
estimated mean value reported in the marketing research project.
 Research Design
Potential Sources Of Error

The total error includes both random sampling error and non-
sampling error.
• Random sampling error happens when the chosen sample
selected is not representative of the population being
studied.
• Non-sampling errors can be random or nonrandom. Possible
causes include poor problem definition, an inappropriate
approach to the problem, errors in the design of
measurement scales, questionnaires, and interviews, or
errors in preparing and analyzing data. Non-response errors
and response errors are also forms of non-sampling error.
 Research Design
Potential Sources Of Error
• Whenever some of the respondents in the sample do not
respond, the study is subject to some measure of non-
response error. Non-response in consumer research usually
occurs when respondents refuse to participate or are not at
home Non-response will make the resulting sample different
from the intended sample (as defined by demographics such
as age, gender, income, etc.). If non-respondents differ from
the respondents in terms of their potential responses, the
data are biased. If large enough, non-response error can
lower validity to such an extent that conclusions based on the
data are likely to be wrong.
• If respondents give inaccurate answers, if interviewers
erroneously record answers, or if researchers analyze the
answers improperly, response errors result.
 Research Design
Reliability and Validity in Marketing Research
RELIABILITY
• Reliability is the degree to which a procedure produces the
same results if repeated. Measurements are reliable to the
extent they are free from random error (e.g., measurement
error which occurs from random differences between
respondents or measurement situations).
• Techniques for evaluating reliability include the test-retest
and alternative forms methods.
 Research Design
Reliability and Validity in Marketing Research
VALIDITY
• A measurement instrument is said to be valid when
differences in measurements reflect true differences among
the objects being measured, rather than random error or
systematic error (e.g., error that affects the observed score in
the same way each time the measurement is made).
• In research design, the most important consideration is often
construct validity, which addresses the question of what
construct or characteristic the study is, in fact, measuring. In
other words, are the variables as defined in the study actually
useful measures of meaningful characteristics of the
population?
 Research Design
Relationship Between Reliability and Validity
• A measure must be reliable to be valid, but just because it is
reliable does not automatically make it valid. Consider this
example: You step on your bathroom scale three times in
succession and you receive three different measures of your
weight. The scale is neither reliable nor valid—it is not
accurately measuring your weight. But if each time you
weighed yourself it read 10 pounds, then the scale is reliable
but not valid (unless you truly weighed 10 pounds).
 Research Design
Cross-Cultural And International Marketing Research
• Due to cultural and other environmental differences a given
research design may not be appropriate in every country,
even when the same problem is being investigated.
• For example, different methods are required to determine
household attitudes toward major appliances in the United
States and Saudi Arabia. In the United States, exploratory
research could involve focus groups that included both male
and female heads of households. However, given the
traditional culture of Saudi Arabia, mixed-sex focus groups
could produce response errors because women might not
express themselves freely in the presence of men. A better
design for this situation would use one-on-one interviews
with both male and female heads of households.