Research Methodology

Unit 1

1. Introduction to Research Methodology

1.1 What is Research?

The term research is derived from the French word recherche, meaning "to seek out" or "to
investigate thoroughly." Research is a systematic, scientific, and objective process of inquiry
aimed at discovering, interpreting, and revising facts, events, behaviors, or theories.

Definitions:

Source Definition

"Research comprises defining and redefining problems, formulating hypothesis or

suggested solutions; collecting, organizing and evaluating data; making
Clifford Woody
deductions and reaching conclusions; and at last carefully testing the conclusions
to determine whether they fit the formulating hypothesis."

"Research is an honest, exhaustive, intelligent searching for facts and their

P.M. Cook
meanings or implications with reference to a given problem."

"Research is the manipulation of things, concepts or symbols for the purpose of

D. Slesinger and
generalizing to extend, correct or verify knowledge, whether that knowledge aids
M. Stephenson
in the construction of theory or in the practice of an art."

Key Characteristics of Research:

• Systematic: Follows a structured process and logical sequence.

• Objective: Based on facts and evidence, not personal biases.

• Empirical: Relies on direct observation or experimentation.

• Replicable: Can be repeated by other researchers to verify findings.

• Cumulative: Builds upon existing knowledge and contributes to theory.

1.2 Research Methodology vs. Research Methods

Term Description

Research The specific techniques, tools, and procedures used to collect and analyze data
Methods (e.g., surveys, interviews, questionnaires, statistical tests).

The broader philosophy and framework that guides the research. It

Research
explains why a particular method is chosen, the assumptions behind it, and
Methodology
how it fits the research problem. It is the science of how research is conducted.

2. Importance of Research

Research is essential in business and management for informed decision-making, risk

reduction, and competitive advantage.

2.1 For Business and Management

Importance Explanation

Provides data-driven insights rather than intuition or guesswork. Helps

Informed
managers make strategic choices about product launches, market entry, pricing,
Decision-Making
and resource allocation.

Identifies root causes of business problems (e.g., declining sales, employee

Problem Solving
turnover) and suggests effective solutions.

Opportunity Uncovers market gaps, emerging trends, and unmet customer needs, enabling
Identification innovation and growth.

Minimizes uncertainty by providing empirical evidence before committing

Risk Reduction
significant resources.

Performance Assesses the effectiveness of marketing campaigns, operational processes, and

Evaluation employee performance.

Competitive Helps organizations understand competitors, customer preferences, and

Advantage industry dynamics to stay ahead.
2.2 For Society and Academia

Importance Explanation

Knowledge
Contributes to the body of knowledge in business disciplines.
Creation

Policy Provides evidence for government and regulatory policies affecting business
Formulation and the economy.

Research on consumer behavior, ethics, and sustainability promotes

Social Welfare
responsible business practices.

3. Process of a Research

The research process is a systematic sequence of steps that guide the researcher from problem
identification to final reporting. The process is cyclical and iterative.

Step Stage Description

Identify the Select a broad area of interest and narrow it down to a specific,
1
Research Problem feasible, and significant problem. Formulate research questions.

Conduct an extensive review of existing academic journals, books,

2 Review Literature reports, and theories to understand what is already known, identify
gaps, and develop a theoretical framework.

Develop testable statements (hypotheses) that predict the relationship

Formulate
3 between variables. (e.g., "There is a positive relationship between
Hypotheses
advertising spend and sales.")

Create a blueprint for the research. Decide on the type of research

Prepare Research
4 (exploratory, descriptive, causal), sampling method, data collection
Design
tools, and analysis techniques.
Step Stage Description

Design Data
Develop questionnaires, interview guides, observation checklists, or
5 Collection
experimental setups. Ensure validity and reliability.
Instrument

Define the target population and select a representative sample.

6 Sampling Determine sample size and sampling technique (probability or non-
probability).

Execute the data collection plan. Administer surveys, conduct

7 Collect Data
interviews, record observations, or gather secondary data.

Process and Clean, code, and organize data. Apply statistical or qualitative
8
Analyze Data analysis techniques to test hypotheses and derive insights.

Draw conclusions from the analysis. Relate findings back to the

9 Interpret Findings research questions, hypotheses, and literature review. Discuss
implications, limitations, and recommendations.

Document the entire research process and findings in a structured

Prepare Research
10 format (introduction, literature review, methodology, results,
Report
discussion, conclusion).

Part B: Research Design

4. Research Design: Introduction

4.1 What is Research Design?

Research design is the overall plan or blueprint that outlines how the research will be
conducted. It provides structure and direction, ensuring that the research objectives are
achieved efficiently, accurately, and economically.

Definition:
"A research design is the arrangement of conditions for collection and analysis of data in a
manner that aims to combine relevance to the research purpose with economy in procedure."
— Jahoda, Deutsch & Cook

"Research design is the conceptual structure within which research is conducted; it

constitutes the blueprint for the collection, measurement, and analysis of data." — Selltiz,
Wrightsman & Cook

4.2 Key Components of Research Design

Component Description

Purpose The research objective (exploratory, descriptive, or causal).

Population & Sampling Who will be studied and how they will be selected.

How data will be gathered (surveys, experiments, secondary

Data Collection Methods
data).

Measurement &
Tools used to measure variables (questionnaires, scales).
Instrumentation

Data Analysis Plan How data will be processed and analyzed.

Time Horizon Cross-sectional (one point in time) or longitudinal (over time).

Budget & Resources Constraints and available resources.

5. Formulating the Research Problem

The research problem is the foundation of any study. A well-defined problem ensures that the
research stays focused and relevant.

5.1 Steps in Formulating a Research Problem

Step Description

1. Identify a Broad Start with a general field (e.g., "consumer behavior," "employee
Area of Interest motivation," "supply chain efficiency").

Examine existing research to understand what is already known and

2. Review Literature
identify gaps or unresolved issues.

3. Narrow Down the Focus on a specific aspect, context, or population (e.g., "Impact of social
Topic media advertising on purchase intention among Gen Z in urban India").

4. Define the Problem State the problem in a clear, concise, and unambiguous manner. Use a
Clearly declarative or interrogative format.

5. Establish Research Formulate specific questions that the research will answer (e.g., "What is
Questions the relationship between influencer credibility and purchase intention?").

6. Formulate Develop testable statements derived from the research questions (if
Hypotheses quantitative research).

Ensure the problem can be investigated within available time, resources,

7. Assess Feasibility
and access to data.

5.2 Sources of Research Problems

• Personal experience (workplace observations, managerial challenges)

• Literature review (gaps, contradictions, unexplored areas)

• Theories (testing or extending existing theories)

• Social and business trends (digital transformation, sustainability)

• Consulting assignments (organizational problems)

5.3 Criteria for a Good Research Problem (FINER Criteria)

Criteria Description

Feasible Adequate resources, time, expertise, and access to subjects.

Interesting Holds the researcher's interest and is motivating.

Novel Contributes new knowledge or addresses a gap.

Ethical Does not harm participants; respects confidentiality and dignity.

Relevant Has significance for academia, business practice, or society.

6. Choice of Research Design

The choice of research design depends on the research purpose, the nature of the problem,
and the stage of knowledge about the topic.

6.1 Factors Influencing Choice of Design

Factor Consideration

Research Exploratory (to discover) → Qualitative design; Descriptive (to describe) →

Objectives Survey design; Causal (to explain cause-effect) → Experimental design.

Nature of the Vague/unexplored → Exploratory; Well-defined/structured → Descriptive;

Problem Hypothesis testing → Causal.

Availability of Secondary data available → Descriptive; Primary data needed →

Data Surveys/experiments.

Time and Limited time → Cross-sectional; Abundant resources →

Resources Longitudinal/experimental.

Degree of High control over variables → Experimental; Low control → Descriptive or ex

Control post facto.
7. Types of Research Design

Research designs can be broadly classified into three major

categories: Exploratory, Descriptive, and Causal (Experimental) .

7.1 Exploratory Research Design

Aspect Description

To explore a problem or phenomenon when little is known. To generate insights,

Purpose
formulate hypotheses, and identify key variables.

When
Early stages of research; when the problem is vague or unstudied.
Used

Methods Literature review, expert interviews, focus groups, case studies, pilot studies.

Outcome Provides direction for future research; does not provide conclusive answers.

Understanding why a new startup failed; exploring consumer attitudes towards a novel
Example
technology.

7.2 Descriptive Research Design

Aspect Description

To describe characteristics of a population, situation, or phenomenon. Answers "what,"

Purpose
"who," "where," and "how" questions.

When
When the research problem is well-defined and structured.
Used

Methods Surveys, cross-sectional studies, observational studies, secondary data analysis.

Provides a detailed snapshot; describes trends, frequencies, and associations (but not
Outcome
causation).
Aspect Description

Market share analysis; customer satisfaction survey; demographic profile of online

Example
shoppers.

Sub-types of Descriptive Design:

• Cross-sectional: Data collected at a single point in time.

• Longitudinal: Data collected from the same subjects over multiple time periods
(trend studies, panel studies).

7.3 Causal (Experimental) Research Design

Aspect Description

To establish cause-and-effect relationships between variables. Answers "why" and

Purpose
tests hypotheses.

When
When the researcher wants to determine if one variable influences another.
Used

Methods Experiments (laboratory, field), quasi-experiments.

Provides evidence of causality (independent variable causes change in dependent

Outcome
variable).

Testing whether a price discount (IV) increases sales (DV); evaluating the impact of an
Example
advertisement on brand recall.

Essential Elements of Causal Design:

• Independent Variable (IV): The presumed cause (manipulated by researcher).

• Dependent Variable (DV): The presumed effect (measured).

• Control Group: A group that does not receive the treatment.

• Experimental Group: A group that receives the treatment.

• Random Assignment: Subjects are randomly assigned to groups to ensure

comparability.
8. Sources of Experimental Errors

In experimental research, errors can threaten the validity of the findings. These errors can be
categorized into internal validity threats and external validity threats.

8.1 Internal Validity Threats

Internal validity refers to the extent to which the observed effect (DV) is actually caused by
the treatment (IV) and not by other factors.

Threat Description

External events occurring during the experiment that affect the dependent
History
variable (e.g., a competitor's ad campaign during a test).

Changes within the subjects over time (e.g., fatigue, boredom, growing
Maturation
older) that affect results.

The effect of taking a pre-test influencing scores on the post-test (e.g.,

Testing
participants become familiar with the test).

Changes in the measurement instrument or observer over time (e.g., different

Instrumentation
interviewers, calibration issues).

Statistical
Tendency for extreme scores to move closer to the mean upon retesting.
Regression

Non-random assignment of subjects to groups, leading to pre-existing

Selection Bias
differences.

Mortality
Dropout of subjects from the experiment, making groups incomparable.
(Attrition)

Diffusion of Communication between experimental and control groups (e.g., control

Treatment group learns about the treatment).

8.2 External Validity Threats

External validity refers to the extent to which the findings can be generalized to other
populations, settings, and times.

Threat Description

The sample may not represent the target population (e.g., using students
Population Validity
for a study on all adults).

The experimental setting may not represent real-world conditions

Ecological Validity
(artificial lab setting vs. actual shopping environment).

Reactive Effects Subjects change their behavior because they know they are being
(Hawthorne Effect) observed.

Pretest-Treatment The effect of the treatment may only occur if a pre-test is given, limiting
Interaction generalizability to populations that have not been pre-tested.

Time Interaction Results may only hold true for the specific time period of the study.

8.3 Ways to Control Experimental Errors

Method Description

Random assignment of subjects to groups to control for selection bias and

Randomization
extraneous variables.

Matching Pairing subjects with similar characteristics across groups.

Using a group that does not receive the treatment to account for history and
Control Group
maturation.

Single-blind (subjects unaware of group assignment) or double-blind (both

Blinding
subjects and researchers unaware) to reduce bias.

Keeping all conditions (instructions, environment, measurement) consistent

Standardization
across groups.
Method Description

Statistical
Using techniques like ANCOVA to adjust for initial differences.
Controls

Unit 2
Module 1: Sampling and Sampling Design

1.1 Some Basic Terms

Term Definition Example

The entire group of individuals, objects, or

Population
events that possess the characteristics the All smartphone users in India.
(Universe)
researcher wants to study.

Element A single member of the population. One smartphone user in India.

A complete enumeration of every element Surveying every smartphone user

Census
in the population. in India.

A subset of the population selected to 5,000 smartphone users selected

Sample
represent the entire population. from different cities.

A list of all elements in the population from A database of mobile numbers, a

Sampling
which the sample is drawn. It should be telephone directory, or a voter
Frame
representative and complete. list.

An individual smartphone user. In

The basic unit that is selected at each stage
Sampling multi-stage sampling, it could be
of sampling. In simple random sampling, it
Unit a city first, then a locality, then a
is the same as the element.
household.

Sampling The error that arises due to the fact that a

The average satisfaction score
Error sample, not the entire population, is
from the sample (7.5) differs
studied. It is the difference between the
Term Definition Example

sample statistic and the true population from the true average of all
parameter. smartphone users (7.2).

Errors that occur for reasons other than

Non- Respondents misinterpreting a
sampling, such as poor question design,
Sampling question, or some people refusing
data entry mistakes, non-response bias, or
Error to participate.
interviewer bias.

1.2 Advantages and Limitations of Sampling

Advantages Limitations

Sampling Error: There is always a margin of

Cost-Effective: Studying a sample is significantly
error because only a subset is studied. Results
cheaper than conducting a census.
are estimates, not exact values.

Sampling Frame Issues: If the sampling

Time-Saving: Data collection and analysis are much frame is inaccurate or incomplete, the sample
faster, allowing for timely decision-making. may not represent the population, leading to
biased results.

Design Complexity: Designing a truly

Greater Accuracy: Smaller scale allows for more
representative sample requires expertise. A
intensive and accurate data collection (e.g., better-
poorly designed sample can lead to invalid
trained interviewers, thorough supervision).
conclusions.

Accessibility: When a population is infinite (e.g.,

Non-Sampling Errors: Issues like non-
potential customers for a new product) or destructive
response or measurement errors can still affect
testing is required (e.g., testing the lifespan of light
the sample and may be difficult to correct.
bulbs), sampling is the only practical option.
Advantages Limitations

Generalization Risk: If the sample is not

Feasibility: For large or geographically dispersed
truly representative, generalizing findings to
populations, a census is often logistically impossible.
the population can be misleading.

1.3 Sampling Process

The sampling process is a systematic, step-by-step approach to ensure a representative and

unbiased sample.

1. Define the Population: Clearly specify the target population in terms of elements,
sampling units, extent (geographical), and time.

o Example: "All undergraduate students enrolled in BBA programs in Pune city

during the academic year 2025-26."

2. Identify the Sampling Frame: Obtain or construct a list of all elements in the
defined population.

o Example: A list of all BBA students with their enrollment numbers from the
affiliated universities.

3. Select a Sampling Technique (Design): Decide whether to use probability (random)

or non-probability (non-random) sampling. This is a critical decision affecting the
generalizability of results. (See Section 1.4 & 1.5 for details).

4. Determine the Sample Size: Decide how many elements need to be sampled. This
depends on factors like population size, desired precision, confidence level, and
budget. (See Section 1.7 for details).

5. Execute the Sampling Process: Actually select the sample elements according to the
chosen technique and collect data from them. This involves operational planning,
training field workers, and managing the data collection process.

6. Validate the Sample: Check if the final sample truly represents the population.
Compare sample demographics (e.g., age, gender distribution) with known population
parameters. If significant discrepancies exist, weighting or adjustments may be
needed.

1.4 Types of Sampling (Broad Classification)

Sampling techniques are broadly classified into two categories:

Probability Sampling Non-Probability Sampling

Every element in the population has a The probability of selection is unknown. Selection is
known, non-zero chance of being selected. based on convenience or judgment.

Allows for statistical inference and Does not allow for statistical generalization to the
calculation of sampling error. population.

More rigorous, objective, and representative. Less rigorous, subjective, and prone to bias.

Commonly used in exploratory research, pilot

Preferred for conclusive research and when
studies, and when probability sampling is not
generalizability is critical.
feasible.

1.5 Types of Sample Designs

A. Probability Sampling Designs

Design Description Example

Every element in the sampling frame has an

Simple Assigning numbers to all 10,000
equal and independent chance of being
Random customers and using a random
selected. Selection is done using a random
Sampling number generator to select 500.
number generator or lottery method.

From a list of 10,000 customers,

Selecting every kth element from the
Systematic randomly choose a start between 1
sampling frame after a random start. *k* =
Sampling and 20, then select every 20th
Population size / Sample size.
customer.

The population is divided into mutually Divide customers into income

exclusive subgroups (strata) based on a key groups (low, medium, high).
Stratified
characteristic (e.g., age, income). Then, Randomly select 100 from each
Sampling
random samples are drawn from each group in proportion to their
stratum, often proportionally. representation in the population.
Design Description Example

To survey university students

The population is divided into clusters
across India, randomly select 20
(naturally occurring groups, e.g., cities,
universities (clusters) and then
Cluster schools). A random sample of clusters is
survey all students within those
Sampling selected, and then either all elements within
universities (one-stage) or a
those clusters (one-stage) or a random sample
random sample from each (two-
from within them (two-stage) are surveyed.
stage).

B. Non-Probability Sampling Designs

Design Description Example

Selecting elements that are easiest to reach or A researcher standing outside a

Convenience
most convenient. The most common but least mall and interviewing shoppers
Sampling
rigorous method. who pass by.

Selecting only "frequent flyers"

Judgmental Selecting elements based on the researcher's
for a survey on airline service
(Purposive) judgment about who will be most appropriate
quality because they have the
Sampling or representative.
most relevant experience.

The researcher ensures that certain Setting quotas to interview 50

characteristics (e.g., age, gender) are men and 50 women. The
Quota Sampling represented in the sample in proportion to interviewer can choose any 50
their prevalence in the population, but men and 50 women they
selection within quotas is non-random. encounter.

Studying the habits of "digital

Initial respondents are selected, and then they
nomads." The researcher
Snowball refer other potential respondents who meet
interviews one, who then
Sampling the criteria. Useful for hard-to-reach
recommends others in their
populations.
network.

1.6 Testing of Hypothesis

• Definition: Hypothesis testing is a statistical procedure used to determine whether
there is enough evidence from a sample to infer that a certain condition or proposition
is true for the entire population.

• Key Concepts:

o Null Hypothesis (H₀): A statement of no effect, no difference, or no

relationship. It is the hypothesis that the researcher tries to disprove or reject.

▪ Example: "The average customer satisfaction score is 7.0 out of 10."

(H₀: μ = 7.0)

o Alternative Hypothesis (H₁ or Hₐ): A statement that contradicts the null

hypothesis. It is what the researcher wants to prove.

▪ Example: "The average customer satisfaction score is not 7.0." (H₁: μ ≠

7.0) — two-tailed test

▪ Or, "The average customer satisfaction score is greater than 7.0." (H₁:
μ > 7.0) — one-tailed test

o Level of Significance (α): The probability of rejecting the null hypothesis

when it is actually true (Type I error). Common values are 0.05 (5%) or 0.01
(1%).

o Type I Error (α): Rejecting a true null hypothesis (false positive).

o Type II Error (β): Failing to reject a false null hypothesis (false negative).

o p-value: The probability of obtaining the observed sample results (or more
extreme) if the null hypothesis is true. If p-value < α, we reject H₀.

• The Process:

1. State H₀ and H₁.

2. Choose the level of significance (α).

3. Select the appropriate test statistic (e.g., z-test, t-test, chi-square).

4. Calculate the test statistic from the sample data.

5. Compare the calculated value to the critical value (from statistical tables) or
compare the p-value to α.

6. Make a decision: Reject H₀ or fail to reject H₀.

7. Draw a conclusion in the context of the problem.

1.7 Determining the Sample Size

Sample size is a critical decision. A sample too small may not be representative; a sample too
large wastes resources. Key factors influencing sample size:

1. Population Size: For very small populations, a larger proportion is needed. For large
populations, the required sample size plateaus.

2. Desired Precision (Margin of Error - e): The maximum acceptable difference

between the sample estimate and the true population value. Smaller margin of error
requires a larger sample.

3. Confidence Level: The probability that the true population parameter lies within the
confidence interval. Common levels are 90%, 95%, and 99%. Higher confidence
requires a larger sample.

4. Variability (Standard Deviation - σ): The more heterogeneous the population, the
larger the sample needed. If variability is unknown, use p = 0.5 (50%) for proportions,
which maximizes sample size.

• Formula for Sample Size (for estimating a mean):

n=Z2⋅σ2e2n=e2Z2⋅σ2

Where:

o nn = sample size

o ZZ = Z-value corresponding to the confidence level (e.g., 1.96 for 95%)

o σσ = estimated population standard deviation

o ee = desired margin of error

• Formula for Sample Size (for estimating a proportion):

n=Z2⋅p⋅(1−p)e2n=e2Z2⋅p⋅(1−p)

Where:

o pp = estimated proportion of the population with the characteristic

o (1−p)(1−p) = estimated proportion without the characteristic

1.8 Sampling Distribution of the Mean

• Definition: The sampling distribution of the mean is the theoretical distribution of the
means of all possible random samples of a given size (nn) drawn from a population.
 • Central Limit Theorem (CLT): This is a fundamental theorem that states:

1. The mean of the sampling distribution (μxˉμxˉ) is equal to the population

mean (μμ).

2. The standard deviation of the sampling distribution, called the Standard

Error of the Mean (σxˉσxˉ), is equal to the population standard deviation
divided by the square root of the sample size: σxˉ=σnσxˉ=nσ.

3. Most Importantly: Regardless of the shape of the original population

distribution, the sampling distribution of the mean will approximate a normal
distribution as the sample size (nn) becomes sufficiently large
(typically n≥30n≥30).

• Importance for Research: The CLT is the foundation for hypothesis testing and
confidence intervals. It allows researchers to make inferences about a population
mean using the normal distribution (z-test) or t-distribution, even when the population
distribution is unknown, as long as the sample size is adequate.

Module 2: Scaling Techniques

2.1 The Concept of Attitude

• Definition: An attitude is a learned, enduring predisposition to respond consistently in

a favorable or unfavorable manner toward a given object, person, or idea.

• Components of Attitude (ABC Model):

o Affective Component: The emotional or feeling aspect (e.g., "I love this
brand").

o Behavioral Component: The tendency to act in a certain way (e.g., "I always
buy this brand").

o Cognitive Component: The beliefs or knowledge about the object (e.g., "This
brand is known for quality").

2.2 Difficulty of Attitude Measurement

Attitudes are abstract, latent constructs that cannot be directly observed. Measuring them
poses several challenges:

1. Intangibility: Attitudes are internal mental states. Researchers must infer them from
responses to questions or observed behavior.
 2. Social Desirability Bias: Respondents may answer in a way that they believe is
socially acceptable, rather than expressing their true attitude.

3. Lack of Self-Awareness: Individuals may not be fully aware of their own attitudes or
may hold contradictory attitudes (ambivalence).

4. Instability: Attitudes can change over time, influenced by new information or

experiences.

5. Cultural and Language Nuances: The meaning of words and the expression of
attitudes can vary across cultures, requiring careful adaptation of scales.

To overcome these difficulties, researchers use scaling techniques—systematic methods to

assign numbers or symbols to attitudes to capture their intensity and direction.

2.3 Types of Scales

Scales are classified based on the level of measurement they provide. Understanding these
levels is crucial for choosing appropriate statistical tests.

Scale Description Properties Example Statistics

Numbers are
Gender: 1=Male,
used as
2=Female. Brand Frequency,
labels or Classification,
Nominal purchased: 1=Apple, mode, chi-
categories. counting
2=Samsung, square
No order or
3=OnePlus.
magnitude.

Rank your preference

for brands: 1st, 2nd,
3rd. Satisfaction:
Numbers
1=Very Dissatisfied,
indicate rank
2=Dissatisfied, Median,
order, but the
Classification 3=Neutral, percentile,
Ordinal intervals
+ Order 4=Satisfied, 5=Very rank-order
between
Satisfied correlation
ranks are not
(the order matters, but
equal.
the gap between 1 and
2 may not equal the
gap between 4 and 5).
Scale Description Properties Example Statistics

Numeric
Mean,
scales with Temperature in
standard
equal Classification Celsius (0°C does not
deviation, t-
intervals + Order + mean absence of
Interval test,
between Equal temperature). Likert
ANOVA,
points, but Intervals Scale (when treated as
Pearson
no true zero interval data).
correlation
point.

Has all All statistical

properties of Classification techniques,
Sales revenue (₹0
interval + Order + including
means no sales). Age
Ratio scales, plus a Equal geometric
(0 means no age).
meaningful Intervals + mean and
Number of purchases.
true zero True Zero coefficient of
point. variation

2.4 Common Attitude Scaling Techniques

Technique Description Example

Statement: "This smartphone has

Respondents indicate their level of
excellent battery life." Scale:
agreement or disagreement with a series
1=Strongly Disagree, 2=Disagree,
Likert Scale of statements related to the attitude
3=Neutral, 4=Agree, 5=Strongly
object. Typically a 5-point or 7-point
Agree. Scores are summed or
scale.
averaged across statements.

Respondents rate a concept on a set of

Semantic Brand X: Modern _ _ _ _ _ _ _
bipolar adjective pairs (e.g., good-bad,
Differential Traditional; Reliable _ _ _ _ _ _ _
modern-traditional). Often uses a 5-
Scale Unreliable.
point or 7-point scale between the pairs.
Technique Description Example

A unipolar scale (usually 10 points) that

measures the direction and intensity of
Customer Service of Company
an attitude. It uses a single adjective and
Stapel Scale Y: +5 +4 +3 +2 +1 ACCURATE -1 -
asks respondents to rate how accurately
2 -3 -4 -5.
it describes the object, typically from -5
to +5.

A set of statements that are

progressively more extreme. Agreement Would you... 1) Consider buying an
Guttman Scale
with a stronger statement implies electric vehicle? 2) Test drive an
(Cumulative
agreement with all weaker ones. Rarely electric vehicle? 3) Pay a premium
Scale)
used in commercial research due to for an electric vehicle?
complexity.

"Distribute 100 points to show the

Respondents allocate a fixed number of
importance of these features when
Constant Sum points (e.g., 100) among multiple
buying a laptop: Battery Life ___;
Scale attributes to indicate their relative
Processor Speed ___; Screen Quality
importance.
___; Portability ___."

2.5 Criteria for a Good Test (Scale)

A good measurement scale must be reliable, valid, and practical.

1. Reliability: The consistency or stability of the measurement. A scale is reliable if it

produces the same results under consistent conditions.

o Test-Retest Reliability: Consistency over time.

o Internal Consistency: Consistency across items within the scale (measured

by Cronbach's Alpha; a value above 0.70 is generally considered acceptable).

2. Validity: The extent to which a scale measures what it is intended to measure.

o Content Validity: Does the scale adequately cover all aspects of the
construct? (Judgmental, based on expert opinion).

o Criterion Validity: Does the scale correlate with an external, established

measure (criterion)?
 ▪ Concurrent Validity: Correlates with a criterion measured at the same
time.

▪ Predictive Validity: Predicts a future outcome.

o Construct Validity: The degree to which the scale measures the theoretical
construct it claims to measure. This is the most important and comprehensive
form of validity, often assessed through convergent validity (correlates with
measures of related constructs) and discriminant validity (does not correlate
with measures of unrelated constructs).

3. Practicality: The scale should be easy to administer, score, and interpret, and should
be cost-effective within the research budget.

Unit 3
Part A: Methods of Data Collection

Data collection is the systematic process of gathering information from relevant sources to
answer research questions, test hypotheses, and evaluate outcomes. The quality of any
research study is directly dependent on the quality of the data collected.

Data can be broadly classified into two categories: Primary Data and Secondary Data.

1. Secondary Data

Definition: Secondary data refers to data that has already been collected, processed, and
published by someone else for some other purpose, but is being utilized by the researcher for
a new study.

Characteristics:

• It is not original to the current researcher.

• It is readily available and often less time-consuming to obtain.

• It may or may not perfectly fit the current research problem.

Sources of Secondary Data:

Secondary data sources are broadly classified into internal and external sources.
Category Sources Examples

Data generated within the Sales records, customer databases, inventory

Internal
organization for which the research records, previous research reports, financial
Sources
is being conducted. statements, customer feedback logs.

External Data generated outside the

Sources organization.

Published Sources:

Census of India, Economic Survey, RBI

- Government Publications
Bulletins, Ministry of Commerce reports.

World Bank, IMF, UNO, WTO reports and

- International Publications
databases.

FICCI, CII, ASSOCHAM reports; industry-

- Trade/Professional Bodies
specific journals.

Market research reports from Nielsen,

- Commercial Sources Kantar, Statista; business magazines
(Economic Times, Business Today).

Journal of Marketing, Harvard Business

- Academic Journals
Review, Indian Journal of Marketing.

Unpublished Sources:

Theses, dissertations, research papers

presented at conferences, internal records of
other organizations (accessed with
permission).

Advantages of Secondary Data:

• Cost-Effective: Less expensive than collecting primary data.

 • Time-Saving: Readily available, reducing data collection time.

• Broad Scope: Can provide data from a wider geographic area or longer time period.

• Comparative Analysis: Allows comparison with past data or industry benchmarks.

• Often Reliable: Data from reputed sources (e.g., government) may be highly reliable.

Disadvantages of Secondary Data:

• Lack of Relevance: Data may not precisely match the research objectives, questions,
or population.

• Outdated Information: Data may be obsolete, especially in fast-changing industries.

• Questionable Accuracy: Sources may be biased or contain errors.

• Lack of Control: The researcher has no control over how the data was collected or
processed.

• Unit of Measurement Issues: Data may be presented in different units (e.g., sales in
units vs. value) that are not directly comparable.

2. Primary Data

Definition: Primary data refers to data that is collected first-hand by the researcher
specifically for the purpose of the current research study. It is original, fresh, and directly
addresses the research problem.

Advantages:

• Relevance: Tailored to meet the specific objectives of the study.

• Accuracy: The researcher has control over the data collection process.

• Current: Data is up-to-date and reflects the present situation.

• Confidentiality: The researcher can maintain the privacy of sensitive information.

Disadvantages:

• Expensive: Requires significant financial resources.

• Time-Consuming: Planning, collecting, and processing primary data takes

considerable time.

• Effort-Intensive: Requires skilled personnel and careful administration.

3. Collection of Primary Data: Methods

There are three primary methods for collecting primary data: Observation, Questionnaire,
and Interview. A fourth common method is Experimentation, but the focus here is on the
first three as per the topic.

A. Observation Method

Definition: Observation is a method of data collection in which the researcher systematically

watches, records, and analyzes the behavior, actions, or phenomena of subjects in their
natural setting without directly interacting with them.

Types of Observation:

Type Description Example

Structured: Predetermined checklist; used Structured: Counting number of

Structured for descriptive customers entering a
vs. research. Unstructured: No store. Unstructured: Observing
Unstructured predetermined plan; used for exploratory general customer behavior in a
research. shopping mall.

Participant: Researcher works as a

Participant: Researcher becomes part of store employee to study customer
Participant
the group being observed. Non- interactions. Non-
vs. Non-
Participant: Researcher remains detached Participant: Researcher stands in a
Participant
and observes from outside. corner of a store and observes
shopping patterns.

Controlled: Observation in a controlled Controlled: Observing consumer

Controlled vs. environment (e.g., reactions in a simulated
Uncontrolled laboratory). Uncontrolled: Observation in store. Uncontrolled: Observing
a natural, real-life setting. consumers in an actual supermarket.

Disguised: Using a one-way

Disguised: Subjects are unaware they are
Disguised vs. mirror. Undisguised: A researcher
being observed. Undisguised: Subjects
Undisguised with a clipboard clearly observing
know they are being observed.
shoppers.

Advantages:

• Collects data on actual behavior, not self-reported intentions.

 • No reliance on respondent memory or willingness to answer.

• Captures real-time, natural context.

Disadvantages:

• Cannot observe attitudes, opinions, or motivations.

• Time-consuming and may be expensive.

• Observer bias may influence what is recorded.

• Ethical concerns regarding privacy (especially in disguised observation).

B. Questionnaire Method

Definition: A questionnaire is a structured research instrument consisting of a set of written

questions designed to collect information from respondents. It is one of the most widely used
methods in survey research.

Designing a Questionnaire:

A well-designed questionnaire is critical for obtaining reliable and valid data. The process
involves several steps:

1. Define Research Objectives: Clearly outline what information is needed and how it
will be used.

2. Determine Question Types:

o Open-Ended Questions: Allow respondents to answer in their own words.

Useful for exploring opinions and getting rich, qualitative data. (e.g., "What
factors influenced your purchase decision?")

o Closed-Ended Questions: Provide a set of predetermined response options.

Easier to analyze and code.

▪ Dichotomous: Two options (e.g., Yes/No, Male/Female).

▪ Multiple Choice: Several options, select one or more.

▪ Likert Scale: Measures intensity of agreement/disagreement (e.g.,

Strongly Agree to Strongly Disagree).

▪ Semantic Differential: Uses bipolar adjectives to rate an object (e.g.,

"Inexpensive — Expensive," "Modern — Traditional").

3. Word Questions Clearly:

 o Use simple, unambiguous language.

o Avoid leading questions (e.g., "Don't you agree that our product is the best?").

o Avoid double-barreled questions (e.g., "How satisfied are you with the
product's quality and price?" — two separate issues).

o Avoid loaded or emotionally charged language.

4. Determine Question Sequence:

o Start with easy, non-threatening questions (screening/filter questions).

o Use funnel approach: broad questions first, then narrow, specific ones.

o Group similar topics together.

o Place sensitive or demographic questions at the end.

5. Provide Clear Instructions: Guide respondents on how to answer (e.g., "Select only
one," "Circle your response").

6. Design Layout: Ensure the questionnaire is visually appealing, uncluttered, and easy
to navigate.

7. Pre-Test (Pilot Testing): Test the questionnaire on a small sample to identify

ambiguous questions, confusing instructions, or technical issues before full-scale
administration.

C. Interview Method

Definition: An interview is a method of data collection involving direct, face-to-face (or

telephone/video) interaction between the interviewer and the respondent, where the
interviewer asks questions and records responses.

Types of Interviews:

Type Description Advantages Disadvantages

Follows a High consistency;

predetermined Lacks flexibility;
Structured easy to analyze;
questionnaire with cannot explore
Interview reduces
fixed wording and unexpected answers.
interviewer bias.
order. The interviewer
Type Description Advantages Disadvantages

reads questions
verbatim.

No fixed questions.
The interviewer Provides deep, Time-consuming;
Unstructured explores topics rich, qualitative difficult to analyze;
Interview broadly, allowing the insights; highly high risk of
conversation to guide flexible. interviewer bias.
the direction.

Uses a loose guide or Balances structure Requires skilled

Semi-
topic list, but allows with depth; interviewer; analysis
Structured
flexibility to probe common in is moderately
Interview
and explore. business research. complex.

Generates diverse
A group interview (6- perspectives; Group dynamics can
Focus Group 12 participants) led by interactive; dominate; not
Interview a moderator to discuss efficient for generalizable to a
a topic. exploratory larger population.
research.

Key Skills for Effective Interviewing:

• Probing: Asking follow-up questions to clarify or expand on a response (e.g., "Can

you tell me more about that?").

• Active Listening: Paying close attention and showing genuine interest.

• Neutrality: Avoiding leading questions and maintaining an unbiased demeanor.

• Rapport Building: Establishing trust and comfort with the respondent.

Part B: Data Processing and Tabulation

Once data is collected, it must be processed and organized to make it meaningful and ready
for analysis. This stage involves editing, coding, and tabulation.
1. Editing

Definition: Editing is the process of reviewing and adjusting raw data to ensure accuracy,
consistency, completeness, and uniformity before coding and analysis. It is the first step in
data processing.

Objectives of Editing:

• To ensure data is accurate and consistent.

• To identify and correct omissions, illegible entries, or errors.

• To standardize data for analysis.

• To make decisions on how to handle questionable responses.

Types of Editing:

• Field Editing: Preliminary editing done in the field by the interviewer or supervisor
shortly after data collection to catch obvious errors or incomplete responses while
memory is fresh.

• Central Editing: In-depth editing conducted at a central office where all

questionnaires are reviewed thoroughly by a team of editors.

Problems in Editing:
Editors often encounter several issues while reviewing data:

Problem Description Possible Action

If the omission is minor, the editor may

attempt to contact the respondent (if possible)
Incomplete Respondent leaves some to complete it. If not, the question may be
Responses questions unanswered. coded as "no response" or the entire
questionnaire may be rejected if critical data is
missing.

Illegible Handwritten responses are The editor may interpret if possible; otherwise,
Entries unclear or cannot be read. the response is treated as missing.

Contradictory answers (e.g., The editor checks for logic errors. If

Inconsistent
a respondent says they are resolvable, corrections are made; if not, the
Responses
"unemployed" but also conflicting data may be discarded or flagged.
Problem Description Possible Action

provides their "annual

salary").

Vague or unclear responses The editor may code the response as

Ambiguous
(e.g., "sometimes," "often") "unspecified" or attempt to interpret based on
Answers
that lack specificity. context.

A response falls outside the

Out-of-Range The response is treated as an error and
permissible range (e.g., age =
Values corrected if possible, or deleted.
150 years).

Patterns like "straight-lining"

Response (selecting the same option for The editor flags such questionnaires for
Bias all questions) or "yea-saying" potential exclusion or special handling.
(tendency to agree).

2. Coding

Definition: Coding is the process of assigning numerical or alphanumeric codes to responses

to facilitate classification, tabulation, and analysis. It converts raw, often qualitative, data into
a format that can be processed by statistical software.

Process of Coding:

• For closed-ended questions, codes are pre-assigned (e.g., 1=Male, 2=Female).

• For open-ended questions, codes are developed after reviewing a sample of responses
to identify common themes or categories (post-coding). A codebook is created to
define all variables and their corresponding codes.

Example:

• Question: "What is your primary reason for choosing this brand?"

• Responses: "Good quality," "Trustworthy," "Friend recommended," "Price."

• Codes: 1=Quality, 2=Trust, 3=Recommendation, 4=Price.

3. Tabulation

Definition: Tabulation is the systematic arrangement of data in rows and columns (tables) to
summarize information and facilitate further statistical analysis. It is the final step before
analysis and interpretation.

Types of Tabulation:

Type Description Example

Summarizes data for a single variable. A table showing the number and
Simple (One-
Provides frequency counts and percentage of respondents by age
Way) Tabulation
percentages. group.

Simultaneously summarizes two or A contingency table showing the

Cross (Two-Way
more variables to examine number of respondents by Age
or Multi-Way)
relationships. It displays the joint Group (rows) and Brand
Tabulation
distribution of variables. Preference (columns).

Parts of a Statistical Table:

A well-constructed table typically includes:

Part Description

Table Number For identification and reference.

A clear, concise description of what the table contains (including place,

Title
time, and nature of data).

Stub (Row Headings) The left-hand column listing the categories for the row variable.

Caption (Column
The top row listing the categories for the column variable.
Headings)

Body The cells containing the data (frequencies, percentages, etc.).

Part Description

Footnote Explanatory notes about specific parts of the table.

Source Note Indicates the source from which the data was obtained.

Example of Cross-Tabulation:
Table 1: Brand Preference by Age Group

Brand A Brand B Brand C Total

Age Group
(%) (%) (%) (%)

18-25 45 30 25 100

26-35 30 50 20 100

36-50 20 35 45 100

Source: Primary Data

Survey, 2025

Advantages of Tabulation:

• Simplifies Complexity: Condenses large volumes of raw data into a comprehensible

format.

• Facilitates Comparison: Makes it easy to compare data across categories or groups.

• Reveals Patterns: Helps identify trends, relationships, and outliers.

• Saves Space: Presents information efficiently.

• Provides Basis for Analysis: Serves as the foundation for statistical tests and
interpretation.

Unit 4

DATA ANALYSIS
Data analysis involves organizing, summarizing, and interpreting data to extract meaningful
insights for decision-making.

1. Measurement of Central Tendency

Central tendency refers to a single value that represents the entire dataset.

(a) Mean (Average)

• Formula: Mean = ΣX / N

• Most commonly used measure

• Sensitive to extreme values (outliers)

Example:
Marks: 50, 60, 70 → Mean = (50+60+70)/3 = 60

(b) Median

• Middle value when data is arranged in order

• Not affected by outliers

Example:
Marks: 10, 20, 100 → Median = 20

(c) Mode

• Most frequently occurring value

• Useful in categorical data

Example:
Data: 2, 3, 3, 5 → Mode = 3

Importance

• Simplifies large data

• Helps in comparison

• Basis for further statistical analysis

2. Measurement of Dispersion

Dispersion shows how spread out the data is.

(a) Range

• Range = Maximum – Minimum

• Simple but ignores distribution

(b) Variance

• Measures average squared deviation from mean

(c) Standard Deviation (SD)

• Square root of variance

• Most widely used measure

(d) Coefficient of Variation (CV)

• CV = (SD / Mean) × 100

• Used to compare variability across datasets

Importance

• Indicates consistency

• Helps assess risk (important in finance/marketing)

3. Univariate Analysis

Analysis of one variable only

Techniques:

• Frequency distribution

• Mean, median, mode

• Graphs (bar chart, histogram, pie chart)

Example:

Analyzing students’ marks in one subject

4. Bivariate Analysis

Analysis of two variables to study relationships

Techniques:

• Correlation analysis

• Regression analysis

• Cross-tabulation

Example:

Relationship between advertising expenditure and sales

5. Multidimensional Analysis I

Also called multi-variable analysis (basic level)

• Involves more than two variables

• Focus on understanding patterns across multiple dimensions

Techniques:

• Pivot tables

• Data cubes

• OLAP (Online Analytical Processing)

Example:

Sales data by product, region, and time

6. Multivariate Analysis II (Advanced Techniques)

These are advanced statistical techniques used for deeper insights:

(a) Factor Analysis

 • Reduces large number of variables into fewer factors

• Identifies hidden patterns

Example:

Customer preferences → reduced into factors like price sensitivity, quality, brand loyalty

(b) Cluster Analysis

• Groups similar data into clusters

• No predefined categories

Example:

Market segmentation (grouping customers)

(c) Multidimensional Scaling (MDS)

• Represents data visually in 2D/3D space

• Shows similarity/dissimilarity

Example:

Brand positioning in consumer perception

(d) Conjoint Analysis

• Determines how customers value product features

Example:

Customer preference for:

• Price

• Quality

• Brand

Importance of Multivariate Analysis

• Better decision-making
 • Useful in marketing, finance, HR

• Helps in prediction and segmentation

REPORT WRITING

Research report is a structured document presenting research findings.

1. Types of Research Reports

(a) Technical Report

• For experts

• Detailed methodology and analysis

(b) Popular Report

• For general audience

• Simple language and visuals

(c) Interim Report

• Submitted during research

• Shows progress

(d) Summary Report

• Brief version of full report

• Highlights key findings

(e) Research Paper/Journal Article

• Academic publication

• Structured format (Abstract, Methodology, etc.)

2. Structure of a Research Report

(a) Preliminary Section

• Title page

• Acknowledgement

• Abstract/Executive Summary

• Table of contents

(b) Main Body

1. Introduction

2. Literature Review

3. Research Methodology

4. Data Analysis & Interpretation

5. Findings

(c) End Section

• Conclusion

• Recommendations

• Bibliography

• Appendices

3. Guidelines for Writing a Report

Writing Style

• Use simple and clear language

• Avoid jargon

• Maintain logical flow

Presentation
 • Use tables, charts, graphs

• Proper headings and subheadings

Accuracy

• Ensure data correctness

• Avoid bias

Referencing

• Proper citation (APA/MLA format)

• Avoid plagiarism

Time Management

• Plan before writing

• Follow deadlines

Clarity & Objectivity

• Stick to facts

• Avoid personal opinions unless required

Importance of Report Writing

• Communicates research findings

• Helps decision-making

• Acts as a permanent record

• Useful for academic and business purposes