Quantitative Research Methods:

Revision
PIYUSA DAS, ASSOCIATE PROFESSOR (KSOM)
 Corporate Experience Courses I teach and learn from

Business Analytics
•Business Research Methodology (2nd Sem MBA)
•Analytics Tool Box (TABLEAU & POWERBI) (3rd Sem MBA)
•Predictive Analytics Using SAS (3rd Sem MBA)
•Marketing Analytics (3rd Sem MBA)
•Building Generative AI Applications(4th Sem MBA)
•Supply Chain Analytics (4th Sem MBA)
•Multivariate Data Analysis (4th Sem MBA)

PIYUSH DAS Marketing

•B2B Marketing (3rd Sem MBA)
Education •Marketing Research (4th Sem MBA)

Operations
•Managing Service Operations (3rd Sem MBA)
•Project Management (4th Sem MBA)

General Management
•Blockchain Applications in Business (3rd Sem MBA)
Rules for the Sessions
 Attendance: Do join the class before the scheduled time of class start.You won't get
attendance in case you join late.
 Data Analysis Tools: Bring your fully charged laptop to the class as we will be
carrying out exercises in excel/SPSS.
 Internal Evaluation: There will be one assignment for 5 marks, class participation for
5 marks and a project for 15 marks. Details of the assignment and project will be
uploaded in google classroom in due time.
 @ Class Representatives: Send me the details of groups (Name, Roll numbers) in a excel file
by 17.03.25. A maximum of 6 students can be in a group.
 Study Material: I will share the class discussion slides and excel files in the google
classroom.
 Class Discipline: I have zero tolerance for class disturbance. Any class disturbance
will result in loss of attendance and penalty in internal marks in increasing order (e.g
for 1st offence penalty is -2 and for 2nd offense -4 and so on)
 Quantitative Research Methods
 Session 1: Revision of Concepts, Statistical Hypothesis Testing (When
to use which test)
 Session 2: Product Development Research - Conjoint Analysis
 Session 3: Product Development Research - Conjoint Analysis
 Session 4: STP Research – Cluster Analysis
 Session 5: STP Research – Logistics Regression, Discriminant Analysis
 Session 6: STP Research – Logistics Regression, Discriminant Analysis
 Session 7: Project Presentation
 Session 8: Project Presentation
A Primer to Scales in Research Methodology
Nonparametric
Parametric tests Scales tests assume that
assume that the
the variables are
variables of interest are
measured on a
measured on at least an
nominal or ordinal
interval scale.
Metric Categorical scale

Interval Ordinal Nominal

Ratio Scale
Scale Scale Scale

All the three property of Can I measure the By using the responses , Any scale will have minimum
Nominal, Ordinal and Interval. difference between Can I put the responses in nominal (identification
And have an absolute zero. groups particular order/groups property)
Hypothesis
Hypothesis: Hypotheses are usually defined in terms of interrelations between
variables and are often based on a substantive theory. They are tentative answers
to research questions.
• In the classical paradigm, a statistical hypothesis is a statement about probability
distributions that potentially govern the experimental data.
 For instance, the statement ‘‘this coin is not biased’’ is typically transformed into
‘‘P(this coin turns up head ) = 0.5

Research hypothesis or Alternate Hypothesis: (H1 or Ha ): A statement reflecting the

substantive hypothesis. It is always expressed in terms of population parameters, but its specific
form varies from test to test.

Null hypothesis (H0): A statement of “no difference” that contradicts the research
hypothesis and is always expressed in terms of population parameters
Alternate Hypotheses: One- and Two-Tailed Tests
One-tailed test
In a one-tailed test, the alternate hypothesis is directional; that is, it
specifies that a population mean is either less than (<) or greater than
(>)some specified value.
• Right-tailed test
 When a one-tailed test specifies that the population mean is greater than
some specified value, because we will evaluate the outcome at the right tail of
the sampling distribution
 H1 : µ > some specified value
• Left-tailed test
 If the research hypothesis specifies that the population mean is less than
some specified value, it is called a left-tailed test because the outcome will be
evaluated at the left tail of the sampling distribution.
 H1 : µ < some specified value
Two-tailed test
 When we have no theoretical reason for specifying a direction in
the research hypothesis, we conduct a two-tailed test. The research
hypothesis specifies that the population mean is not equal to some
specified value
Farm Loan
Research Objective: To understand
whether the average debt per
agricultural household in Odisha is
greater than INR 1.04 Lakh

Ha: Average loan amount by

farmers in Odisha > INR 104000
( µ > 104000)

H0: Average loan amount by

farmers in Odisha <= INR 104000
( µ <= 104000)
Average tuition fees in higher education
Research Objective: To understand
whether the average spend on
tuition fees for higher education in
Odisha is less than INR 361975(year)

Ha: Average spend on tuition fees is

less than INR 361975 (µ < 361975)

H0: Average spend on tuition fees

is greater than or equal to INR
361975 (µ >= 361975)
 Statistical Inference
• The process making an inference about the population based on what is measured in the
sample is called statistical inference
• Three forms of Statistical Inference are
 Point Estimation
•In point estimation, we estimate an unknown parameter using a single number that is calculated from
the sample data.
•Based on sample results, we are 95% confident that that average marks secured by students in BRM will be 73
 Interval Estimation
• In interval estimation, we estimate an unknown parameter using an interval of values that is likely to contain
the true value of that parameter (and state how confident we are that this interval indeed captures the true
value of the parameter)
• Based on sample results, we are 95% confident that final marks of students will be between 65 to 81
 Hypothesis Testing
• In hypothesis testing, we begin with a claim about the population (we will call the null hypothesis), and we
check whether or not the data obtained from the sample provide evidence AGAINST this claim.
Ha: There is a relationship between number of classes attended and final marks scored
Ho: There is no relationship between number of classes attended and final marks
 CENTRAL LIMIT THEOREM
• The central limit theorem states that if you have a i.i.d
(independent and identically distributed) population
with mean μ and standard deviation σ and take sufficiently
large random samples from the population with
replacement, then the distribution of the sample means will
be approximately normally distributed.
• This will hold true regardless of whether the source
population is normal or skewed, provided the sample size is
sufficiently large (usually n > 30).

Standard deviation of sampling distribution is known as standard error

Statistical Hypothesis Testing
Statistical hypothesis testing : A procedure that allows us to evaluate hypotheses
about population parameters based on sample statistics

ASSUMPTIONS OF STATISTICAL HYPOTHESIS TESTING

• Assumptions include considerations of the level of measurement of the variable, the
method of sampling, the shape of the population distribution, and the sample size.
• All statistical tests assume random sampling
• Tests of hypotheses about means also assume interval-ratio level of measurement
• Population under consideration be normally distributed or that the sample size be larger than 30

• Test Statistic - The distance between the data and the model prediction is measured
using a test statistic (such as a t-statistic or a Chi squared statistic)
Which Statistical Test to Use ?
Variable Type Prepared by:
Hypothesis Testing Data Analysis method PROF PIYUSA DAS
Study Aim Assistant Professor- KIIT School of Managment
Ha has '>' symbol : Right tailed test
Comparison with a Metric One-sample t-test Ha has '<' symbol : Left tailed test
UNI-VARIATE expected value Ha has 'not equal to' : Two-tailed test
Dependent Independent
Categorical Categorical Cross-tabulation
Relationship
Metric Metric Correlation

Dependancy Metric Metric Regression

Comparison Metric Categorical 2-Sample t-test

(between 2 groups) data of two samples are statistically independent Independence

BI-VARIATE/MULTI- Equal Variances

F-test to test variances are equal or not.
VARIATE
While testing, take the larger varainces in numerator
Unequal Variances

Paired

Comparison Metric Categorical One-Way ANOVA

(more than 2 groups) (one category)
Metric Categorical Two-Way ANOVA without replication
(2 categories, no interaction)
Metric Categorical Two-Way ANOVA with Replication
(2 categories, with interaction)

Hypothesis Testing Approach FOR SOLVING PROBLEMS FOR PROJECT/INTERNSHIP WORK

Step 1 Find out number of variables and type of variables Step 1 Carry out secondary research to form research objectives
Find out study aim from hypothesis (Comparison, relationship,
Step 2 dependency) Step 2 If required, convert research objectives to hypothesis (alternate and null)
Step 3 Select the appropriate data analysis method Step 3 Decide on scale type (for questions) based on variables hypothesis (create a research framework)
Step 4 Carry out data analysis and provide your insights Step 4 Design questionnaire and carry out pilot testing
Step 5 Decide on sample size and data collection method (on-line, offline)
Step 6 Collect data , clean data and tabulate data for data analysis
Step 7 Carry out data analysis and provide your insights
 Dataset
Body Mass Insurance You have joined a health insurance
Index Number of City Premium Claimed Claim
Gender Age (BMI) Children Smoker Type charges Insurance Amount company as a summer intern.
Female 25 28 0 No Tier I 5000 Yes 2000 Insurance companies primarily
Male 32 24 3 No Tier III 6000 No 0 generate revenues by selling
Male 28 33 1 Yes Tier III 6000 Yes 4000 insurance policies and receiving
Male 33 35 0 Yes Tier II 5000 Yes 4000 payment as a premium. Using
Male 32 31 0 No Tier II 4000 Yes 3000 historical data and statistical analysis,
Female 31 26 0 No Tier I 4000 No 0 the insurance company can predict
Female 46 23 3 Yes Tier I 8000 No 0
the loss events that may occur in the
Female 37 28 2 No Tier II 10000 Yes 3000
Male 37 33 3 No Tier II 10000 Yes 3000
future and, thus, the approximate
Female 60 27 1 No Tier II 5000 No 0 amount they may have to pay out in
Male 25 26 0 Yes Tier I 10000 Yes 8000 claims. The primary way that an
Female 62 22 0 Yes Tier I 10000 Yes 4000 insurance company makes a profit is
Male 36 24 1 No Tier III 2000 No 0 by ensuring the premiums received
Female 52 22 1 No Tier III 5000 No 0 are greater than any claims made
Male 23 24 0 No Tier III 2000 No 0 against the policy. This is known as
the underwriting profit
Q1(HOMEWORK for next class)
 Your manager believes that customers with children will care
for their own health and might have fewer health issues.
Hence, those customers may be more profitable for your
insurance company. He wants to understand whether profit
depends on the number of customers' children.
 For the above research objective, answer the following
 Alternate Hypothesis and Null Hypothesis:
 Variable Type Identification (Dependent, Independent):
 Scale Type Identification(Nominal, Ordinal, Interval, Ratio) :
 Identify and carry out appropriate Statistical Test:
Q2 (HOMEWORK for next class)
 Your manager thinks there might be profit differences
across different tier cities (Tier I, Tier II, Tier III).
 For the above research objective, answer the following
 Alternate Hypothesis and Null Hypothesis:
 Variable Type Identification (Dependent, Independent):
 Scale Type Identification(Nominal, Ordinal, Interval,
Ratio) :
 Identify and carry out appropriate Statistical Test:
Q3 (HOMEWORK for next class)
 Based on historical data, the manager thinks Female
customers are more profitable (At least by INR 500) than
Male customers
 For the above research objective, answer the following
 Alternate Hypothesis and Null Hypothesis:
 Variable Type Identification (Dependent, Independent):
 Scale Type Identification(Nominal, Ordinal, Interval,
Ratio) :
 Identify and carry out appropriate Statistical Test:
Q4 (HOMEWORK for next class)
 Body mass index (BMI) is a person's weight in kilograms divided by
the square of height in meters. BMI is an inexpensive and easy
screening method for weight categories—underweight (BMI below
18.5), healthy weight (BMI between 18.5 to 24.9), overweight (BMI
between 25 to 29.9), and obesity (BMI 30 or higher).The manager
thinks a person's weight category will be related to whether they
claimed insurance.
 For the above research objective, answer the following
 Alternate Hypothesis and Null Hypothesis:
 Variable Type Identification (Dependent, Independent):
 Scale Type Identification(Nominal, Ordinal, Interval, Ratio) :
 Identify and carry out appropriate Statistical Test:
 Q5 (HOMEWORK for next class)
 Another intern in your office
had conducted a multiple linear
regression with profit as the
dependent variable and 'Age,'
'BMI,' & 'number of children' as
independent variables. The
partial regression output is
shown here
i) Calculate t-statistics and find out the confidence interval of coefficients (lower 95%, upper 95%.
ii) Based on your findings, provide your insights on the effect of independent variables ('Age,' 'BMI,' &
'number of children') on the dependent variable (customer profit).
Q & As