End-Term Exam (PGDM 2019-21), Term-V Introduction To R in Business Applications (Open Book and Online) Max. Marks - 40 Max. Time - 4 Hours

Download as pdf or txt
Download as pdf or txt
You are on page 1of 2

End-term Exam (PGDM 2019-21), Term-V

Introduction to R in Business Applications (Open book and Online)

Max. Marks -40

Max. Time – 4 hours

Instructions –

 There are four questions based on a case. Download the case dataset in your
working directory. All four questions are compulsory.
 Work on a .Rmd file, knit the output as .html file.
 All other instructions are as per the official guidelines

Case – Adorable Brand

Case dataset – Adorable.xlsx

Hundred management graduate students were interviewed to check their


perception on adorable brands i.e. which traits lead to adoration for a brand.

They were asked to think about an adorable brand in any category they choose
and rate the brand on four parameters on a 10-point scale. They were allowed to
use upto one decimal digit to precisely record their perception.

Similarly, they were asked to think about a not-adorable brand in any category
they choose and rate the brand on the same four parameters on a 10-point scale.

They were also asked, if they need to repurchase the same brand, how much
efforts, they would make to repurchase the same brand. Efforts are in term of
visiting multiple stores, checking multiple e-comm sites, waiting for some time till
the brand is available, willingness to pay some additional money etc. Their
responses were captured on a 10-point scale.

The parameters are listed below -

1. I feel connected with this brand as it reflects my personality (CONNECT)


2. I feel delighted when others see me using this brand (DELIGHT)
3. I feel excited while talking about this brand to my friends (EXCITED)
4. I feel gratified over my choice when I compare it with other competing brands
(GRATIFIED)
5. Whether the brand is adorable or not (ADORABLE)
6. How much efforts I would like to make to repurchase the brand (EFFORT)

The one outcome variable is binary, ADORABLE captured as Adorable and Not-
adorable. The other outcome variable, EFFORT is continuous.

The data is compiled and saved as adorable.xlsx. Attempt the following questions.

1|Page
1. Build a classification model using linear discriminant analysis [10 marks]
a. Check assumptions before carrying the analysis. Comment on the
results.
b. Create two scatter plots, one using CONNECT, DELIGHT as variables and
two, using EXCITED, GRATIFIED as variables. Comment on the
difference.
c. Create a model using LDA. Write the linear discriminant equation from
the model output.
d. Check the accuracy of the model

2. Build a classification model using logistic regression [10 marks]


a. Create a model using logistic regression. Write the logistic regression
equation from the model output.
b. Interpret logistic regression equation in terms of odds ratio
c. Check the accuracy of the model
d. Compare the accuracy from both models (1) and (2) and comment on
the difference.

3. Build a segmentation model using cluster analysis [10 marks]


a. Create a segmentation model using a non-hierarchical clustering and
check the accuracy of the model
b. Using a hierarchical clustering model, arrive at an optimal number of
clusters. Then, create a segmentation model using a hierarchical
clustering.
c. Compare the accuracy from both models created in a) and b) and
comment on the difference.
d. Run two different hierarchical clustering models using two different
linkage methods and compare the accuracy from both models and
comment on the difference.

4. Build a linear regression model to explain EFFORT [10 marks]


a. Create a step-wise regression model with EFFORT being the dependent
variable.
b. Explain the regression output. Write the regression equation with the
significant variable only.
c. Check assumptions with the help of residual plots. Comment on the
results. (Hint – create residual plots for step-wise regression output)
d. Create a scatter plot with EFFORT as the dependent variable and the
most significant variable from a) above. Also create a trendline and
comment.

2|Page

You might also like