Chapter 3 Parts
You can add parts to organize one or more book chapters together. Parts can be inserted at the top of an .Rmd file, before the first-level chapter heading in that same file.
Add a numbered part: # (PART) Act one {-} (followed by # A chapter)
Add an unnumbered part: # (PART\*) Act one {-} (followed by # A chapter)
Add an appendix as a special kind of un-numbered part: # (APPENDIX) Other stuff {-} (followed by # A chapter). Chapters in an appendix are prepended with letters instead of numbers.
3.1 “Group Assignment 2 @ MKTR”
Class,
Namaste, Salaam and Satsriakal.
Group assignment 2 is below. Involves using factor- and cluster-an apps we learnt in class.
The total assignment weightage is 10% of course grade. There are 3 tasks worth 3, 3 and 4 percentage points respectively of the course grade.
3.1.1 Task 1: Factor-An of Shopping Likerts.
Perform factor-an on ShoppingAttributes_co2022.csv. Drop variables with high (say >0.85) uniqueness.
Hint: Since the number of variables is high, may help to use the sort facility in the Loadings tab rather than the table in the Summary tab.
On the remaining variables, run factor-an. Then name and interpret the factors. Present your results in tabulated form.
3.1.2 Task 2: Cluster-An of Shopping Likerts.
Cluster-Analyze all variables in the dataset ShoppingAttributes_co2022.csv. How many segments are optimal? Give an informative to the segments that emerge.
Say Reliance Jio plans to enter the e-comm category in a big way to compete with the likes of Amazon and Flipkart. Which two segments should they target? Why? Write a short, speculative description of these two segments (in the spirit of what we saw for ‘Analogs and Wannabes’ in class).
3.1.3 Task 3: Cluster-An on Factor Bases.
Scan the dataset ‘hw2 ott prefs co2022.csv’. It combines variables from the OTT genre and OTT attributes datasets.
First factor-analyze the combined dataset. Drop variables with high (say, > 0.85) uniqueness. Name and interpret the factors that emerge.
Hint: Since the number of variables is high, may help to use the sort facility in the Loadings tab rather than the table in the Summary tab.
Download the factor-scores from the app and upload into the cluster-an wala app. You can choose to add back the dropped variables in the factor-an stage at this point.
Find how many clusters are optimal. Name and interpret the clusters.
Say Reliance Jio plans to enter the desi OTT category in a big way to compete with the likes of AmazonPrime and Zee5. Which two segments should they target? Why? Write a short, speculative description of these two segments (in the spirit of what we saw for ‘Analogs and Wannabes’ in class).
3.1.4 Deliverable: a zipped folder containing the following:
- A PPT with <15 slides (excluding Title slide, separator slides and annexures) detailing your results / recommendations / responses to Tasks 1-3 above.
- Pls use tables, animations, screenshots of results etc to make your PPT as self-explanatory as possible.
- Ensure your group number, group membership composition (names and PGIDs) are written in the Title slide.
- the actual, final datasets as csv files that you used to run the analyses (if different from the initial input datasets).
Deadline: 28-Nov Sunday Midnight. Dropbox will be made for the purpose and shared.
Any Qs etc, let us in Team MKTR know. Have fun with the assignment!
Ciao
Sudhir Voleti
