BB108 Business Statistics
Word count: no more than 1,500-2500
Total marks: 100 (70 written report, 30 presentation) Weight: 40%
The project is the culmination of your business analytics skills and is designed to give exposure to the real-world environment of business analytics. It is also a creative project!
In the project you select the dataset (no boring data at all!), explore it and find some useful “insights”. Those “insights” need to answer questions that you pose Yourself.
This is where you can “unleash” your creative potential. Make sure you show how your insights are helpful There are several parts in your project and you are required to describe each stage in the report that you develop for submission.
• Setup the “Research Assignment” project in RStudio Cloud: you are working with your group member, so you need to make it accessible and “share” it, so both of you can work on it
• Setup a RMarkdown document for your report
• Select your dataset and load it to your RStudio
• Set up the structure of the report and start working!
1. Data exploration
a.Look at your data and come up with a MAJOR question to answer with your data.
b.You need further split your major question into steps (i.e. subquestions of the major question). They will be your “insights”. You need to write how your insights may be useful for people (and interesting as well)
c.Your project needs to have at least 7 subquestions related to descriptive analytics in section 2 (see lecture 1 for definition) and 1 subquestion related to predictive analytics in section 3 (this may be the answer to your Major question).
2. Data wrangling - descriptive insights
a.The focus is on using tidyverse` package and show your Mastery in getting descriptive insights from the data: see our tutorials for insights.
Type of questions: what are averages for.
b.what most expensive/used/delayed.
c. what is the least.
d.what are major groupings of.
e. and Many More
f. You need to show your understanding of the dataset (=variables there) as well as the problem to find descriptive insights in your data
g.Your descriptive insights need to come with at least 1 data viz. They need to be appropriate for the insight and add value = “tell a story”
h. You are WELCOME to use other packages as well! Google is welcomed.
3. Predictive insights
a. The focus is on using `tidymodels` package and show your Mastery in getting predictive insights from the data: see our tutorials for insights
b. Type of questions:
e. Use of “extra” packages is welcome, but stay within the `tidymodels` approach.
f. You need to supply data viz for your predictive insight to show the “validity” of your approach (=model evaluation)
Note: the same result can be achieved in million ways and the same question can be “done” using different approaches.
The use of data analysis tools needs to be explained by its suitability to address the question/subquestions. Each step needs to be accompanied with at least one data viz that supports your analysis and suggests next step in it.
The focus is on answering the question/subquestions you posed.
What learned about the problem by completing this project.
Present your data analysis in class, 10 minute oral presentation, recorded and uploaded as a video.
Upload link to VUCollaborate will be provided.
Structure of the report
• Motivation and background (appr 400 words) of the dataset
• Research question and subquestions (100 words) – this section needs to have a strong logical connection to the previous discussion
Descriptive analytics and insights (600-700 words)
• Your description of data analysis steps, dot-points style with justification are welcome. Make it visual using data viz.
Predictive analytics and insights (600-700 words)
• Your description of data analysis steps, dot-points style with justification are welcome.
Make it visual using data viz. Why to use this particular method to address this particular question/subquestion.
How to submit:
• Student declaration signed by group members
• Provide a link to your project at Rstudio.cloud
Please ensure that you enable sharing by clicking logging to your RStudio.cloud and selecting And then copying the link
What to submit:
• Shared link to the Rstudio cloud project that includes
• a RMarkdown file displaying your report and R code to perform the task above
• a copy of the dataset at the final stage (section 3)
• In your message specify:
• A copy/link to the original dataset used
• A link to the Rstudio project
• Name of the RMarkdown file with the report
• A copy of your presentation
• Video for the presentation.