Q1: AB Testing (20 pts)
Background
A wholesale company sells food products (fresh produce, dairy products, and frozen products) primarily to two types of customers: retail customers (such as grocery stores) and hotels/restaurants. A few months ago, it launched a new promotion program designed to increase sales, and would like to see if the program was effective. The dataset “WholeSaleData.txt” has the following information:
• Consumer_Type: HoRe for sales to hotels/restaurants; Retail for other sales.
• Old/New: Old if sales before the new program was introduced and New otherwise.
• Fresh_Produce: Sales of fresh produce
• Dairy_Products: Sales of dairy products
• Frozen: Sales of frozen products

Q1a) (6 pts) Compare the average sales of Fresh product before and after the program was introduced. Is there statistical evidence (difference, z-statistics and p-value) that the program is effective for fresh product, at a 5% level? Does your conclusion hold for both consumer types separately?

Q1b) (6 pts) Repeat your analysis in part Q1a), but for dairy products,

Q1c) (6 pts) Consider the two statements:
i) The promotion increases sales of frozen products to retail consumers.
ii) The promotion increases sales of frozen products to hotels and restaurants.
Do you agree with one, both or neither statement? Which statement do you agree with more confidently? Justify your response quantitatively and precisely.

Q1d) (2 pts) Given this data (and, in particular, your findings in a), b), c)), what is your overall recommendation to this company if the new program has to be applied to all customers in the same product category? What if different customer type can have different program in each product category? Your recommendation should draw on your quantitative analysis for evidence.

Q2: KPI and Dashboarding (30 pts)
Background
Greenhouse gas emissions are considered the main cause of global warming. The common measurement of total greenhouse gas emissions is “kt (kilotonnes) of CO2 equivalent.” Under the United Nations Framework Convention for Climate Change (UNFCCC), all Parties are invited to submit their emissions of Greenhouse Gases (GHG). With the Paris Agreement of December 2015, all countries are encouraged to report frequent updates of emissions and nationally determined mitigation options.

The dataset “Total_GHG.txt” is collected through The World Bank and contains total greenhouse gases emissions in kt of CO2 equivalent for different countries/regions from year 1994 to 2012. Please use this together with “CountryCode.txt” and “GDP.txt” from BB to answer the following questions.

Q2a) (8 pts) In 2012, which five countries had the most GHG emission? Provide a line plot to compare the total GHG emission from 1994 to 2012 for these five countries. Briefly explain your observations.

Q2b) (6 pts) Import “CountryCode.txt” on a separate worksheet. For each country in the the emission dataset, find the corresponding Region and IncomeGroup. Which region has the highest total GHG emission in the High Income group? What is the corresponding value?

Q2c) (10 pts) Please import “GDP.txt” on a separate worksheet. For each country in the GDP data set, calculate the total GHG emission per GDP$ by the following formula for each year.
GHG (in metric tons) per GDP$ = 1000 * Total GHG in kt / GDP in $
(Hint: You may make a copy of the GDP worksheet and fill the content with the above formula. Please note that the list of countries in the emission data is not exactly the same as the one in the GDP data. You may want to use Vlookup to find the GHG and/or GDP for different countries in different years.)

For the same five countries in Q2a), develop a line plot from 1994 to 2012 for this new metric. Briefly explain your observation. Compare to your answer in Q2a.)

Q2d) (6 pts) Consider the two metrics
• Total GHG Emissions (from Q2a)
• GHG per GDP$ (from Q2c)
If you were the UNFCCC and were interested in identifying countries that were not making significant steps towards combatting climate change, which metric would you prefer to use and track? Justify your answer using the properties of a good metric discussed in class.

Q1: AB Testing (20 pts) Background A wholesale company sells food...