This is a 24 case study on Hypothesis Testing given mid Statistics Module (Module 2) at the Flatiron School.
You are a DS working for Northwind, a supplier company. Your job is find interesting relationships in their database.
For this case study I used the northWind.sqlite data set. This included tables on Products, Orders, Employees, and Customers.
This case study used statistical A/B testing, ttest, pvalue, and power anaylsis methods to examine a given hypothesis.
For this question I want to look at one specific product sold by one distributor to a deliverable insight recommendation on how that producer should price their product. After some general exploration I decided to focus on Outback Lager sold at 15$ unit price by Pavlova, Ltd.
I conducted Welch's Ttest, Bootstrapping and calcualted pvales for the quantity sold of this product at to 0.00% discount rate, the 0.33% discount rate, the 1.33% discount rate and the 1.67% discount rate.
For each experiments I falied to reject the null hypothesis.
I failed to reject the null hypothesis at 0.00, 0.33, 1.33, and 1.67 % discount rate. This leads me to believe there is little correlation between discount rate and quantity sold I turn I would provide the recommendation that Pavlov, Ltd should stop discounting their Outback Lager as it does not lead to higher sales.
Given addional time I would pursue the following:
- Apply this analysis to other beers with in the data set- what beer do show an increase in sales given a discount?
- Examine what factors led the supplier changed the price of this been from 15$ to 12$.
- Analyze sale by region- does a particular group of customers buy at the discounted rate?