Your boss has just acquired data about ice cream sales in different cities of two countries, and she is interested in learning how to increase sales. She is also interested in learning something new about ice cream sales. So, she wants you to figure it all out.
As part of this project, you will complete exploratory data analysis, hypothesis testing, modelling, and prediction.
You are given a dataset on ice cream sales. The data is collected by surveying ice cream shops in two countries.
The dataset you are given has the following variables with brief descriptions.
Variable Description
shopID Shop ID for surveying purposes
icecream_sales Ice cream sales at the shop on the day the data is collected (£)
income The average income of people in the area where the shop is located (£ per year)
price The average price per portion of ice cream in the shop on the day the data is collected (£)
temperature The temperature in the area the shop is located on the day the data is collected (Celsius)
season Spring, Summer, Autumn, Winter
country Country A and Country B
You can load the data from here: http://bit.ly/2OMHgFi. (Links to an external site.)Using this dataset, perform the following analyses.
A. Exploratory data analysis (EDA): (30 points) Perform exploratory data analysis (EDA) and explain your variables numerically and graphically. Please do not replicate the same investigation in numeric and visual explorations. A brief interpretation should accompany your R output and plot. (575 words max)
B. Hypothesis testing: (10 points) Construct your hypotheses for testing the average income in a location in Country A relative to the average income in a location in Country B. Test the hypothesis and explain the result. (100 words max)
C. Modelling:(40 points) Develop a multiple linear regression model to predict ice cream sales using all explanatory variables. The outcome variable and the explanatory variables can be existing variables in the dataset or new variables you might create based on existing variables. (1,150 words max)
Specifically, answer the following questions:
What is your regression equation?
What are the interpretations of all your coefficients?
All else being equal, what is the predicted difference between ice cream sales in a location in Country A with an average income of £13,000 compared to a location in Country B with an average income of £20,000?
All else being equal, what is the predicted change in ice cream sales if the price goes up by £0.75 and temperature goes up by 0.5 degrees at the same time?
What is the R squared of the model? Interpret it.
Is this model statistically significant at a 1% significance level?
What are the confidence intervals of coefficients on explanatory variables at a 99% confidence level? Explain what they mean.
Test and explain if your data meets the regression conditions.