Companies marketing to consumers are very interested in statistics on retail sales. Statistics Canada puts out a monthly publication called “Retail Trade,” which is based on information from a sample of Canadian retailers who contribute to the Monthly Retail Trade Survey. Other commercial organizations assemble their own data and make their own projections about trends in retail sales. For instance, the Financial Post publishes “FP Markets—Canadian Demographics,” which contains estimates of retail sales for the current year together with “buying power indices” such as income per capita. A major question is which buying power indices are related to which retail sales data.
A selection from “FP Markets—Canadian Demographics,” is available in the data file ch07_MCSP_Retail_Sales_Canada for Canadian cities with over 500,000 population. It includes data based on households (people living under the same roof) and also data based on individuals (per capita data).3 In order to estimate total retail sales, we need to find some strong relationships—for example, regression lines with high R2. (a) Examine a regression of “sales per household” based on the explanatory variable “income per household,” and then examine a regression of “sales per capita” based on the explanatory variable “income per capita.” What happens if we remove the two cities at the top right of the scatterplot?
We will write a
specifically for you for only
805 certified writers online
(b) Next focus on clothing sales per capita and examine the strength of its relationship to income per capita. What happens if we remove the two cities that we removed in (a)?