Regression Quiz - Statistics.com: Data Science, Analytics & Statistics Courses

You are an analyst for a chain restaurant company, and would like to build a regression model for predicting profit at new restaurant locations. The predictor variables are: Cov = number of covers or customers served (in thousands), Fco = food costs (in thousands of dollars), Oco = overhead costs (in thousands of dollars), Lco = labor costs (in thousands of dollars), Region = geographical location (Mountain, Southwest, or Northwest). What is the most appropriate treatment of the "Region" variable?

a. It can be used as is (a text categorical variable)

b. It should be turned into 3 binary variables (Mountain – 0/1, Southwest – 0/1, Northwest – 0/1)

c. It should be turned into 2 binary variables (any two of: Mountain – 0/1, Southwest – 0/1, Northwest – 0/1)

d. It should be handled as an ordered categorical variable

A multiple linear regression was built to try to predict customer expenditures based on 200 independent variables (behavioral and demographic). 10,000 rows of data were fed into a stepwise regression, each row representing one customer. 1,000 customers were male, and 9,000 customers were female. The final model had an adjusted R-squared of 0.27 and seven independent variables. Increasing the number of rows of data to 100,000 and rerunning the stepwise regression will most likely:

a. Reduce the R-squared

b. Increase the R-squared

c. Reduce heteroskedasticity

d. none of the above

A regression model is built to predict how long a web visitor will spend at a site, based on visitor characteristics and the initial two pages viewed. This information will be used dynamically to show the visitor a different web experience, depending on the prediction. If you had to choose one of the following measures to evaluate the model, what would it be?

a. Root-mean-squared-error (RMSE)

b. R-squared

c. F-test

d. coefficient p-values

First Name

Last Name

Email(Required)