Transform the suggested retail price to log (Price) and √Price. Transforming data using roots, logarithms, or reciprocals can often reduce heteroskedasticity and right skewness. (Transformations are discussed in Chapter 2.)
Create regression models and residual plots for these transformed response variables using the explanatory variables selected in Question 5.
a. Which transformation did the best job of reducing the heteroskedasticity and skewness in the residual plots? Give the R2 values of both new models.
b. Do the best residual plots correspond to the best R2 values? Explain.
While other transformations could be tried, throughout this investigation we will refer to the log-transformed response variable as TPrice.