Multiplying an estimating equation by a correcting factor to correct heteroskedasticity may cause extra correlation to enter the model, which raises the R^2. This renders our ultimate regression results meaningless. True or false, or uncertain? Please provide an explanation.
The answer is false. A regression model is useless until all the regression assumptions are satisfied. So a regression model where Heteroskedasticity is not satisfied is a useless model. We must make necessary correcting adjustments to remove Heteroskedasticity.
Multiplying a regression ...
Explains why a regression model where Heteroskedasticity is not satisfied is a useless model.