The problem with the typical R-squared (aka, coefficient of variation) is that it can be artificially increased by simply adding more independent variables (aka, regressors). A higher R^2 does not necessarily imply that the model is a better fit. The "adjusted R-squared" addresses this problem: the adjusted R^2 will always be less than the unadjusted R^2, but it will not necessarily increase merely because an independent variable is added to the model.
David's XLS is here: https://trtl.bz/2I01sOF
David's XLS is here: https://trtl.bz/2I01sOF
Last edited:
