The adjusted r-squared is helpful for multiple regression and corrects for erroneous regression, giving you a more accurate correlation coefficient. If you look at the multiple regression we did, ...

Using a hypothetical data set, the authors provide concrete examples to illustrate that canonical correlation analysis is a general linear model, subsuming other parametric procedures as special cases ...