Hello Anthony,

I don't understand how your question has apparently been unanswered for four years!

The two statistics each quantify something different, and therefore, each has utility.

R-squared quantifies the proportion (or, if you prefer, percent, after multiplying by 100) of variance in the dependent variable (DV) that can be explained by or accounted for by score differences in the independent variable(s). Higher implies a stronger, more dependable, linear association (with an upper limit of 1.00 or 100%); whereas lower implies little or no shared variance as a function of the linear relationship (with a lower limit of zero).

Standard error of estimate quantifies a "typical" degree of precision (or imprecision) in estimation/prediction of a case's score on the DV, as compared with that case's observed score on the DV. Formally, it is the standard deviation of all such discrepancies (residuals or errors of estimate). Smaller implies better precision (with a lower limit of zero), and larger implies poorer precision (with an upper limit of the SD of the DV).

Good luck with your work!