Multivariate Multiple Regression (ViSta-MulReg) is a data analysis technique for predicting the values of a many response variables simultaneously from the values of two or more predictor variables. If you have just one response variable, you should use ViSta-Regres, not ViSta-MulReg. 

ViSta-MulReg uses OLS (ordinary least squares) regression techniques to find the strongest possible linear relationship between a linear combination of the predictors and one of the response variables. A different linear combination of the predictors is obtained for each response variable. This is the same as repeatedly doing many (univariate) multiple regressions. Multivariate Regression then combines all of the separate (univariate) multiple regressions together to obtain an overall multivariate test of the significance of simultaneously predicting all of the response variables.

If you  choose the Redundancy option you fit a Redundancy Analysis model to your data. This model obtains the single linear combination of the predictor variables which simultaneously predicts all of the response variables optimally, in the sense of maximizing the mean squared correlation between the linear combination and each response. This mean squared correlation is higher than that obtained by Multivariate Regression.

The ViSta-MulReg visualization shows the relationship between responses, the linear combinations of predictors, and the redundancy variables. No regression diagnostices are shown (use ViSta-Regres for diagnostics).
