When you have one independent variate and several dependent variates, you can fit each dependent variate separately to the independent variate using simple linear regression. This module makes the process more convenient by having a scroll button going through each dependent variate.
The module expects two or more columns of measured data, with the independent in the first column and the dependents in consecutive columns.
In addition, an overall MANOVA test of multivariate regression significance is provided.
Missing data supported by column average substitution.
Regression for geometric morhpometrics
For Procrustes-fitted landmarks or Elliptic Fourier coefficients as the dependent variables, see the Geometry menu for regression with visualization of shape change.
For mathematical details, see the Past manual.