Hi there!
My very limited understanding is that when there are more than one dependent variable and a set of independent variables, i.e., (y1,y2,...,yn) = f(x1, x2, ..., xm) is called multivariate and the case where y = f(x1, x2, ..., xm) is called multiple. In the first case I can MANOVA while in the second case I can run Multiple Regression, for example. Suppose all y1, y2, ..., yn are ordinal; does something like MULTIVARIATE ORDINAL LOGISTIC REGRESSION exist to handle this? If yes, what if y1, y2, ..., yn are mix of ordinal and nominal? The case y = f(x1, x2, ..., xm) seems to be called MULTIVARIATE LOGISTIC REGRESSION on Google; I would call it MULTIPLE LOGISTIC REGRESSION. Am I right? Please refer/suggest book or article on MULTIVARIATE LOGISTIC REGRESSION, if any (according to my remark) with example involving numbers to help understanding what's going on.
Thanks,
Ray
|