search for: multivaruate

...iates, would fit the 2x2 table). This issue becomes more interesting when there are covariates as well. In fact, it seems to simply fit the marginals. So I'm wondering what is to be gained by using vglm instead of simple glm? Or is there something I've overlooked, which would cause a true multivaruate fit to to be done by vglm? If course, I can use brute force to fit the 2x2 table: summary(vglm(cbind(x*y,x*(1-y),(1-x)*y,(1-x)*(1-y))~1, family=binomialff(mv=TRUE))) ##Coefficients: ## Value Std. Error t value ##(Intercept):1 -1.1527 0.33113 -3.4810 ##(Intercept):2 -1.26...