Hi everyone,   I'm conducting a multinomial logistic regression model using  proc logistic  in SAS with around 3.6 million observations, an outcome with 5 levels, and dozens of categorical predictors. I had no issue running both univariate and multivariate models when setting  param = ref .   However, once I tried  param = glm , it started giving the warning message of "The information matrix is singular and thus the convergence is questionable. specifying a larger SINGULAR= value." in multivariate models. After doing some research, I found this message suggesting a multicollinearity issue in the model. I then tried to use only 2 predictors and it still gave the message while the correlation matrix showed no correlation between the two predictors. As far as I know, the only difference of  param = ref  and  param = glm  is that  param = glm  uses less-than-full-rank reference coding, meaning that it will create  k-1  dummy variables given  k  levels in the categorical predictor. These two parametrization methods should generate the same log-likelihood and estimates given the same reference level. To confirm this, I also compared the result of the two models using only 2 predictors. While  param = glm  throwing a warning, the result is identical to  param = ref  (Except a bunch of zeros in the estimates of reference levels for each predictor in  param = glm , is it the cause?).   My question is, why did the  param = glm  model throwing a warning while  param = ref  did not. And more importantly, in this situation, should I trust the result of the  param = ref  even though no warning was displayed. I appreciate any advice and suggestions. Thank you in advance. ... View more