Hi, I have run a gologit2 command on my dependent variabel comprising of four catergoies with autofit. It is however difficult to interpret the odds ratio and therefore I attempted to get the predicted probabilities for the same. I am using Stata 14.0 version and my independent variables are a mix of factor and continous variables along with some dummies.
I tried to first run the following command to find out the marginal effects
margins, dydx(*) atmeans post
The margins had negative values as well. However, I wanted to calculate the predicted probabilities for my model. The dependent variable has four categories so here as well the interpretation is confusing for me. I calculated the predicted probabilties using the following command
margins, dydx (*) atmeans predict (outcome(1)) predecit (outcome(2)) predecit (outcome(3)) predecit (outcome(4))
The results using both the commands are the same.
I understand that for prediction, the predict command has to be used, so after running the gologit model, i ran the following command as well
predict p1 p2 p3 p4
However, upon tabulation it says too many values.
Can someone please help me out
Thanks in advance
I tried to first run the following command to find out the marginal effects
margins, dydx(*) atmeans post
The margins had negative values as well. However, I wanted to calculate the predicted probabilities for my model. The dependent variable has four categories so here as well the interpretation is confusing for me. I calculated the predicted probabilties using the following command
margins, dydx (*) atmeans predict (outcome(1)) predecit (outcome(2)) predecit (outcome(3)) predecit (outcome(4))
The results using both the commands are the same.
I understand that for prediction, the predict command has to be used, so after running the gologit model, i ran the following command as well
predict p1 p2 p3 p4
However, upon tabulation it says too many values.
Can someone please help me out
Thanks in advance
Comment