ApolloChoiceModelling forum

Hi Stephane and community,

I've run a multinomial logit (MNL) model using my dataset as the base model. This initial model, without any interaction terms, resulted in a disappointing model fit, with a Rho-square of less than 0.1. Even after increasing the sample size, there has been no improvement in this model. I have not compared this model to other models yet.

I'm wondering that
1. Is it possible for the base model to show such a low Rho-square, especially lower than 0.1? I'm afraid that I might do something wrong during model estimation.

2. Could you suggest any possible reasons for this outcome? Moreover, I would appreciate any advice on how to enhance the model's fit.

Thank you in advance for your guidance.

Hi

again, you would find answers to this in standard textbooks. In choice modelling, our emphasis is much more on relative fit by comparing different models on the same data, rather than on fit per se. Base models often have a low rho2

Stephane