The article discusses four different models used to predict people's choices in travel modes. The researchers tested these models and found that if the data follows a certain pattern, the Multinomial Logit model is the best choice for predicting mode choices. The Nested Logit and Heteroscedastic Extreme Value models didn't perform significantly better than the Multinomial Logit model. The Mixed Logit model, while fixing a flaw in the Multinomial Logit model, is harder to estimate.