Common mistakes in discrete choice modeling

This series of videos discusses various mistakes that are commonly done in the context of discrete choice modeling. The mistakes are identified, and the correct way to deal with the issue is presented.

Click on the title or the YouTube icon to access the video.

Episode 1:To be or not to be significantRemoving coefficients that have a t-test less than 1.96 (in absolute value) is not necessarily correct
Episode 2:Survival of the fittest... or notConcluding that the model with the best fit is the best model is not necessarily correct
Episode 3:Aggregate elasticitiesAggregate elasticities are not the sum of disaggregate elasticities.