Actually i'm doing research on modal shift, so, i'm considering objective as, develop the models to study the mode choice behavior of travelers, to investigate the forecasting performance of mode shift models, to assess the impacts of the different government policies to be incorporated based on the model.

I read papers regarding these, but i ended up in confusion like which logit model i have to apply either multinomial or nested logit model.

Can anyone suggest me which model should i use to satisfy my objectives.

Thanks in advance.