The P-RRM model is one of the most recent members of the RRM model
family. This model has a cornerstone meaning: it postulates the
strongest random regret minimization behaviour possible within the
RRM modelling framework. In fact, it is one of the two special
limiting cases of the µRRM model (the other one is linear-additive
RUM). The model has a number of favourable properties. Most
notably, it is very fast to estimate in MNL form - which is
especially beneficial in the context of large choice sets.
The µRRM model generalizes the Classical RRM model by allowing the
variance of the error term to be estimated. More precisely, in the
µRRM model we estimate the scale parameter µ, which is
definitionally linked to the error variance.
This model is another recently proposed generalization of the
Classical RRM model. It has not been proposed by me but by Caspar
). For reasons of completeness, I also included codes for this
model on my website.
Not recently proposed, but of course not to be forgotten here: the
RRM2010 model. This model is proposed by Caspar Chorus (see
). So far, most RRM applications in the literature have used this
model. See Chorus et al. 2014 for a recent overview for
applications of this model.
Latent class (LC) models are increasingly used in choice analysis
and are particularly suitable to investigate the existence of
decision rule heterogeneity. In the context of advanced RRM
models, it is particularly interesting to define classes
corresponding to different decision models. In the latent class
software page 3 examples latent class models are presented:
1) a two-class model comprising of a RUM class
and a P-RRM class.
2) a two-class model comprising of two μRRM
3) a three-class model comprising of a RUM
class, a P-RRM class and a μRRM class.