More RRM Methodoly
A measure of profundity of regret
The degree of regret minimization behaviour (or profundity of regret) imposed by the RRM models (except for the P-RRM model) is not constant, but depends on the size of the estimated parameter as well as on the distribution of the attribute levels in the data. Therefore, the parameter estimates on their own are not very informative for the imposed behaviour by the RRM models. To acquire insight on the behaviour imposed by these models, a formal measure of the profundity of regret is proposed. With this measure it is possible to pinpoint the overall degree of regret behaviour for each attribute.Click here to go to the Profundity of regret software page.
Accouting for variation in choice set size in RRM models
In many choice situations the choice set size, i.e. the number of alternatives which are available to the decision-makers, varies across choice observations. In RRM models such variation in choice set size is consequential for the model predictions. To account for variation in the choice set size when estimating RRM models a simple, but effective correction factor can be used. This correction factor scales the overall regret.Click here to go to the page on Variation in choice set size in RRM models.
Robustness of RRM modelling outcomes towards omitted attributes
As discrete choice models may be misspecified, it is crucial for choice modellers to have knowledge on the robustness of their modelling outcomes towards misspecification. One type of model misspecification occurs when not all attributes that are relevant for the choice are included in the choice model. To investigate the robustness of RRM modelling outcomes towards the omission of a relevant attribute several Monte Carlo experiments are conducted.Click here to go to the page on the Robustness of RRM modelling outcomes towards omitted attributes.
Decision rule robust experimental designs
Despite compelling evidence that decision-makers use a wide range of decision rules when making choices, the design of Stated Choice experiments has exclusively been based on the (often implicit) assumption that decision-makers make choices using (linear-additive) Random Utility Maximization (RUM) rules. In a recent study I show that efficient experimental designs can also be created for RRM. One particularly important result of this study is that designs that are efficient for estimating RUM models can be highly inefficient for estimating RRM models, and vice versa. Therefore, it is appealing to take multiple decision rules into account when creating efficient experimental designs.Click here to go to the page on the Decision rule robust experimental designs.