Choice model based on revealed preference is under certainty, risk or uncertainty

[zh.kev...@gmail.com]

Dear all,

I have read a few papers, but still a little bit unclear on whether a choice model based on revealed preference should be considered under the condition of certainty, risk or uncertainty?

Economically, certainty means individuals has absolute understanding of each choice's outcome. Risk is the situation that individuals have clear understanding of the probability distribution of outcomes of each choice, while the uncertainty means one may not aware the outcome of a choice.

From what I understand that though random utility theory (RUM) do have a error part, it mainly used to describe what the modeller does not know, and that means subjectives fully know the characters of each choice. if I understand this right, for example, if I am conducting a choice model based on some revealed preference, this model should be considered under certainty. And transport modeller usually do not specify the different situations and adopt the certainty to describe the uncertainty. (excepted research intending to describe create uncertainty in some stated preference studies).

Am I understand this right?

Thanks,

Kevin

[Jason Hawkins]

Hi Kevin,

You are correct that the error term generally refers to uncertainty on the part of the modeller due to omitted variables. It is generally assumed that the revealed preference data are for rational decision-makers. I know that there has been work looking at risk and uncertainty through the lens of rank dependence and applications of prospect theory but, to the best of my knowledge, there are few mainstream applications in operational transport demand models. I found an article by some of the eminent researchers in the field from 2008:

 https://link.springer.com/article/10.1007/s11002-008-9047-0

I am sure there are more recent sources, but that seems like a good place to start to me.

Jason