New trends in urban systems establish the need for more complicated behavioral models that better describe traveler behavior and investigate unknown mobility patterns. Adding context and accounting for conditions of uncertainty constitute a major challenge in that regard. The current paper contributes to the further development of regret-based behavioral models under uncertainty by modeling context dependency and respondent heterogeneity. Context defined in terms of activity type, number of people involved and the nature of the relationship among the people in the travel party is used to specify class membership as opposed to the common practice of using only social demographic variables. A stated choice experiment of route choice behavior ...