Cognitive cost in route choice with real-time information: An exploratory analysis

Abstract

Real-time traffic information is increasingly available to support route choice decisions by reducing the travel time uncertainty. However it is likely that a traveler cannot assess all available information on all alternative routes due to time constraints and limited cognitive capacity. This paper presents a model that is consistent with a general network topology and can potentially be estimated based on revealed preference data. It explicitly takes into account the information acquisition and the subsequent path choice. The decision to acquire information is assumed to be based on the cognitive cost involved in the search and the expected benefit defined as the expected increase in utility after the search. A latent class model is proposed, where the decision to search or not to search and the depth of the search are latent and only the final path choices are observed. A synthetic data set is used for the purpose of validation and ease of illustration. The data are generated from the postulated cognitive-cost model, and estimation results show that the true values of the parameters can be recovered with enough variability in the data. Two other models with simplifying assumptions of no information and full information are also estimated with the same set of data with significantly biased path choice utility parameters. Prediction results show that a smaller cognitive cost encourages information search on risky and fast routes and thus higher shares on those routes. As a result, the expected average travel time decreases and the variability increases. The no-information and full-information models are extreme cases of the more general cognitive-cost model in some cases, but not generally so, and thus the increasing ease of information acquisition does not necessarily warrant a fullinformation model.

The influence of affective cues on positive emotion in predicting instant information sharing on microblogs: Gender as a moderator

“…Unlike sharing professional knowledge or past experience, information sharing on microblogs tends to be temporally and cognitively constrained, and to function in a relatively unplanned, spontaneous, and emotional manner (i.e., sharing news anywhere and anytime) (Lee & Ma, 2012;Wang et al, 2015). Time constraints limit one's attention to the elements of the environment (Howard & Sheth, 1969) and consequently lead a decision maker to choose a satisfying solution instead of an optimal one (Gao et al, 2011). Hence, when an incident occurs, information is delivered immediately on microblogs without a reduction in the timeliness of the information provided (Tobias, 2011).…”

“…This -reactiveness‖ is aligned with the impulsivity of human behavior, i.e., the need to express emotions predominantly precedes its anticipated utility (Strack et al, 2006). Therefore, instant information sharing on microblogs tends to be a stochastic behavior rather than a deterministic event (Gao et al, 2011) and, hence, can be conceptualized as a reactive and impulsive behavior (Wang et al, 2015). Based on this conceptualization, theories of impulsive behavior are appropriate in capturing and understanding the underlying mechanisms of instant information sharing in microblogs.…”

Entangled communities and spatial synchronization lead to criticality in urban traffic

“…Alongside traffic flow theory, the impact of different travel behaviours and traffic control schemes have been studied in depth10, for example in connection with the effect of pre-travel and real time travel information and variable travel demands1112, real and perceived information uncertainty1314 and its cognitive cost15, inequality among drivers16 or different user responses to travel information17.…”

Searching for empirical evidence on traffic equilibrium

“…There is a vast literature of discrete choice models developed for modelling the perception error and reaching SUE conditions [ 20 – 25 ]. Apart from deterministic and stochastic equilibrium definitions, a large body of literature focuses on bounded rationality to address the lack of accurate information and to model the inertia to switch decisions [ 26 – 28 ]. Nevertheless, given that the deterministic UE conditions can be established by virtually parameter-free algorithms and are empirically supported in controlled settings [ 29 , 30 ], they form the basis of many traffic assignment models, which are essential to estimate travel demand under hypothetical scenarios [ 31 ].…”