The hypothetical bias in pedestrian route choice behaviors: experimental perspectives

Hypothetical Bias (HB) has rarely been examined in pedestrian route choice studies, as models are often difficult to validate and compare across experiments. This study investigates HB by analyzing how different scenario formats (Numbers, Diagrams, and Pictures) affect pedestrian route choice behavior within a choice experiment framework. We conducted two paired Stated Preference (SP) surveys and Revealed Preference (RP) experiments, designed three metrics to quantify HB, and tested five simple calibration methods representing classic approaches together with their external validity. In addition, participants' perceptions of route attributes were tracked through open-ended questions, providing qualitative insights into the decision process. The results show that (1) formats significantly influence participants’ perception of route attributes; (2) Diagram and Picture formats, due to their stronger visual salience, induce more bias than the Number format; (3) systematic HB exists, with both the type of scenario format and the way scenarios are constructed identified as major sources; and (4) two calibration methods are effective: recoding uncertain choices to the scenario-advantage route, and incorporating personality scores (Big Five Inventory). This study demonstrates how information formats shape HB through underlying cognitive mechanisms. It also emphasizes that experimental design itself can be a source of HB requiring careful consideration.