Search Papers | Poster Sessions | All Posters
Poster C120 in Poster Session C - Friday, August 9, 2024, 11:15 am – 1:15 pm, Johnson Ice Rink
Reconstruction of real-time planning processes in a multi-goal task using eye tracking
Sanghyun Park1 (), Hun S. Choi2, Byeong Jun Park1, Won Mok Shim1,2; 1Sungkyunkwan University, 2Institute for Basic Science
Given a dynamically changing environment, it is essential for humans to establish diverse plans in order to flexibly adapt to such changes. However, examining the planning processes has proven difficult as they are not manifested in observed actions. Previous studies on planning utilized tree search models but they often implemented tasks in a stable environment that were too simple to effectively induce complexity in the process of planning. In this study, we designed a new arithmetic calculation paradigm that requires considering multiple plans before making decisions. To examine the real-time planning process, we reconstructed the planning process of each participant by an eye-tracking model and compared it with the result from a parameterized tree search model trained using their reported plans. Our eye-tracking model calculated the probability of possible plans in real-time based on the gaze location. The correlations observed between the likelihood of each possible plan derived from the eye-tracking and tree search models confirmed that the considered plans could be reliably inferred from the eye-tracking data. Beyond explaining the planning process through behavioral models, we demonstrate a new method for revealing the real-time process of plan generation.
Keywords: Eye-tracking Plan reconstruction tree search model