Transformations in visual working memory impose fidelity costs
Poster Presentation: Sunday, May 18, 2025, 2:45 – 6:45 pm, Pavilion
Session: Visual Memory: Capacity and encoding of working memory
Schedule of Events | Search Abstracts | Symposia | Talk Sessions | Poster Sessions
Eva Lout1 (), Jarrod Lewis-Peacock1; 1University of Texas at Austin
In our dynamic world, working memory (WM) plays a critical role in storing, updating, and using goal-relevant information to enable adaptive behavior. Previous research on WM updating has demonstrated that updating slows significantly when a transformation of the item is required before substitution (e.g., “add 4” to the “25” stored in WM vs. “replace 25 with 29”). Transformation processes require the original information to be retained throughout the updating process. Critically, recent EEG findings suggest that this original information may linger in WM even after the transformation is completed. Retaining the original copy may help stabilize the newly transformed representation, but it can also tax the limited capacity of WM storage. Performing multiple transformations may impose an accumulative burden, ultimately degrading the fidelity of the representation. This study aimed to test this hypothesis by varying the number of mental rotations participants performed on the visual orientation of a remembered teardrop shape (zero, one, three, or five rotations of either 30°, 60°, or 90° in either direction). We conducted two experiments with the same task but different memory reports: In Experiment 1, participants reported the updated teardrop shape’s precise orientation, while in Experiment 2, they made a same/different judgment between the remembered orientation and a probe orientation. Our preliminary results revealed that as the number of transformations increased, both memory precision and recognition accuracy progressively declined towards chance levels, while reaction times remained relatively stable. These results suggest that the amount of uncertainty in the reported memory representation increases with each successive transformation. Updating the contents of WM through transformations of existing representations appears to be a lossy process that reduces memory fidelity. Fidelity costs may arise from interference caused by the continued storage of outdated representations or from imprecise transformation processes that degrade sharp representations into blurred ones.