Manipulation Type Interacts with Scene-Object Recognition Performance and Memory-Based Eye-Movement Behavior

Poster Session E - Monday, April 15, 2024, 2:30 – 4:30 pm EDT, Sheraton Hall ABC

Dana Slabbekoorn¹, Deborah Hannula¹; ¹University of Wisconsin - Milwaukee

Previous work suggests that eye movements are sensitive to changes in encoded scenes (e.g., addition of an object) that go undetected in explicit reports. However, other studies indicate that these memory-based viewing effects are only evident when participants report the change. These results have important implications for memory systems theories that propose hippocampus-dependent memory is exclusively declarative (i.e., accessible to awareness). In our experiment, we examined whether manipulation type (object additions versus deletions) interacts with awareness and viewing patterns (e.g., greater viewing and recognition of the change when a new object is present in a scene, decreased viewing and recognition when an object is removed from a scene). We addressed this question by comparing viewing effects across scene types (manipulated, repeated) subdivided by whether the critical scene region was empty or filled. We also investigated the influence of test type (indirect, direct) on memory-based viewing effects. As predicted, more time was spent looking at objects than empty space, and object additions were more often reported than object deletions. Memory-based viewing effects (manipulated > repeated AOI viewing) were also larger when the critical region was filled, especially when memory was tested directly. Preliminary results suggest that these effects may depend on awareness (i.e., knowledge of the change), though analyses are still underway. Our results indicate that manipulation type should be considered in studies that address questions about memory, awareness, and eye-movement behavior. Future studies should examine object identity and/or position changes (e.g., left-right shifts), which was not done here.

Topic Area: LONG-TERM MEMORY: Other