Scene and object false memory in a photo-realistic paradigm

Poster Presentation 43.449: Monday, May 22, 2023, 8:30 am – 12:30 pm, Pavilion
Session: Visual Working Memory: Space, features, objects

Shaela Jalava1 (); 1Queen's University

Encoding of false visual memories can have severe negative consequences (e.g. incorrectly identifying foraged berries; mistaken eyewitness testimony). This phenomenon remains unclear, as most research focuses on verbal episodic stimuli, or nonrepresentational drawings when studied in the visual domain. Here, we complement this prior work by probing for whether false memories could be induced for unusual object-scene pairings in realistic photographs. Participants (N = 60) studied photographs of household scenes (SCEGRAM database: Öhlschläger & Võ, 2017). Later, we added unusual objects (e.g. ketchup in shower) to some scenes and removed them from others, and participants indicated whether each scene was an exact match with a studied scene. We investigated whether participants would report false memories for scenes where the unusual object was added and whether the frequency of these endorsements would differ from other false alarms (e.g., to brand-new, object-removed scenes). The hit rate for exact-match scenes was high, indicating strong memory formation. Interestingly, the false memory rate for object-added scenes was greater than for brand-new scenes. However, participants were also highly likely to false alarm to object-removed scenes, indicating a reliance on gist-based memory that overlooks incongruous objects. We also investigated the downstream fate of memories for the changed objects themselves. We found a boost to memory for objects that, via their addition or removal, violated predictions from prior memories. In contrast to previous work, we showed that false memories can be generated even for out-of-place objects, where the error cannot be attributed to scene schemas. Though participants endorsed schema violations, they were also likely to overlook these violations when the object was removed at test, highlighting the malleability of visual memory. Current follow-ups identify periods of off-task thought in real-time to trigger trials that gauge the association between susceptibility to false memory and ongoing attentional states.

Acknowledgements: This work was supported by an NSERC Discovery Grant.