Schedule of Events | Search Abstracts | Symposia | Invited Symposia | Poster Sessions | Data Blitz Sessions

Poster E31

Quantifying schemas in future narratives

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

Isaac Kinley1,2 (isaac.kinley@gmail.com), Yang Xu3,4, Reece Roberts5,6, Donna Rose Addis1,2,5,6,7; 1Rotman Research Institute, Baycrest Health Sciences, 2Data Sciences Institute, University of Toronto, 3Department of Computer Science, University of Toronto, 4Cognitive Science Program, University of Toronto, 5School of Psychology, The University of Auckland, 6Centre for Brain Research, The University of Auckland, 7Department of Psychology, University of Toronto

Schemas are mental representations of common structures of experience, and their importance in human memory has been recognized since the time of Bartlett. Recently, it has been proposed that schemas also play a role in structuring our imagination of the future. However, tools for automatically measuring the schematic content of written and spoken event narratives are underdeveloped. We report a preliminary investigation on the development of quantitative methods for measuring schematicity in people’s narratives about the future. We analyzed narratives of imagined future events in an experimental paradigm designed to induce differences in schematic content (Roberts et al., 2017). Specifically, healthy young adults (N = 30) each imagined 8 novel events involving a person, place and object from either the same autobiographical schema (e.g., “working at company A”; congruent condition) or from three distinct schemas (incongruent condition). We found that metrics including dispersion in semantic embedding space (computed using the Universal Sentence Encoder; Cer et al, 2018) and average distances in word association networks (created using the Small World of Words dataset; De Deyne et al., 2019) differentiated between narratives from these two conditions. Our finding suggests that schemas manifest in narratives that are relatively compact in both semantic and associative spaces. These results highlight how quantitative tools can help us understand schema scaffolding in the construction of imagined events, which carries the potential for applications to other cognitive paradigms and patient populations.

Topic Area: LONG-TERM MEMORY: Episodic

 

CNS Account Login

CNS2024-Logo_FNL-02

April 13–16  |  2024