Abstract
Event Segmentation Theory (EST) explains the perceptual organization of an ongoing activity into meaningful events. Classical event segmentation task involves watching an online video and indicating with keypress, the event boundaries i.e., when one event ends and the next one begins. The resulting hierarchical organization of object-based coarse-events and action-based fine-events provide insight about various cognitive processes. We studied event perception in the domain of assistance and training systems development for assembly workers in industrial settings. We investigated event segmentation performance of 32 intellectually-disabled people (mean IQ = 64.4) compared to 30 controls (IQ>100). Stimuli: One "breakfast-making" video and three assembly task videos. 38% Intellectually-disabled participants defined more coarse-events than fine-events, indicating misconception of higher- and lower-level content. The remaining 62% showed diminished hierarchical alignment compared to controls. We suggest that simple event segmentation task can serve as diagnostic assessment for intellectually disabled workers' cognitive potential. Furthermore, we investigated how repeated practice of sequential assembly tasks in virtual training influences learning of the task's coarse and fine assembly steps in car door assembly for domain experts (N = 18) and novices (N = 19). Stimuli: A video of car door assembly task. The video stopped at time points associated with the coarse-event boundaries; either before a coarse-event boundary – this tested memory for coarse-events - or after a coarse-event boundary – this tested memory for fine-events. We found virtual training enhances ability to predict object-based coarse-event from nearest fine-event boundary. It did not improve memory for action-based fine-event from the nearest coarse-event boundary. Expertise had a positive effect on memory for fine assembly steps. Our research in event perception can help improve the quality of life of workers at various levels- experts, new hires, and intellectually-disabled workers, by identifying gaps in existing assistance systems for the different user groups.
Meeting abstract presented at VSS 2016