Effects of Construction Safety Signs and Training Interventions on Memory Encoding: Evidence from Neuroscientific Data

Abstract Construction safety signs are critical for accident prevention, but their effectiveness often is limited due to insufficient comprehension and memory of these signs among individuals. Existing research has paid insufficient attention to the neural mechanisms of memory encoding for construction safety signs and lacks targeted training designs. This study determined the memory encoding mechanisms of construction safety signs (concrete versus abstract) and investigated how task types (image-based versus text-based) and training interventions affect safety sign memory. An event-related potential (ERP) experiment combining within-subject (sign type and task type) and between-subject (training intervention) factors was conducted, capturing both behavioral (accuracy and reaction time) and electrophysiological data (the ERP components N200 and P300). The training intervention was designed based on memory encoding theory and evaluated whether scenario-based learning could enhance cognitive processing of safety signs. The behavioral results showed that concrete signs significantly outperformed abstract signs in memory encoding efficiency, response time, and recognition accuracy; similarly, image-based tasks were superior to text-based tasks. The ERP results revealed that concrete signs and image-based tasks elicited greater N200 amplitudes and smaller P300 amplitudes, indicating deeper memory encoding and lower cognitive resource consumption. Training interventions significantly improve memory encoding for construction safety signs. This study theoretically provides the knowledge regarding the neural mechanisms of construction safety sign memory encoding and the cognitive impacts of scenario-based training interventions. Practically, the findings offer potential implications for designing more-effective safety signs and training programs, which may contribute to improving safety outcomes in construction settings.