Abstract
Asynchronous online engineering education requires instructional materials that present complex mathematical and visuospatial concepts without exceeding students' working memory capacities. Grounded in Dual-Coding Theory and Cognitive Load Theory, this study examined the interactive effects of visual modality presentation (annotated static diagrams vs. dynamic step-by-step animations) and auditory narration pacing (standard continuous delivery vs. student-regulated segmented narration) on learning outcomes and cognitive load. A 2 × 2 factorial experimental design was conducted with undergraduate engineering students (N = 184) enrolled in an asynchronous thermodynamics module. Participants completed pre-tests, experimental learning sessions, post-test retention and structural transfer assessments, and multidimensional cognitive load surveys. Quantitative analysis revealed a statistically significant interaction effect between visual modality and auditory pacing on transfer performance and extraneous cognitive load. Dynamic animations combined with student-regulated segmented narration yielded the highest transfer scores and lowest reported extraneous mental effort. Conversely, continuous unsegmented audio paired with dynamic animations induced severe cognitive overload, illustrating the visual transient information effect. Static diagrams performed optimally under continuous narration compared to segmented delivery. These findings demonstrate that dual-coding optimization in engineering e-learning depends critically on temporal auditory regulation to prevent cognitive friction during complex visual processing. Practical guidelines for asynchronous STEM course design are provided.