<![CDATA[Digital mathematics learning may increase cognitive load when interdependent text, symbols, diagrams, and solution steps are presented separately, while audio may introduce redundancy and transient information effects. This study synthesized empirical evidence on the split-attention and modality effects, identified conditions influencing these effects, and derived implications for digital mathematics learning design. A Systematic Literature Review (SLR) was conducted following PRISMA 2020, with narrative synthesis guided by SWiM. Searches across Scopus, DOAJ, Google Scholar, and ScienceDirect identified 35 records. After removing 10 duplicates and screening titles and abstracts, 22 articles underwent full-text assessment, of which 11 primary studies met the eligibility criteria. The findings show that studies were predominantly conducted in Indonesia, frequently employed research and development designs, and commonly used multimedia, while direct measurement of cognitive load remained limited. Empirical evidence specifically addressing split-attention and modality effects in digital mathematics learning was also scarce. Available evidence suggests that spatial-temporal integration and audio may support learning under certain conditions, but their effectiveness depends on prior knowledge, material complexity, narration pace, learner control, redundancy, and device characteristics. Thus, the modality effect should not be considered a universal principle for reducing cognitive load. Based on these findings, this study proposes the Conditional SAM-Math Model, integrating design decisions, boundary conditions, cognitive load measurement, and learning outcomes to guide more context-sensitive digital mathematics learning design.]]>
Copyrights © 2026
