Multimodal configuration in Indonesia’s online English Academic Ability Test: Implications for L2 reading assessment

Digital L2 reading assessment increasingly combines written language with images, spatial organization, and interface controls, yet the contribution of these resources to the reading construct remains underexamined in national online assessment. This qualitative Multimodal Discourse Analysis examines Indonesia’s official Tes Kemampuan Akademik (TKA; Academic Ability Test) simulation platform, using 105 screenshots representing 30 English items (66 screenshots/20 Compulsory items; 39 screenshots/10 Advanced items). Kress and van Leeuwen’s visual grammar guides analysis of representational, interactive, and compositional meanings, alongside coding of linguistic, visual, spatial, gestural/interactional, and aural resources. To address unequal corpus sizes, cross-level frequency summaries are normalized within each screenshot corpus. Linguistic resources occur throughout both corpora, and spatial organization is similarly prevalent (89.4% Compulsory; 87.2% Advanced). Visual resources are proportionally more prevalent in Advanced screenshots (69.2% vs. 28.8%), as are gestural/interactional resources (38.5% vs. 16.7%). Qualitative analysis indicates that Compulsory items more often use multimodal resources as representational scaffolding, whereas Advanced items more often foreground interactive positioning and compositional integration. As the study contains no test-taker process or performance data, these patterns are interpreted as design-implied reading demands rather than measured cognitive effects. The findings have implications for L2 reading pedagogy, teacher assessment literacy, multimodal literacy, and construct-focused digital test design.