The mosaik model: an expert-validated model integrating toulmin argumentation and multimodal literacy for vocational writing instruction

Abstract

This study develops and validates the MOSAIK Model, a five-phase digital-literacy-based learning model for argumentative multimodal writing among vocational high school students. Named after the Indonesian word for “mosaic,” the model integrates verbal, visual, and digital elements into coherent, evidence-based arguments. Using the ADDIE framework, the study covered the analysis, design, development, and implementation stages, with data collected through writing tests, questionnaires, expert validation, and classroom observations. The analysis of 373 students showed a generally good baseline writing ability (M = 14.60, 73.0%), although evidence-based argumentation was the weakest component. This indicated the need for a structured instructional model rather than remedial teaching. A digital-literacy self-efficacy questionnaire demonstrated good reliability (Cronbach's α = 0.849). Expert validation of the MOSAIK instructional products yielded a mean feasibility score of 86.5% (“Highly Feasible”), with inter-rater agreement ranging from weak to good (ICC = 0.32–0.66). The model was implemented in limited (n = 67) and broad (n = 306) trials. The broad trial revealed significant differences in implementation across study programs, F(2, 303) = 20.44, p < .001, η² = 0.12, while multimodal production remained the weakest phase. Overall, the findings indicate that the MOSAIK Model is a feasible and structurally valid instructional model, although further refinement is needed to strengthen multimodal production support and standardize group assessment procedures.