A Core-Semantic-Feature Scoring System for Drawing-Based Augmentative and Alternative Communication: Development and Initial Psychometric Evaluation in Mandarin-Speaking Adults with and Without Aphasia

Background/Objectives: Drawing is a long-established non-verbal channel of augmentative and alternative communication (AAC) for people with aphasia (PWA), yet drawings are usually scored only as recognised or not. Using the symbol-oriented (SO)/realistic-oriented (RO) framework, we adapted 30 drawing objects for Mandarin Chinese and derived an object-level rubric from 60 healthy adults’ drawings. Methods: Features depicted in ≥75% of an object’s drawings were its core semantic features (53: 31 SO, 22 RO); semantic accuracy was the proportion retained. Fifty-nine independent participants (20 controls drawing with the non-dominant hand, 10 left- and 17 right-brain-damaged patients without aphasia; 12 PWA) completed the paradigm; analyses comprised a mixed ANOVA, item-adjusted trial-level models, reliability (ICC, SEM, MDC95) and blind third-party identification. Results: SO drawings scored higher than RO drawings in all groups (F(1,55) = 85.40, η2p = 0.608); patient groups scored below controls (η2p = 0.230), with no interaction detected (p = 0.790). Reliability was high (ICC(2,1) = 0.938; SEM = 0.115; MDC95 = 0.318; negligible rater variance). Semantic accuracy was strongly associated with blind third-party identification in a mixed model (β = 0.656, 95% CI 0.620–0.692). Conclusions: The score quantifies the object-specific semantic information a drawing retains and yields a difficulty gradient for material selection.