Multimodal 3D Supports Improve Observational Drawing from Life: A Quasi-Experimental Study of First-Year Art Students

Abstract

The persistent challenge of transferring Descriptive Geometry and Perspective (DGP) knowledge into practical control action in the Basics of Academic Drawing (BAD) arises within first-year art teacher education. Science has proven that students can learn geometric and perspectival principles conceptually, yet be unable to apply them consistently when creating observational drawings from life. To address this issue, this exploratory quasi-experimental study examined whether improved drawing process control and product quality, leading to reduced common structural errors, could be associated with an integrated instructional package comprising AutoCAD-based 3D supports, multimodal prompts, and the Integrated Control Operations Checklist (ICO_norm). A total of 46 first-year students from two art education programs completed a progression of three syllabus-oriented still-life modules with increasing complexity. Participants in the experimental group received the combined 3D multimodal support package, and those in the control group received standard BAD instruction. Three independent raters, blinded to condition, evaluated drawing outcomes using the IQD-5 rubric (Indicator of the Quality of Drawing), a binary error matrix, and ICO_norm coding for control operations during S1-S3. Across M1-M3, ICO_norm completion, IQD-5 scores, and error reduction, these same results showed larger gains in the experimental group than in the control group. Exploratory mediation analysis indicated that changes in integrated control operations were statistically consistent with a process-mediated improvement in drawing quality and error reduction. However, because this study used a non-randomized design, a relatively small sample size, and a bundled intervention, the findings should be considered preliminary evidence requiring replication and component-level analysis. Overall, the study conceptualizes multidimensional 3D scaffolding as a Self Regulated Learning (SRL) mechanism, in which planning, monitoring, correction, and verification connect digital supports to drawing quality and error reduction.