University Entrepreneurship: AI-Based Teaching to Enhance Advertising Communication Competence of the MSME Community Outreach: Sequential Exploratory Mixed Methods

Abstract

This study evaluates the role of university entrepreneurship in training the use of Artificial Intelligence to improve the skills of the Micro, Small, and Medium Enterprises (MSMEs) community in developing mass communication through advertising. Using an explanatory mixed-methods approach involving MSMEs in Sukoharjo, this study applies the Kirkpatrick four-stage model in four phases. Phase 1 (Quantitative) used surveys measuring Reaction, Learning, and Behavior (Cici AI, Canva, Leonardo AI, Suno AI), and PLS-SEM for structural analysis. Phase 2 (Qualitative) utilized semi-structured interviews and observations, analyzed through thematic analysis. Phase 3 (Integration) used joint display analysis to connect findings. The Wilcoxon test showed significant improvement across all participants. PLS-SEM confirmed that AI Usability significantly predicts Advertising Competency and Training Effectiveness, with all constructs meeting validity criteria. Qualitative analysis identified four explanatory themes: accessibility (zero design-entry for audio/text AI versus skill barriers for visual AI), confidence transformation, infrastructure constraints, and text/audio AI as collaborative partners versus visual AI requiring expertise. The sequential explanatory design provides both statistical evidence that AI training significantly improves MSME competency and mechanistic understanding of why. Audio and text AI tools suit MSME beginners best, while visual AI requires extended training. This framework offers a template for evidence-based training evaluation in MSME empowerment programs.