Development of a training model for enhancing artificial intelligence skills of teachers in primary schools in Yunnan Province

This study aimed to (1) assess the professional-development needs of primary school teachers in Yunnan Province regarding artificial intelligence (AI) competencies, (2) develop an expert-informed 4T AI Training Model, and (3) compare achievement-test scores immediately before and after a preliminary training implementation. The sequential three-phase study included a needs survey of 400 teachers selected through a reported two-stage school-and-teacher sampling procedure, a focus-group review involving nine purposively selected specialists, and a one-group pre-test/post-test evaluation of 32 teachers. Instruments included a five-point needs-assessment questionnaire, a focus-group discussion form, and a 50-item achievement test. Descriptive statistics, the modified Priority Needs Index (PNI_modified), content analysis, and a paired-samples t-test were used. The overall current-condition rating was low (M = 2.37, SD = 0.65), whereas the desired-condition rating was high (M = 3.99, SD = 1.12); the reported overall PNI_modified was 0.68. Priority needs included AI Collaboration, Classroom Management with AI, Problem-Solving with AI Tools, AI Assessment and Evaluation, and Instructional Integration. The 4T model comprised Training Content, Training Methods, Training Resources, and Training Evaluation, structured across Basic, Application, and Development stages. Overall achievement-test scores rose from a reported pre-test mean of 30.69 (SD = 1.64) to a post-test mean of 41.19 (SD = 2.95), t(31) = 25.36, p < .001. This represents higher short-term test scores after training; without a comparison group or delayed follow-up, the findings do not establish a causal effect or sustained improvement in classroom practice.