Teacher Development for Creating Instructional Media Using Artificial Intelligent in Primary Schools

Abstract

This research article aimed to: 1) study teachers' needs for creating AI-powered learning media at Ban Chiang Dao School 2) develop teachers' capabilities in creating such media 3) evaluate teachers' capabilities. This is a mixed-methods research study. The target group consisted of 51 teachers and administrators from Ban Chiang Dao School, selected through purposive sampling. Three experts evaluated the appropriateness of the research instruments. The instruments used included questionnaires, appropriateness evaluation forms, teacher development project plans, PLC activity record forms, and teacher capability assessment forms, with an IOC evaluation result of 1. Data analysis was conducted using means, standard deviations, and content analysis.

The research found that:

1. Teachers have needs for creating media according to the 4M management concept: In terms of personnel, they need skills in creating text content, images, multimedia, and require mentor teachers for guidance. Regarding budget, they need continuous budget support. For materials and equipment, they need additional media storage systems. In terms of management, they need learning exchange networks and media evaluation systems.
2. The teacher development process involved designing practical training workshops focusing on AI tools including Perplexity (text), Midjourney (images/multimedia), Botnoi Voice (audio), and Kling AI (video). The project was then evaluated by experts for appropriateness, with results showing the project was highly appropriate. The school subsequently implemented the project for 9 weeks, finding that PLC activities systematically promoted teachers' knowledge, understanding, and problem-solving abilities in media creation.
3. Teacher capability evaluation results showed that assessment based on systems theory across all three aspects revealed teachers had overall capabilities at the highest level. Input factors had a mean of 4.72, and output had a mean of 4.58, both at the highest level, while processing aspects (mean 4.23) were at a high level.