Applying Artificial Intelligence into Early Childhood Math Education: Lesson Design and Course Effect

Abstract

A growing trend has been observed in applying artificial intelligence (Al) for educational purposes. With emerging initiatives in adapting Al in early childhood education, greater effort can be made in providing empirical evidence to theoretically support the use of Al techniques for young children. This study aims to introduce an instructional unit of math instruction for children aged 3-8. The course was delivered by a learning app, Zebra Al, where an intelligent tutoring system was applied. Students were invited from social networking sites and 66 children from India agreed to participate. A four-week course (five lessons each week) was delivered, covering basic math concepts including recognizing patterns, identifying 2D shapes, dissecting 2D shapes, and counting. The effect of the course was assessed in both cognitive and attitudinal facets. The results indicated that the course was effective in improving math performance and sustaining math attitudes.