A Self-Efficacy Theory-Based Study on the Teachers’ Readiness to Teach Artificial Intelligence in Public Schools in Sri Lanka

Objectives . This article explores teacher readiness for introducing artificial intelligence (AI) into Sri Lankan schools, drawing on self-efficacy theory. Similar to some other countries, Sri Lanka plans to integrate AI into the school curriculum soon. However, a key question remains: Are teachers prepared to teach this advanced technical subject to schoolchildren? Assessing teacher readiness is crucial, as it is intricately linked to the overall success of this initiative and will inform the development of appropriate policies. Participants . This study surveyed over 1,300 Sri Lankan public school teachers who teach Information and Communication Technology (ICT) using the snowball sampling approach. The respondents represent approximately 20% of the total ICT teacher population in Sri Lanka. Their readiness to teach AI was assessed using a general teacher self-efficacy scale specifically developed based on the well-established Self-Efficacy Theory. While key demographic factors like gender, education level and educational background were also collected, their impact analysis is not included in this article. Study Method . The study was conducted based on the premise that teachers’ readiness to teach AI hinges on their self-efficacy towards teaching AI in the classroom. This premise was substantiated through a review of existing research, and a conceptual model of teachers’ self-efficacy for AI instruction was developed. To assess this model, a nationwide survey targeting school ICT teachers was conducted. The questionnaire used in the survey was based on existing research on evaluating teacher self-efficacy. Data analysis, focusing on testing and validating the conceptual model, primarily employed the PLS-SEM approach. Findings . This study identified several key findings: (1) Teachers generally reported low self-efficacy regarding their ability to teach AI; (2) Teachers’ self-efficacy was most influenced by their emotional and physiological states, as well as their imaginary experiences related to teaching AI; (3) Surprisingly, mastery experiences had a lesser impact on their self-efficacy for teaching AI; and (4) Neither vicarious experiences (observing others teach AI) nor verbal persuasion had a significant impact on teachers’ self-efficacy. Additionally, the study revealed that the teachers’ own level of expertise in AI, along with their social capital, is insufficient to deliver a standard AI curriculum. Conclusions . The analysis of the results found that Sri Lankan teachers currently lack the readiness to teach AI in schools effectively. Potential lapses in certain sources of self-efficacy were also identified. It further revealed the need for a more systemic approach to teacher professional development. Therefore, the study recommends further research exploring the potential of incorporating a socio-technical systems perspective into the government’s teacher training initiatives.