Large Language Models in Education: Insights from Nigerian In-Service Teachers

Strategic Insights -

A recent study published in F1000Research has shed light on the adoption of Large Language Models (LLMs) like ChatGPT in education, particularly in resource-constrained settings such as Nigeria. The research, led by Kayode A. Adewale and colleagues from Tai Solarin University of Education, explored the perceptions and intentions of 260 Nigerian in-service teachers regarding ChatGPT after participating in structured training. The study employed a hybrid approach, combining Partial Least Squares Structural Equation Modelling (PLS-SEM) and Artificial Neural Networks (ANN), to examine the factors influencing the teachers' behavioral intention to adopt ChatGPT.

Key Takeaways:

  • The study identified Perceived Usefulness (PUC), Technology Anxiety (TA), Your Colleagues and Your Use of ChatGPT (YCC), and Perceived Ease of Use (PEU) as significant predictors of behavioral intention, explaining 15.8% of the variance in BI.
  • The ANN analysis highlighted PEU, Attitude Towards ChatGPT (ATC), and PUC as the most critical factors, demonstrating substantial predictive accuracy with an RMSE of 0.87.
  • The research concluded that targeted professional development initiatives are necessary to enhance teachers' digital competencies, reduce technology-related anxiety, and build trust in AI tools like ChatGPT.
  • The study offers actionable insights for policymakers and educational stakeholders, emphasizing the importance of fostering an inclusive and ethical AI ecosystem.
  • The research aims to empower teachers and support AI-driven educational transformation in resource-limited environments by addressing contextual barriers.

Statistics:

  • 260 Nigerian in-service teachers participated in the study.
  • The study employed a hybrid approach, combining PLS-SEM and ANN, to examine the factors influencing the teachers' behavioral intention to adopt ChatGPT.
  • The PLS-SEM results explained 15.8% of the variance in BI.
  • The ANN analysis demonstrated substantial predictive accuracy with an RMSE of 0.87.
  • The study identified four significant predictors of behavioral intention: PUC, TA, YCC, and PEU.

Sources:

  • Adewale, K. A. et al. (2025). Large language models and GenAI in education: Insights from Nigerian in-service teachers through a hybrid ANN-PLS-SEM approach [version 1; peer review: 2 approved]. F1000Research, 14.
  • NewsRx (2025). Researchers at Tai Solarin University of Education Target Life Sciences (Large language models and GenAI in education: Insights from Nigerian in-service teachers through a hybrid ANN-PLS-SEM approach [version 1; peer review: 2 approved]). Health & Medicine Week. May 23, 2025; p 4421.