A cross-sectional survey investigation of secondary school information technology teachers in China has demonstrated that an extended behavioral model explains 64.2 percent of the variance in educators’ intentions to teach artificial intelligence. Researchers examined how cognitive, environmental, and emotional factors interact to influence classroom instructional intentions as artificial intelligence education expands rapidly within K–12 educational systems. The empirical study utilized covariance-based structural equation modeling alongside supervised machine learning classification analyses to rigorously evaluate the hypothesized structural relationships and predictive patterns.
Amid the rapid global expansion of artificial intelligence, AI education has become an increasingly integral component of modern K–12 educational systems. Teachers’ instructional intentions are recognized as instrumental in shaping classroom implementation and guiding long-term professional development trajectories. However, empirical research has previously paid limited attention to teachers’ perspectives within artificial intelligence education contexts. In response to this specific research gap, the authors expanded the classical Theory of Planned Behavior framework by integrating artificial intelligence literacy, facilitating conditions, and artificial intelligence anxiety into an integrative motivational model.
Using cross-sectional survey data collected from 218 secondary school information technology teachers, the researchers employed covariance-based structural equation modeling to test their hypothesized theoretical relationships. In addition, machine learning classification analyses were conducted to examine the predictive performance of the theoretically relevant variables. This dual methodological approach enabled the investigators to evaluate whether statistically significant associations observed across aggregate survey responses also translated into stable, dependable individual-level classifications when tested with machine learning algorithms.
The resulting structural equation model demonstrated that the extended Theory of Planned Behavior explained 64.2 percent of the variance in secondary teachers’ intentions to teach artificial intelligence. The analysis established that teacher attitude, subjective norms, and perceived behavioral control were each significantly and positively associated with behavioral intention. Furthermore, subjective norms were positively associated with teacher attitude, whereas artificial intelligence anxiety was negatively associated with attitude. At the same time, facilitating conditions showed a strong positive association with perceived behavioral control across the sample.
The findings revealed a nuanced and complex role for artificial intelligence literacy among the surveyed information technology instructors. Statistical modeling showed that artificial intelligence literacy was not directly associated with either teacher attitude or perceived behavioral control. However, literacy was significantly associated with subjective norms and exhibited a significant positive association with artificial intelligence anxiety.
The supplementary machine learning classification analyses provided important empirical boundaries regarding how well theoretical variables forecast individual teacher outcomes. In out-of-sample testing, the machine learning models achieved moderate to good discrimination when predicting behavioral intention, teacher attitude, perceived behavioral control, and subjective norms. Conversely, the classification models exhibited near-chance performance when evaluating artificial intelligence anxiety. These machine learning findings demonstrated that statistically significant structural associations identified in large-sample structural equation models do not necessarily translate into stable individual-level predictive accuracy.
In their conclusions, the investigators emphasized that future teacher professional development initiatives must address more than technical knowledge and subject-matter skills. Effective educational interventions must simultaneously attend to teachers’ emotional concerns and their perceived environmental support.
Subject of Research: Factors influencing secondary school teachers' intentions to teach artificial intelligence
Article Title: This study examines the factors associated with Chinese secondary school teachers’ intentions to teach artificial intelligence using structural equation modeling and machine learning
Article References: Gu, Y., Li, R., Yu, J., Pan, Z., & Pan, Y. (2026). This study examines the factors associated with Chinese secondary school teachers’ intentions to teach artificial intelligence using structural equation modeling and machine learning. BMC Psychology. https://doi.org/10.1186/s40359-026-05791-w
Image Credits: AI Generated
DOI: 10.1186/s40359-026-05791-w
Keywords: Theory of Planned Behavior, artificial intelligence education, secondary school teachers, structural equation modeling, machine learning classification, AI literacy, AI anxiety, facilitating conditions, perceived behavioral control, subjective norms, teacher attitudes, instructional intentions, informed anxiety, information technology teachers, teacher professional development, out-of-sample prediction, K-12 education
Cite Scienmag News
Blake Davidson. (November 19, 2026). Extended Behavioral Model Explains Teachers’ Intentions to Teach Artificial Intelligence. Scienmag. https://scienmag.com/extended-behavioral-model-explains-teachers-intentions-to-teach-artificial-intelligence/
Blake Davidson. "Extended Behavioral Model Explains Teachers’ Intentions to Teach Artificial Intelligence." Scienmag, 19 November 2026, https://scienmag.com/extended-behavioral-model-explains-teachers-intentions-to-teach-artificial-intelligence/. Accessed 22 November 2026.
Blake Davidson. "Extended Behavioral Model Explains Teachers’ Intentions to Teach Artificial Intelligence." Scienmag. November 19, 2026. https://scienmag.com/extended-behavioral-model-explains-teachers-intentions-to-teach-artificial-intelligence/








