From Acceptance to Satisfaction: Understanding AI-Driven Chatbots on English Learning Through Technology Acceptance Model (TAM) and Self-Determination Theory (SDT)
AI developers are encouraged to design chatbots that can enhance EFL learners’ engagement, autonomy, competence, and relatedness for sustainable adoption and effective learning outcomes.
Abstract
Adopting AI-driven chatbots in English as foreign language (EFL) learning has significantly gained scholarly attention. However, studies primarily adopted technology acceptance model (TAM) and self-determination theory (SDT) separately. An integration of both theoretical models is crucial to develop a comprehensive understanding, specifically the roles of technological perceptions and psychological needs. A cross-sectional online survey was conducted among 552 EFL learners who were at least aged 18 and undergraduate students using AI chatbots at least once a week for over 10 weeks. Covariance-based structural equation modeling (CB-SEM) was implemented to test the hypothesised relationships. Findings revealed that perceived ease of use (PEOU) is a significant predictor of perceived usefulness (PU) but not a significant predictor of attitude (AT). PU predicts AT significantly. Subsequently, AT serves as a significant predictor of behavioural intention (BI), while PU does not. For mediating effect, BI is a significant mediator for the relationship between AT and actual use (AU). Need satisfaction and need frustration are the significant outcomes of AU. AI developers are encouraged to design chatbots that can enhance EFL learners’ engagement, autonomy, competence, and relatedness for sustainable adoption and effective learning outcomes.
The rapid integration of conversational AI tools in higher education necessitates a deeper understanding of learners’ sustained engagement with chatbot systems. This study investigated the cognitive, affective, and value-based determinants of higher education learners’ intention to continue using AI chatbot assistance with the extended Technology Continuance Theory (TCT). Data were collected from 600 Indian higher education learners and analyzed using Covariance-based structural equation modeling (CB-SEM). The results demonstrate that perceived usefulness (PU) and perceived ease of use (PEU) play central roles in shaping learners’ attitudes (ATT), satisfaction (SAT), and continuance intention (CI). Confirmation (CON) emerged as a critical post-adoption factor, significantly influencing PU, SAT, and affective support (AS). Notably, AS and academic value (AV) contributed uniquely to learners’ CI, highlighting the AI chatbot’s socio-emotional and instrumental dimensions. ATT and SAT were also identified as strong predictors of sustained usage, and AI Self-Efficacy (AISE) positively influenced PEU. The model explained substantial variance in key endogenous constructs with CI (
R
2
= 0.740), thereby underpinning the model’s strong explanatory power. The findings extend TCT by integrating affective and academic value dimensions within chatbot-mediated learning contexts. Implications emphasize responsible pedagogical integration, balancing efficiency and emotional support with critical engagement, transparency, and meaningful human interaction in higher education.
K. Kavitha, V. P. Joshith· Journal of educational compu...· 0 citations
Feedback is a core mechanism of formative assessment, yet its effect on learning depends on how students conceive of and act upon it. This study tests an AI-enhanced student conceptions of feedback model in English as a Foreign Language (EFL) writing, relating six feedback conceptions to academic self-efficacy and self-regulation. Survey responses from 300 undergraduate students at a state university in Indonesia were analysed with partial least squares structural equation modelling (PLS-SEM) using 5,000 bootstrap subsamples. The measurement model met every reliability and validity criterion: outer loadings ranged from 0.743 to 0.888, composite reliability from 0.857 to 0.931 and average variance extracted from 0.609 to 0.750, while the Fornell-Larcker criterion, cross-loadings and HTMT ratios all confirmed discriminant validity. All seven structural paths were significant. Active use of AI feedback was the strongest antecedent of self-efficacy, followed by enjoyment, whereas ignoring feedback exerted a significant negative effect. Self-efficacy in turn strongly predicted self-regulation and accounted for 43.5% of its variance. The results indicate that what students do with feedback matters more than the source from which it originates, and that self-efficacy is the pivotal mechanism translating feedback experience into self-regulated learning.
Virgiawan Adi Kristianto, Sucipto, W. Sumbodo et al.· MATEC Web of Conferences· 0 citations
Artificial Intelligence (AI) chatbots are increasingly reshaping learning practices in higher education. This study extends the Technology Acceptance Model (TAM) by incorporating critical thinking as a key cognitive antecedent of behavioral intention to use AI chatbots for learning. Using Partial Least Squares Structural Equation Modeling (PLS-SEM), data collected from Vietnamese university students (n = 100) were analyzed to examine the proposed relationships. The findings indicate that critical thinking is predictively associated with students’ self-reported behavioral intention to use AI chatbots and shows a larger standardized association than traditional TAM factors, including perceived usefulness, perceived ease of use, and attitude. Attitude toward AI chatbot use also directly predicts behavioral intention and is positively associated with critical thinking, suggesting a potential mediation mechanism. The extended model demonstrates substantial explanatory power in explaining variance in students’ self-reported behavioral intention. Overall, the results highlight critical thinking as an important cognitive factor associated with more reflective and responsible use of AI chatbots in university education. Given the cross-sectional and exploratory design, the findings should be interpreted as predictive rather than causal.
My-Duyen Thi Nguyen, Diep-Ngoc Le· International Journal of Inf...· 0 citations
The proliferation of Generative AI (GAI) in K-12 education makes AI Explanation (AIE) style a critical yet underexplored factor in learning engagement. Grounded in Dual-Process Theory and Cognitive Load Theory, this study examines how Perceived Technical Explanation (PTE) and Perceived Analogical Explanation (PAE) affect Learning Behavioral Intention (LBI) via Germane Cognitive Load (GCL), Self-Efficacy (SE), and Learning Interest (LI), and whether Subject Preference (SP) moderates these effects. Using a single-group dual-stimulus perception design, 258 Grade 8 students across three Chinese middle schools simultaneously received both AIE styles and rated their perceptions under joint exposure conditions. Data were analyzed via PLS-SEM. PTE significantly predicted LBI directly and indirectly via SE. PAE influenced LBI only through LI. GCL did not predict LBI. SP moderated the PTE-LBI path but not PAE-LBI. Findings offer preliminary evidence that subject preference conditions the effectiveness of technical explanation on behavioral intention, suggesting that AIE style differentiation warrants consideration in GAI system design for K-12 learners.
Zhou Jin, Zhongtian Lv, Yingxin Li et al.· Acta Psychologica· 0 citations
Theoretically, the current research offers a new lens through which accelerated technology adoption can be analysed and guides tailored digital sustainability interventions in Confucian models of education.
Jingyi Li, Lianyun Huang, F. Furuoka· Frontiers in Psychology· 0 citations
Growing use of generative AI technologies like ChatGPT has changed online learning and increased student motivation. This study explores online learning motivation and ChatGPT using Self-Determination Theory (SDT) to examine competence, autonomy, and relatedness in online learners. 189 academics from various fields participated in a quantitative survey. A five-point Likert scale-based 52-item questionnaire was derived from Ryan and Deci (2000), Fowler (2018), and Youssef et al. (2024). Competence, autonomy, and relatedness were not gender-specific across academic groupings. In the descriptive study, students rated the AI system's function in critical thinking, academic accomplishment, engagement, and learning motivation positively. The greatest competency item was students' practice of cross-checking ChatGPT knowledge with independent study (M = 4.06), whereas the most autonomous item was achieving good grades (M = 4.59). Relatedness was strong in social engagement and teacher support. They liked class discussions (M = 4.00) and found course materials meaningful (M = 4.28). Positive correlations were found between competence, autonomy (r =.550, p <.001), and competence and relatedness (r =.551, p <.001). The results support the Self-Determination Theory as a valid framework for online learning motivation and show that ChatGPT can promote learners' competence, autonomy, and relatedness if responsibly integrated into online learning settings. The work has major theoretical, pedagogical, and practical consequences for higher education AI-assisted learning.
E. S. Mohandas, Aini Faridah Azizul Hassan, Nor Azyyati Md Saad et al.· International journal of res...· 0 citations