Sage Journals: Discover world-class research

Abstract

This study examines the factors influencing distance education and online English learning in AI-driven education by integrating flow experience theory with the Technology Acceptance Model. Using a quantitative research approach, data was collected from 289 learners using AI-assisted platforms and analyzed through structural equation modeling (SEM). The findings revealed that flow was significantly associated with continuous intention. All the antecedents, including intrinsic motivation, immediacy, and feedback, significantly impacted flow, except telepresence and competition. Furthermore, flow significantly impacted online engagement, perceived ease of use, academic performance, and perceived usefulness. Lastly, online engagement, perceived usefulness, academic performance, and perceived ease of use significantly impacted CI. These findings offer theoretical insights into digital learning engagement and provide practical implications for designing AI-enhanced educational platforms that foster sustained learner commitment through optimized user experiences, personalized feedback mechanisms, and instructional strategies that promote flow.

Plain Language Summary

This research examines how students remain engaged with AI-powered English learning platforms and what motivates them to continue using these tools for learning. The study examined 289 students who use AI-assisted learning platforms to understand what motivates them and enables them to perform well. The researchers found that when students experience "flow" - a state in which they are entirely absorbed and engaged in their learning - they're more likely to continue using these platforms. This flow state is triggered by students' internal motivation to learn and how quickly they receive feedback from the system. Interestingly, while factors such as competition and virtual presence weren't as important as expected, the study revealed that when students experience flow, they find the platforms easier to use, perform better academically, and are more engaged with their learning. The research offers valuable insights for educators and developers designing AI-powered learning platforms, suggesting that they should focus on creating experiences that promote a state of flow through personalized feedback and engaging content.

Keywords

flow experience online engagement academic performance technology acceptance model Distance education and online learning

Introduction

Advancements in artificial intelligence (AI)-assisted learning technologies have transformed digital education, significantly enhancing accessibility, engagement, and effectiveness in higher education (C.-T. Chen, Chen, et al., 2024). Integrating AI-powered learning systems into educational platforms has helped mitigate traditional learning constraints, including time limitations, geographical barriers, and disparities in learning pace (Onesi-Ozigagun et al., 2024). Given the flexibility and adaptive learning opportunities offered by these technologies, AI-assisted learning has become an increasingly preferred instructional approach across universities worldwide. As digital knowledge drives the modern education economy, academic institutions leverage AI-driven platforms to disseminate knowledge, personalize learning experiences, and foster lifelong learning opportunities (X. Wang et al., 2024). Despite these advantages, empirical research indicates that AI-powered learning platforms face persistent challenges, particularly high dropout rates and fluctuating learner engagement levels (Onesi-Ozigagun et al., 2024).

The flow experience (FLE), a psychological state characterized by deep immersion, heightened concentration, and optimal engagement, is influenced by several key antecedents, including intrinsic motivation (IM), immediacy (IMME), competition (COMP), telepresence (TP), and feedback (FB). These factors collectively shape the conditions for learners to enter and sustain a FLE state, particularly in AI-assisted online learning environments. IM, which involves engaging in learning activities for the inherent satisfaction they provide rather than for external rewards, is fundamental to fostering sustained engagement and cognitive absorption (Sadoughi & Eskandari, 2024). IMME, defined as the psychological closeness learners feel toward their educational content and instructors, has enhanced perceived engagement and interactivity, critical precursors to FLE (T.-Y. Chen, Li, & Yang, 2024). When structured appropriately, COMP is a motivational driver that introduces challenging yet attainable goals, encouraging active participation and perseverance. Well-designed competitive elements within AI-driven educational platforms stimulate engagement and task persistence, reinforcing FLE (Y.-C. Lin & Hou, 2024). TP, or the perception of being fully immersed in a virtual learning environment, further strengthens cognitive involvement by creating an illusion of presence within the digital space, enhancing sustained attention and deep learning (Muhammad Sohail Jafar et al., 2024). Additionally, FB mechanisms clarify task expectations, reinforce competence, and help learners navigate academic challenges, fostering optimal engagement and prolonged FLE (Bae, 2014). Recognizing the pivotal role these antecedents play in shaping FLE, this study aims to examine their impact within AI-driven digital learning environments systematically. Furthermore, this research also aims to find the relationship between FLE and continuous intention (CI) to use AI-assisted learning platforms.

In the rapidly evolving domain of digital education, the FLE has emerged as a pivotal factor influencing learners’ online engagement (OE) to use AI-assisted learning environments (Y. Wei et al., 2024). Research indicates that learners who attain FLE demonstrate heightened interaction with AI-powered platforms, leading to greater OE (Sadoughi & Eskandari, 2024). Integrating flow-inducing elements, such as gamification, real-time FB, and immersive TP, has significantly enhanced students’ willingness for OE with online platforms, thus fostering a more profound commitment to self-directed learning (Arghashi & Yuksel, 2022). Given the persistent challenge of high dropout rates in online education, examining the relationship between FLE and OE is essential for developing more effective AI-assisted learning environments (Hamari et al., 2016). This study aims to contribute to this body of knowledge by analyzing how FLE fosters OE.

The technology acceptance model (TAM) asserts that an individual’s adoption and sustained use of technology are primarily influenced by their perceived usefulness (PU) and perceived ease of use (PEU; Davis, 1989). TAM has evolved significantly since Davis’s (1989) seminal work, particularly in AI-enhanced educational contexts (Abu-AlSondos et al., 2023; Ali et al., 2025). Recent research has demonstrated that traditional TAM constructs—PU and PEU—interact distinctly with FLE in AI-assisted environments (Ruangkanjanases et al., 2024). While classic TAM emphasizes cognitive evaluation, contemporary research reveals that AI-specific factors, such as algorithmic transparency and adaptive learning capabilities, significantly impact these relationships (Ali et al., 2025). When learners enter FLE, characterized by deep concentration and intrinsic motivation, they are more likely to view educational technology as an effective tool for enhancing learning outcomes, thereby reinforcing PU. Additionally, seamless navigation contributes to FLE, resulting in a heightened PEU, as learners find AI-powered educational tools more intuitive and enjoyable to engage with (Yan et al., 2024). Empirical studies suggest that the interplay between FLE and TAM constructs has a significant impact on learners’ CI when using AI-assisted platforms. Understanding the relationship between FLE, PU, and PEU is essential for optimizing AI-driven learning environments (Ruangkanjanases et al., 2024). Hence, this study aims to explore the relationships of FLE with PEU and PU.

Research has consistently demonstrated that students who frequently experience FLE within AI-assisted learning platforms achieve higher learning outcomes, demonstrate stronger critical thinking skills, and exhibit greater academic performance (AP; Sadoughi & Eskandari, 2024). Furthermore, FLE significantly fosters self-regulated learning behaviors, leading to enhanced AP (Y. Wei et al., 2024). The integration of adaptive AI technologies, which provide interactive, personalized, and immersive learning experiences, has been found to minimize cognitive barriers and sustain student motivation, both of which are critical for AP (Kaya & Ercag, 2023). Given these empirical insights, cultivating FLE in AI-assisted educational platforms is essential for maximizing student engagement and improving AP (Jahedizadeh et al., 2021). Hence, this research aims to examine the association between FLE and AP.

OE plays a pivotal role in shaping CI, reflecting a learner’s commitment to using AI-assisted learning platforms over time (Kumar & Kumar, 2024). OE, defined by active participation, meaningful interaction, and deep immersion in digital learning environments, contributes to positive learning experiences, reinforcing a learner’s CI to use the platform (Cheng, 2023). Empirical research highlights a strong correlation between higher OE levels and lower dropout rates in online learning environments, as learners who perceive value in their educational experience are more likely to develop CI to use learning platforms (Shao & Chen, 2021). Furthermore, AI-driven learning platforms, OE, and peer collaboration enhance learner interaction and commitment, strengthening their CI to persist in digital learning (Cheng, 2023). Understanding the relationship between OE and CI is essential for designing AI-powered learning systems that capture initial learner interest and foster long-term retention, motivation, and academic success (Kumar & Kumar, 2024). Consequently, this research aims to explore the association between OE and CI.

The TAM has been extensively utilized to examine the factors influencing CI’s use of educational technologies, particularly AI-assisted learning platforms. PU and PEU are key determinants of this framework, significantly shaping learners’ long-term engagement and technology adoption decisions (Ruangkanjanases et al., 2024). PU has a direct positive effect on CI, as students are more likely to persist in using an AI-powered system if they perceive it as beneficial to their academic success (Gani et al., 2024). Additionally, PEU plays a crucial role in shaping CI, as learners are less likely to engage with systems perceived as complex or challenging to navigate (Basuki et al., 2022). When AI-assisted learning tools feature intuitive interfaces, seamless interactivity, and user-friendly navigation, learners experience lower cognitive load, resulting in greater satisfaction and prolonged engagement (Liesa-Orús et al., 2023). Since sustained engagement with AI-driven educational platforms is critical for maximizing learning outcomes and ensuring long-term academic benefits, a deeper understanding of the interplay between PU, PEU, and CI is critical (Zhao & Khan, 2021). Hence, this research aims to explore the relationships of PEU and PU with CI. Additionally, this study will also examine the significance of PEU on PU.

AP is a key determinant in shaping learners’ CI to engage with educational technologies, particularly in AI-assisted learning environments (Maqableh et al., 2021). According to the Expectancy-Value Theory (Man et al., 2024), students are more inclined to continue using an educational platform when they recognize that their academic achievements are directly enhanced through sustained engagement. Empirical studies indicate that higher AP reinforces positive attitudes toward AI-driven learning systems, fostering a stronger CI (Wandira et al., 2024). Research in self-regulated learning further suggests that students who experience measurable AP through personalized AI-driven feedback and adaptive learning paths are more likely to maintain CI by using these platforms (Maqableh et al., 2021). Consequently, this research aims to explore the association between AP and CI.

Researchers have applied various behavioral and technology adoption theories to address the factors influencing the adoption and continued usage of AI-assisted learning. Notable frameworks such as the TAM, the expectation-confirmation model (ECM; Ruangkanjanases et al., 2024), and the theory of planned behavior (TPB; A. Khan et al., 2020) provide critical insights into learner engagement. While these models effectively explain the determinants of AI adoption, they often neglect the role of IM in sustaining long-term engagement within AI-driven e-learning ecosystems. Furthermore, CI with AI-powered platforms is critical for ensuring that students and educational institutions derive long-term benefits (Onesi-Ozigagun et al., 2024). This research significantly advances the understanding of AI-assisted learning platforms by integrating constructs from the TAM, FLE, and its antecedents, including IM, IMME, COMP, TP, and FB. It also includes AP and OE as significant mediators between FLE and CI, thereby constructing a comprehensive framework. This study uniquely investigates the interrelationships between FLE and CI by addressing existing gaps. Secondly, it investigates the impacts of antecedents of FLE on FLE. Thirdly, this research assesses the impacts of FLE on AP, OE, PEU, and PU. Lastly, this research measures the impacts of OE, PEU, PU, and AP on CI. Practically, this research provides actionable insights for optimizing AI-driven learning platforms, enhancing user engagement and academic outcomes.

Theoretical Background and Hypotheses Development

Flow and Continuous Intention

The impact of FLE on learning outcomes and CI has been widely recognized in digital education research. Learners who achieve FLE become fully immersed in the learning process, allowing them to engage deeply with the educational content (Cao et al., 2024). The level of user involvement in AI-powered learning platforms enhances concentration and cognitive absorption, leading to a more seamless learning experience (Yao et al., 2024). Several studies have employed the stimulus-organism-response (S-O-R) model to explore the relationship between learners’ FLE and their CI of learning (Gao, 2023; Zhao & Khan, 2021). It is indicated that students who experience enjoyment and satisfaction while interacting with AI-driven educational platforms demonstrate stronger learning CI and motivation (Gao, 2023). Similarly, research suggests that FLE significantly influences cognitive engagement, knowledge retention, and CI of platform usage in AI-assisted learning environments. Learners who find AI-driven learning interactive, adaptive, and stimulating are more likely to remain fully engaged and motivated to continue their online education (Zhao & Khan, 2021). Given these findings, learners are likelier to exhibit CI when using AI-assisted learning platforms if their educational experiences are engaging, immersive, and enjoyable. Positive interactions within adaptive AI-driven learning systems enhance FLE. The following hypothesis is proposed.

H1: FLE significantly impacts CI.

Antecedents of Flow

Intrinsic Motivation and Flow

IM motivation is pivotal in promoting engagement and enjoyment in AI-assisted online learning environments, particularly where direct instructor interaction and conventional classroom structures are absent (Rheinberg & Engeser, 2018). In these digital learning platforms, learners are not merely influenced by external rewards but are also driven by internal goals, intellectual curiosity, and the inherent satisfaction of acquiring knowledge (Hong et al., 2017). AI-powered educational systems further enhance IM by tailoring instructional content to meet individual learning needs and competencies (Kong & Wang, 2021). In AI-enhanced learning environments, achieving FLE has been linked to improved knowledge retention, enhanced problem-solving abilities, and overall academic success (Sadoughi & Eskandari, 2024). Empirical findings in educational technology and cognitive psychology indicate that IM is a key determinant of FLE, as it fosters an alignment between students’ interests and structured academic activities. This alignment reduces cognitive overload, supporting sustained attention and engagement (X.-M. Wang et al., 2023). By creating highly interactive, personalized, and adaptable learning experiences, AI-driven educational platforms increase FLE states’ probability, leading to higher retention rates, deeper conceptual understanding, and a more sustained commitment to lifelong learning (Hong et al., 2017; Kaya & Ercag, 2023). The following hypothesis is proposed.

H2: IM significantly impacts FLE.

Immediacy and Flow

AI-assisted learning technologies have revolutionized online education, enabling learners to access educational materials instantly, regardless of time or location. The concept of IMME pertains to the ease with which students can engage with AI-powered learning platforms as needed (Pedrero-Esteban & Barrios-Rubio, 2024). AI-driven educational systems, incorporating personalized learning pathways, automated tutoring, and real-time feedback, enhance accessibility by eliminating conventional constraints such as fixed schedules, physical classroom limitations, and instructor availability (T.-Y. Chen, Li, & Yang, 2024). Studies on digital engagement and learning motivation indicate that ubiquitous access to educational resources significantly fosters FLE states in technology-enhanced environments (Hew et al., 2024). Similarly, IMME to digital services strengthens user immersion, promoting sustained cognitive engagement. When applied to AI-assisted education, the ability to seamlessly initiate a learning session at any moment heightens the probability of FLE, allowing students to become deeply engrossed in their educational tasks (He et al., 2024). Features such as adaptive content delivery, interactive learning activities, and instant information retrieval further reinforce this effect, ensuring that learners maintain focus, engagement, and intrinsic motivation throughout their educational experience (Hew et al., 2024). The following hypothesis is proposed.

H3: IMME significantly impacts FLE.

Competition and Flow

COMP is a critical psychological mechanism that fosters cognitive engagement, motivation, and sustained learning behavior in AI-assisted online learning environments (Hew et al., 2024). Rooted in self-determination theory (Waterman et al., 2024) and goal-setting theory (Locke & Latham, 2019), COMP has been identified as a key driver of performance and persistence in technology-enhanced learning. When learners are placed in competitive learning contexts, they experience a heightened sense of challenge and achievement, which enhances IM and engagement with the learning material (Y.-C. Lin & Hou, 2024).

While AI-assisted learning environments utilize adaptive leaderboards and performance analytics, the integration of VR/AR simulations, achievement systems, and gamification elements creates a multi-dimensional competitive landscape (Alkhwaldi, 2024; bayu dani Nandiyanto & Sidik, 2026). For instance, VR-enabled competitive scenarios provide immersive experiences, while gamification platforms offer achievement badges and competitive challenges, collectively enhancing FLE (Cinar et al., 2024). This technology-enhanced competition creates an engaging environment that promotes sustained FLE (Li et al., 2024). Through gamified assessments, AI-driven leaderboards, and real-time performance analytics, AI-assisted learning platforms introduce COMP that encourages learners to remain actively engaged, ultimately fostering FLE (C.-H. Chen et al., 2018). Studies in educational gamification suggest that AI-enhanced COMP mechanisms, such as adaptive difficulty scaling and dynamic feedback systems, significantly increase the likelihood of learners achieving and maintaining FLE (Hamari et al., 2016). Furthermore, gamified AI-driven learning environments, which integrate ranking systems, achievement-based rewards, and adaptive challenges, have been found to intensify learner engagement and increase their time-on-task, ultimately reinforcing FLE (Li et al., 2024).

This suggests that when learners engage in AI-mediated COMP, they are more likely to become fully immersed in their educational activities, demonstrating higher levels of FLE (Hew et al., 2024). The following hypothesis is proposed.

H4: COMP significantly impacts FLE.

Telepresence and Flow

TP refers to the extent to which learners feel present in a digital environment despite being physically distant from an instructor or traditional classroom setting (Muhammad Sohail Jafar et al., 2024). This construct has been widely studied in human-computer interaction and educational technology, where it has been linked to higher levels of engagement, deeper cognitive processing, and improved learning performance (Doğan et al., 2024).

Within AI-driven educational platforms, TP is enhanced through interactive virtual environments, intelligent tutoring systems, and adaptive content delivery mechanisms, creating a seamless and immersive learning experience (Muhammad Sohail Jafar et al., 2024). TP demonstrates differently across various educational technologies, with each platform offering unique immersive experiences. While AI-assisted learning platforms offer adaptive interactions, VR environments foster a profound sense of spatial presence through immersion and feedback (Hilty et al., 2020). AR connections enhance real-world learning spaces with interactive digital elements. Gamification in AI-assisted learning platforms utilizes achievement systems and narrative elements to deepen engagement (Addas et al., 2024). This multi-technology approach creates a rich TP that significantly contributes to FLE by fostering deeper cognitive engagement and sustained attention (Muhammad Sohail Jafar et al., 2024).

Empirical research suggests that when learners experience a strong sense of TP, they are more likely to engage deeply with the material, sustaining focused attention and cognitive involvement, contributing to the emergence of FLE states (Q. Wang et al., 2025). Moreover, studies on technology-enhanced learning have demonstrated that TP mitigates cognitive load by creating a psychologically engaging environment, reducing distractions, and enabling learners to remain absorbed in complex learning tasks (H. Kim et al., 2023). This effect is further amplified in AI-powered learning environments, where customized AI-driven interactions and responsive digital interfaces create a dynamic and engaging educational experience (Parahiyanti & Dimara, 2024). The following hypothesis is proposed.

H5: TP significantly impacts FLE.

Feedback and Flow

Within technology-enhanced education, FB refers to the reciprocal exchange of information between learners and AI-driven learning systems, allowing for real-time guidance and self-regulation (Liu & Hwang, 2024). Through natural language processing, automated error detection, and predictive analytics, AI-powered platforms personalize learning pathways by offering timely and context-specific FB, helping learners refine their understanding and performance (Chien et al., 2024). Educational psychology and cognitive science studies highlight that well-structured FB fosters perceived competence, facilitating FLE (Buzady et al., 2024). Learners receiving prompt and targeted FB can effectively adjust their approach, set appropriate learning goals, and maintain focus, all of which are conducive to deep engagement and sustained FLE (Xu et al., 2021). Conversely, lacking FB can lead to uncertainty and disengagement, preventing learners from fully immersing in AI-assisted learning environments (Cao et al., 2024). AI-powered platforms that integrate machine learning algorithms, dynamic FB loops, and intelligent tutoring systems enable learners to develop a progressive sense of mastery, reinforcing FLE (Bae, 2014). The following hypothesis is proposed.

H6: FB significantly impacts FLE.

Flow and Online Engagement

The FLE, defined by deep immersion and sustained concentration, fosters learner OE within AI-assisted online learning environments (H. E. Park & Yap, 2024). The alignment of task complexity with learner skill levels in digital education platforms encourages optimal OE, creating an interactive and rewarding learning experience (Hsiao & Chang, 2024). The direct impact of FLE on OE is reflected in extended participation, increased cognitive involvement, and greater interaction with learning platforms (Sadoughi & Eskandari, 2024). This heightened engagement is instrumental for knowledge acquisition and skill development, particularly vital in technology-mediated learning contexts (K.-Y. Lin & Huang, 2024). Additionally, FLE generates a positive reinforcement cycle, where increased OE improves learning outcomes, motivating students to immerse themselves more deeply in the content (Sadoughi & Eskandari, 2024). Beyond increasing interaction frequency, FLE-driven OE enhances the quality of learning experiences. This depth of OE is essential for maximizing the effectiveness of AI-assisted learning systems, underscoring the need for instructional strategies that promote sustained FLE states to optimize educational outcomes (Y. Wei et al., 2024). The following hypothesis is proposed.

H7: FLE significantly impacts OE.

Flow and Perceived Ease of Use

Within AI-assisted online learning environments, the FLE, characterized by deep engagement and heightened immersion, is crucial in shaping learners’ perceptions of technology usability (S. Khan & Khan, 2021). The intense focus and IM associated with FLE may diminish users’ awareness of interface complexities, increasing their PEU toward the learning platform (Xu et al., 2021). This suggests that learners who attain FLE in AI-assisted learning environments may find the interface easier to navigate and develop greater competence in managing technological challenges (Q. Wang et al., 2025). Consequently, their PEU may be enhanced as they become more confident interacting with the system (Ruangkanjanases et al., 2024). Moreover, FLE has been associated with increased familiarity and proficiency with a platform’s functionalities, further reinforcing its perceived usability. As learners engage with the system more frequently, they develop FLE, which may lead to a progressive shift in their perception of the platform as user-friendly (Lu et al., 2022). FLE is not solely an outcome of individual engagement but is also influenced by how well the digital environment supports PEU. Once engaged in FLE, users may attribute their seamless interaction with the platform to its PEU, further reinforcing positive perceptions of the technology (Xu et al., 2021; Zhao & Khan, 2021). The following hypothesis is proposed.

H8: FLE significantly impacts PEU.

Flow and Perceived Usefulness

Research suggests that when learners experience deep cognitive absorption in an FLE state (X. Wang et al., 2022), they evaluate AI-driven learning platforms as more intuitive and user-friendly, reinforcing PU and their technology acceptance (Wu & Xie, 2024). Self-perception theory posits that individuals adjust their attitudes and behaviors to maintain internal consistency, reducing any dissonance between their expectations and experiences with technology (Bem, 1972). This sense of cognitive control and familiarity enhances their PU, making complex technological interactions appear more seamless and intuitive. Moreover, FLE in AI-enhanced learning contexts has been linked to higher task engagement, reduced cognitive load, and greater technology adoption rates (X. Wang et al., 2022). Consequently, the initial learning curve of AI-driven education systems diminishes, reinforcing positive PU and system effectiveness (Kong & Wang, 2021). Furthermore, empirical studies emphasize the role of environmental structuring in sustaining FLE states, as AI-driven platforms optimize content delivery, engagement strategies, and task complexity alignment to foster an uninterrupted learning experience (Ruangkanjanases et al., 2024). When learners attribute their smooth and immersive interaction with an AI system to its design efficiency, their PU is further strengthened (Zhao & Khan, 2021). The following hypothesis is proposed.

H9: FLE significantly impacts PU.

Flow and Academic Performance

The FLE, characterized by deep cognitive absorption, sustained concentration, and IM, plays a vital role in shaping AP in AI-assisted online learning environments (Shkëmbi & Treska, 2023). FLE facilitates improved comprehension, problem-solving abilities, and task persistence within AI-assisted educational platforms, leading to enhanced AP (Shernoff, 2010). Empirical research highlights that students who experience FLE in learning environments are more likely to demonstrate higher levels of engagement, cognitive flexibility, and sustained effort, all of which contribute to AP (Jahedizadeh et al., 2021). AI-driven learning systems, incorporating adaptive learning pathways, personalized FB, and real-time content adjustments, optimize task complexity to align with learners’ skill levels, fostering an environment conducive to FLE (Elareshi et al., 2022). As learners experience higher levels of IM through FLE, they are more likely to exhibit greater AP and increased problem-solving skills in assessments and coursework. The cognitive load theory further supports the impact of FLE on AP by explaining how FLE minimizes extraneous cognitive load, enabling learners to allocate cognitive resources toward meaningful learning activities effectively (Jahedizadeh et al., 2021; Sweller, 2023). AI-assisted learning platforms facilitate this process by offering real-time scaffolding, automated assistance, and predictive analytics, which enhance learning retention and efficiency (Shernoff, 2010; Shkëmbi & Treska, 2023). By reducing cognitive overload and promoting structured knowledge acquisition, these platforms enable students to achieve higher AP scores more easily. The following hypothesis is proposed.

H10: FLE significantly impacts AP.

Perceived Ease of Use and Perceived Usefulness

When learners find AI-assisted educational platforms easy to navigate, their confidence in the system's ability to enhance learning performance increases, ultimately strengthening their perception of its usefulness (Yan et al., 2024). User-friendly AI-driven learning interfaces, characterized by adaptive recommendations and automated content structuring, reduce cognitive effort, reinforcing the perception that the technology is valuable for academic success (D. Y. Park & Kim, 2023). When AI-driven learning platforms offer intuitive interfaces, automated guidance, and personalized learning trajectories, learners require less effort to engage with instructional materials, leading to an increased perception that the technology enhances their learning efficiency (Lu et al., 2022). Furthermore, studies indicate that PEU strengthens PU by fostering a sense of control, essential for maintaining engagement and long-term system adoption (Liesa-Orús et al., 2023). Additionally, the self-regulated learning model suggests that when learners experience minimal barriers in navigating AI-assisted learning platforms, they internalize their interactions as efficient and productive, reinforcing their PU (Yan et al., 2024). As AI-driven adaptive learning models continuously optimize the user experience, learners develop greater digital literacy, enhancing their PEU and PU (D. Y. Park & Kim, 2023). The following hypothesis is proposed.

H11: PEU significantly impacts PU.

Online Engagement and Continuous Intention

OE, defined as the intensity and consistency of a learner’s interaction with an AI-assisted learning platform, is a crucial determinant of CI’s use of such systems (Cheng, 2023). Engagement Theory (Kearsley & Shneiderman, 1998) underscores that meaningful OE with digital learning resources fosters a deeper CI in educational activities. In AI-assisted learning environments, OE is a behavioral phenomenon and a cognitive and emotional process that directly influences the learner’s CI utilizing the platform (Shao & Chen, 2021). Within AI-driven educational platforms, OE promotes an enhanced connection to the learning environment, fostering a more profound sense of belonging and investment in the platform. This sense of involvement plays a pivotal role in transitioning initial adoption behaviors into long-term commitment as learners develop habitual OE patterns and integrate the system into their regular learning routines (Sokro et al., 2025). These theoretical perspectives and empirical findings suggest that OE within AI-assisted online learning platforms is a key driver of CI (Kumar & Kumar, 2024). The following hypothesis is proposed.

H12: OE significantly impacts CI.

Perceived Ease of Use and Continuous Intention

In AI-assisted online learning platforms, the PEU with which learners navigate and engage with the system plays a crucial role in shaping their CI (Yan et al., 2024). When users experience PEU in an educational platform, their learning experience is enhanced, reducing cognitive load and fostering a more positive attitude toward CI for technology adoption. Empirical research in educational technology consistently highlights PEU as a determining factor in the CI’s use of digital learning environments (Filieri et al., 2021). Studies indicate that when learners experience PEU in a system, they are more likely to recognize its value and develop a higher inclination toward its CI to use (Gupta et al., 2021). This relationship is particularly pronounced in AI-assisted learning settings, where the PEU can efficiently navigate system features, access personalized content, and interact with AI-driven tools, which can significantly impact learning outcomes and overall CI to engage (Basuki et al., 2022; Ruangkanjanases et al., 2024). Furthermore, PEU indirectly enhances CI’s use of AI-assisted learning platforms by positively influencing perceived usefulness (Yan et al., 2024). When a system is user-friendly, learners are more inclined to explore its advanced functionalities, leading to a deeper appreciation of its educational benefits (Zhao & Khan, 2021). The following hypothesis is proposed.

H13: PEU significantly impacts CI.

Perceived Usefulness and Continuous Intention

PU shapes learners’ OE and CI within AI-assisted online learning platforms (Gani et al., 2024). When students recognize that a platform effectively contributes to their educational progress, they increase their CI in using the technology over time (Yan et al., 2024). Empirical research has consistently demonstrated a strong correlation between PU and the likelihood of CI for technology adoption (Basuki et al., 2022). This relationship is particularly pronounced in educational technology, where a learner’s PU directly influences the depth of overall satisfaction and CI (Wu & Xie, 2024). In AI-driven learning environments, the ability of adaptive algorithms to personalize instructional content and cater to individual learning needs significantly enhances the PU of the platform, thereby increasing the probability of long-term CI (X. Wang et al., 2022). Additionally, PU is a mediating factor between other key determinants and learners’ CI to engage with AI-assisted platforms (Zhao & Khan, 2021). While a user-friendly interface may encourage initial adoption, the platform’s PU in improving learning outcomes ultimately solidifies a learner’s CI to use it (Yan et al., 2024). The following hypothesis is proposed.

H14: PU significantly impacts CI.

Academic Performance and Continuous Intention

AP, often measured by learning outcomes, student progress, and assessment results, serves as a strong predictor of a learner’s CI to use AI-driven educational platforms (Maqableh et al., 2021). Empirical research in technology-mediated learning supports the premise that positive AP reinforces a learner’s CI using digital learning tools (Tawafak et al., 2021). According to self-determination theory, students who perceive their AP as improving through AI-assisted platforms experience greater intrinsic motivation and self-efficacy, contributing to their CI to use the platform (Luarn et al., 2023). Additionally, AP serves as a mediator between learner satisfaction and TAM. When students achieve higher grades, better comprehension, and improved problem-solving skills through AI-driven instruction, they are more likely to view the platform as indispensable to their learning process (Tawafak et al., 2018). The model further explains that perceived AP fosters technology acceptance, reinforcing the learner’s belief in the platform’s value and effectiveness (O. Chen, Paas, & Sweller, 2023).

AP, measured through learning outcomes and assessment results, significantly predicts learners’ CI in AI-assisted learning platforms, particularly when facilitated through personalized learning paths and intelligent tutoring systems (Jiao et al., 2022). When students experience improved performance through AI-enabled features such as automated grading feedback, adaptive learning algorithms, and personalized content recommendations, they develop stronger motivation to continue using these platforms (Adewale et al., 2024). The integration of specific AI functionalities, including real-time performance analytics and automated progress tracking, reinforces students’ perceived value of the platform, thereby strengthening their CI (Lee et al., 2022). This efficiency encourages students’ CI to use AI-driven systems, as they recognize the platform’s role in enhancing AP. Integrating AI-based assessment tools and interactive educational simulations further strengthens AP, ensuring a continuous engagement cycle (Tawafak et al., 2018). The following hypothesis is proposed.

H15: AP significantly impacts CI.

The theoretical framework of this research is indicated in Figure 1.

Figure 1.

Theoretical framework.

Methodology

Data Collection

This study examined online students in China who actively participate in AI-assisted English learning platforms. A sample of 289 students was selected, and data collection was facilitated through an online survey using a purposive sampling technique. To ensure sample representativeness, this research implemented several screening criteria. According to the study’s criteria, the respondents must have used AI-assisted English learning platforms within the past 6 months. Furthermore, the respondents were required to have completed at least one full English learning module or course using these platforms. This sampling approach helped ensure that the data collected was representative of the target population of AI-assisted English learning platform users in China.

The surveys were mailed to users of online English learning platforms with the assistance of SOJUMP in China over 2 months from March to April 2024. Sojump was selected due to its widespread use in academic research and its ability to reach a diverse online learner population. The survey was distributed through multiple channels, including direct invitations to university students via institutional email lists at five major universities in different regions of China, social media platforms such as WeChat groups focused on online learning, and AI-assisted English learning platform user communities.

Research Instrument

To ensure precise measurement, the study employed Likert scale items, allowing respondents to express their level of agreement or disagreement on a seven-point scale, ranging from 1 (strongly disagree) to 7 (strongly agree), with four serving as a neutral category.

To ensure content validity, the questionnaire was first developed in English based on validated scales from previous studies. It was then translated into Chinese using a back-translation procedure involving two bilingual experts in educational technology.

Various constructs were measured using previously validated scales. The FLE, PEU, PU, and FB were adapted from the study by Xu et al. (2021). Items assessing CI were sourced from Cheng’s (2014) research. The measurement of COMP and IMME was modified and adapted from the study by Hew et al. (2024). The construct of IM was assessed based on the scale proposed by Hong et al. (2017), while TP was measured using items from N. Wei and Li’s (2021) research. Additionally, OE was measured using the framework developed by Taghizade et al. (2024). AP was evaluated using the approach suggested by Hashemi (2021). The details regarding the research items are provided in the Appendix A.

Before launching the full-scale survey, a pilot test was conducted with 40 randomly selected participants to assess the validity and reliability of the questionnaire items. The final sample consisted of 289 respondents who had experience using AI-assisted learning platforms. The research hypotheses were tested using partial least squares structural equation modeling (PLS-SEM) to analyze the relationships among the study variables and determine the robustness of the proposed model. The distributed close-ended questionnaires yielded a 93.78% response rate, which is considered optimal for research reliability. The demographic composition of the sample is provided in Table 1.

Table 1.

Demographic Profile of Respondents (N = 289).

Characteristics	Categories	Frequency	Percentage (%)
Gender	Male	133	46.0
Gender	Female	156	54.0
Age	18–22	127	44.0
	23–27	98	33.9
	28–32	42	14.5
	33–35	22	7.6
Education level	Undergraduate	188	65.0
	Master’s	87	30.0
	Doctoral	14	5.0
Experience with AI learning platforms	Less than 6 months	67	23.2
	6–12 months	98	33.9
	1–2 years	85	29.4
	More than 2 years	39	13.5

Data Analysis

This study employed PLS-SEM to examine the dataset and test the hypothesized relationships. PLS-SEM is widely used in research because it can handle complex models and diverse data samples (J. Hair et al., 2023). The evaluation process followed a two-stage approach. Initially, the reliability and validity of the model were assessed to ensure robustness. Subsequently, path analysis was conducted to test the proposed hypotheses. The statistical analysis was carried out using SMARTPLS software (Ringle et al., 2015), which facilitated examining structural relationships within the model.

Convergent and Discriminant Validity

Convergent validity evaluates how multiple measures accurately reflect the same underlying construct. Interpreting research findings can lead to ambiguity when alternative measures exhibit weak convergent validity. To ensure internal reliability, this study utilized Cronbach’s alpha, a widely accepted measure of internal consistency. The reliability of the measurement model was assessed based on Cronbach’s alpha values.

Additionally, convergent validity was examined using Average Variance Extracted (AVE), Rho_A, and composite reliability (CR). As shown in Table 2, the Cronbach’s alpha and Rho_A values exceeded 0.7 (Taber, 2018). The AVE values surpassed 0.5 (Fornell & Larcker, 1981). The CR values were above 0.7, indicating strong convergent validity and a reliable measurement model.

Table 2.

Convergent Validity.

Constructs	Cronbach’s alpha	Rho_A	CR	AVE
AP	.874	.904	.905	0.655
COMP	.831	.846	.898	0.746
CI	.843	.846	.897	0.688
FB	.769	1.237	.709	0.404
FLE	.836	.895	.903	0.759
IMME	.943	.944	.959	0.855
IM	.899	.903	.930	0.770
OE	.964	.967	.973	0.902
PEU	.793	.790	.879	0.709
PU	.875	.923	.919	0.746
TP	.893	.910	.920	0.699

Note. AP = academic performance; COMP = competition; CI = continuous intention; FB = feedback, FLE = flow experience, IMME = immediacy, IM = intrinsic motivation; OE = online engagement; PEU = perceived ease of use; PU = perceived usefulness; TP = telepresence.

In research involving latent variables and multiple indicators, assessing discriminant validity is essential to confirm that the constructs measuring causal relationships remain conceptually distinct. This study employed both the Fornell-Larcker criterion and cross-loadings to evaluate discriminant validity.

The Fornell-Larcker criterion assesses discriminant validity by comparing the square root of the AVEs with the correlation coefficients. In this approach, the diagonal values represent the square root of the AVEs, while the off-diagonal elements denote the correlation coefficients between constructs. As shown in Table 3, the diagonal values exceed the correlation coefficients, confirming strong discriminant validity within the measurement model.

Table 3.

Fornell and Larcker Criterion.

Constructs	AP	COMP	CI	FB	FLE	IMME	IM	OE	PEU	PU	TP
AP	0.809
COMP	0.225	0.864
CI	0.423	0.375	0.830
FB	0.087	0.172	0.091	0.635
FLE	0.215	0.237	0.417	0.409	0.871
IMME	0.427	0.368	0.614	0.063	0.381	0.925
IM	0.395	0.349	0.678	0.135	0.405	0.699	0.877
OE	0.094	0.127	0.240	0.006	0.307	0.109	0.119	0.949
PEU	0.308	0.329	0.480	0.261	0.199	0.412	0.382	−0.077	0.842
PU	0.494	0.308	0.542	0.127	0.271	0.551	0.531	−0.025	0.572	0.864
TP	0.472	0.370	0.560	0.103	0.276	0.470	0.416	0.208	0.407	0.437	0.836

Note. AP = academic performance; COMP = competition; CI = continuous intention; FB = feedback; FLE = flow experience; IMME = immediacy; IM = intrinsic motivation; OE = online engagement; PEU = perceived ease of use; PU = perceived usefulness; TP = telepresence.

As illustrated in Table 4, each indicator demonstrates acceptable discriminant validity (italicized) as its loading value is higher than other indicators within the latent structure (J. F. Hair et al., 2021). These findings confirm that the measurement model effectively distinguishes between different constructs, ensuring statistical robustness.

Table 4.

Cross-Loadings.

Constructs	AP	COMP	CI	FB	FLE	IMME	IM	OE	PEU	PU	TP
AP1	0.782	0.220	0.437	0.096	0.284	0.412	0.425	−0.062	0.319	0.515	0.370
AP2	0.846	0.118	0.335	0.050	0.185	0.376	0.315	0.107	0.299	0.387	0.353
AP3	0.818	0.203	0.344	0.087	0.124	0.314	0.246	0.123	0.302	0.384	0.416
AP4	0.791	0.156	0.240	0.057	0.120	0.266	0.282	0.158	0.094	0.292	0.364
AP5	0.807	0.194	0.265	0.037	0.060	0.291	0.257	0.156	0.123	0.318	0.413
COMP1	0.198	0.852	0.302	0.149	0.219	0.255	0.300	0.092	0.243	0.276	0.263
COMP2	0.112	0.857	0.236	0.190	0.162	0.272	0.223	0.145	0.237	0.226	0.253
COMP3	0.250	0.881	0.410	0.118	0.223	0.415	0.360	0.102	0.359	0.285	0.424
CI17	0.364	0.327	0.922	0.032	0.317	0.509	0.502	0.228	0.409	0.434	0.516
CI18	0.301	0.222	0.794	0.055	0.273	0.396	0.380	0.249	0.342	0.400	0.454
CI19	0.367	0.345	0.896	0.076	0.359	0.537	0.477	0.191	0.454	0.453	0.517
CI20	0.358	0.333	0.685	0.135	0.418	0.573	0.858	0.131	0.373	0.495	0.361
FB1	−0.051	0.110	0.008	0.948	0.434	0.024	0.104	0.014	0.151	0.007	0.012
FB2	0.346	0.200	0.160	0.503	0.072	0.043	0.082	−0.065	0.401	0.327	0.227
FB3	0.358	0.133	0.101	0.484	0.005	0.072	0.107	−0.077	0.250	0.287	0.166
FB4	0.375	0.225	0.293	0.477	0.104	0.170	0.152	0.022	0.332	0.344	0.284
FLE1	0.130	0.171	0.287	0.148	0.702	0.221	0.267	0.295	0.092	0.158	0.192
FLE2	0.200	0.223	0.386	0.429	0.937	0.375	0.381	0.245	0.188	0.254	0.263
FLE3	0.218	0.222	0.405	0.433	0.952	0.373	0.395	0.284	0.217	0.278	0.259
IMME1	0.391	0.361	0.586	0.073	0.357	0.924	0.683	0.066	0.395	0.531	0.407
IMME2	0.400	0.383	0.558	0.058	0.357	0.916	0.638	0.125	0.393	0.525	0.486
IMME3	0.405	0.309	0.584	0.040	0.347	0.938	0.626	0.106	0.380	0.515	0.416
IMM34	0.382	0.307	0.544	0.062	0.346	0.920	0.638	0.107	0.355	0.463	0.427
IM1	0.341	0.239	0.570	0.145	0.368	0.515	0.882	0.126	0.269	0.386	0.337
IM2	0.338	0.299	0.586	0.123	0.373	0.560	0.916	0.112	0.291	0.439	0.347
IM3	0.342	0.334	0.630	0.117	0.352	0.614	0.918	0.112	0.363	0.515	0.394
IM4	0.370	0.360	0.598	0.086	0.327	0.786	0.788	0.062	0.431	0.536	0.385
OE1	0.112	0.153	0.245	0.011	0.288	0.142	0.110	0.946	−0.031	0.007	0.213
OE2	0.088	0.149	0.246	0.024	0.317	0.115	0.113	0.958	−0.057	−0.017	0.222
OE3	0.083	0.078	0.197	−0.015	0.280	0.065	0.099	0.946	−0.127	−0.058	0.162
OE4	0.075	0.096	0.218	0.000	0.279	0.087	0.128	0.949	−0.087	−0.032	0.187
PEU1	0.217	0.243	0.291	0.230	0.189	0.372	0.350	−0.070	0.876	0.533	0.263
PEU2	0.199	0.271	0.297	0.237	0.185	0.394	0.342	−0.135	0.878	0.495	0.251
PEU3	0.343	0.307	0.588	0.191	0.131	0.279	0.273	0.000	0.766	0.416	0.485
PU1	0.553	0.122	0.270	0.114	0.113	0.227	0.243	0.021	0.359	0.573	0.379
PU2	0.430	0.327	0.526	0.130	0.267	0.529	0.495	−0.010	0.557	0.952	0.417
PU3	0.415	0.292	0.514	0.126	0.274	0.533	0.518	−0.014	0.536	0.948	0.365
PU4	0.389	0.281	0.513	0.077	0.249	0.545	0.524	−0.069	0.502	0.924	0.375
TP1	0.366	0.612	0.440	0.089	0.304	0.519	0.401	0.160	0.285	0.304	0.710
TP2	0.412	0.242	0.547	0.102	0.217	0.371	0.366	0.142	0.408	0.428	0.864
TP3	0.394	0.158	0.493	0.072	0.193	0.348	0.314	0.136	0.344	0.411	0.877
TP4	0.407	0.149	0.383	0.095	0.197	0.312	0.311	0.205	0.320	0.332	0.848
TP5	0.363	0.176	0.435	0.057	0.171	0.295	0.267	0.225	0.334	0.340	0.868

Additionally, the Heterotrait-monotrait ratio (HTMT) analysis in Table 5 reveals several significant relationships between the constructs in the AI-driven educational context. The results demonstrate discriminant validity among the constructs, as most HTMT values fall below the conservative threshold of 0.85, suggesting adequate construct distinction (Ab Hamid et al., 2017).

Table 5.

Heterotrait-Monotrait Ratio.

Constructs	AP	COMP	CI	FB	FLE	IMME	IM	OE	PEU	PU
AP
COMP	0.247
CI	0.463	0.433
FB	0.439	0.275	0.245
FLE	0.218	0.279	0.490	0.258
IMME	0.448	0.409	0.684	0.118	0.419
IM	0.424	0.397	0.775	0.175	0.462	0.767
OE	0.163	0.144	0.267	0.069	0.352	0.113	0.126
PEU	0.326	0.395	0.568	0.467	0.236	0.479	0.459	0.112
PU	0.570	0.346	0.620	0.426	0.302	0.591	0.592	0.051	0.687
TP	0.523	0.362	0.633	0.282	0.297	0.480	0.445	0.222	0.466	0.506

The goodness of fit (GOF) assessment for this research was conducted based on Tenenhaus et al.’s (2005) proposed methodology. It is calculated with the following formula.

GOF = \sqrt{\bar{AVE}} x \sqrt{\bar{R^{2}}} = \sqrt{0.721 x 0.213} = 0.391

The GOF values can be classified as small (0.1), medium (0.25), and large (0.36). The obtained GOF value of 0.391 exceeds the large effect size threshold of 0.36, indicating a robust global validation of the PLS model (Wetzels et al., 2009). This suggests that the model has satisfactory explanatory power and predictive relevance (Hair et al., 2019). Lastly, according to Tenenhaus et al. (2005), this result would be considered acceptable for complex structural models.

Empirical Results

The path analysis in this study was performed using the SMARTPLS statistical software (Ringle et al., 2024). The p-value and t-value criteria were applied to determine the significance of the proposed relationships.

According to the findings presented in Table 6 and Figure 2. FLE was significantly associated with CI (β = .193, T-value = 3.411). All the antecedents including IM (β = .194, T-value = 2.321), IMME (β = .166, T-value = 2.109), and FB (β = .366, T-value = 2.474) significantly impacted FLE, except TP (β = .079, T-value = 0.792), and COMP (β = .022, T-value = 0.227). Furthermore, FLE significantly impacted OE (β = .302, T-value = 4.041), PEU (β = .199, T-value = 2.575), PU (β = .163, T-value = 2.208), and AP (β = .214, T-value = 2.433). PEU was discovered to impact PU (β = .545, T-value = 8.870). Lastly, OE (β = .188, T-value = 3.076), PEU (β = .257, T-value = 2.939), PU (β = .272, T-value = 3.458), and AP (β = .152, T-value = 2.136) were found to significantly impact CI.

Table 6.

Empirical Results.

Hypotheses	Path coefficient (β)	T values	p Values	Results
H1: FLE → CI	0.193	3.411	.001	Supported
H2: IM → FLE	0.194	2.321	.020	Supported
H3: IMME → FLE	0.166	2.109	.035	Supported
H4: COMP → FLE	0.022	0.227	.820	Not supported
H5: TP → FLE	0.079	0.792	.428	Not supported
H6: FB → FLE	0.366	2.474	.013	Supported
H7: FLE → OE	0.302	4.041	.000	Supported
H8 FLE → PEU	0.199	2.575	.010	Supported
H9: FLE → PU	0.163	2.208	.027	Supported
H10: FLE → AP	0.214	2.433	.015	Supported
H11: PEU → PU	0.545	8.870	.000	Supported
H12: OE → CI	0.188	3.076	.002	Supported
H13: PEU → CI	0.257	2.939	.003	Supported
H14: PU → CI	0.272	3.458	.001	Supported
H15: AP → CI	0.152	2.136	.033	Supported

Figure 2.

Empirical results.

Discussions

Comparison With Other Studies

According to the findings of this study, FLE was found to impact CI directly. This result is similar to a recent study by Muhammad Sohail Jafar et al. (2024). Their study proposed that the evolution of digital commerce is reaching new frontiers with the rapid expansion of the Metaverse ecosystem. Their results indicated that FLE strengthened the users’ CI to engage in Metaverse-based shopping. Their research contributed practical insights for technology firms and businesses seeking to leverage opportunities within this virtual space.

The present study found that IM was not a significant antecedent of FLE. The result can be somewhat comparable to a study by Mehta and Vyas (2022). Using standardized search queries, their research data were sourced from three peer-reviewed academic databases—Scopus, PubMed, and JSTOR. Their findings revealed a strong positive correlation between IM and the FLE. Their results provided valuable insights for educators and instructional designers aiming to enhance student engagement and optimize learning environments by fostering conditions that encourage IM and sustained attention.

Furthermore, according to this study’s findings, IMME had a significant impact on FLE; however, COMP was not significantly associated with FLE. The results of this study were somewhat similar to a study conducted by Hew et al. (2024). Their study was built on SOR and FLE theories and targeted the gaming industry. Their findings indicated that COMP and IMME have a direct and positive influence on FLE. Their results shed light on the dual nature of FLE in gaming as both a driver of engagement and a potential pathway to excessive dependency. The non-significant relationship between COMP and FLE deviates from the findings of Hew et al. (2024) in gaming contexts. This distinction likely stems from fundamental differences in motivational dynamics between educational and gaming environments. While gaming inherently thrives on competitive elements, educational settings may prioritize command-oriented goals over competition. Achievement Goal Theory (Urdan & Kaplan, 2020) suggests that COMP might actually delay FLE in learning contexts where the primary goals are targeted toward skill development and conceptual understanding, rather than outperforming others.

Additionally, regarding the findings of another FLE antecedent, TP was not found to be significantly associated with FLE. The finding differs from a previous study by N. Wei and Li (2021). Their study examined an English learning application through an online survey among university students. Their findings indicated that TP was significantly associated with FLE. Their research offered valuable implications for companies seeking to enhance system functionalities and optimize user experience. The non-significant relationship between TP and FLE diverges from the findings of N. Wei and Li (2021) in language learning contexts. This discrepancy may be attributed to the distinct cognitive demands of the AI-assisted learning tasks in the present study. While language acquisition inherently benefits from environmental immersion and social presence, the current study’s context emphasizes analytical processing and conceptual understanding, where TP might be less critical.

In terms of antecedents of FLE, FB was significantly associated with FLE. This result is similar to an earlier study by Xu et al. (2021). Their study developed a theoretical framework integrating key elements from the FLE and TAM to examine the antecedents of FLE and their impact on social media CI to purchase among senior shoppers. Xu et al.’s (2021) findings revealed that FB is positively associated with FLE.

Additionally, the present study also discovered that FLE significantly impacted OE. The result is somewhat comparable to an earlier study by B. Kim et al. (2020). Their research provided a valuable framework for understanding customer engagement in digital environments. Their findings indicated that FLE significantly enhances users’ attitudes and strengthens their OE, highlighting the pivotal role of FLE in fostering customer engagement.

Furthermore, according to the current study, FLE significantly impacted PEU and PU. These results are similar to a study by Yan et al. (2024). Their study identified key dimensions of learner motivation in E-learning. Their findings confirmed that FLE was a critical cognitive factor in shaping PEU and PU. Their research offered valuable implications for E-learning system designers.

The present research discovered that FLE and AP are significantly associated. The result is somewhat different from Bian et al.’s earlier study (2018). Their findings indicated that the impact of FLE on performance is insignificant in virtual environments where the primary task and the operation of interactive elements lack alignment. In such cases, interactive elements may independently induce FLE unrelated to task performance, weakening the overall effect.

In addition, the present study indicated a significant association between OE and CI. This finding is similar to an earlier study by Kumar and Kumar (2024). Their study examined the mediating role of user attitude in shaping the relationship between online service convenience and OE, as well as their CI using e-resources. Their findings indicated a significant association of OE with CI.

The present study also indicates that the TAM constructs PEU and PU had significant associations with CI. This finding is consistent with a previous study by Tao et al. (2022). Their study aimed to examine the key determinants of user acceptance by incorporating three critical massive open online courses (MOOCs) characteristics—usability, perceived quality, and perceived enjoyment—within the framework of the TAM. Their findings indicated that PEU and PU significantly impacted CI.

Lastly, the present study indicated a significant association between AP and CI. This result is similar to Tawafak et al. (2018). Their paper investigated whether the universities’ communication model and the TAM simultaneously enhance student satisfaction and improve the teaching method and AP level. Their findings indicated a significant impact of AP on CI.

Theoretical Implications

This study advances theoretical understanding in AI-assisted online learning by integrating multiple theoretical frameworks to examine learner OE (Taghizade et al., 2024) and AP in digital environments. It extends the TAM by demonstrating that FLE is a key mediator influencing PU and PEU, strengthening learners’ CI using AI-driven platforms (Wandira et al., 2024). The findings align with prior research, emphasizing the importance of usability, interactive feedback, and intuitive system design as critical factors for sustaining long-term engagement with educational technologies. The study further reinforces FLE by validating its role in digital learning, showing that TP and adaptive learning mechanisms contribute to sustained engagement (Muhammad Sohail Jafar et al., 2024). AI-enhanced learning tools are crucial in maintaining FLE by reducing cognitive overload and balancing task complexity with learner skills. The findings build on existing literature by emphasizing the importance of personalized learning in fostering cognitive absorption and IM, thereby strengthening the connection between FLE and digital learning persistence (Sadoughi & Eskandari, 2024). Applying EVT, the study establishes a direct link between AP and CI. Hence, it illustrates that students who experience academic improvements through AI-driven feedback and adaptive learning pathways are more likely to remain engaged with the platform (Wandira et al., 2024). This supports existing research highlighting the influence of perceived academic benefits and self-efficacy on long-term technology adoption. The research also contributes to self-regulated learning theories by demonstrating how AI-assisted environments support learners in monitoring progress, refining strategies, and engaging with personalized feedback. Students who engage in self-regulated learning behaviors within AI-driven platforms exhibit higher AP and a stronger commitment to digital learning. These findings highlight the need for instructional strategies incorporating adaptive learning pathways, intelligent tutoring systems, and interactive content delivery, ensuring a more effective and engaging digital learning experience.

Practical Implications

This study offers significant practical insights for various stakeholders, including practitioners, policymakers, administrators, and learners, by providing strategies to optimize AI-assisted online learning environments for enhanced engagement, usability, and AP. For practitioners, the findings emphasize the role of FLE in influencing CI, OE, and PEU. Hence, instructional designers should focus on developing interactive and adaptive learning pathways. Leveraging AI-powered content recommendations, personalized feedback mechanisms, and intuitive system interfaces can help educators foster higher engagement and sustained motivation among learners. Additionally, for practitioners, the findings emphasize the importance of developing engaging environments through intelligent tutoring systems, automated assessment tools, and personalized learning paths (Abdulmuhsin et al., 2025). Instructional designers should integrate AI-assisted features such as adaptive content sequencing and automated grading with detailed feedback to enhance FLE.

The study provides evidence for policymakers to support integrating adaptive learning technologies into formal curricula. Given the demonstrated impact of FLE on AP and sustained engagement, educational policies should promote investment in digital infrastructure, teacher training, and AI-driven learning solutions to ensure that technology-enhanced education remains inclusive, engaging, and effective. Additionally, data-driven monitoring systems should be implemented to track learner engagement and improve digital interventions through real-time analytics (Alkhwaldi et al., 2025). Policymakers should prioritize implementing AI technologies that support personalized learning courses, while administrators should focus on developing comprehensive solutions that combine automated grading, intelligent content delivery, and predictive analytics.

For managers and administrators, the study underscores the importance of user experience design in AI-assisted learning platforms. Enhancing PEU and PU through intuitive navigation, clear content organization, and AI-driven personalization can improve learner retention and reduce dropout rates. Educational institutions should also invest in faculty development programs, equipping instructors with the necessary skills to integrate AI-driven teaching strategies effectively. For learners, the study highlights the benefits of self-regulated learning strategies in maximizing AP within AI-assisted educational environments. Students who experience FLE demonstrate higher problem-solving abilities, improved knowledge retention, and long-term engagement. Using AI-enhanced tools, such as adaptive feedback, self-paced learning modules, and gamified assessments, enables learners to create a more immersive and productive educational experience.

Research Limitations and Future Research Directions

This study has several limitations that provide opportunities for future research. First, while this research focused on AI-assisted English learning platforms, the sample was limited to Chinese learners. Future studies should consider cross-cultural comparisons to validate the model’s applicability in diverse cultural settings, providing in-depth comparative insights.

Second, the cross-sectional nature of this study provides an analysis of online learners at a single point in time, limiting the understanding of how FLE and engagement patterns evolve. Hence, future researchers should employ a longitudinal research method to track changes in FLE and its impact on CI over the course of an academic term or program.

Third, this study primarily examined quantitative measures of FLE and its antecedents, potentially missing insights into learners’ experiences. Future research could employ mixed-method approaches, incorporating qualitative interviews or observational studies to provide deeper insights into how learners achieve and maintain FLE in AI-assisted learning environments.

A significant limitation of this study is that the hypotheses regarding TP and COMP significantly influencing FLE learning outcomes across all educational contexts are not supported. Empirical results did not support these results because TP might vary substantially between different learning scenarios, such as theoretical versus practical courses, or synchronous versus asynchronous learning environments. Future research should investigate how various contextual factors (course type, learning objectives, and delivery mode) moderate the effect of TP on FLE. Additionally, COMP’s effect on FLE could differ between collaborative project work and individual skill development within AI-assisted environments. Future research should employ contextual analysis to examine how various learning situations (individual vs. group-based) moderate the relationship between COMP and FLE.

Footnotes

Appendix

Appendix A

Constructs	Research items
Flow experience	• I am absorbed in what I am doing while using the online AI-assisted learning platform. • I am often unable to keep track of the passage of time while using the online AI-assisted learning platform. • I find using the online AI-assisted learning platform to be enjoyable.
Feedback	• I was aware of how well I was performing. • It was really clear to me that I was doing well. • I had a good idea about how well I was doing. • I could tell by the way I was using the online AI-assisted learning platform how well I was doing.
Perceived usefulness	• Using the online AI-assisted learning platform enhances my learning effectiveness. • Using the online AI-assisted learning platform can improve my learning performance. • Using the online AI-assisted learning platform gives me greater control over learning. • I find the online AI-assisted learning platform to be useful in my learning.
Perceived ease to use	• Learning to use the online AI-assisted learning platform for my English learning needs is easy for me. • The process of using the online AI-assisted learning platform with regard to English learning is clear and understandable. • I find the online AI-assisted learning platform easy to use with regard to English learning.
Continuous intention	• I intend to continue using the online AI-assisted learning platform in the future. • I will use the online AI-assisted learning platform on a regular basis in the future. • I will frequently use the online AI-assisted learning platform in the future. • My intentions are to continue using the online AI-assisted learning platform than use any alternative means (traditional learning).
Intrinsic motivation	• I often attempt to use online AI-assisted learning platforms to enhance my English proficiency • I use English to chat with my friends to improve my English proficiency. • I watch TV programs with English subtitles to improve my ability to recognize more English words. • I work hard on my English homework in order to improve my English.
Immediacy	• Online AI-assisted learning platforms is accessible at any time and place. • Online AI-assisted learning platforms enables me to have a match in real time. • Online AI-assisted learning platforms allows me to have a match at the best moment for me. • I can use online AI-assisted learning platforms anytime, anywhere.
Telepresence	• When using the online AI-assisted learning platforms, I forget about myself. • After using the online AI-assisted learning platforms, I feel I just had a journey. • When using the online AI-assisted learning platforms, it seems that I was put in the world the application created. • When I stopped using the online AI-assisted learning platforms, the application environment suddenly disappears. • When using the online AI-assisted learning platforms, my body is in the room, but my mind is in the virtual world.
Online engagement	• I stay focused during online learning activities. • I talk about online learning topics even when I am offline. • I complete all online learning tasks on time. • I feel inspired to improve my online learning skills.
Academic performance	• Performing well in online AI-assisted learning platforms made me feel good about myself. • I felt that online AI-assisted learning platforms were good to advance my studies and that there were dramatic changes in my academic performance. • Using online AI-assisted learning platforms moved me closer to attaining my career goals. • I feel able to perform well in online AI-assisted learning platforms. • I have acquired more knowledge by using online AI-assisted learning platforms.

ORCID iDs

Jing Zhao

Li Liu

Ethical Considerations

The manuscript was exempted by the ethics review committee of Shaanxi Normal University because this research is not for minors, pregnant women, those with physical and mental disabilities, or those with spiritual disabilities. In addition, this research has low risk, and the risk to the study subjects is not higher than that of those who did not participate. After the evaluation by the ethical review committee members, the exemption from review and verification is issued. The minimum risk mentioned in the preceding paragraph refers to the probability or intensity of the harm or discomfort suffered by the research subject and is not higher than the harm or discomfort suffered in daily life.

Consent to Participate

The participants provided their written informed consent to participate in this study. “We welcome your involvement in our investigation of online learning platforms and AI learning experiences. Our research examines how technological features shape learning outcomes and intentions to use online learning platforms. Eligible participants must have utilized online learning platforms and be of legal age (18+). The questionnaire requires roughly a quarter-hour to finish and explores your encounters with online learning systems. Please note that your engagement is entirely optional—you maintain the right to discontinue at any point without consequences. We implement stringent privacy protocols to safeguard your information. All responses undergo secure digital encryption; only authorized investigators can access the anonymized dataset.”

Author Contributions

W.R., J.Z., and L.L.: Conceptualization; W.R., J.Z., and L.L.: formal analysis; W.R., J.Z., and L.L.: methodology; W.R., J.Z., and L.L.: writing – original draft; W.R., J.Z., and L.L.: writing – review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study is supported by the following projects: Research Project on Integration of Production and Education in Undergraduate Universities in Henan Province (Project name: Comprehensive Reform and Application of Multiple Collaborative Practice Teaching Mode of Design Major under AI Enabling; Project number: 2023348073); Research and Practice Project on Undergraduate Education and Teaching Reform of Henan Agricultural University (Project name: Research and Practice on Teaching Reform of General Courses of Public Art in Colleges and Universities in the New Era of “Educating People with Aesthetics and Infiltrating Integration” Project number: 2024XJGLX002); Henan Provincial Department of Education, 2022 Research and Practice Project on Research-based Teaching Reform in Undergraduate Colleges and Universities (2022SYJXLX097).

Declaration of Conflicting Interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Unlocking Continuous Engagement in AI-Driven Education: An Integrated Model of Flow Experience, Academic Performance, and Technology Acceptance

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors on request.

References

Abdulmuhsin

A. A.

Owain

H. O.

Dbesan

A. H.

Bhat

M. A.

Rehman

S. U.

Alkhwaldi

A. F.

(2025). Acceptance of KM-driven metaverse technology in higher education institutions: Are educators ready to be immersed? Information Discovery and Delivery. Advance online publication. https://www.emerald.com/idd/article-abstract/doi/10.1108/IDD-11-2024-0188/1251740/Acceptance-of-KM-driven-metaverse-technology-in?redirectedFrom=fulltext

Ab Hamid

M. R.

Sami

Sidek

M. M.

(2017, September). Discriminant validity assessment: Use of Fornell & Larcker criterion versus HTMT criterion. Journal of Physics: Conference Series, 890(1), 012163.

Abu-AlSondos

I. A.

Salameh

A. A.

Alkhwaldi

A. F.

Mushtaha

A. S.

Shehadeh

Al-Junaidi

(2023). Evaluating mobile E-learning systems acceptance: An integrated model. International Journal of Interactive Mobile Technologies, 17(16), 1–18.

Addas

Naseer

Tahir

Khan

M. N.

(2024). Enhancing higher-education governance through telepresence robots and gamification: Strategies for sustainable practices in the AI-driven digital era. Education Sciences, 14(12), 1324.

Adewale

M. D.

Azeta

Abayomi-Alli

Sambo-Magaji

(2024). Impact of artificial intelligence adoption on students’ academic performance in open and distance learning: A systematic literature review. Heliyon, 10(22).

Ali

Warraich

N. F.

Butt

(2025). Acceptance and use of artificial intelligence and AI-based applications in education: A meta-analysis and future direction. Information Development, 41(3), 859–874.

Alkhwaldi

A. F.

(2024). Understanding learners’ intention toward Metaverse in higher education institutions from a developing country perspective: UTAUT and ISS integrated model. Kybernetes, 53(12), 6008–6035.

Alkhwaldi

A. F.

Abdulmuhsin

A. A.

Masa’deh

R. e.

Abu-AlSondos

I. A.

(2025). Business intelligence adoption in higher education: the role of data-driven decision-making culture and UTAUT. Journal of International Education in Business, 18(4), 484–504.

Arghashi

Yuksel

C. A.

(2022). Interactivity, inspiration, and perceived usefulness! How retailers’ AR-apps improve consumer engagement through flow. Journal of Retailing and Consumer Services, 64, Article 102756.

10.

Bae

J.-H.

(2014). Effectiveness of learning flow and academic achievement on learning activities with real-time feedback utilizing a smart clicker app in higher education. Journal of the Korea Academia-Industrial Cooperation Society, 15(9), 5543–5552.

11.

Basuki

Tarigan

Z. J. H.

Siagian

Limanta

L. S.

Setiawan

Mochtar

(2022). The effects of perceived ease of use, usefulness, enjoyment and intention to use online platforms on behavioral intention in online movie watching during the pandemic era. Petra Christian University.

12.

bayu dani Nandiyanto

Sidik

N. A. C.

(2026). Revolutionizing education: Exploring technological innovations, opportunities and challenges across I0T, AR, VR, AI, LMS, ML, gamification and emerging. Journal of Advanced Research Design, 136(1), 108–137.

13.

Bem

D. J.

(1972). Self-perception theory. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 6, pp. 1–62). Elsevier.

14.

Bian

Yang

Zhou

Liu

Gai

Meng

Tian

Shen

(2018). Exploring the weak association between flow experience and performance in virtual environments. In Proceedings of the 2018 CHI conference on human factors in computing systems, New York, NY, USA.

15.

Buzady

Wimmer

Csesznak

Szentesi

(2024). Exploring flow-promoting management and leadership skills via serious gaming. Interactive Learning Environments, 32(2), 757–771.

16.

Cao

Chen

(2024). Metaverse application, flow experience, and gen-zers’ participation intention of intangible cultural heritage communication. Data Science and Management, 7(2), 144–153.

17.

Chen

C.-H.

Liu

J.-H.

Shou

W.-C.

(2018). How competition in a game-based science learning environment influences students’ learning achievement, flow experience, and learning behavioral patterns. Journal of Educational Technology & Society, 21(2), 164–176.

18.

Chen

C.-T.

Chen

S.-C.

Khan

Lim

M. K.

Tseng

M.-L.

(2024). Antecedents of big data analytics and artificial intelligence adoption on operational performance: The ChatGPT platform. Industrial Management & Data Systems, 124, 2388–2413.

19.

Chen

Paas

Sweller

(2023). A cognitive load theory approach to defining and measuring task complexity through element interactivity. Educational Psychology Review, 35(2), 63.

20.

Chen

T.-Y.

P.-F.

Yang

Y.-L.

(2024). Investigating the immediacy and flow factors to influence purchase intention in live video streaming shopping. Commerce & Management Quarterly, 25(1), Article 104415.

21.

Cheng

Y. M.

(2014). Extending the expectation-confirmation model with quality and flow to explore nurses’ continued blended e-learning intention. Information Technology & People, 27(3), 230–258.

22.

Cheng

Y.-M.

(2023). Which quality determinants cause MOOCs continuance intention? A hybrid extending the expectation-confirmation model with learning engagement and information systems success. Library Hi Tech, 41(6), 1748–1780.

23.

Chien

C.-C.

Chan

H.-Y.

Hou

H.-T.

(2024). Learning by playing with generative AI: design and evaluation of a role-playing educational game with generative AI as scaffolding for instant feedback interaction. Journal of Research on Technology in Education, 57(7), 1–20.

24.

Cinar

O. E.

Rafferty

Cutting

Wang

(2024). AI-powered VR for enhanced learning compared to traditional methods. Electronics, 13(23), Article 4787.

25.

Davis

F. D.

(1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13, 319–340.

26.

Doğan

Şahin

Y. L.

Kobak

Okur

M. R.

(2024). Enhancing clinical law education through immersive virtual reality: A flow experience perspective. Learning and Instruction, 94, Article 101989.

27.

Elareshi

Habes

Youssef

Salloum

S. A.

Alfaisal

Ziani

(2022). SEM-ANN-based approach to understanding students’ academic-performance adoption of YouTube for learning during Covid. Heliyon, 8(4), e09236.

28.

Filieri

Acikgoz

Ndou

Dwivedi

(2021). Is TripAdvisor still relevant? The influence of review credibility, review usefulness, and ease of use on consumers’ continuance intention. International Journal of Contemporary Hospitality Management, 33(1), 199–223.

29.

Fornell

Larcker

D. F.

(1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39–50.

30.

Gani

M. O.

Rahman

M. S.

Bag

Mia

M. P.

(2024). Examining behavioural intention of using smart health care technology among females: dynamics of social influence and perceived usefulness. Benchmarking: An International Journal, 31(2), 330–352.

31.

Gao

(2023). Understanding smart education continuance intention in a delayed benefit context: An integration of sensory stimuli, UTAUT, and flow theory. Acta Psychologica, 234, Article 103856.

32.

Gupta

Prashar

Vijay

T. S.

Parsad

(2021). Examining the influence of antecedents of continuous intention to use an informational app: The role of perceived usefulness and perceived ease of use. International Journal of Business Information Systems, 36(2), 270–287.

33.

Hair

J. F.

Jr. Sarstedt

Ringle

C. M.

Gudergan

S. P.

(2023). Advanced issues in partial least squares structural equation modeling. Sage Publications.

34.

Hair

J. F.

Risher

J. J.

Sarstedt

Ringle

C. M.

(2019). When to use and how to report the results of PLS-SEM. European Business Review, 31(1), 2–24.

35.

Hair

J. F.

Jr. Hult

G. T. M.

Ringle

C. M.

Sarstedt

Danks

N. P.

Ray

Hair

J. F.

(2021). An introduction to structural equation modeling. In Partial least squares structural equation modeling (PLS-SEM) using R: A workbook (pp. 1–29). Springer.

36.

Hamari

Shernoff

D. J.

Rowe

Coller

Asbell-Clarke

Edwards

(2016). Challenging games help students learn: An empirical study on engagement, flow and immersion in game-based learning. Computers in Human Behavior, 54, 170–179.

37.

Hashemi

(2021). Effects of COVID-19 on the academic performance of Afghan students’ and their level of satisfaction with online teaching. Cogent Arts & Humanities, 8(1), Article 1933684.

38.

Sukpasjaroen

Chankoson

(2024). An empirical model of impulse purchases in live streaming commerce on the Chinese Economy. Kurdish Studies, 12(2), 6044–6065.

39.

Hew

J.-J.

Lee

V.-H.

T’ng

S.-T.

Tan

G. W.-H.

Ooi

K.-B.

Dwivedi

Y. K.

(2024). Are online mobile gamers really happy? On the suppressor role of online game addiction. Information Systems Frontiers, 26(1), 217–249.

40.

Hilty

D. M.

Randhawa

Maheu

M. M.

McKean

A. J.

Pantera

Mishkind

M. C.

Rizzo

A. S.

(2020). A review of telepresence, virtual reality, and augmented reality applied to clinical care. Journal of Technology in Behavioral Science, 5(2), 178–205.

41.

Hong

J.-C.

Hwang

M.-Y.

Tai

K.-H.

Lin

P.-H.

(2017). Intrinsic motivation of Chinese learning in predicting online learning self-efficacy and flow experience relevant to students’ learning progress. Computer Assisted Language Learning, 30(6), 552–574.

42.

Hsiao

J.-C.

Chang

J. S.

(2024). Enhancing EFL reading and writing through AI-powered tools: Design, implementation, and evaluation of an online course. Interactive Learning Environments, 32(9), 4934–4949.

43.

Jahedizadeh

Ghonsooly

Ghanizadeh

(2021). A model of language students’ sustained flow, personal best, buoyancy, evaluation apprehension, and academic achievement. Porta Linguarum Revista Interuniversitaria de Didáctica de las Lenguas Extranjeras (35), 257–275.

44.

Jiao

Ouyang

Zhang

Alavi

A. H.

(2022). Artificial intelligence-enabled prediction model of student academic performance in online engineering education. Artificial Intelligence Review, 55(8), 6321–6344.

45.

Kaya

O. S.

Ercag

(2023). The impact of applying challenge-based gamification program on students’ learning outcomes: Academic achievement, motivation and flow. Education and Information Technologies, 28(8), 10053–10078.

46.

Kearsley

Shneiderman

(1998). Engagement theory: A framework for technology-based teaching and learning. Educational Technology, 38(5), 20–23.

47.

Khan

Yang

A. J.-F.

Chen

Y. J.

(2020). Impact of buying counterfeit brands on the brand image of original brands. Asia Proceedings of Social Sciences, 6(1), 89–94.

48.

Khan

(2021). Consumer E-loyalty for E-grocery shopping in a metro city of India: Role of flow and TAM antecedents. In Information Resources Management Association (Ed.), Research anthology on E-commerce adoption, models, and applications for modern business (pp. 1612–1632). IGI Global.

49.

Kim

Yoo

Yang

(2020). Online engagement among restaurant customers: The importance of enhancing flow for social media users. Journal of Hospitality & Tourism Research, 44(2), 252–277.

50.

Kim

S.-E.

Park

Tennessee

(2023). Exploring the role of flow experience and telepresence in virtual reality (VR) concerts. Journal of Travel & Tourism Marketing, 40(7), 568–582.

51.

Kong

S. C.

Wang

Y. Q.

(2021). The influence of parental support and perceived usefulness on students’ learning motivation and flow experience in visual programming: Investigation from a parent perspective. British Journal of Educational Technology, 52(4), 1749–1770.

52.

Kumar

(2024). Examining the impact of online service convenience on user engagement and intention to continuously use e-resources: Mediating role of attitude. Global Knowledge, Memory and Communication. Advance online publication. https://doi.org/10.1108/GKMC-05-2024-0272

53.

Lee

Y.-F.

Hwang

G.-J.

Chen

P.-Y.

(2022). Impacts of an AI-based cha bot on college students’ after-class review, academic performance, self-efficacy, learning attitude, and motivation. Educational Technology Research and Development, 70(5), 1843–1865.

54.

Yang

(2024). Exploring the ineffectiveness of gamification health management: a U-shaped relationship between competition and technological exhaustion. Information Technology & People, 37(3), 1229–1250.

55.

Liesa-Orús

Latorre-Cosculluela

Sierra-Sánchez

Vázquez-Toledo

(2023). Links between ease of use, perceived usefulness and attitudes towards technology in older people in university: A structural equation modelling approach. Education and Information Technologies, 28(3), 2419–2436.

56.

Lin

K.-Y.

Huang

T. K.

(2024). Shopping in the digital world: How augmented reality mobile applications trigger customer engagement. Technology in Society, 77, Article 102540.

57.

Lin

Y.-C.

Hou

H.-T.

(2024). The evaluation of a scaffolding-based augmented reality educational board game with competition-oriented and collaboration-oriented mechanisms: Differences analysis of learning effectiveness, motivation, flow, and anxiety. Interactive Learning Environments, 32(2), 502–521.

58.

Liu

L. A.

Hwang

G. J.

(2024). Effects of metalinguistic corrective feedback on novice EFL students’ digital game-based grammar learning performances, perceptions and behavioural patterns. British Journal of Educational Technology, 55(2), 687–711.

59.

Locke

E. A.

Latham

G. P.

(2019). The development of goal setting theory: A half century retrospective. Motivation Science, 5(2), 93.

60.

Deng

Huang

Song

Shen

Fan

Zhang

(2022). The roles of mobile app perceived usefulness and perceived ease of use in app-based Chinese and English learning flow and satisfaction. Education and Information Technologies, 27(7), 10349–10370.

61.

Luarn

Chen

C.-C.

Chiu

Y.-P.

(2023). Enhancing intrinsic learning motivation through gamification: A self-determination theory perspective. The International Journal of Information and Learning Technology, 40(5), 413–424.

62.

Man

Liu

(2024). Motivational beliefs as predictors of academic achievement of translation students: A situated expectancy-value theory perspective. Journal of Multilingual and Multicultural Development, 1–15.

63.

Maqableh

Jaradat

Azzam

A. A.

(2021). Exploring the determinants of students’ academic performance at university level: The mediating role of internet usage continuance intention. Education and Information Technologies, 26, 4003–4025.

64.

Mehta

Vyas

(2022). A systematic literature review on the experience of flow and its relation to intrinsic motivation in students. Indian Journal of Positive Psychology, 13(3), 299–304.

65.

Muhammad Sohail Jafar

Ahmad

Chen

(2024). Metaverse in human behavior: The role of telepresence and flow experience on consumers’ shopping behavior in the Metaverse. SAGE Open, 14(2), Article 21582440241261256.

66.

Onesi-Ozigagun

Ololade

Y. J.

Eyo-Udo

N. L.

Ogundipe

D. O.

(2024). Revolutionizing education through AI: A comprehensive review of enhancing learning experiences. International Journal of Applied Research in Social Sciences, 6(4), 589–607.

67.

Parahiyanti

C. R.

Dimara

N. I.

(2024). Impulsive buying in TikTok live streaming: Enhancing the role of telepresence, brand trust, and flow state. Innovation, Technology, and Entrepreneurship Journal, 1(1), 42–54.

68.

Park

D. Y.

Kim

(2023). Determinants of intentions to use digital mental healthcare content among university students, faculty, and staff: motivation, perceived usefulness, perceived ease of use, and parasocial interaction with AI Chatbot. Sustainability, 15(1), 872.

69.

Park

H. E.

Yap

S.-F.

(2024). A dualistic approach to harmonious and obsessive passion: The impact of game mechanics on mobile gaming engagement. Electronic Commerce Research, 25, 4923–4946.

70.

Pedrero-Esteban

L. M.

Barrios-Rubio

(2024). Digital communication in the age of immediacy. Digital, 4(2), 302–315.

71.

Rheinberg

Engeser

(2018). Intrinsic motivation and flow. In Motivation and action (pp. 579–622). Springer International Publishing.

72.

Ringle

Da Silva

Bido

(2015). Structural equation modeling with the SmartPLS. Brazilian Journal of Marketing, 13(2), 56–73.

73.

Ringle

C. M.

Wende

Becker

J.-M.

(2024). SmartPLS 4. SmartPLS. https://www.smartpls.com

74.

Ruangkanjanases

Khan

Sivarak

Rahardja

Chen

S.-C.

(2024). Modeling the consumers’ flow experience in E-commerce: The integration of ECM and TAM with the antecedents of flow experience. SAGE Open, 14(2), Article 21582440241258595.

75.

Sadoughi

Eskandari

(2024). The relationship between intrinsic motivation and academic flow among medical students: The mediating role of academic engagement. Research in Medical Education, 16(1), 22–32.

76.

Shao

Chen

(2021). Understanding individuals’ engagement and continuance intention of MOOCs: The effect of interactivity and the role of gender. Internet Research, 31(4), 1262–1289.

77.

Shernoff

D. J.

(2010). Engagement in after-school programs as a predictor of social competence and academic performance. American Journal of Community Psychology, 45, 325–337.

78.

Shkëmbi

Treska

(2023). A review of the link between self-efficacy, motivation and academic performance in students. European Journal of Social Science Education and Research, 10(1 Suppl), 23–31.

79.

Sokro

Obuobisa-Darko

Okpattah

(2025). Learner attitudes, satisfaction and success in online learning environments: A mediation-moderation model for higher educational institutions. International Journal of Educational Management, 39(3), 747–780.

80.

Sweller

(2023). The development of cognitive load theory: replication crises and incorporation of other theories can lead to theory expansion. Educational Psychology Review, 35(4), 95.

81.

Taber

K. S.

(2018). The use of Cronbach’s alpha when developing and reporting research instruments in science education. Research in Science Education, 48(6), 1273–1296.

82.

Taghizade

Musavian

S. S.

Hosseininik

S. S.

(2024). Psychometric properties of the Persian Version of the Online Student Engagement Questionnaire: A transcultural adaptation and psychometric study. Interdisciplinary Journal of Virtual Learning in Medical Sciences, 15(3), 226–240.

83.

Tao

Wang

Zhang

(2022). Key characteristics in designing massive open online courses (MOOCs) for user acceptance: An application of the extended technology acceptance model. Interactive Learning Environments, 30(5), 882–895.

84.

Tawafak

R. M.

Malik

S. I.

Alfarsi

(2021). Impact of technologies during the COVID-19 pandemic for improving behavioral intention to use e-learning. International Journal of Information and Communication Technology Education (IJICTE), 17(3), 137–150.

85.

Tawafak

R. M.

Romli

A. B.

Arshah

R. B. A.

(2018). Continued intention to use UCOM: Four factors for integrating with a technology acceptance model to moderate the satisfaction of learning. IEEE Access, 6, 66481–66498.

86.

Tenenhaus

Vinzi

V. E.

Chatelin

Y.-M.

Lauro

(2005). PLS path modeling. Computational Statistics & Data Analysis, 48(1), 159–205.

87.

Urdan

Kaplan

(2020). The origins, evolution, and future directions of achievement goal theory. Contemporary Educational Psychology, 61, Article 101862.

88.

Wandira

Fauzi

Nurahim

(2024). Analysis of factors influencing behavioral Intention to use cloud-based academic information system using extended technology acceptance model (TAM) and expectation-confirmation model (ECM). Journal of Information Systems Engineering and Business Intelligence, 10(2), 179–190.

89.

Wang

Yan

(2025). When the mindful ones experience flow: a moderated-mediation model of purchase intention in live commerce. Information Technology & People. Advance online publication. https://doi.org/10.1108/ITP-04-2023-0377

90.

Wang

Lin

Liu

Song

Huang

Jiang

(2022). Perceived usefulness predicts second language learners’ continuance intention toward language learning applications: A serial multiple mediation model of integrative motivation and flow. Education and Information Technologies, 27, 1–17.

91.

Wang

Pang

Wallace

M. P.

Wang

Chen

(2024). Learners’ perceived AI presences in AI-supported language learning: A study of AI as a humanized agent from community of inquiry. Computer Assisted Language Learning, 37(4), 814–840.

92.

Wang

X.-M.

Q.-N.

Hwang

G.-J.

X.-H.

(2023). Learning with digital technology-facilitated empathy: An augmented reality approach to enhancing students’ flow experience, motivation, and achievement in a biology program. Interactive Learning Environments, 31(10), 6988–7004.

93.

Waterman

A. S.

Schwartz

S. J.

Conti

(2024). Resolving the paradox of work: Generalization of the roles of self-determination, the balance of challenges and skills, and self-realization values in intrinsic motivation across activity domains. The Journal of Positive Psychology, 19(2), 356–368.

94.

Wei

(2021). Telepresence and interactivity in mobile learning system: Its relation with open innovation. Journal of Open Innovation: Technology, Market, and Complexity, 7(1), 78.

95.

Wei

Yin

Shi

Yang

H. H.

(2024). Research on the impact of social presence, academic self-efficacy, and flow experience on students’ learning engagement in blended synchronous learning. 2024 International Symposium on Educational Technology (ISET). IEEE (Institute of Electrical and Electronics Engineers Inc).

96.

Wetzels

Odekerken-Schröder

Van Oppen

(2009). Using PLS path modeling for assessing hierarchical construct models: Guidelines and empirical illustration. MIS Quarterly, 33, 177–195.

97.

Xie

(2024). Research on the influence of perceived quality on users’ continuance usage intention of online live streaming class platforms: the mediating role of flow experience and the moderating impact of perceived usefulness. Frontiers in Psychology, 15, Article 1455597.

98.

Wang

Khan

Zhao

(2021). Consumer flow experience of senior citizens in using social media for online shopping. Frontiers in Psychology, 12, Article 732104.

99.

Yan

Eng

L. G.

Seong

L. C.

(2024). Influencing factors of continuous intention to use E-learning system of undergraduates in Guangxi, China: The mediating role of perceived ease of use and perceived usefulness. SAGE Open, 14(4), Article 21582440241305231.

100.

Yao

Fan

Duan

(2024). The effect of mindfulness on the promotion of graduate students’ scientific research creativity: The chain mediating role of flow experience and creative self-efficacy. Journal of Intelligence, 12(3), 24.

101.

Zhao

Khan

(2021). The students’ flow experience with the continuous intention of using Online English Platforms. Frontiers in Psychology, 12, 807084.

Unlocking Continuous Engagement in AI-Driven Education: An Integrated Model of Flow Experience,Academic Performance,and Technology Acceptance

Abstract

Plain Language Summary

Keywords

Introduction

Theoretical Background and Hypotheses Development

Flow and Continuous Intention

Antecedents of Flow

Intrinsic Motivation and Flow

Immediacy and Flow

Competition and Flow

Telepresence and Flow

Feedback and Flow

Flow and Online Engagement

Flow and Perceived Ease of Use

Flow and Perceived Usefulness

Flow and Academic Performance

Perceived Ease of Use and Perceived Usefulness

Online Engagement and Continuous Intention

Perceived Ease of Use and Continuous Intention

Perceived Usefulness and Continuous Intention

Academic Performance and Continuous Intention

Methodology

Data Collection

Research Instrument

Data Analysis

Convergent and Discriminant Validity

Empirical Results

Discussions

Comparison With Other Studies

Theoretical Implications

Practical Implications

Research Limitations and Future Research Directions

Footnotes

Appendix

ORCID iDs

Ethical Considerations

Consent to Participate

Author Contributions

Funding

Declaration of Conflicting Interests

Data Availability Statement

References