Next Article in Journal
Health Decisions Under Uncertainty: The Roles of Conspiracy Beliefs and Institutional Trust
Previous Article in Journal
Cognitive–Linguistic Profiles of German Adults with Dyslexia
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Exploring the Dynamics of Artificial Intelligence Literacy on English as a Foreign Language Learners’ Willingness to Communicate: The Critical Mediating Roles of Artificial Intelligence Learning Self-Efficacy and Classroom Anxiety

1
School of Foreign Languages, Huainan Normal University, Huainan 232001, China
2
School of Foreign Studies, China University of Political Science and Law, Beijing 100088, China
3
School of Foreign Studies, Soochow University, Suzhou 215006, China
*
Authors to whom correspondence should be addressed.
Behav. Sci. 2025, 15(4), 523; https://doi.org/10.3390/bs15040523
Submission received: 12 February 2025 / Revised: 29 March 2025 / Accepted: 3 April 2025 / Published: 14 April 2025
(This article belongs to the Section Cognition)

Abstract

:
The increasing incorporation of artificial intelligence (AI) in English as a foreign language (EFL) instruction has garnered much attention on the importance of technological elements in language instruction. However, while AI in education (AIED) is still in its early development, research on how learners’ AI literacy affects their language learning outcomes is insufficient. Furthermore, studies examining the impact of learners’ emotional states within the context of AIED are remarkably few. This study examines the interplay between AI literacy and EFL learners’ willingness to communicate (WTC), emphasizing the mediating roles of learners’ AI learning self-efficacy and foreign language classroom anxiety. This study utilizes structural equation modeling, analyzing data from 517 university students in China to construct a prediction model for WTC in AI-enhanced EFL contexts. The findings indicate that AI literacy improves self-efficacy in AI learning and diminishes classroom anxiety, both of which are significant mediators in the relationship between AI literacy and willingness to communicate. The study highlights the imperative of integrating AI literacy into EFL instruction to enhance learners’ expressive confidence and mitigate fear. The findings improve understanding of the interplay between AI literacy, psychological factors, and language learning outcomes, offering practical insights for the integration of AI in EFL education.

1. Introduction

The breadth and depth of technology-enhanced English as a foreign language (EFL) instruction have increased in recent decades due to the quick development of computer and Internet technologies. The incorporation of contemporary educational technology in EFL teaching has elevated the technological proficiency demanded from both educators and learners. Research has identified various literacies—information literacy, computer literacy, ICT literacy, media literacy, digital literacy, and AI literacy—each emphasizing specific facets of learners’ technological competence within the context of technology-enhanced EFL instruction (Hernández-Marín et al., 2024; Nichols & Stornaiuolo, 2019; H. Park et al., 2021; Zhang et al., 2024). The paramount importance of technology in EFL research and education is increasingly acknowledged (Lee, 2019; Lee & Hsieh, 2019; Reinders & Wattana, 2015; Zare et al., 2022).
In conventional EFL instruction, particularly in listening and speaking, willingness to communicate (WTC) has been a prominent topic of scholarly investigation. Empirical research conducted in various global locations over the past five years demonstrates that EFL learners’ WTC is influenced by numerous factors, including the learning environment, intrinsic motivation, self-efficacy, emotional states, and personal backgrounds (Minguez-Alcaide & Perez-Tejada, 2023; Y. L. Wang et al., 2024). The thorough integration of modern educational technology, especially computer-assisted tools, into EFL instruction has garnered significant attention from scholars about the relationship between technological settings and learners’ communicative objectives (Hsieh et al., 2023; H. Huang & Li, 2024). Over the past decade, the advent of artificial intelligence in education (AIED) has emerged as a trend in EFL instruction, addressing learners individualized educational needs and promoting their autonomy and independence (Cacho, 2024; Y. Park & Doo, 2024). The impact of AI literacy, defined as EFL learners’ ability to understand and effectively employ AI technologies, on learners’ WTC has not been extensively examined.
Conventional educational frameworks in EFL research have recognized individual learner emotions and psychological attributes, such as self-efficacy, enjoyment, resilience, and anxiety, as important factors influencing WTC (Al-Saidat et al., 2023; Lee & Chiu, 2023). Nevertheless, within an AIED framework, the influence of technological factors, such as learners’ AI literacy, on their emotional states and the subsequent effect on WTC remains little examined and lacks empirical evidence.
This research, based on self-efficacy theory, formulates a predictive model for EFL learners’ WTC inside an AIED framework. It investigates the correlation among learners’ AI literacy, self-efficacy, and anxiety within foreign language classrooms, as well as the impact of individual psychological emotions on AI-enhanced EFL education. This research intends to enhance our understanding of the impact of technological factors on WTC in AI-enhanced EFL learning and to clarify the importance of learners’ particular psychological states. The study’s conclusions hold empirical significance, highlighting the crucial role of learners in AIED environments.

2. Literature Review

2.1. Willingness to Communicate

Willingness to communicate refers to an individual’s readiness to enter into conversation with specific people under certain conditions using a language (Baker & Watson, 2015). It is considered a critical component in language learning and communication contexts, as it reflects a person’s intention to initiate and engage in discourses (Zare et al., 2022). Initially, it was developed to account for individual differences in first language communication (McCroskey & Richmond, 1990). Over time, it has been expanded to explain second language learning and communication, particularly in non-English-speaking countries where English is often taught as a foreign language (Baker & MacIntyre, 2000; M. Liu, 2017). Previous research has identified that the willingness of learners to engage in communicative language activities is primarily influenced by sociocultural and individual psychological factors. These factors are predominantly manifested in aspects such as individual cultural background, cognitive abilities, teacher–student interaction, and learners’ personal emotions (Jelinková et al., 2023; Pawlak & Mystkowska-Wiertelak, 2015; Weda et al., 2021). Despite the extensive research on WTC in the context of EFL instruction, studies examining WTC within the context of AIED remain relatively limited, especially in the era where AI technologies are increasingly integrated into EFL education (Fathi et al., 2024; F. Huang & Zou, 2024).

2.2. AI Literacy

AI literacy (AIL), as delineated in contemporary scholarly discourse, refers to an individual’s capacity to harness AI-related knowledge and skills, thereby fostering a robust comprehension of AI’s foundational principles and practical applications (Belda-Medina & Calvo-Ferrer, 2022; Dai et al., 2020). Within the educational sphere, the integration of AI literacy into curricula is essential for equipping students to navigate an AI-augmented future. This preparation entails the cultivation of a comprehensive understanding of AI, encompassing its underlying mechanisms, the critical evaluation of its societal implications, and the development of competencies for effective collaboration with AI systems (Allen & Kendeou, 2024). In the realm of EFL education, AIL is becoming crucial for learners, enabling them to critically understand and assess AI-driven applications in educational platforms and tools. Nonetheless, the relationship between EFL learners’ AI literacy and their language learning outcomes has received limited attention in current research, with even less exploration of the pathways through which this relationship operates.

2.3. Foreign Language Classroom Anxiety

The emotional landscape of foreign language learning has long intrigued second language acquisition scholars, with initial research predominantly focusing on the detrimental role of negative emotions in the language learning process. MacIntyre (1999) conceptualizes foreign language classroom anxiety (FLCA) as “the worry and negative emotional reaction aroused when learning or using a second language,” which is shaped by the specific challenges and experiences inherent to learning a language within an educational setting (Dewaele et al., 2024).
FLCA is intrinsically linked to feelings of apprehension, trepidation, and nervousness that learners often encounter during the process of acquiring a foreign language. It fosters a sense of insecurity and the perception of psychological threats, potentially stifling linguistic development and classroom participation (Alberth, 2022). This constellation of negative emotions can significantly impede the optimal performance of language learners, particularly their willingness to communicate (Amirian et al., 2022; Baran-Lucarz, 2014; Bensalem, 2022; Dewaele, 2019).
With the widespread application of AI in EFL education, the impact of AI technology on learners’ FLCA and the influence of FLCA on EFL learning outcomes in the context of AI in AIED warrant further exploration.

2.4. Self-Efficacy Theory and AI Learning Self-Efficacy

As a fundamental concept within educational theory, self-efficacy, initially conceptualized by Bandura (Bandura & Wessels, 1997) as part of social cognitive theory, refers to an individual’s fundamental belief in their capabilities to navigate and triumph over difficult situations. This overarching belief system further differentiates into task-specific self-efficacy, which provides a more detailed assessment of an individual’s confidence in their ability to excel in specific tasks or environments (Q. Liu et al., 2020). Self-efficacy is influenced by a range of factors, including mastery experiences, vicarious experiences, social persuasion, psychological and physiological states, goal setting, social support, individual differences, and educational environment (Calaguas & Consunji, 2022; D. Shen et al., 2013; Yesilyurt et al., 2016). The self-efficacy theory posits that individuals’ beliefs in their capabilities to perform specific tasks significantly influence their motivation, behavior, and performance outcomes (H. Liu et al., 2024).
In the domain of AI, the concept of self-efficacy has been specialized to address a particular facet: AI learning self-efficacy (ALSE). This construct differs from the broader notion of self-efficacy by focusing on an individual’s confidence in their ability to perform specific tasks or navigate particular situations that are centered around AI technologies. ALSE evaluates how well-equipped and assured individuals feel when it comes to applying AI tools and systems effectively to achieve their goals (Y.-Y. Wang & Chuang, 2024). This tailored measure of self-efficacy is crucial as AI becomes more integrated into various fields, influencing not only the performance but also the learning and adoption of AI technologies (D. Chen et al., 2024). Nevertheless, the empirical investigation of self-efficacy’s influence on the acquisition of novel AI technologies remains scarce (Y.-M. Wang et al., 2024).

2.5. The Present Study and Hypotheses

This study, guided by self-efficacy theory, explores the dynamic relationships among learners AIL, AISE, FLCA, and WTC in AI-assisted EFL teaching contexts (see Figure 1).
Literacy, as an indicator of individual cognitive ability, has been shown to positively predict self-efficacy (Niu et al., 2021; C. Park, 2024). This relationship is particularly pronounced in the context of digital literacy. For instance, Gumus et al. (2024) found that digital literacy significantly enhances secondary school students’ self-efficacy in both computer programming and computational thinking. Similarly, Dai (2024) demonstrated that AI literacy among upper elementary students (ages 10–12) can substantially mitigate classroom tension. Although the existing literature offers limited evidence regarding the predictive role of AI literacy on communicative willingness, X. Shen et al. (2024) empirically confirmed that digital literacy positively influences the communicative willingness of second language learners. Building on these findings, the following hypotheses are proposed for this study:
H1. 
EFL learners’ AI literacy positively influences their AI learning self-efficacy.
H2. 
EFL learners’ AI literacy negatively influences their foreign language classroom anxiety.
H3. 
EFL learners’ AI literacy positively influences their willingness to communicate.
Learners’ self-efficacy reflects their confidence in their own learning behaviors. In the context of AI in education (AIED), D. Chen et al. (2024) found that second language learners’ AI self-efficacy negatively impacts their anxiety, which in turn affects their use of AI tools. Similarly, Li et al. (2024) empirically demonstrated in their study on Chinese learners’ willingness to use ChatGPT-3.5 that AI learning self-efficacy positively predicts communicative willingness. Furthermore, the mediating role of self-efficacy in AIED contexts has been confirmed in prior research (B.-J. Kim & Kim, 2024; B.-J. Kim & Lee, 2024). Based on the above analysis, the following hypotheses are proposed for this study:
H4. 
EFL learners’ AI learning self-efficacy negatively influences their foreign language classroom anxiety.
H5. 
EFL learners’ AI learning self-efficacy positively influences their willingness to communicate.
H6. 
EFL learners’ AI learning self-efficacy acts as a mediator between EFL learners’ AI literacy and their willingness to communicate.
Prior research has conclusively demonstrated that foreign language classroom anxiety significantly hampers learners’ willingness to communicate (Amirian et al., 2022; Dewaele, 2019). This adverse relationship has also been empirically confirmed in the context of AI in education, where studies have shown that learner anxiety negatively predicts foreign language communicative willingness (H. Liu & Fan, 2025; Zhang et al., 2024). Additionally, the mediating role of anxiety has been extensively examined in prior studies. Based on these findings, the following research hypotheses are proposed in this study:
H7. 
EFL learners’ foreign language classroom anxiety negatively influences their willingness to communicate.
H8. 
EFL learners’ foreign language classroom anxiety acts as a mediator between EFL learners’ AI literacy and their willingness to communicate.
H9. 
EFL learners’ AI learning self-efficacy and their foreign language classroom anxiety sequentially mediate the relationship between EFL learners’ AI literacy and their willingness to communicate.

3. Method

3.1. Participants and Procedure

The study population consisted of 517 non-English major undergraduates from four higher educational institutions in Anhui Province, China. Participants were enrolled in university English-speaking courses that employed artificial intelligence (AI)-assisted software. Teachers encouraged students to utilize AI software across various terminal devices, including smartphones and desktop computers. In the domain of English language acquisition, participants frequently employed AI technology for voice correction training, vocabulary retention, automatic translation, corpus development, information retrieval, speaking practice with AI companions, and voice transcription. The respondents consisted of 318 females and 199 males. The demographic data are provided in Table 1.
The survey employed random sampling, focusing on the communicative willingness of learners in AI-augmented EFL learning contexts. Questionnaires were distributed by EFL instructors utilizing AI technology in their various classrooms. Before the release of the questionnaires, participants were briefed on the study’s objectives and guaranteed the confidentiality of their information. They were directed to respond to the questions with accuracy but were also notified that they might withdraw at any moment if they chose not to participate. Participants received a canvas tote bag as a token of appreciation upon completing the survey. Over a three-week duration, 650 questionnaires were distributed, yielding 588 returns and 517 valid responses.

3.2. Instruments

To construct a model for predicting the willingness to communicate in foreign languages among EFL learners in an AIED context and to explore the interactive relationships between learners’ artificial intelligence literacy, self-efficacy in AI learning, foreign language classroom anxiety, and willingness to communicate in foreign languages, this research conducted a survey on the subjects using relevant questionnaires. The participants’ replies on the relevant problems are assessed using a 5-point Likert scale, where 1 signifies “strongly disagree” and 5 denotes “strongly agree”.
The study utilized Chai et al. (2020)’s AIL scale, which assesses the degree to which EFL learners comprehend the operational mechanisms and cognitive processes underlying AI-driven learning technologies. The scale encompasses six items, including “I know the processes through which deep learning enables AI to perform voice recognition tasks”. In this study, a Cronbach’s α coefficient of 0.914 was reported for AIL, signifying substantial internal consistency.
The assessment of AI learning self-efficacy utilized the scale created by Y.-M. Wang et al. (2024), comprising five items (e.g., “I do not think I lack the foundation for AI-related skills learning.”). The scale is designed to measure learners’ self-efficacy in learning AI technologies, which aligns with the context established for this study. In the present study, the reliability index of ALSE estimated through Cronbach’s alpha was 0.871, signifying adequate internal consistency.
The current research employed the instrument created by Dewaele and MacIntyre (2014) to evaluate participants’ perceptions of anxiety in foreign language educational settings. The scale in question is a classic instrument for measuring foreign language learners’ classroom anxiety and has been widely employed by numerous scholars across different educational settings. This scale comprises eight items (e.g., “I get nervous and confused when I am speaking in my foreign language class”). The analysis revealed FLCA enjoyed good internal consistency (Cronbach’s α = 0.942).
This study assessed EFL learners’ willingness to communicate using a modified scale by Dewaele (2019). Since its introduction, the scale has been widely employed to assess foreign language learners’ willingness to communicate. It consists of six items (e.g., “I can imagine a situation where I am speaking English with foreigners.”). In the current study, the reliability of learners’ WTC scale estimated via Cronbach’s α was 0.946, indicating substantial internal consistency.

3.3. Data Analysis

The data were analyzed utilizing SPSS version 26.0 and AMOS version 24.0. Initial descriptive and correlational analyses were performed to summarize the data. Subsequently, confirmatory factor analysis was utilized to validate the measurement model. Subsequently, structural equation modeling was employed to assess the model’s fit and to investigate the correlations via regression and mediation analysis. The goodness-of-fit indices conformed to the following criteria: χ2/df < 3, RMSEA ≤ 0.08, SRMR ≤ 0.10, GFI, CFI, TLI, and AGFI > 0.90 (Kline, 2023).

4. Results

4.1. Test for Common Method Variance

To investigate common method bias, the present study employs the Harman single-factor test (Podsakoff et al., 2003). All four factors were included in the maximum likelihood extraction method. The value of Harman’s single factor is 32.135%, well below the critical threshold of 40% (Hair et al., 1998). The common method bias was not a significant issue in the present study.

4.2. Measurement Analysis

This study utilized SPSS 26.0 to evaluate the measurement model’s reliability and validity. The eigenvalue method detected four prominent constructs, each with a loading beyond 0.65. The internal consistency among items within each dimension is robust, since the Cronbach’s alpha for each construct was above 0.7. Furthermore, as Table 2 indicated, the composite reliability (CR) and average variance extracted (AVE) of each dimension were higher than the threshold criteria of 0.7 and 0.5, respectively. The square root values of AVE for each construct surpassed the correlation between these constructs (refer to Table 3), meeting the requirement of the discriminant validity.

4.3. Test for Structural Model

This study utilized AMOS 24.0 to do a goodness-of-fit examination of the theoretical research model. According to Kline’s guidelines, the optimal model fit indices are χ2/df < 3, with CFI, GFI, IFI, AGFI, and TLI ≥ 0.90, and RMSEA and SRMR ≤ 0.08. In this study, χ2 equals 379.080 with degrees of freedom (df) equal to 269, yielding a χ2/df ratio of 1.409. The other fit indices, including GFI (0.945), AGFI (0.934), CFI (0.988), TLI (0.987), RMSEA (0.028), and SRMR (0.036), satisfied the requirements (see Figure 2). The results demonstrate strong data and model alignment, facilitating additional empirical investigation.
Table 4 illustrates that in an AI-augmented EFL learning context, learners’ artificial intelligence literacy positively impacts their self-efficacy on AI learning (β = 0.420, p < 0.001, SE = 0.033) and negatively predicts their foreign language classroom anxiety (β = −0.183, p < 0.001, SE = 0.041), hence corroborating hypotheses H1 and H2. However, the study revealed that learners’ AI literacy did not substantially affect their preparedness to communicate (β = 0.075, p = 0.147, SE = 0.041), leading to the dismissal of hypothesis H3. Concurrently, learners’ self-efficacy in AI learning negatively predicted their anxiety in foreign language classrooms and positively influenced their communication preparedness (β = −0.178, p < 0.001, SE = 0.062; β = 0.151, p = 0.004, SE = 0.063), thereby substantiating hypotheses H4 and H5. The results indicated that learners’ anxiety in foreign language classrooms negatively impacted their readiness to speak (β = −0.216, p < 0.001, SE = 0.049), hence supporting hypothesis H7.

4.4. Mediation Analysis

The findings indicate an insignificant direct effect of AIL on WTC in the proposed model (p = 0.161), suggesting full mediation. The results show that AIL has a significant indirect effect on WTC through both ALSE (β = 0.051, SE = 0.020, 90% CI [0.015, 0.093]) and FLCA (β = 0.032, SE = 0.014, 90% CI [0.010, 0.065]), thereby supporting the establishment of hypotheses H6 and H8. Moreover, the study demonstrates that ALSE and FLCA together exert a serial mediating effect on the relationship between AIL and WTC (β = 0.013, SE = 0.005, 90% CI [0.005, 0.025]), confirming hypothesis H9. Among the three mediating pathways, the sequence “AIL → ALSE → WTC” accounts for the highest proportion of the indirect effect, at 53.2% (see Table 5).

5. Discussion

The study employed variables such as EFL undergraduates’ AI literacy, foreign language classroom anxiety, and AI learning self-efficacy to develop a predictive model for their willingness to communicate in an AI-enhanced educational environment.
Firstly, the finding that EFL learners’ AIL positively influences their ALSE aligns with the work of Yao and Wang (2024), who demonstrated that digital literacy significantly impacts self-efficacy in the context of AIED for preservice special-education teachers. This consistency underscores the pivotal role of technological proficiency in bolstering learners’ confidence when utilizing AI tools for educational purposes. In the current study, this positive effect is attributed to the empowering nature of AI tools, which provide learners with greater control over their learning process and offer immediate, personalized feedback. This enhanced control and feedback mechanism reinforces learners’ belief in their capacity to effectively employ AI for language learning, thereby augmenting their self-efficacy. Moreover, the inverse relationship between AIL and FLCA is corroborated by D. Chen et al. (2024), who found that higher AI learning self-efficacy among Chinese college students learning a second language was associated with lower AI anxiety and more positive perceptions of AI technology utilization. In the current study, this negative correlation can be explained by the fact that increased AI literacy empowers learners with a better understanding and control over AI tools, which in turn reduces their anxiety levels. As learners become more proficient in using AI for language learning, they feel more confident and less anxious about engaging with these technologies in the classroom setting.
Secondly, the finding that EFL learners’ ALSE negatively impacts their FLCA is consistent with previous research showing a negative correlation between self-efficacy and anxiety in various educational contexts (Awofala et al., 2019; Chou, 2001; Kutuk et al., 2022; X. Wang et al., 2023). This suggests that higher ALSE can reduce FLCA. In AI-enhanced EFL settings, AI tools provide immediate feedback and personalized learning, which can boost learners’ confidence and reduce anxiety. Thus, fostering AI literacy and self-efficacy can create a more positive learning environment. The present study also indicated that AIL does not directly predict WTC. This contrasts with X. Shen et al. (2024), who empirically demonstrated a positive impact of digital literacy on communicative willingness among second language learners. The discrepancy likely arises because AIL, while boosting technical proficiency, does not directly address the psychological and emotional factors influencing WTC. Essentially, WTC is driven by subjective psychological emotions rather than solely by learners’ basic cognitive abilities. Thus, AIL alone cannot significantly impact WTC without mediating factors addressing these psychological aspects.
Finally, the study found that learners’ self-efficacy in AI learning mediates the relationship between AI literacy and their readiness to communicate in English, echoing Ibrahim and Aldawsari (2023), who showed that self-efficacy affects the link between digital competence and academic performance. This indicates that self-efficacy is key in translating AI literacy into communicative readiness. Confidence in using AI for language learning enhances learners’ willingness to communicate, highlighting the need to boost both technical skills and self-efficacy. The study also reveals that in AI contexts, learners’ FLCA significantly mediates the relationship between their AIL and WTC. This finding is consistent with previous research indicating that anxiety acts as a strong moderator influencing learners’ readiness to interact in both online and offline settings in non-AI contexts (Brauer et al., 2023; Hejazi et al., 2023; Piechurska-Kuciel et al., 2021). The mediating role of anxiety in the current study suggests that even with enhanced AI literacy, learners’ WTC is still heavily influenced by their emotional state. High levels of anxiety can inhibit the positive effects of AI literacy on communication, highlighting the importance of addressing emotional barriers alongside technological proficiency in AI-enhanced language learning environments. Furthermore, the study finds that ALSE and FLCA successively mediate the relationship between AIL and WTC. This aligns with prior EFL research highlighting self-efficacy and anxiety as key mediators in language learning (Jin et al., 2024; X. Wang et al., 2023). In AI-enhanced EFL contexts, higher AIL boosts ALSE, as learners gain confidence in using AI tools. This increased self-efficacy then reduces FLCA, making the learning environment less intimidating. Thus, addressing both psychological factors (self-efficacy and anxiety) alongside AIL is crucial for enhancing communicative competence.

6. Conclusions

This study investigated Chinese non-English major university students and used structural equation modeling to analyze the interplay between AI literacy, AI learning self-efficacy, and foreign language classroom anxiety in AI-enhanced EFL environments. It affirmed that AI literacy improves self-efficacy and diminishes classroom anxiety, hence affecting communication willingness—positively through self-efficacy and negatively through anxiety. The study revealed that AI literacy does not significantly influence communicative willingness, suggesting that self-efficacy and classroom anxiety completely mediate this relationship in AI-supported EFL educational environments.
This study contributes theoretically by elucidating the intricate interplay between AI literacy, psychological factors (self-efficacy and anxiety), and communicative willingness in AI-enhanced EFL contexts, and practically by offering insights for integrating AI literacy into EFL education to optimize learning outcomes and mitigate anxiety. To begin with, universities must align with the AI education landscape by leveraging the positive role of AI technology in foreign language learning while recognizing that mere literacy education has limited direct influence on learners’ communicative willingness. In response, higher education institutions can embed AI foundational theory and practical application modules into existing foreign language courses (X. Liu & Xiao, 2025; Namaziandost & Rezai, 2024), helping students understand AI’s role in language learning and emphasizing the importance of enhancing personal AI literacy during language training.
Next, educators should harness the impact of AI technology on foreign language teaching by using AI tools (e.g., chatbots) to support classroom interaction and group discussions (Jiang, 2025; Kong et al., 2025). They can employ human–AI interactive games to reduce classroom anxiety and utilize intelligent learning companions to boost students’ AI learning self-efficacy and adaptive learning abilities, enabling learners to more effectively express their viewpoints in a foreign language (M.-R. A. Chen, 2024; Xin & Derakhshan, 2025; Zhang et al., 2024).
Lastly, software developers should prioritize enhancing the interactivity of AI software to lower the learning threshold and establish good user habits (Ng et al., 2024). AI software should provide emotional support to help students overcome psychological barriers such as anxiety and lack of confidence, offering positive feedback and encouragement during human–AI interactions to enhance learners’ self-efficacy and improve their willingness and skills to communicate (F. Huang et al., 2024; Ishizu et al., 2024).
The study provides insights into the relationship between AIL and WTC among EFL learners but has notable limitations. First, the sample of Chinese non-English major undergraduates limits the generalizability of the findings to other regions or educational contexts. Future research should include more diverse samples to enhance applicability. Second, the cross-sectional design and self-reported data capture attitudes at a single point in time but do not track changes in psychological states or communicative goals over time. Longitudinal studies using latent growth models could provide a more nuanced understanding of these dynamics. Third, self-reported data may be biased by social desirability, potentially affecting accuracy. Future research should integrate qualitative methods, such as interviews and classroom observations, to validate findings and offer deeper insights into learners’ experiences.

Author Contributions

Conceptualization: Q.Z. and H.L.; formal analysis: Q.Z., H.N. and J.F.; funding acquisition: J.F.; investigation: J.F. and H.L.; methodology: Q.Z., H.N. and J.F.; software: J.F.; supervision: H.L.; writing—original draft: Q.Z., H.N., J.F. and H.L.; writing—review and editing: Q.Z., H.N. and H.L. All authors have read and agreed to the published version of the manuscript.

Funding

This study was funded by the Chinese Society for TAOXINGZHI Studies (grant number: zhongtaohui202411838AH) and the Anhui Provincial Educational Science Research Project (JKT24107). The research was supported by the HNNU Teaching Quality project (grant number: 2024hsjyxm06).

Institutional Review Board Statement

All methods followed the guidelines and regulations of the 1964 Helsinki Declaration and its later amendments. Since this study does not involve intervention and is low-risk, ethical review and approval were waived according to the institutional review boards at the School of Foreign Languages at Huainan Normal University.

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study.

Data Availability Statement

Data will be made available upon reasonable request.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Alberth, A. (2022). Predicting sense of classroom community from foreign language enjoyment and anxiety in an English as a foreign language classroom. International Journal of Emotional Education, 14(1), 38–52. [Google Scholar] [CrossRef]
  2. Allen, L. K., & Kendeou, P. (2024). ED-AI Lit: An interdisciplinary framework for AI literacy in education. Policy Insights from the Behavioral and Brain Sciences, 11(1), 3–10. [Google Scholar] [CrossRef]
  3. Al-Saidat, E., Buragohain, D., & Amer, F. (2023). Role of anxiety in willingness to communicate in the Jordanian EFL context. 3L: Language, Linguistics, Literature—The Southeast Asian Journal of English Language Studies, 29(2), 133–145. [Google Scholar] [CrossRef]
  4. Amirian, S. M. R., Daneshrah, A., & Mehrabadi, H. (2022). Examining the role of foreign language anxiety and attachment style in willingness to communicate. Psychological Studies, 67(4), 480–487. [Google Scholar] [CrossRef]
  5. Awofala, A. O. A., Olabiyi, O. S., Awofala, A. A., Arigbabu, A. A., Fatade, A. O., & Udeani, U. N. (2019). Attitudes toward computer, computer anxiety and gender as determinants of pre-service science, technology, and mathematics teachers’ computer self-efficacy. Digital Education Review, 36, 51–67. [Google Scholar] [CrossRef]
  6. Baker, S. C., & MacIntyre, P. D. (2000). The role of gender and immersion in communication and second language orientations. Language Learning, 50(2), 311–341. [Google Scholar] [CrossRef]
  7. Baker, S. C., & Watson, B. M. (2015). How patients perceive their doctors’ communication: Implications for patient willingness to communicate. Journal of Language and Social Psychology, 34(6), 621–639. [Google Scholar] [CrossRef]
  8. Bandura, A., & Wessels, S. (1997). Self-efficacy. Cambridge University Press. [Google Scholar]
  9. Baran-Lucarz, M. (2014). The link between pronunciation anxiety and willingness to communicate in the foreign-language classroom: The Polish EFL context. Canadian Modern Language Review, 70(4), 445–473. [Google Scholar] [CrossRef]
  10. Belda-Medina, J., & Calvo-Ferrer, J. R. (2022). Using chatbots as AI conversational partners in language learning. Applied Sciences, 12(17), 8427. [Google Scholar] [CrossRef]
  11. Bensalem, E. (2022). The impact of enjoyment and anxiety on English-language learners’ willingness to communicate. Vivat Academia, 155, 91–110. [Google Scholar] [CrossRef]
  12. Brauer, K., Barabadi, E., Aghaee, E., Alrabai, F., Shirvan, M. E., Sendatzki, R., & Vierow, L. M. (2023). Impostor phenomenon and L2 willingness to communicate: Testing communication anxiety and perceived L2 competence as mediators. Frontiers in Psychology, 13, 1060091. [Google Scholar] [CrossRef]
  13. Cacho, R. M. (2024). Integrating generative AI in university teaching and learning: A model for balanced guidelines. Online Learning, 28(3), 55–81. [Google Scholar] [CrossRef]
  14. Calaguas, N. P., & Consunji, P. M. P. (2022). A structural equation model predicting adults’ online learning self-efficacy. Education and Information Technologies, 27(5), 6233–6249. [Google Scholar] [CrossRef] [PubMed]
  15. Chai, C. S., Wang, X., & Xu, C. (2020). An extended theory of planned behavior for the modelling of Chinese secondary school students’ intention to learn artificial intelligence. Mathematics, 8(11), 2089. [Google Scholar] [CrossRef]
  16. Chen, D., Liu, W., & Liu, X. (2024). What drives college students to use AI for L2 learning? Modeling the roles of self-efficacy, anxiety, and attitude based on an extended technology acceptance model. Acta Psychologica, 249, 104442. [Google Scholar] [CrossRef]
  17. Chen, M.-R. A. (2024). The AI chatbot interaction for semantic learning: A collaborative note-taking approach with EFL students. Language Learning & Technology, 28(1), 1–25. Available online: https://hdl.handle.net/10125/73586 (accessed on 28 March 2025).
  18. Chou, H. W. (2001). Effects of training method and computer anxiety on learning performance and self-efficacy. Computers in Human Behavior, 17(1), 51–69. [Google Scholar] [CrossRef]
  19. Dai, Y. (2024). Dual-contrast pedagogy for AI literacy in upper elementary schools. Learning and Instruction, 91, 101899. [Google Scholar] [CrossRef]
  20. Dai, Y., Chai, C.-S., Lin, P.-Y., Jong, M. S.-Y., Guo, Y., & Qin, J. (2020). Promoting students’ well-being by developing their readiness for the artificial intelligence age. Sustainability, 12(16), 6597. [Google Scholar] [CrossRef]
  21. Dewaele, J.-M. (2019). The effect of classroom emotions, attitudes toward English, and teacher behavior on willingness to communicate among English foreign language learners. Journal of Language and Social Psychology, 38(4), 523–535. [Google Scholar] [CrossRef]
  22. Dewaele, J.-M., Ferrer, M. S., Ahmed, I. K., & Albakistani, A. (2024). Does multilingualism affect foreign language classroom anxiety and foreign language enjoyment? Porta Linguarum, IX, 13–29. [Google Scholar] [CrossRef]
  23. Dewaele, J.-M., & MacIntyre, P. D. (2014). The two faces of Janus? Anxiety and enjoyment in the foreign language classroom. Studies in Second Language Learning and Teaching, 4(2), 237–274. [Google Scholar] [CrossRef]
  24. Fathi, J., Rahimi, M., & Derakhshan, A. (2024). Improving EFL learners’ speaking skills and willingness to communicate via artificial intelligence-mediated interactions. System, 121, 103254. [Google Scholar] [CrossRef]
  25. Gumus, M. M., Kukul, V., & Korkmaz, O. (2024). Relationships between middle school students’ digital literacy skills, computer programming self-efficacy, and computational thinking self-efficacy. Informatics in Education, 23(3), 571–592. [Google Scholar] [CrossRef]
  26. Hair, J., Anderson, R., Tatham, R., & Black, W. (1998). Multivariate data analysis (5th ed.). Prentice Hall. [Google Scholar]
  27. Hejazi, S. Y., Sadoughi, M., & Peng, J. E. (2023). The structural relationship between teacher support and willingness to communicate: The mediation of L2 anxiety and the moderation of growth language mindset. Journal of Psycholinguistic Research, 52(6), 2955–2978. [Google Scholar] [CrossRef]
  28. Hernández-Marín, J. L., Castro-Montoya, M. D., & Figueroa-Rodríguez, S. (2024). Media, information and digital literacy: Assessment instrument analysis. Investigación Bibliotecológica, 38(99), 55–73. [Google Scholar] [CrossRef]
  29. Hsieh, W.-M., Yeh, H.-C., & Chen, N.-S. (2023). Impact of a robot and tangible object (R&T) integrated learning system on elementary EFL learners’ English pronunciation and willingness to communicate. Computer Assisted Language Learning, 1–26. [Google Scholar] [CrossRef]
  30. Huang, F., Wang, Y., & Zhang, H. (2024). Modelling generative AI acceptance, perceived teachers’ enthusiasm and self-efficacy to English as a foreign language learners’ well-being in the digital era. European Journal of Education, 59(4), e12770. [Google Scholar] [CrossRef]
  31. Huang, F., & Zou, B. (2024). English speaking with artificial intelligence (AI): The roles of enjoyment, willingness to communicate with AI, and innovativeness. Computers in Human Behavior, 159, 108355. [Google Scholar] [CrossRef]
  32. Huang, H., & Li, M. (2024). The impact of technology-enhanced language learning environments on second language learners’ willingness to communicate: A systematic review of empirical studies from 2012 to 2023. Language Learning & Technology, 28(1), 1–22. Available online: https://hdl.handle.net/10125/73593 (accessed on 28 March 2025).
  33. Ibrahim, R. K., & Aldawsari, A. N. (2023). Relationship between digital capabilities and academic performance: The mediating effect of self-efficacy. BMC Nursing, 22(1), 434. [Google Scholar] [CrossRef]
  34. Ishizu, N., Yeoh, W. L., Okumura, H., & Fukuda, O. (2024). The effect of communicating AI confidence on human decision making when performing a binary decision task. Applied Sciences, 14(16), 7192. [Google Scholar] [CrossRef]
  35. Jelinková, J., Petrus, P., & Laue, A. (2023). The impact of feedback on students’ willingness to communicate in foreign language learning: Systematic review. Journal of Teaching English for Specific and Academic Purposes, 11(3), 663–686. [Google Scholar] [CrossRef]
  36. Jiang, Y. (2025). Interaction and dialogue: Integration and application of artificial intelligence in blended mode writing feedback. Internet and Higher Education, 64, 100975. [Google Scholar] [CrossRef]
  37. Jin, F. Y., Gu, C. F., & Li, Y. (2024). Trait emotional intelligence and foreign language performance: Associations with academic self-efficacy and foreign language anxiety. Frontiers in Education, 9, 1419328. [Google Scholar] [CrossRef]
  38. Kim, B.-J., & Kim, M.-J. (2024). How artificial intelligence-induced job insecurity shapes knowledge dynamics: The mitigating role of artificial intelligence self-efficacy. Journal of Innovation & Knowledge, 9(4), 100590. [Google Scholar] [CrossRef]
  39. Kim, B.-J., & Lee, J. (2024). The mental health implications of artificial intelligence adoption: The crucial role of self-efficacy. Humanities & Social Sciences Communications, 11(1), 1561. [Google Scholar] [CrossRef]
  40. Kline, R. B. (2023). Principles and practice of structural equation modeling (5th ed.). Guilford Press. [Google Scholar]
  41. Kong, X., Liu, Z., Chen, C., Liu, S., Xu, Z., & Tang, Q. (2025). Exploratory study of an AI-supported discussion representational tool for online collaborative learning in a Chinese university. Internet and Higher Education, 64, 100973. [Google Scholar] [CrossRef]
  42. Kutuk, G., Putwain, D. W., Kaye, L. K., & Garrett, B. (2022). Relations between gender stereotyping and foreign language attainment: The mediating role of language learners’ anxiety and self-efficacy. British Journal of Educational Psychology, 92(1), e12446. [Google Scholar] [CrossRef]
  43. Lee, J. S. (2019). EFL students’ views of willingness to communicate in the extramural digital context. Computer Assisted Language Learning, 32(7), 692–712. [Google Scholar] [CrossRef]
  44. Lee, J. S., & Chiu, M. M. (2023). Modeling EFL learners’ willingness to communicate: The roles of face-to-face and digital L2 communication anxiety. Annual Review of Applied Linguistics, 43, 64–87. [Google Scholar] [CrossRef]
  45. Lee, J. S., & Hsieh, J. C. (2019). Affective variables and willingness to communicate of EFL learners in in-class, out-of-class, and digital contexts. System, 82, 63–73. [Google Scholar] [CrossRef]
  46. Li, N., Zhang, L., Lau, K. L., & Liang, Y. (2024). Predicting Chinese language learners’ ChatGPT acceptance in oral language practices: The role of learning motivation and willingness to communicate. Journal of Technology and Chinese Language Teaching, 15(1), 25–48. [Google Scholar]
  47. Liu, H., Chen, B., Li, X., & Zhou, X. (2024). Exploring the predictive role of self-efficacy in engagement among EFL teachers in online teaching: The mediation of buoyancy. The Asia-Pacific Education Researcher, 33(4), 879–888. [Google Scholar] [CrossRef]
  48. Liu, H., & Fan, J. (2025). AI-mediated communication in EFL classrooms: The role of technical and pedagogical stimuli and the mediating effects of AI literacy and enjoyment. European Journal of Education, 60(1), e12813. [Google Scholar] [CrossRef]
  49. Liu, M. (2017). Adult Chinese as a second language learners’ willingness to communicate in Chinese: Effects of cultural, affective, and linguistic variables. Psychological Reports, 120(3), 423–442. [Google Scholar] [CrossRef]
  50. Liu, Q., Liu, J., Cai, J., & Zhang, Z. (2020). The relationship between domain- and task-specific self-efficacy and mathematical problem posing: A large-scale study of eighth-grade students in China. Educational Studies in Mathematics, 105(3), 407–431. [Google Scholar] [CrossRef]
  51. Liu, X., & Xiao, Y. (2025). Chinese university teachers’ engagement with generative AI in different stages of foreign language teaching: A qualitative enquiry through the prism of ADDIE. Education and Information Technologies, 30(1), 485–508. [Google Scholar] [CrossRef]
  52. MacIntyre, P. (1999). Language anxiety: A review of the research for language teachers. In D. J. Young (Ed.), Affect in foreign language and second language learning (pp. 24–45). McGraw-Hill. [Google Scholar]
  53. McCroskey, J. C., & Richmond, V. P. (1990). Willingness to communicate: Differing cultural perspectives. Southern Journal of Communication, 56(1), 72–77. [Google Scholar]
  54. Minguez-Alcaide, X., & Perez-Tejada, J. (2023). Predictors of willingness to communicate in Basque. Rivista di Psicolinguistica Applicata, 23(1), 79–92. [Google Scholar] [CrossRef]
  55. Namaziandost, E., & Rezai, A. (2024). Interplay of academic emotion regulation, academic mindfulness, L2 learning experience, academic motivation, and learner autonomy in intelligent computer-assisted language learning: A study of EFL learners. System, 125, 103419. [Google Scholar] [CrossRef]
  56. Ng, D. T. K., Chen, X. Y., Leung, J. K. L., & Chu, S. K. W. (2024). Fostering students’ AI literacy development through educational games: AI knowledge, affective and cognitive engagement. Journal of Computer Assisted Learning, 40(5), 2049–2064. [Google Scholar] [CrossRef]
  57. Nichols, T. P., & Stornaiuolo, A. (2019). Assembling “digital literacies”: Contingent pasts, possible futures. Media and Communication, 7(2), 14–24. [Google Scholar] [CrossRef]
  58. Niu, Z., Willoughby, J., & Zhou, R. (2021). Associations of health literacy, social media use, and self-efficacy with health information-seeking intentions among social media users in China: Cross-sectional survey. Journal of Medical Internet Research, 23(2), e19134. [Google Scholar] [CrossRef] [PubMed]
  59. Park, C. (2024). Electronic health literacy as a source of self-efficacy among community-dwelling older adults. Clinical Gerontologist, 48(2), 289–298. [Google Scholar] [CrossRef]
  60. Park, H., Kim, H. S., & Park, H. W. (2021). A scientometric study of digital literacy, ICT literacy, information literacy, and media literacy. Journal of Data and Information Science, 6(2), 116–138. [Google Scholar] [CrossRef]
  61. Park, Y., & Doo, M. Y. (2024). Role of AI in blended learning: A systematic literature review. International Review of Research in Open and Distributed Learning, 25(1), 164–196. [Google Scholar] [CrossRef]
  62. Pawlak, M., & Mystkowska-Wiertelak, A. (2015). Investigating the dynamic nature of L2 willingness to communicate. System, 50, 1–9. [Google Scholar] [CrossRef]
  63. Piechurska-Kuciel, E., Ozanska-Ponikwia, K., & Skalacka, K. (2021). Can the neuroticism-willingness to communicate relationship across languages be explained by anxiety? Moderna Språk, 115(4), 176–199. [Google Scholar] [CrossRef]
  64. Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879–903. [Google Scholar] [CrossRef]
  65. Reinders, H., & Wattana, S. (2015). Affect and willingness to communicate in digital game-based learning. ReCALL, 27(1), 38–57. [Google Scholar] [CrossRef]
  66. Shen, D., Cho, M.-H., Tsai, C.-L., & Marra, R. (2013). Unpacking online learning experiences: Online learning self-efficacy and learning satisfaction. Internet and Higher Education, 19, 10–17. [Google Scholar] [CrossRef]
  67. Shen, X., Shen, X., & Peng, J.-E. (2024). Digital storytelling: A literacy-building tool to promote willingness to communicate in a second language. RELC Journal, 55(3), 819–826. [Google Scholar] [CrossRef]
  68. Wang, X., Liu, Y. L., Ying, B., & Lin, J. (2023). The effect of learning adaptability on Chinese middle school students’ English academic engagement: The chain mediating roles of foreign language anxiety and English learning self-efficacy. Current Psychology, 42(8), 6682–6692. [Google Scholar] [CrossRef]
  69. Wang, Y. L., Wu, H. W., & Wang, Y. S. (2024). Engagement and willingness to communicate in the L2 classroom: Identifying the latent profiles and their relationships with achievement emotions. Journal of Multilingual and Multicultural Development, 1–17. [Google Scholar] [CrossRef]
  70. Wang, Y.-M., Wei, C.-L., Lin, H.-H., Wang, S.-C., & Wang, Y.-S. (2024). What drives students’ AI learning behavior: A perspective of AI anxiety. Interactive Learning Environments, 32(6), 2584–2600. [Google Scholar] [CrossRef]
  71. Wang, Y.-Y., & Chuang, Y.-W. (2024). Artificial intelligence self-efficacy: Scale development and validation. Education and Information Technologies, 29(4), 4785–4808. [Google Scholar] [CrossRef]
  72. Weda, S., Atmowardoyo, H., Rahman, F., Said, M. M., & Sakti, A. E. F. (2021). Factors affecting students’ willingness to communicate in EFL classroom at higher institution in Indonesia. International Journal of Instruction, 14(2), 719–734. [Google Scholar] [CrossRef]
  73. Xin, Z., & Derakhshan, A. (2025). From excitement to anxiety: Exploring English as a foreign language learners’ emotional experiences in the artificial intelligence-powered classrooms. European Journal of Education, 60(1), e12845. [Google Scholar] [CrossRef]
  74. Yao, N., & Wang, Q. (2024). Factors influencing pre-service special education teachers’ intention toward AI in education: Digital literacy, teacher self-efficacy, perceived ease of use, and perceived usefulness. Heliyon, 10(14), e34894. [Google Scholar] [CrossRef]
  75. Yesilyurt, E., Ulas, A. H., & Akan, D. (2016). Teacher self-efficacy, academic self-efficacy, and computer self-efficacy as predictors of attitude toward applying computer-supported education. Computers in Human Behavior, 64, 591–601. [Google Scholar] [CrossRef]
  76. Zare, M., Shooshtari, Z. G., & Jalilifar, A. (2022). The interplay of oral corrective feedback and L2 willingness to communicate across proficiency levels. Language Teaching Research, 26(6), 1158–1178. [Google Scholar] [CrossRef]
  77. Zhang, H., Zhu, C., Sang, G. Y., & Questier, F. (2024). Effects of digital media literacy course on primary school students’ digital media literacy: An experimental study. International Journal of Technology and Design Education, 34(1), 1–17. [Google Scholar] [CrossRef]
Figure 1. Hypothesized model.
Figure 1. Hypothesized model.
Behavsci 15 00523 g001
Figure 2. SEM test model.
Figure 2. SEM test model.
Behavsci 15 00523 g002
Table 1. Individual’s demographic characteristics.
Table 1. Individual’s demographic characteristics.
CategoryN%
Gender
Female31861.5
Male19938.5
Grade
Freshmen10219.7
Sophomore8917.2
Junior16632.1
Senior6011.6
Age
≤1910520.3
2014427.9
2119537.7
≥227214.1
Field of study
Natural science11021.3
Engineering and technical study17233.2
Humanities23545.5
L (Length of weekly AI use)
L ≤ 1 h9518.40%
1 h < L ≤ 3 h19036.8
3 h < L ≤ 7 h11722.6
7 h < L ≤ 10 h5610.8
10 h < L ≤ 15 h122.3
L > 15 h479.1
Total517
Table 2. Factor analysis, construct reliability and convergent validity.
Table 2. Factor analysis, construct reliability and convergent validity.
ConstructsItemsEstimateS.E.Z ValuepFactor LoadingαC.R.AVE
FLCAFLCA011.000 ***0.8540.9420.9420.671
FLCA020.8890.03823.462***0.817
FLCA030.8780.03723.671***0.821
FLCA040.9090.04022.823***0.804
FLCA050.8480.03623.475***0.817
FLCA060.9350.03526.661***0.878
FLCA070.8530.03822.401***0.794
FLCA080.8290.04020.989***0.763
AILAIL11.000 ***0.8160.9140.9230.667
AIL20.9810.04920.277***0.781
AIL30.9790.04919.869***0.778
AIL41.0210.04721.535***0.825
AIL50.9310.04620.171***0.787
AIL60.9880.04720.855***0.906
ALSEALSE11.000 ***0.7980.8710.8730.580
ALSE20.9970.04920.277***0.842
ALSE30.8860.05017.597***0.744
ALSE40.8860.05117.475***0.740
ALSE50.7590.04815.664***0.674
WTCWTC11.000 ***0.8830.9460.9470.750
WTC21.0300.03430.566***0.904
WTC30.8440.03127.212***0.857
WTC40.9640.03428.771***0.880
WTC50.8970.03327.333***0.859
WTC60.7340.03024.396***0.810
Note: *** p < 0.001.
Table 3. Correlation matrices and discriminant validity.
Table 3. Correlation matrices and discriminant validity.
ConstructsAILALSEFLCAWTC
AIL0.817
ALSE0.420.762
FLCA−0.258−0.2550.819
WTC0.1940.237−0.2740.866
Note: AIL denotes artificial intelligence literacy, ALSE signifies AI learning self-efficacy, FLCA represents foreign language classroom anxiety, and WTC indicates willingness to communicate.
Table 4. Path analyses.
Table 4. Path analyses.
HypothesisPathBSEZpβHypotheses
H1AIL → ALSE0.2810.0338.553***0.420Supported
H2AIL → FLCA−0.1440.041−3.529***−0.183Supported
H3AIL → WTC0.060.0411.4510.1470.075Rejected
H4ALSE → FLCA−0.2090.062−3.372***−0.178Supported
H5ALSE → WTC0.1810.0632.8670.0040.151Supported
H7FLCA → WTC−0.2220.049−4.54***−0.216Supported
Note: *** p < 0.001.
Table 5. Mediation Analysis.
Table 5. Mediation Analysis.
Pathβ 95% CI
SEZ-ValueLLULp
AIL → ALSE → FLCA → WTC0.0130.0052.6000.0050.0250.001
AIL → ALSE → WTC0.0510.0202.5500.0150.0930.008
AIL → FLCA → WTC0.0320.0142.2860.010.0650.002
Total Mediation0.0960.0244.0000.0530.1470.000
Direct Mediation0.0600.0441.363-0.0250.1480.161
Total Effect0.1560.0394.0000.0780.2320.001
Note: SE—standard error; LL represents the lower limit of the bias-corrected 95% confidence interval, whereas UL indicates the upper limit of the bias-corrected 95% confidence interval.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Zhang, Q.; Nie, H.; Fan, J.; Liu, H. Exploring the Dynamics of Artificial Intelligence Literacy on English as a Foreign Language Learners’ Willingness to Communicate: The Critical Mediating Roles of Artificial Intelligence Learning Self-Efficacy and Classroom Anxiety. Behav. Sci. 2025, 15, 523. https://doi.org/10.3390/bs15040523

AMA Style

Zhang Q, Nie H, Fan J, Liu H. Exploring the Dynamics of Artificial Intelligence Literacy on English as a Foreign Language Learners’ Willingness to Communicate: The Critical Mediating Roles of Artificial Intelligence Learning Self-Efficacy and Classroom Anxiety. Behavioral Sciences. 2025; 15(4):523. https://doi.org/10.3390/bs15040523

Chicago/Turabian Style

Zhang, Qinqing, Hua Nie, Jiqun Fan, and Honggang Liu. 2025. "Exploring the Dynamics of Artificial Intelligence Literacy on English as a Foreign Language Learners’ Willingness to Communicate: The Critical Mediating Roles of Artificial Intelligence Learning Self-Efficacy and Classroom Anxiety" Behavioral Sciences 15, no. 4: 523. https://doi.org/10.3390/bs15040523

APA Style

Zhang, Q., Nie, H., Fan, J., & Liu, H. (2025). Exploring the Dynamics of Artificial Intelligence Literacy on English as a Foreign Language Learners’ Willingness to Communicate: The Critical Mediating Roles of Artificial Intelligence Learning Self-Efficacy and Classroom Anxiety. Behavioral Sciences, 15(4), 523. https://doi.org/10.3390/bs15040523

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop