Juan Yang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Juan Yang has a total of 50 co-authors. Here are the top co-authors by number of shared publications:
Bo Sun
6 shared publicationsLing Wang
Department of Educational Technology Sichuan Normal University Chengdu China
6 shared publicationsZhijie Liang
College of Computer Science, Sichuan Normal University, Chengdu, 610101, China
5 shared publicationsDaifa Wang
4 shared publicationsHongtao Liu
College of Computer Science and Technology Chongqing University of Posts and Telecommunications Chongqing China
4 shared publicationsZhixing Huang
3 shared publications
Latest Publications
- 10 May 2026
- Published by Wiley in Journal of Computer Assisted Learning
- Vol. 42 (3)
An Interdisciplinary Method of Applying AI ‐Empowered Chatbots in Chinese Quatrains' Learning for Chinese Young Pupils
Background: AI‐empowered chatbot has been widely and longitudinally applied in second language (L2) education, but few of them were applied in first language (L1) education. Chinese ancient Quatrains, containing high‐density semantic information with minimalist word combinations, are the learning targets of the most advanced level of native language learning for Chinese pupils. Objectives: Designing and implementing AI‐empowered chatbots to explore the efficacy of their application in L1 Quatrains' learning for young Chinese pupils. Samples: Four hundred and thirty four second‐grade pupils from Qingtaishan Primary School participated in this training program, including 220 boys and 214 girls. Participants' average age was 8.2 years (SD = 0.47). Two hundred and eight of them were assigned to the experimental class while the remaining 226 pupils were assigned to the control group. Methods: A between‐participants design was applied to 434 second‐grade pupils from a primary school. AI‐empowered chatbots were designed and developed with agents and a knowledge base to implement the pedagogical instructional designs. An ANOVA comparison was conducted between learners with and without the intervention, measuring their behavioural outcomes. Five‐point Likert scales were used to collect learners' self‐perceptions about the training, and interviews were conducted with teachers. Structural Equation Models (SEMs) were built to explore how the chatbots affected learners' learning outcomes. Results: Results showed that learners' sentence comprehension ability was significantly enhanced compared to their peers without intervention, but their ability to extract keywords from the Quatrains showed no superiority to their peers. SEMs based on learners' self‐perceptions and their performance suggested that their positive attitude toward new technology had no causal relationship with their learning outcomes. Conclusions: We summarised the training's advantages and disadvantages based on a SWOT analysis; practitioner notes were made for policymakers, school leaders and L1 teachers seeking to integrate AI‐supported chatbots in their classrooms. Moreover, by using interaction, constructiveness, and activeness of the learning activity to evaluate the gains of applying AI‐empowered chatbots in L1 learning, drawbacks and limitations were further diagnosed to evolve instructional designs, which may lead to better application outcomes.
- 10 May 2026
- Published by Wiley in Journal of Computer Assisted Learning
- Vol. 42 (3)
An Interdisciplinary Method of Applying AI ‐Empowered Chatbots in Chinese Quatrains' Learning for Chinese Young Pupils
Background: AI‐empowered chatbot has been widely and longitudinally applied in second language (L2) education, but few of them were applied in first language (L1) education. Chinese ancient Quatrains, containing high‐density semantic information with minimalist word combinations, are the learning targets of the most advanced level of native language learning for Chinese pupils. Objectives: Designing and implementing AI‐empowered chatbots to explore the efficacy of their application in L1 Quatrains' learning for young Chinese pupils. Samples: Four hundred and thirty four second‐grade pupils from Qingtaishan Primary School participated in this training program, including 220 boys and 214 girls. Participants' average age was 8.2 years (SD = 0.47). Two hundred and eight of them were assigned to the experimental class while the remaining 226 pupils were assigned to the control group. Methods: A between‐participants design was applied to 434 second‐grade pupils from a primary school. AI‐empowered chatbots were designed and developed with agents and a knowledge base to implement the pedagogical instructional designs. An ANOVA comparison was conducted between learners with and without the intervention, measuring their behavioural outcomes. Five‐point Likert scales were used to collect learners' self‐perceptions about the training, and interviews were conducted with teachers. Structural Equation Models (SEMs) were built to explore how the chatbots affected learners' learning outcomes. Results: Results showed that learners' sentence comprehension ability was significantly enhanced compared to their peers without intervention, but their ability to extract keywords from the Quatrains showed no superiority to their peers. SEMs based on learners' self‐perceptions and their performance suggested that their positive attitude toward new technology had no causal relationship with their learning outcomes. Conclusions: We summarised the training's advantages and disadvantages based on a SWOT analysis; practitioner notes were made for policymakers, school leaders and L1 teachers seeking to integrate AI‐supported chatbots in their classrooms. Moreover, by using interaction, constructiveness, and activeness of the learning activity to evaluate the gains of applying AI‐empowered chatbots in L1 learning, drawbacks and limitations were further diagnosed to evolve instructional designs, which may lead to better application outcomes.
- Open Access
- 18 March 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16
An early detection framework for young Chinese learners at risk of reading difficulty using fNIRS and deep learning
Reading difficulty (RD), a neurodevelopmental disorder affecting language acquisition in children, necessitates early screening for effective educational interventions. This study proposes the RD-risk Classifier (RDr-C), a novel framework integrating functional near-infrared spectroscopy (fNIRS) with deep learning, specifically combining a dual-layer graph convolutional network (GCN), a bidirectional long short-term memory network (BiLSTM), and multi-head self-attention mechanisms (MSA) for 7-8-year-old children's literacy assessment. The model was validated using fNIRS signals from 30 participants (16 experimental group, 14 control group) during the visual sign recognition test and phonetic discrimination test, with performance evaluated through 5 runs × 5-fold cross-validation experiments. Results show that RDr-C achieved a mean classification accuracy of 99.60% and 99.66% in visual and auditory tests, respectively, significantly outperforming traditional convolutional neural networks (CNN), long short-term memory networks (LSTM), and existing fNIRS classification models (e.g., fNIRS-T, fNIRSNet). Furthermore, leave-one-subject-out cross-validation demonstrates that RDr-C achieves global accuracies of 89.33% and 87.93% on visual and auditory tasks, respectively, with corresponding Kappa coefficients of 0.78 and 0.76, confirming its robustness across individuals. Feature shuffling and wavelet transformation visualizations further confirm robust feature separation, highlighting the model's ability to capture distinct hemodynamic patterns associated with RD. By integrating the spatial feature extraction of GCN, the temporal modeling of BiLSTM, and the global dependency capture of MSA, this work establishes a non-invasive neuroimaging paradigm for educational neuroscience. The high-precision classification lays a technical foundation for early screening tools, with future applications extending to multimodal brain-computer interfaces and longitudinal intervention monitoring.
- 18 March 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16
Open AccessAn early detection framework for young Chinese learners at risk of reading difficulty using fNIRS and deep learning
Reading difficulty (RD), a neurodevelopmental disorder affecting language acquisition in children, necessitates early screening for effective educational interventions. This study proposes the RD-risk Classifier (RDr-C), a novel framework integrating functional near-infrared spectroscopy (fNIRS) with deep learning, specifically combining a dual-layer graph convolutional network (GCN), a bidirectional long short-term memory network (BiLSTM), and multi-head self-attention mechanisms (MSA) for 7-8-year-old children's literacy assessment. The model was validated using fNIRS signals from 30 participants (16 experimental group, 14 control group) during the visual sign recognition test and phonetic discrimination test, with performance evaluated through 5 runs × 5-fold cross-validation experiments. Results show that RDr-C achieved a mean classification accuracy of 99.60% and 99.66% in visual and auditory tests, respectively, significantly outperforming traditional convolutional neural networks (CNN), long short-term memory networks (LSTM), and existing fNIRS classification models (e.g., fNIRS-T, fNIRSNet). Furthermore, leave-one-subject-out cross-validation demonstrates that RDr-C achieves global accuracies of 89.33% and 87.93% on visual and auditory tasks, respectively, with corresponding Kappa coefficients of 0.78 and 0.76, confirming its robustness across individuals. Feature shuffling and wavelet transformation visualizations further confirm robust feature separation, highlighting the model's ability to capture distinct hemodynamic patterns associated with RD. By integrating the spatial feature extraction of GCN, the temporal modeling of BiLSTM, and the global dependency capture of MSA, this work establishes a non-invasive neuroimaging paradigm for educational neuroscience. The high-precision classification lays a technical foundation for early screening tools, with future applications extending to multimodal brain-computer interfaces and longitudinal intervention monitoring.
- Open Access
- 11 February 2026
- Published by Elsevier in Thinking Skills and Creativity
- Vol. 61
A reading model of English as a foreign language accommodated by programming-oriented computational thinking: An interdisciplinary approach
- 11 February 2026
- Published by Elsevier in Thinking Skills and Creativity
- Vol. 61
Open AccessA reading model of English as a foreign language accommodated by programming-oriented computational thinking: An interdisciplinary approach
- Open Access
- 21 May 2025
- Published by Elsevier in Brain Research Bulletin
- Vol. 227, 111398
Unraveling the neural dynamics of mathematical interference in english reading: A novel approach with deep learning and fNIRS data
- 21 May 2025
- Published by Elsevier in Brain Research Bulletin
- Vol. 227, 111398
Open AccessUnraveling the neural dynamics of mathematical interference in english reading: A novel approach with deep learning and fNIRS data
- Open Access
- 20 June 2024
- Published by Elsevier in Learning and Instruction
- Vol. 93
Mobile application-based phonetic training facilitates Chinese-English learners’ learning of L2
- 20 June 2024
- Published by Elsevier in Learning and Instruction
- Vol. 93
Open AccessMobile application-based phonetic training facilitates Chinese-English learners’ learning of L2
- 15 May 2024
- Published by Taylor & Francis in Innovation in Language Learning and Teaching
- Vol. 19 (2) , 155-173
How music facilitates second language (L2) learning: a systematic review from 2002 to 2022
- 15 May 2024
- Published by Taylor & Francis in Innovation in Language Learning and Teaching
- Vol. 19 (2) , 155-173
How music facilitates second language (L2) learning: a systematic review from 2002 to 2022
- Open Access
- 21 September 2023
- Published by Springer Nature in Education and Information Technologies
- Vol. 29 (8) , 9843-9863
Employing animation-based TE to activate Chinese children’ cognitive-involvement in L2 feature acquisition
A substantial amount of literature has shown that the input flooding and enhancement (Textual enhancement, TE) technique helps learners to perceive and acquire target features implicitly. Previous studies explored the efficacy of computer-assisted TE interventions in Second Language Acquisition with diverse techniques, but animation has not yet been sufficiently taken into consideration. In this paper, we developed an animation-based TE intervention to promote the primary school children’ acquisition of the linguistic features of English question sentences. The results suggested that learners in experimental group improved their performances in producing questions and statements with a large effect size, as they were found to spontaneously put extra efforts in exploring the related knowledge of the visually enhanced forms. The study presumed that the animation-based TE intervention would stimulate participants’ awareness and curiosity of the target forms, which consequently led to improved performance in tasks both involving precursor knowledge and newly acquired knowledge.
- 21 September 2023
- Published by Springer Nature in Education and Information Technologies
- Vol. 29 (8) , 9843-9863
Open AccessEmploying animation-based TE to activate Chinese children’ cognitive-involvement in L2 feature acquisition
A substantial amount of literature has shown that the input flooding and enhancement (Textual enhancement, TE) technique helps learners to perceive and acquire target features implicitly. Previous studies explored the efficacy of computer-assisted TE interventions in Second Language Acquisition with diverse techniques, but animation has not yet been sufficiently taken into consideration. In this paper, we developed an animation-based TE intervention to promote the primary school children’ acquisition of the linguistic features of English question sentences. The results suggested that learners in experimental group improved their performances in producing questions and statements with a large effect size, as they were found to spontaneously put extra efforts in exploring the related knowledge of the visually enhanced forms. The study presumed that the animation-based TE intervention would stimulate participants’ awareness and curiosity of the target forms, which consequently led to improved performance in tasks both involving precursor knowledge and newly acquired knowledge.
- Open Access
- 25 January 2023
- Published by Elsevier in Thinking Skills and Creativity
- Vol. 47
Influence of high-level mathematical thinking on L2 phonological processing of Chinese EFL learners: Evidence from an fNIRS study
- 25 January 2023
- Published by Elsevier in Thinking Skills and Creativity
- Vol. 47
Open AccessInfluence of high-level mathematical thinking on L2 phonological processing of Chinese EFL learners: Evidence from an fNIRS study
- Open Access
- 1 January 2023
- Published by INSTICC
- p. 454-457
The Design of High Speed Wideband Receiver Based on FPGA
- 1 January 2023
- Published by INSTICC
- p. 454-457
Open AccessThe Design of High Speed Wideband Receiver Based on FPGA
- Open Access
- 30 October 2022
- Published by Australasian Society for Computers in Learning in Tertiary Education in Australasian Journal of Educational Technology
- Vol. 38 (6) , 21-33
A technology-enhanced scaffolding instructional design for fully online courses
The COVID-19 pandemic has forced teachers to implement fully online teaching. This study reviewed the popular technologies that are used in online learning, as well as the advantages and difficulties of applying fully online courses for formal education. Based on this research background, this study proposed a nested scaffolding design of an online course for 215 college students in China with the help of six technological tools, which effectively replaced face-to-face interactions and significantly improved the usage of the supporting learning platform. The inner-outer learning cycles supported by the technological tools improved the quality of the scaffolding conversations, reduced the scaffolding time cost that teachers had to expend and enhanced the effectiveness of the individualised scaffolding instructions. Implications for practice or policy: First-year students’ learning outcomes can be improved by the scaffolding support from Web 2.0 resource URLs, a small private online course, and EducCoder resources. Course leaders should construct at least 3–5 stage-wise evaluations to deconstruct the big learning process into several observable learning cycles, making the Kolb (1984) cycles controllable. Assessors may need to consider involving various exercises, such as quizzes, online experiments and synthesised tasks to facilitate students’ learning.
- 30 October 2022
- Published by Australasian Society for Computers in Learning in Tertiary Education in Australasian Journal of Educational Technology
- Vol. 38 (6) , 21-33
Open AccessA technology-enhanced scaffolding instructional design for fully online courses
The COVID-19 pandemic has forced teachers to implement fully online teaching. This study reviewed the popular technologies that are used in online learning, as well as the advantages and difficulties of applying fully online courses for formal education. Based on this research background, this study proposed a nested scaffolding design of an online course for 215 college students in China with the help of six technological tools, which effectively replaced face-to-face interactions and significantly improved the usage of the supporting learning platform. The inner-outer learning cycles supported by the technological tools improved the quality of the scaffolding conversations, reduced the scaffolding time cost that teachers had to expend and enhanced the effectiveness of the individualised scaffolding instructions. Implications for practice or policy: First-year students’ learning outcomes can be improved by the scaffolding support from Web 2.0 resource URLs, a small private online course, and EducCoder resources. Course leaders should construct at least 3–5 stage-wise evaluations to deconstruct the big learning process into several observable learning cycles, making the Kolb (1984) cycles controllable. Assessors may need to consider involving various exercises, such as quizzes, online experiments and synthesised tasks to facilitate students’ learning.