English Language Centre (ELC) Practitioners at the 5th XEC Best Practices Sharing event
On Wednesday, February 26, 2025, six English for Academic Purposes (EAP) practitioners (Duncan James Roulstone, Martina Dorn,
Matthew Wallwork, Xiaoqing Bi, Wei Guo, and Jonathan Culbert) gathered at XJTLU Entrepreneur College Taicang (XEC) to share their best teaching and learning practices. These lecturers, who support both undergraduate (UG) and postgraduate (PG) students, provided valuable reflections and insights on their innovative efforts to integrate Generative Artificial Intelligence (GenAI) tools into their EAP modules while aligning with the Syntegrative Education (SE) framework. The following three reflective summaries of the practices the EAP101TC, Summer Pre-sessional English (PSE) and EAP403TC, and EAP117 practitioners implemented demonstrate the English Language Centre’s (ELC) commitment to the positive student experience at XEC and the effective use of GenAI in EAP classrooms.
By Duncan James Roulstone
EAP education aims to equip students with the communication skills necessary for success in university a remit that must evolve alongside the academic environment itself. For EAP101TC, which supports Year 2 undergraduate(UG) students in the Intelligent Supply Chain with Contemporary Entrepreneurialism programme at XJTLU Entrepreneur College Taicang (XEC), the rise of GenAI since early 2023 presented such an evolution. The EAP102TC team identified an immediate need to address GenAI use, integrating instruction on effective and responsible practices into the module. The technology's continuous advancement confirms that this literacy remains a critical and growing requirement for students.
Addressing this increasing need requires moving beyond the common misconception that younger, 'digital native' students instinctively possess the skills to use these tools effectively. Practical observations consistently reveal that without structured support, students often struggle with the effective and ethical GenAI applications. Crucially, interacting successfully with most GenAI platforms depends heavily on natural language proficiency and critical interpretation, skills central to EAP, rather than purely technical expertise. This positions the EAP classroom as an ideal environment for fostering GenAI literacy, leading us to collaborate closely with the student's home academic school to embed this vital teaching throughout the EAP101TC course content.
Our instructional approach begins by establishing foundational principles for interacting with GenAI. This initial phase covers essential knowledge, starting with familiarising oneself with different GenAI platforms and models. We investigate the specifics of prompt engineering, crafting effective instructions to elicit desired outputs, and emphasise the critical importance of correctly acknowledging GenAI use in academic work.
To make these principles tangible, we employ activities designed to foster critical awareness. One such activity involves students creating self-portraits using various AI image-generation tools. Comparing the outputs demonstrates the capabilities, biases, and prompt sensitivities of different models. This exercise effectively highlights both the strengths (e.g., idea generation, speed) and weaknesses (e.g., bias, potential inaccuracy) of GenAI. It facilitates discussions on the appropriate academic use of GenAI.
A core distinction we emphasise is conceptualising GenAI as either a source of information or a tool for working with information. Our firm position is that GenAI should not be treated as a primary source. The provenance of information generated by models like ChatGPT is opaque; even developers cannot pinpoint exact data sources for a specific output. While the generated text often appears plausible, its factual accuracy is not guaranteed, and instances of ‘hallucination’, that is generating fabricated information can occur. Furthermore, citing information derived solely from a GenAI output is academically problematic. A useful analogy is a bespoke Wikipedia page generated on demand. This can be incredibly helpful for gaining initial background understanding or identifying potential avenues for further research. However, like Wikipedia, it requires critical evaluation and should be considered less reliable, especially for specialized topics. Students are consistently reminded that any information incorporated into assessed work must be verified and cited from a credible, identifiable source.
While GenAI falls short as a reliable source, it excels as a powerful tool when used effectively and responsibly. We actively integrate GenAI tuition into EAP classroom activities, focusing on skills directly relevant to the students' academic discipline. This necessitates ongoing consultation with the students' academic school, as perspectives on acceptable student use of GenAI vary considerably across fields.
Within EAP101TC, our teaching has specifically targeted leveraging GenAI to support students in tasks such as: brainstorming ideas, clarifying complex concepts, narrowing down broad research topics, generating relevant search terms for database queries, summarising lengthy texts, assisting with writing (grammar correction, translation checks, generating content from notes), and proofreading final drafts. These skills are developed through a variety of methods, including hands-on workshops where students practice using different tools, review activities analysing the relationship between prompt variations and output quality, and the provision of example prompt banks tailored to common academic tasks.
Reflecting on the past two years of integrating GenAI into the EAP101TC curriculum, several key conclusions have emerged. Firstly, the assumption that students automatically possess the skills to use GenAI effectively simply because they are young is demonstrably false; explicit instruction is essential. Secondly, as university educators, we have a responsibility to equip students with the competencies needed to navigate this technology effectively and ethically. Given that these competencies are predominantly communicative, EAP instructors are particularly well-positioned to deliver this crucial training. Finally, while certain general principles for GenAI use are universally applicable, the specific ways in which these tools can be appropriately leveraged vary significantly between academic disciplines. EAP tutors must remain cognizant of these differing academic expectations and ensure their teaching prepares students accordingly.
By Martina Dorn
The PSE team delivers courses to students with conditional offers admitted to XJTLU's more than 55 PG programmes. The historical profile of these learners is strikingly similar, and the student intake for the Summer 2024 and AY 24-25 Semester 1 PSE courses was no different. Nearly all students completed their UG degrees at Chinese universities and their English language proficiency was 0.5 to 1 IELTS band below the requisite level for the two respective PSE courses. Furthermore, the students' academic skills, encompassing areas such as academic integrity and various writing and speaking conventions, fell short of XJTLU's standards. Additionally, it was evident that these students had limited or no experience in utilising GenAI within their UG studies.
In the period from July 2024 to December 2024, the PSE team decided to adopt the approach of encouraging practitioners to utilise GenAI to generate teaching and learning materials with comprehensible input (i + 1), that is, language that is slightly beyond students’ proficiency level but still understandable through context and scaffolding (Krashen, 1982). The other aim was for practitioners to attempt to design and carry out activities that emphasise the role of social interaction, application of higher-order thinking skills, namely analysis, evaluation and language production, and scaffolded learning within the Zone of Proximal Development (ZPD), that is, the range of abilities learners can achieve with assistance (Vygotsky, 1978). Furthermore, the team wished to guide students to help them become GenAI users who employ the tools to advance their writing through an iterative process (Ingley & Pack, 2023).
In exploring the integration of GenAI in the creation of teaching and learning materials, two particularly effective practices emerged that warrant sharing.
First, GenAI was employed for the instruction of summarising skills when multiple versions of text summaries, with intentionally embedded inaccuracies, such as misleading information, factual errors, and mischaracterisations, and the inclusion of synonyms and varied sentence structures, were produced. The efficiency of existing GenAI tools meant that these summaries required only minor adjustments from educators before being implemented in classroom activities.
Practitioners observed that students engaged deeply with the texts, actively working to identify and justify the inaccuracies without relying on dictionaries. This approach compelled students to return to the original reading materials, analyse the content, interpret specific sections, and critically compare them to the erroneous summaries. As a result, this practice naturally guided students to notice various language features, infer the meaning of new vocabulary, and evaluate organisational and cohesive elements within the summaries. Subsequently, students were assigned a writing task where they produced, initially, group summaries, progressing to individual summaries later in the course.
Contrary to typical apprehension associated with such tasks, practitioners noted that students approached these challenges enthusiastically and without resorting to translation tools or other aids. This positive response can be attributed to their familiarity with the original texts and relevant vocabulary and syntax. Furthermore, providing multiple summary examples served as essential scaffolding, building their confidence as they engaged in the writing process. These insights highlight the potential of GenAI in fostering effective student engagement and developing critical analytical skills in an academic setting.
Next, GenAI was employed in the development of educational materials for the instruction of paraphrasing skills. During the summer course, practitioners observed that students frequently resorted to GenAI to obtain paraphrases of sentences. While this tendency was expected, it proved counter-productive to the students' acquisition of this essential academic skill. It posed challenges for practitioners aiming to foster effective paraphrasing techniques with the assistance of GenAI.
Therefore, in Semester 1, a more scaffolded and inductive approach was implemented. GenAI-produced materials were utilised, and students were encouraged to apply previously introduced linguistic strategies while actively engaging in the practice of paraphrasing. Instead of asking students to independently paraphrase sentences, which risked their reliance on GenAI-generated output, they were presented with four alternative paraphrases for each sentence. Students were then tasked with identifying lexical and structural distinctions between the original sentences and the provided paraphrases, relating these differences to the recognised paraphrasing strategies. Following this, students referred to the original text to assess which paraphrase best captured the intended meaning. Finally, they composed their own paraphrased versions, compared them with their peers’, and justified their linguistic choices.
This pedagogical intervention resulted in an engaging and productive learning experience wherein practitioners effectively integrated GenAI into material production. At the same time, students actively honed their paraphrasing skills without succumbing to the temptation of using the tool as a shortcut. Practitioners emphasised to students that while it is permissible to utilise GenAI in paraphrasing, it is imperative to accompany this usage with informed linguistic decision-making.
The PSE team embarked on an initial exploration of the integration of GenAI in their course instruction. While some may view these efforts as modest in scope, the insights gained have proven exceptionally valuable. It was encouraging to witness the team actively engaging with GenAI, investigating its capabilities, and identifying meaningful and impactful ways to incorporate it into their pedagogical practices. This positive engagement has inspired the team to broaden their vision for leveraging GenAI to support PSE students in developing other essential academic competencies, such as research and citation skills.
Such endeavours support the University’s overarching vision and its ongoing investment in and enhancement of GenAI tools. In the current educational landscape, the technical and other support afforded to XJTLU practitioners to explore GenAI applications in practice are abundant. Thus, the PSE team intends to capitalise on this momentum by integrating the feature into the design of sessions focused on research skills development during the Summer 2025 three PSE courses (6-week full-time, 12-week full-time and 12-week part-time), and after that during the 13-week full-time EAP403TC course Semester 1 in AY 25-26. The team wants to train students to use this tool to enable them to quickly extract relevant information from academic papers and research articles. Also, they want to encourage them to experiment with asking specific questions about the content and consider the responses that AI generates. This is likely to lead to enhanced understanding and learning because students will interactively engage with the material. Additionally, they plan to develop specific AI agents to assist PSE students with citation practices. The team’s objective is to guide students through citation formats and styles, encourage ethical writing behaviours and reinforce the value of giving credit to original authors. Once set up, AI agents will be able to provide instant feedback and examples, helping students with the nuances of different citation styles. This should result in saving students time, and enhancing their overall writing skills and confidence in academic work. As some colleagues in the ELC are already experimenting with these two tools, the PSE team will have a rich repository of expertise to draw upon. Fueled by their enthusiasm, the team is poised to create and implement new opportunities for exploration and innovation in their teaching methodologies.
By Wei Guo
In our investigation of in-sessional English language support models, we compared two key approaches: English for Specific Academic Purposes (ESAP) and Content and Language Integrated Learning (CLIL). The parallel ESAP model (Anderson, 2014; 2017), implemented in Year 2 EAP modules (e.g., EAP117, EAP118), helps students achieve CEFR B2+ proficiency while developing discipline-specific academic skills. Its interdisciplinary design, as exemplified by EAP117’s adaptation for the Academy of Films and Creative Technology (AFCT), proves effective for programmes with shared content modules or skill demands. In contrast, CLIL excels in Year 4 modules (e.g., CCS305, INS302), merging professional knowledge acquisition with advanced language outcomes (Coyle et al., 2010). While CLIL requires closer faculty-language lecturer collaboration to align content and language objectives, ESAP offers a flexible framework for targeted skill-building.
EAP117 was designed through a rigorous curriculum mapping process. We analysed assessments (see Figure 1) and consulted stakeholders across the four programmes within the AFCT to identify shared academic requirements, such as story/ script writing, project documentation, and oral communication. By focusing on these transferable competencies, the module equips students with language and soft academic skills directly applicable to their disciplines. Annual reviews ensure ongoing alignment with disciplinary evolution, and student feedback consistently affirms the module's practicality - a testament to ESAP's efficacy in bridging language and content learning.
Figure 1: Common Assessment Items in AFCT (AY24/25)
The module’s project-based tasks (see Figure 2), like the entrepreneurial team meeting simulation, epitomise ESAP’s interdisciplinary approach. Students conduct independent research on companies, their target markets, and marketing strategies, synthesising cross-disciplinary insights to create innovative solutions. Beyond applying theoretical or technical knowledge, this task fosters collaboration, adaptive communication, and critical thinking, aligning with XJTLU’s SE framework, which emphasises real-world application - a principle equally central to CLIL but achieved here through ESAP’s skill-focused design.
Figure 2: EAP117 Project-based Tasks
To address challenges in reading authentic materials, particularly in decoding cultural references and intertextuality in texts like TED talks and film scripts, we integrated GenAI as a personalised tutor. This approach allows students to engage with complex materials in a low-pressure environment, where they can address questions privately without the fear of embarrassment. The success of this method aligns with Krashen’s “affective filter hypothesis” (Krashen, 1982), which emphasises the importance of a relaxed and supportive learning atmosphere, as well as Vygotsky’s “expert-apprentice model” (Vygotsky, 1978), which highlights the value of guided interaction in learning. Students report positive experiences with GenAI, highlighting its potential to enhance personalised and engaging language learning (Galaczi & Luckin, 2024). Their feedback also indicates increased analytical confidence - an outcome aligning with CLIL’s dual-focused objectives but achieved through ESAP’s flexible, technology-enhanced approach.
ESAP's ability to serve multiple programmes through a single module demonstrates its scalability, yet this strength comes with inherent challenges. Maintaining alignment with constantly evolving disciplinary curricula and designing assessments that complement rather than duplicate content-course evaluations remain persistent difficulties. However, the shift toward AI and entrepreneurship in the Y2 curriculum presents an opportunity to address these challenges through an innovative synthesis of approaches. By combining ESAP's skill-building framework with CLIL's content-based approach, enhanced by GenAI simulated real-world tasks, we can create a synergistic model that advances both language proficiency and subject mastery. Although implementing such a model would require enhanced faculty collaboration (a defining feature of CLIL), the potential payoff is significant: a more unified and effective approach to language support that responds dynamically to disciplinary needs.
Our ESAP implementation in EAP117 demonstrates how discipline-tailored language support, enhanced by technology, can meet diverse programme needs. Future steps could explore blending ESAP’s efficiency with CLIL’s immersion, using GenAI to create a ‘third space’ for language- and content-integrated learning. Such innovation would align with XJTLU’s educational vision, preparing graduates for an increasingly interconnected world.
Anderson, R. (2014). ‘A Parallel Approach to ESAP Teaching’, Procedia - Social and Behavioral Sciences, 136, pp.194–202
Anderson, R. (2017). ‘Parallel ESAP courses: What are they? Why do we need them?’, International Journal of Language Studies, 11, pp.13–30
Coyle, D., Hood, P., & Marsh, D. (2010). CLIL: Content and Language Integrated Learning. Cambridge University Press. doi: 978-0-521-13021-9
Galaczi, E. & Luckin, R. (2024). Generative AI and Language Education: Opportunities, Challenges and the Need for Critical Perspectives [Webinar], Cambridge Papers in English Language Education. Cambridge University Press & Assessment. Retrieved from https://www.cambridge.org/sites/default/files/media/documents/CPELE_Generative%20AI%20and%20Language%20Education%20Opportunities%20Challenges%20and%20the%20Need%20for%20Critical%20Perspectives_FINAL%20%281%29.pdf
Ingley, S.J. & Pack, A. (2023). Leveraging AI Tools to Develop the Writer Rather Than the Writing. Trends in Ecology and Evolution, 38(9), 785–787. http://doi.org/10.1016/j.tree.2023.05.007
Krashen, S. (1982). Principles and Practice in Second Language Acquisition. Pergamon Press.
Vygotsky, L. S. (1978). Mind in Society: The Development of Higher Psychological Processes. Harvard University Press.
This needs to be spelt as students’ (with the apostrophe after the s because it is the possessive form of a plural noun).
This number accurately reflects the number of PG programmes, not the previously stated one.
I have added something here. We have not explored the use of the LibAI ChatPDF tool yet, because we only learnt about it in December 2024. However, we plan to start including it in our practice from this summer.
Same as above but related to AI agents. They will be part of the PSE courses from Summer 2025.
