Artificial Intelligence (AI) is increasingly becoming a crucial part of modern education, especially in higher education. Its application across various academic disciplines offers significant potential, with English language education emerging as one of the primary beneficiaries. The incorporation of AI in higher education is not just a technological advancement; it represents a crucial move towards a more personalized and effective learning experience. The importance of AI integration in higher education, particularly in language teaching and learning, cannot be overstated, as it has the potential to transform both students’ learning and educators’ teaching. This article will explore the various uses of XIPU AI in educational practices, facilitating both educators and students.

 

Keywords: XIPU AI, EAP Learning, Class Activities, Assessment Creation

 

In constructivism, learning is driven by students themselves. In this model, students are engaged in an active learning journey, in which they process and interpret the information, explore the learning environment, and develop various skills along the learning path (Rasul, Nair, Kalendra, Robin, de Oliveira Santini, Ladeira, & Heathcote, 2023). This approach is particularly crucial in the higher education context, as students are expected to have full autonomy over their learning, which is conducive to their knowledge and skill acquisition in the long term (Kim & Adlof, 2024). As one of the most important yet difficult language skills, speaking skills have long been considered an obstacle in students' English language learning. Research has shown that learners who speak English as their second language (ESL learners) tend to experience a range of learning difficulties, including low confidence levels in their English capabilities, speaking anxiety, fear of making mistakes, and reported English knowledge deficiency (Kang, 2022). Against this backdrop, balancing the provision of full autonomy while adequately supporting students in their English language learning remains a key challenge.

 

 

 

In Year 1 semester 1, the first assessment obstacle to many students was an important English-speaking Coursework. This task was divided into two parts. The first part required students to deliver a speech lasting around 1-2 minutes, based on a given topic/prompt starting with "Describe ...". In the second part, students were expected to answer four questions, covering question types such as "describe", "explain", "suggest" and "argue". These two tasks were challenging for students in the sense that, firstly, many students lacked the necessary skills to create long and full sentences on the given topic, which made it difficult for them to meet the core task requirements. Secondly, the questions could cover a wide range of topics, including health, design, animals, and history, which overwhelmed many students due to the complexity and variety of the topics. Some reported that they worried about not being able to generate relevant ideas on the given topic, while others feared that they lacked sufficient topic-specific vocabulary and grammar to articulate their thoughts clearly. Given the challenges students encountered, it was imperative to create a mechanism that could help students prepare and practice for the speaking tasks.

 

As there were no question banks available for practice, students were encouraged to use XIPU AI to generate potential questions and practice their responses. Specifically, when students would like to practice on a given topic, they could input prompts in XIPU AI such as " Can you please create 5 questions on the topic of ‘design’? These questions are used to test students’ ability to produce a 1-2-min speech on the given topic. All the questions should start with “describe”. In response, the platform would generate some potential questions, but they seemed more complicated and contained multiple purposes, which might not be similar to authentic exam questions. Therefore, students could request XIPU AI to simplify these questions, ultimately providing practice questions more closely aligned with actual exam formats (Figure 1).

 

 

(Figure 1)

 

Given the difficulty level of the questions, students could utilise AI and generate ideas and acquire topic-specific vocabulary. As mentioned above, since many students struggled with finding sufficient and appropriate aspects to answer a question for an extended amount of time (e.g. 1 to 2 minutes), XIPU AI was able to provide specific perspectives on the given topic. For instance, when prompted with a design-related question, XIPU AI could generate multiple relevant points that can help the student brainstorm and organise their ideas (Figure 2). Throughout this student-initiated learning process, learners begin by engaging prior knowledge, drawing inspiration and knowledge points from AI, and subsequently constructing new knowledge, therefore creating a more personalised and active learning experience (Taber, 2012).

 

 

(Figure 2)

 

In addition to using AI for speaking preparation and practice, some students used XIPU AI to facilitate their writing. Writing in the English language tends to be one of the most challenging skills for many learners, and the reason lies in twofold. Firstly, writing in English language requires strong logic and coherence, the two skills many students have not fully practiced prior to their university education. Secondly, when composing an essay, such as an argumentative essay, students often find it difficult to generate relevant ideas or examples as supporting details for their argument. Against this backdrop, XIPU AI could be used to help students address these challenges. To address the first difficulty, learners might copy and paste the paragraphs they have trouble with regarding logic and coherence, specifically emphasising the need for checking logic and coherence and asking for some explanation for the provided responses from XIPU AI. This step is crucial because building logic and coherence is a long-term undertaking, which requires learners to understand the 'link' between ideas and sentences in different texts, establishing logical thinking. In other words, they need to see the "why" behind the provided responses from the AI. To tackle the second difficulty, students could specify the detailed requirements for the ideas they need. For example, they could provide the AI with the specific topic they are writing on, the argument they are trying to support, the current idea(s) they already developed, as well as the writing style they wish to employ, which is also closely related to the concrete examples offered by the AI (Xiao and Zhi, 2023). While AI is capable of providing specific examples, data and statistics, it is important to recognise that AI, as a large language model, uses its modeling technique to generate texts based on its large data volume (Taecharungroj, 2023). Therefore, it is most likely to generate lines of texts based on the most frequent or likely "collocations", according to its data sets. With this in mind, learners should always fact-check against the responses provided by the AI, navigating AI tools with their critical thinking and evaluative skills.

 

 

A core aspect of my teaching philosophy is rooted deeply in Communicative Language Teaching (CLT) theory. Different from traditional teaching practices, it emphasises the authentic use of language in the learning environment (Sabrina, 2020). CLT has three prominent features. Firstly, communication is highly prioritised in teaching practice. Secondly, teaching and practice materials are designed to be authentic. Thirdly, learning and teaching activities are usually organised in small groups, which also echoes the communicative purpose of these activities (Richards & Rodgers, 2003). Within this framework, I tried to use XIPU AI to create teaching materials that can provide students with an interactive and authentic learning environment.

 

As a regularly used speaking activity, group discussion serves as a communicative and functional classroom activity in students' learning (Crisianita & Mandasari, 2022). As students engage with the topic at hand, they have to take the initiative to explore the topic, garner topical knowledge, organise their language and finally, exchange ideas with their peers (Harmer, 2001). However, inappropriately designed discussion questions can not only discourage students from sharing their opinions, but also tend to undermine their willingness to engage in the ensuing activities (Craven and Hogan, 2001). Therefore, creating and using level-appropriate and well-designed discussion questions is of prominent importance. In my teaching practice, I have occasionally resorted to XIPU AI for ideas. The AI’s ability to produce various possible questions within a short amount of time is particularly beneficial. Meanwhile, as the discussion questions should be relatively approachable on language, XIPU AI could meet the demands on the language level simultaneously. For instance, prior to a key speaking activity on the topic of "environment", it was crucial to spark students' interest and knowledge on this topic. In this case, I used XIPU AI to generate discussion questions by entering the prompt "Can you come up with three discussion questions related to “environment”? They should be relatively easy for university students to discuss, and not require subject-specific knowledge on the topic. The language level should be around CEFR B2", and then XIPU AI generated relevant discussion questions correspondingly (Figure 3). Although the questions seemed relevant to the topic, they all appeared to be slightly lengthy and difficult to answer, especially for weaker students. To address this, I refined the prompt by asking "Can you reduce the sentence lengths for all of them? Two of them should be relatively personal, i.e. related to students’ own lives." This modification aimed to reduce the complexity of the questions, increasing the engagement levels of the discussion activity. As a result, XIPU AI adjusted the previous questions, making them more approachable and suitable for the class (Figure 3).

 

 

(Figure 3)

 

Under the guidance of the Communicative Language Teaching (CLT) theory, several principles or features need to be considered when designing language learning activities. Firstly, these activities should incorporate an interactive component that encourages the use of the target language. Secondly, authentic materials should be employed to facilitate students' learning experience. Thirdly, learners' personal experiences could be embedded in the learning activities, increasing the level of engagement in the classroom (Nunan, 1989). These elements collectively aim to develop communicative competence, which includes: 1) the ability to use language for various purposes; 2) the capacity to adjust their language use to fit a particular setting and to communicate with a particular group of interlocutors; 3) the skills to comprehend and generate various types of texts; 4) the ability to maintain communication in the face of their limitations in language (Richards, 2005). Within this framework, I have found that XIPU AI can be an effective tool for creating interactive materials for classroom use. For instance, one of the frequently used activities in my classes is the interview scenario, where students are tasked with creating dialogues based on a given context, allowing them to engage and release their imagination, process the information in the context, create their own dialogues, and practice their language use. To provide an example, I used XIPU AI to generate an interview setting between a Fortune 500 snack company and a potential candidate. It is important to note that the interview scenario should be relatively approachable to learners, which provides them with the opportunity to have the space and ability to create their own interview conversations. As Figure 4 shows, upon entering my first prompt, "Can you please provide an interview conversation between an employer and a potential employee? The conversation should be no more than 1 minute. The employer is a Fortune 500 company selling snacks. The interviewee is a potential candidate applying for this sales job. Please create an interview session between them, the main focus should be on asking the interviewee’s previous job experiences and how her personality fits this position." XIPU AI produced a relatively satisfactory interview conversation, based on the prompt. However, some phrases and expressions were overly formal, so I requested further on "adjusting the formality and language difficulty level of the conversation". XIPU AI then correspondingly produced a revised conversation that met the majority of my requirements (Figure 5).

 

 

(Figure 4)

 

 

(Figure 5)

 

 

 

One of the most promising applications of AI in English language teaching is its potential to expand and enhance class activities. For instance, when creating pre-listening activities for one of my Continuing Support sessions, I was intrigued to explore how XIPU AI might generate and prepare differently from my teaching repertoire. To test this, I entered the following prompt into XIPU AI (Figure 6), "Can you please help me create some pre-listening vocabulary activities that are suitable for CEFR B1 students? The vocabulary activities should aim to be fun and interactive. Please include the target vocabulary below: break into (phrasal v.) thermometer (n.) porch (n.) fumble (v.) locksmith (n.)?" In response to my request, XIPU AI offered me several classic and interactive activities that could be incorporated into my lessons. However, I was also interested to know if the AI could help me one step further and create the ready-made materials they mentioned, which could potentially save me significant time. Specifically, I kept on requesting "Can you please create the charade paper for me, using the provided five target words? This way, it is easier for me to cut them out and use them directly." It came as little surprise that the AI could not create the charade cards for me. However, when I proceeded to request the similar action on the vocabulary matching game, XIPU AI provided detailed explanations for each target word, which saved me time in creating the definitions for the vocabulary (Figure 7).

 

 

(Figure 6)

 

 

(Figure 7)

 

The previous examples highlight the crucial importance of creating and adjusting prompts when making requests on AI tools. To achieve the desired outcomes, prompts should be as comprehensive and accurate, which reflect the precise aim of the task. Sometimes, when the AI tool fails to execute the prompt, it is often beneficial to slightly adjust the prompt, which may yield more suitable responses. In addition to incorporating various content and language proficiency level requirements when giving prompts, it is worth mentioning that integrating students' interests into class activity design is also of great importance. In my regular teaching practices, given my familiarity with my students, I frequently integrate elements relevant to students' hobbies or interests, which aim to engage them to a greater extent. By expanding the range of class activities and making them more personalized, AI contributes to sustaining students’ motivation and engagement, which is crucial for fostering effective learning.

 

 

 

Listening skills are a fundamental part of language learning; however, creating effective listening assessments can be a challenging task for teaching professionals, especially under tight time constraints. AI can be instrumental in this regard by assisting in developing assessments that are both comprehensive and tailored to the students' proficiency levels. For example, AI can generate listening exercises from a vast variety of audio sources, including podcasts, news broadcasts, and conversational snippets, ensuring that students are exposed to a variety of accents, speech rates, and contexts. For instance, in preparation for one of my listening workshops, I utilised XIPU AI to generate a set of listening comprehension questions, based on a TED Talk to help students practice their listening comprehension skills. It is worth noting again that when asking AI to perform a task, it is important to provide comprehensive and accurate prompts. As students attending my workshops generally needed more listening scaffolding and support, I specifically asked XIPU AI to create questions that aligned with a particular English proficiency level. In addition, to assess students' understanding effectively, the multiple-choice questions were designed to examine students’ understanding of the listening content, and multiple distractors were also intentionally added (Figure 8 & Figure 9). In some cases, if certain students request more practice questions, XIPU AI can also be used to generate additional sets, with varying difficulty levels being applied.

 

 

(Figure 8)

 

 

(Figure 9)

 

 

 

The integration of AI into higher education, particularly in language learning, offers valuable opportunities that can significantly enhance the learning experience for students. AI's ability to personalize education, expand classroom activities, and facilitate the creation of assessments ensures that students receive a more tailored, engaging, and effective learning environment. The benefits of AI extend beyond student benefits; educators also stand to gain from the increased efficiency and insights provided by AI tools. As AI technology continues to evolve, its role in education will undoubtedly expand, offering even more innovative ways to support teaching and learning. It is crucial for educators to embrace AI and harness its potential to create learning environments that are both technologically advanced and responsive to students’ needs. Meanwhile, as Generative AI is increasingly embraced and applied in teaching and learning settings, several considerations must be addressed. Firstly, with the prevalent application of AI, it is crucial for educators to guide students on the effective use of these AI platforms, including advice on effective prompt design, evaluating responses or texts provided by AI and cultivating learner autonomy with the proper use of AI. Secondly, as AI keeps evolving, more functions will be devised to cater to students' learning needs. In this context, the onus is on educators to develop assessments that accurately measure students' learning. As more and more work and tasks could be completed by AI, it is also increasingly crucial to facilitate students' overall development, such as creativity and critical thinking. In conclusion, the incorporation of AI in higher education represents a transformative shift that promises to revolutionize the way we approach teaching and learning, making education more accessible, personalized, and effective for all.

 

 

 

Craven, J. A., & Hogan, T. (2001). Assessing participation in the classroom. Science Scope, 25(1), 36.

 

Crisianita, S., & Mandasari, B. (2022). THE USE OF SMALL-GROUP DISCUSSION TO IMPROVE STUDENTS’SPEAKING SKILL. Journal of English Language Teaching and Learning, 3(1), 61-66.

 

Harmer, J. (2001). Book review: The practice of English language teaching. RELC Journal, 32(1), 135-136.

Kang, H. (2022). Effects of Artificial Intelligence (AI) and Native Speaker Interlocutors on ESL Learners' Speaking Ability and Affective Aspects. Multimedia-Assisted Language Learning, 25(2).

 

Kim, M., & Adlof, L. (2024). Adapting to the future: ChatGPT as a means for supporting constructivist learning environments. TechTrends, 68(1), 37-46.

 

Nunan, D. (1989). Designing tasks for the communicative classroom. Cambridge University Express.

Rasul, T., Nair, S., Kalendra, D., Robin, M., de Oliveira Santini, F., Ladeira, W. J., ... & Heathcote, L. (2023). The role of ChatGPT in higher education: Benefits, challenges, and future research directions. Journal of Applied Learning and Teaching, 6(1), 41-56.

 

Richards, J. C. (2005). Communicative language teaching today. Singapore: SEAMEO Regional Language Centre.

 

Richards, J. C., & Rodgers, T. S. (2014). Approaches and methods in language teaching. Cambridge university press.

 

Sabrina, A. (2020). The implementation of communicative language teaching in English practice. NOTION: Journal of Linguistics, Literature, and Culture, 2(1), 17-27.

 

Taber, K. S. (2012). Constructivism as educational theory: Contingency in learning, and optimally guided instruction. In Educational theory (pp. 39-61). Nova.

 

Taecharungroj, V. (2023). “What can ChatGPT do?” Analyzing early reactions to the innovative AI chatbot on Twitter. Big Data and Cognitive Computing, 7(1), 35.

 

Xiao, Y., & Zhi, Y. (2023). An exploratory study of EFL learners’ use of ChatGPT for language learning tasks: Experience and perceptions. Languages, 8(3), 212.