To survive or to thrive? Synthesizing the narrative trajectories of students’ self-regulated listening practice in an EMI transnational higher education context

Abstract

English-medium instruction (EMI) has burgeoned alongside two decades of L2 self-regulated learning research. In both areas, listening remains under-researched, longitudinal designs are under-employed, and in-depth studies are necessary to unpack learner development. In a context believed to initiate self-regulated learning, the current study employed a longitudinal qualitative design to explore the out-of-class, self-regulated listening practice of 34 Chinese students across their first term transitioning from Chinese-medium high schools to an EMI transnational university in China. Guided by Zimmerman’s social-cognitive model of self-regulated learning, 102 transcripts generated from interviewing students at the beginning, middle, and end of their transition term were analyzed using thematic analysis and narrative analysis. The findings are presented chronologically as a synthesis of the students’ narrative trajectories. They highlight the students’ initial enthusiasm for self-regulated listening practice to ‘survive’ EMI lectures and a watershed moment at the midterm, when many ‘dropouts’ stopped practicing as they perceived their proficiency had reached an abstract threshold. The ‘continuers’ developed sustainable practices informed by the pursuit of long-term, future-oriented goals that outreached the immediate EMI context, guided by metacognitive awareness and emotion regulation strategies. Pedagogical implications to support future cohorts of students to thrive during transition and beyond are provided.

Keywords:

• English medium instruction (EMI)
• learner autonomy
• listening
• motivation
• proficiency
• self-regulation

1. Introduction

The expansion of transnational higher education (TNHE) is one of the most prominent global trends in the twentieth century (Mok and Han 2016; Phan 2016). TNHE has been defined as ‘the mobility of a higher education institution/provider between countries’ (Knight 2016, p. 36). Its growth runs in parallel to the rapidly expanding landscape of English-medium instruction (EMI) in higher education (De Costa et al. 2021). In China, TNHE refers to higher education providers cooperatively run by Chinese and foreign institutions (State Council 2003). This educational format allows for collaboration across borders via branch campuses or joint degree programs. TNHE leverages internationalization schemes to offer EMI programs to students from over 30 Chinese provinces (Ou and Gu 2021). EMI-TNHE curricula consist of lecture-based courses that require high listening proficiency. Consequently, many students who transition from Chinese-medium high schools face tremendous challenges during their initial experiences listening to lengthy academic lectures conducted in English (Evans and Morrison 2018). Unfortunately, the policy-driven nature of EMI-TNHE initiatives means that empirical research has lagged behind rapid implementation aims in this area.

Regarding EMI research more broadly, a central question is whether this instructional context is indeed conducive to students’ language development in addition to the development of subject-content knowledge (Macaro et al. 2018). Although some researchers have reported a positive role of EMI in improving students’ English-language proficiency (e.g. Rogier 2012; Yang 2015), others have found the impact to be incidental (e.g. Lei and Hu 2014). Gao (2008) cautions that although EMI provides initial incentives for students to learn English, it might be insufficient ‘to sustain their ongoing language learning efforts’ (p. 607). This lack of a sustained incentive is notable since language development takes considerable time, and noticeable gains in listening proficiency can be particularly time-intensive to realize (Rost 2015). As such, it is often necessary for students to engage in their own self-regulated learning and persist over an extended period to improve their listening proficiency. Such improvement enables them to understand EMI lectures and engage in other communicative activities via English.

Self-regulated learning refers to learners’ processes to activate, manage, and sustain cognition, affect, and behaviors for achieving personal learning goals (Zimmerman and Schunk 2011). What underpins longer-term self-regulated language learning in EMI contexts seems to be learner-related variables such as motivational orientations rather than the EMI context per se (Jiang and Zhang 2019; Zhou and Thompson 2023b). Therefore, research that investigates how students strategically adapt to listening in EMI-TNHE contexts in relation to their individual differences is needed. We refer to the focus of the current study as students’ out-of-class, self-regulated listening practice (SRLP).

Given that the development of listening proficiency requires significant time and individualized student engagement, qualitative studies that adopt in-depth, longitudinal lenses are equipped to explore how students’ self-regulated learning unfolds during their initial transition period into EMI university programs (Evans and Morrison 2011). However, the longitudinal trajectories of students’ self-regulated language learning beyond the classroom remains under-researched, with even fewer attempts to disentangle the factors underpinning the continuation or suspension of students’ learning. Therefore, to contribute to the field’s understanding of how students’ SRLP unfolds during their initial transition term from Chinese-medium high schools to an EMI transnational university, the current study draws on an in-depth, longitudinal design that heeds recent calls for more rigorous qualitative research on strategic language learning (see Thomas et al. 2022). It also explores, inductively, the factors that seem to mediate students’ effort investment. The findings contribute to the limited longitudinal research on students’ self-regulated learning in EMI contexts, the paucity of research on self-regulated listening in general, and the especially under-researched area of out-of-class SRLP.

2. Literature review

2.1. Self-regulation and learner strategies in EMI research

Research exploring students’ strategic learning in EMI contexts has emerged in recent years, offering important perspectives on the agentive behaviors of students in response to the multitude of academic learning difficulties documented in the existing EMI literature (e.g. Evans and Morrison 2018; Aizawa and Rose 2020; Sahan et al. 2023). A focus of this research has been the language learning and coping strategies that learners employ within or while enrolled in EMI classes to mitigate their challenges (e.g. Soruç and Griffiths 2018; Macaro et al. 2019; Fung and Lo 2023; Zhou et al. 2023). Since learning within EMI classes is often inextricably linked to students’ efforts outside of classrooms, some studies have suggested a more holistic lens to conceptualize strategic choices in relation to students’ self-regulated learning, extending the scope of interest to students’ independent learning efforts (e.g. Evans and Morrison 2011; Ding and Stapleton 2016; Zhou and Rose 2021).

The transition from L1-taught secondary schools to an EMI tertiary context seems to be a critical period for incubating self-regulated learning (see Evans and Morrison 2011; Ding and Stapleton 2016), as students are expected to be more independent learners in such contexts. Drawing on interviews with 28 students at the end of their first two academic terms at an EMI university in Hong Kong, Evans and Morrison (2011, p. 203) found that students engaged in ‘a relentless diet of disciplinary reading and listening’ to develop subject-specific vocabulary and academic literacies in English. Similarly, based on interviews with nine Chinese students embarking on their EMI university studies, Ding and Stapleton (2016) reported that the participants spent considerable time after class watching English online lectures from famous universities to improve their listening skills and alleviate listening anxiety. These studies suggest that students’ first experiences studying via EMI might offer strong incentives for SRLP. However, differences in the learning trajectories of students with varied individual characteristics remain largely unknown.

Jiang and Zhang (2019) described a more detailed picture of students’ self-regulated learning in an EMI university program in China. The authors found that students’ different motivational orientations seemed to mediate their learning efforts. High achievers tended to surpass the impending need for using English to pass exams and appeared more motivated to carry out longer-term language learning for their future personal development. This finding echoes Lasagabaster’s (2016) EMI study in Spain, which identified students’ ideal L2 self (i.e. the imagined self as a proficient speaker of English in future work or life) as a significant predictor of their intended English learning effort. Since EMI practices often align with government policies to cultivate a globalized workforce (Hu et al. 2014; Rose and McKinley 2018), students’ motivational beliefs of viewing English as an international communicative tool to interact with people from diverse cultural backgrounds may influence their learning behavior. This might especially be the case in East Asia, where researchers have found that students may enroll in EMI programs to access international communities (e.g. Iwaniec and Wang 2022) and to be afforded additional opportunities for studying abroad (e.g. Liu et al. 2021). However, thus far, empirical research examining how key learner factors (e.g. proficiency and motivation) play into students’ SRLP while enrolled in EMI contexts is rare, which the present study addresses.

2.2. Self-regulation in language learning

The notion of self-regulation has made inroads into mainstream language education research since the turn of the century. The most prominent voice at the time was Dörnyei (2005), who argued for a shift in attention from learners’ specific strategy use to their self-regulatory capacity. Tseng et al. (2006), among others, were early adopters, while researchers such as Gao et al. (2007) and Rose (2012) expressed more conservative views that recognized the importance of both strategies and self-regulatory capacity. Acknowledging successful self-regulated learning as an educational ideal (rather than a given), Thomas et al. (2021) argue for a more nuanced understanding of students’ sources of learning regulation (i.e. including and beyond the self); however, their argument firmly situates itself in formal education settings, leaving out-of-class learning ripe for new insights while remaining in line with current conceptual thought. In a recent review by Teng and Zhang (2022), the authors indicate that nearly two decades of research on self-regulated learning has largely centered on writing, reading, and vocabulary, whereas skills such as speaking and listening remain under explored. Bowen and Thomas (2022) concur yet add that the paucity of research on self-regulated listening is likely due to the transitory nature of listening and subsequent difficulty of meaningfully researching SRLP in a field dominated by large-scale survey-oriented research.

Extant self-regulated learning studies mainly draw on cross-sectional designs, establishing the relationship between learner-related factors and self-regulation strategies or processes, including language proficiency (Teng and Huang 2019) and motivational factors such as goal-orientation (Kormos and Csizér 2014), task value (Su et al. 2019), growth mindset (Bai and Wang 2023), and self-efficacy beliefs (Yabukoshi 2021). What is lacking, however, is a longitudinal perspective on how students’ self-regulated language learning – and more specifically, SRLP – may unfold in response to contextual and individual factors (see Rose et al. 2018).

Despite the paucity of research on self-regulated listening, evidence from longitudinal self-regulated learning studies in other skill domains have revealed that students’ self-regulation processes might be mediated both by their language proficiency (Rose and Harbon 2013) and their motivation or identity orientations (Bown 2009; Yabukoshi 2021; Schneider 2022). Rose and Harbon’s (2013) study with 12 university students in Japan found that, compared to highly proficiency students, low-proficiency learners tended to set short-term, specific, and achievable goals to help them increase their sense of control over their learning progress. Following seven university students in the US for four years, Schneider (2022) found that when students viewed academic writing competence as important for their imagined professional identities in the future, they tended to adopt investment strategies to connect writing assignments to their personal interest or home culture. These findings may be transferable to EMI-TNHE contexts, where students entering with lower proficiency might experience initial impulses to engage in SRLP outside of classes to cope with understanding lectures. As they secure these basic listening needs, however, their motivation to become members of an international community (Iwaniec and Wang 2022) or to develop a future-oriented self for studying abroad (Liu et al. 2021) might sustain their continued investment long term.

2.3. Self-regulated listening practice: a theoretical framework

The limited L2 self-regulation research investigating students’ listening practice has found self-regulation instruction to be effective in improving L2 listeners’ metacognitive awareness and self-regulation strategies (e.g. Mareschal 2007; Zeng and Goh 2018). It represents a nested area of study in which a significant amount of work has been conducted on cultivating successful listeners through systematic instruction (see Bozorgian and Shamsi 2023). However, in many EMI contexts, explicit strategy instruction is noticeably absent from curricula (Zhou and Thompson 2023a). Therefore, to address the demand of listening comprehension in academic lectures, students often feel an urge to initiate and regulate listening practice on their own.

Among several prominent models of self-regulated learning introduced since the 1980s (see Panadero 2017 for a review), Zimmerman’s (2000) three-phase model is the most cited and continues to guide most new research in this area (e.g. Zhou and Rose 2021; Teng 2022). This model has been applied by Yabukoshi (2021) to examine Japanese EFL learners’ out-of-class listening practice, and most recently by Zhou and Thompson (2023a) to investigate Chinese university students’ lecture listening in EMI classes. It provides the theoretical framework for the current study.

Situated in Bandura’s (1986) social cognitive theory, Zimmerman (1989) proposed that self-regulation involves interactions at personal, behavioral, and environmental levels. Within these levels, self-regulated learning is powered by an individual’s motivation, is guided by that individual’s metacognition, and unfolds cyclically through phases of forethought, performance, and self-reflection (Zimmerman 2000). When these constructs coalesce in the various phases of the learning process, learners become proactive problems solvers of their own learning practices (Zimmerman and Campillo 2003). By viewing self-regulated learning as a cluster of processes (cf. a state or trait), researchers are conceptually equipped to understand why and how learners engage in certain activities (e.g. SRLP), as opposed to, simply, whether they have engaged or not (Zimmerman and Moylan 2009). Thus, in the current study, we view learners who engage in SRLP as proactive problem solvers who leverage self-generated feedback on their prior performance to initiate, maintain, and adjust their SRLP as needed.

Each stage of the current study is framed in Zimmerman’s (2000) cyclical phases of self-regulation: forethought, performance, and self-reflection. The forethought phase occurs before learners begin listening practice. During this phase, learners set goals and plan for how to practice. Then, during the performance phase, learners engage in practice, employ self-control to stay committed to the task and observe their practice via processes such as recording time on task and experimenting with different ways of practicing. Finally, the self-reflection phase occurs after learners finish practicing. During this phase, learners evaluate their progress in listening and make causal attributions that feed into how they approach future practice. Taken together, the three-phase model serves as a series of feedback loops that guide current and future action.

2.4. Rationale and research question

Students in their initial transition period from Chinese-medium high schools to EMI universities form an under-researched population that is especially likely to engage in SRLP. For the most part, previous research on EMI and self-regulated learning has been unable to report on how learning unfolds for individuals over time. Provided that listening development is idiographic and gradual, only longitudinal research that treats individual trajectories as important can offer implications regarding factors that appear to mediate sustained SRLP. To address this research gap, we draw on Zimmerman’s (2000) three-phase social-cognitive model to answer the following research question:

How do students engage in self-regulated listening practice during their transition term from Chinese-medium high schools to an EMI transnational university?

3. Methodology

3.1. Setting and participants

Data were collected at an EMI-TNHE university in southeast China. English is used as the official language for teaching, learning, research, assessment, and administration at the university. The university’s undergraduate programs follow a 4-year curriculum, and students have the option to complete the last 2 years at the partner university in the UK. After applying for approval from the institution’s research ethics committee, the first author was granted permission to collect data from Chinese students enrolled in Business as well as Humanities and Social Sciences courses, so the present study recruited participants from this population. All undergraduate students are required to participate in the Oxford Online Placement Test (OOPT) upon admission to the university. Students’ OOPT listening scores were used in the current study as a standardized measure of their English listening proficiency. Students were informed that such data would be used for research purposes only and anonymized during data analysis, after which they signed a consent form on a voluntary basis. They were also aware that they could withdraw from the study for any or no reason at any time. Upon obtaining consent, the first author accessed students’ OOPT scores through the university’s office for academic affairs.

Following a maximum variation sampling strategy, 35 first-year students who volunteered to participate in the study were invited for interviews to balance the sample across different English listening proficiency levels (indicated by their OOPT scores). Among the participants, 11 students achieved a proficiency level at B2¹ or above (referred to as high proficiency [HP]), 15 students at B1 (medium proficiency [MP]), and 9 at A2 or below (low proficiency [LP]). A total of 23 students majored in Humanities and Social Sciences and 12 in Business. Among them, there were 10 male and 25 female students. All participants had Mandarin Chinese as their L1 and attended Chinese-medium secondary schools prior to university. Twenty-eight of the students had never been abroad, while seven students had participated in short sojourns ranging from one week to one month. A female Business major student with B1 proficiency dropped out of the study at the midterm. The remaining 34 students attended interviews at all three time points. Students’ names were pseudonymized with a unique ID (e.g. Student 1 [S1]), where S1-11 are HP students, S12-25 MP students, and S26-34 LP students.

3.2. Data collection

Employing a longitudinal qualitative design, semi-structured interviews were conducted with all 34 students at the beginning (Week 3, T1), halfway (Week 9, T2), and the end (Week 14, T3) of their first term at the EMI-TNHE university. Before T1 data collection, the researcher explained the purpose of the study to the students and notified them that the interviews would be audio-recorded for research purposes only. Then, each student voluntarily signed a consent form. The students were interviewed in their first language to create a relaxing environment and encourage open participation. The first author prepared a list of questions that aimed to elicit information about the students’ SRLP. The interview protocol (see Appendix 1) included questions that were linked to Zimmerman’s (2000) three-phase model of self-regulated learning and was piloted with five first-year students from the same institution to improve the wording. Questions were asked in an open-ended manner to reduce the possible influence of the protocol on students’ responses about their SRLP. Probing questions were also asked when students’ responses tapped into relevant aspects of the model. The interviews ranged from 30 to 60 minutes and were audio-recorded. A total of 102 transcripts were generated and imported into NVivo11 for data analysis. Excerpts were then translated into English by the first author who holds a BA in translation and cross-checked by asking the students whether the translation matched their intended meaning.

3.3. Data analysis

Analysis of the interview data was guided by Kuckartz’s (2014) description of thematic analysis (Stage 1) and later by Gao’s (2022) description of narrative analysis (Stage 2). Stage 1 began by developing the main categories informed by relevant literature (top down/deductive) while allowing sub-categories and then themes within those sub-categories to be identified from the data itself (bottom up/inductive). The main categories were derived from Zimmerman’s (2000) three phases of self-regulated learning: forethought, performance, and self-reflection. After becoming familiar with the data as a whole, the first author read each transcript closely and coded relevant statements as representing one of the categories mentioned above. Then, within each main category, sub-categories were developed inductively from the data (e.g. goal setting as a sub-category within forethought). Following Schreier’s (2014) principles of mutual exclusiveness and exhaustiveness, the first author further segmented the data into sub-categories. This segmentation ensured that every unit of meaning was coded only once under one main category and that all relevant statements were second-level coded under one sub-category. Some sub-categories were aggregated to make them comparable over time. The first author then re-categorized the data as needed to ensure each sub-category encompassed a central meaning of its own and then indexed the data in a matrix. This was done for T1, T2, and T3, respectively, to identify salient themes across the participant cohort at each data collection point.

In Stage 2 of data analysis, the individual trajectory of each participant’s SRLP from T1 to T3 was analyzed to create a within-participant narrative over time. Participants’ narrative trajectories were then compared and synthesized to form a cohort narrative presented as a narrative analysis (see Gao 2022). Given the longitudinal nature of data collected over three time points, the large sample size for in-depth qualitative research, and the limited word count, key themes are presented first for the group overall – to summarize the content succinctly. Then, a synthesized narrative is provided that stories the cohort’s experiences chronologically. Key factors that seemed to mediate students’ learning trajectories are identified and reported within the overall narrative to allow for a more situated and nuanced understanding of student experiences that appeared to cluster (i.e. unfold in similar patterns) over time. Individual analysis of participants’ narrative trajectories homed in on personal development over the course of the term, while, synthetically, the final reported narrative showcases major group trajectories accentuated by excerpts illustrating individual differences. Three subheadings that correspond with each data collection point and which are illustrative of the content presented in those sections are used to organize the cohort narrative.

4. Findings

Table 1 summarizes the main themes that were identified at each of the three time points, organized according to the main categories of forethought, performance, and self-reflection. A frequency count of how many participants commented on the theme is included in the table to show its prevalence at the participant level within the data (regardless of the number of mentions in each interview). The findings subsections are presented in chronological order (T1, T2, T3) and include only participant responses from the specified time point.

Table 1. A summary of categories and themes for self-regulated listening practice over time.

Time	SRL phase	Themes related to self-regulated listening practice	N = 34
T1	Forethought	Setting listening practice goals to survive in EMI classes	11
		Planning to practice listening on a daily basis	16
	Performance	Preference for an intensive listening practice method	15
	Self-reflection	Attributing listening progress to the EMI environment	13
T2	Forethought	Listening for future development or wider communicative needs	15
		Reduced inclination to set practice goals for listening	9
	Performance	Preference for an extensive listening practice method	8
	Self-reflection	Noting substantial progress in listening abilities	20
T3	Forethought	Listening for future development or wider communicative needs	12
	Performance	Adjusting listening practice to make it more sustainable	7
	Self-reflection	Noting deceleration or a plateau in listening progress	11

Note. The frequency count refers to how many participants provided statements relevant to the theme.

4.1. (T1) upon arrival at EMI-TNHE university: Initial enthusiasm for self-regulated listening practice

Upon arrival at the EMI-TNHE university, students in general reported being enthusiastic about practicing their listening outside of class. They regarded SRLP as necessary for their ‘survival’ in the new learning environment. Students at this time tended to prefer an intensive approach to listening, viewing learning as comprised of multiple sub-tasks addressing discrete linguistic components.

At the beginning of the term (T1), a prevalent theme identified in the interview data was students’ attempts to set goals for listening practice to ‘survive’ (i.e. to understand lecture content) in their EMI classes. These goals were especially prominent among LP and MP students, as they encountered tremendous difficulties in listening to academic lectures in English: ‘The first several days of listening to the courses has been hellish. It’s the overall environment, like, almost all courses are in English, so I have to learn, and I have to practice listening; otherwise, I won’t even know what the assignment is’ (S28, LP). Describing their need ‘to simply adapt to this mode of English-taught courses as soon as possible’ (S26, LP), and ‘to understand the class in a less painstaking way’ (S19, MP), the students tended to make industrious plans for learning. Among 23 LP and MP students, 16 reported scheduling SRLP daily, with others planning to practice less frequently.

What accompanied this initial enthusiasm about SRLP was a noted preference towards an intensive approach to listening, to ‘find out how the spelling of a word is matched with its pronunciation’ (S32, LP), ‘listen sentence by sentence’ (S31, LP), and ‘listen again and again until I understand everything’ (S16, MP). An analytic view of language learning emerged, where students tended to regard English listening as comprised of handling individual linguistic items such as word pronunciation, vocabulary, and grammatical rules. An intensive listening approach, as S29 (LP) described, was considered capable of helping her tackle these different components thoroughly:

I mostly listen to short news passages, and I listen repetitively, focusing on one or two passages per week. It’s not about quantity, but quality that matters. If I can’t understand a sentence, I replay it many times until I can finally distinguish the words clearly and understand it all. That’s progress.

This reliance on intensive listening seemed to be influenced by S29’s (LP) listening practice in secondary school English classes: ‘That’s how we practiced listening in high school, like, the teacher plays the recording and asks each student to repeat, and then fills words into sentences bit by bit’. The hard work of these students seemed to pay off, as 16 out of 23 LP and MP students self-reported progress in their listening ability to varying degrees since their arrival at the university. Commenting on their drive for such progress, 13 attributed it to the EMI environment, which students described as ‘forcing me to get used to the English environment and to progress’ (S28, LP) and ‘having no choice but to adapt and try as hard as possible to listen, as everything is in English’ (S22, MP).

In contrast, their highly proficient peers appeared more at ease. HP students often reported engaging in more casual listening practice: ‘to practice whenever I feel like it’ (S6, HP), ‘listen to news after dinner when I have some time’ (S5, HP), or ‘listen to a podcast on my way to university’ (S7, HP). Even for those HP students who planned for regular practice, they treated it as an auxiliary rather than a necessary activity: ‘I just think of it as something to relax’ (S10, HP). Such individually distinctive needs regarding SRLP for students starting with different proficiency levels seemed to change when they were interviewed at the midterm – a ‘watershed moment’ when their SRLP became increasingly intertwined with evolving personal and motivational dispositions.

4.2. (T2) ‘to continue or not to continue’: a midterm watershed moment

By the midterm, most students recognized marked progress in their listening competence. Concomitantly, a ‘watershed moment’ occurred when some students continued their investment in SRLP while others reduced or stopped practicing. HP students stood out as the group with most ‘dropouts’, many of whom reported an absence of motivation to continue. Among the students who persisted in practice, their learning seemed to be directed at longer-term goals for self-development, with a diversion in the favored approach – from intensive to extensive listening.

Notably, HP students reported the most reduction in practice, where only two out of eight who practiced regularly at the start of the term continued to do so at the midterm. The other six students either stopped SRLP completely or only practiced occasionally. Although most of the participants still recognized the benefits of such practice, it was seen as a low-priority activity. Some HP students seemed to be caught between an absence of the need to practice further and a concern of lagging behind peers in non-EMI institutions. This ambivalent feeling was illustrated by S10 (HP) below:

I tried to practice with the materials on those English mobile apps but gave up not long after. I don’t have such will to persevere. I’m comfortable with where I am now in this university, so my attention gets easily drawn to other activities. But I know in other universities, students are practicing listening and building up their vocabulary because they have to attend the CET Band 4 exam².

A higher ratio of LP and MP students continued their listening practice after class. However, for five students who stopped practicing or began to practice very little, a main reason emerged: by the midterm, as students perceived improvement in their listening ability, they considered themselves more capable of handling listening in EMI classes. Consequently, additional practice outside of class became a lower priority than before to make way for other activities:

Recently, I’ve become very busy, not much free time, so actually very little listening practice after class. But it’s okay, as now I feel my listening has improved quite a lot compared to the beginning of the term, and I feel that I’ve understood much more than before anyway (S26, LP).

Those who persevered in practice, however, all at some point indicated a goal surpassing the immediate need of understanding EMI lectures, which was, instead, directed at a need for their future plans and/or development. This is reflected in S16’s (MP) comment, when she described a necessity for continuing practice:

I think, in general, my listening ability has improved. You know, after all, I’ve always been practicing. Still, there are many hills ahead that I need to climb. For me, there’s a real need in the future. I have expectations for using English not only to understand classes, but also to communicate, to study abroad, and for my development in the long run.

This change in the listening goal was accompanied by a shift in the student’s preferred listening method – from intensive to extensive practice – to foster ‘an ability to grasp the general message from the fast talk of the speaker’ (S16, MP). Similarly, this preference for an extensive listening method also emerged in seven other students’ accounts, which they hoped to allow them to ‘engage more freely in conversations with others’ (S33, LP) and ‘better follow the ideas of international friends when they talk fast’ (S16, MP). As a result, they adapted their practice methods to cater to their communicative needs.

In sum, the midterm appeared to be a watershed moment. It distinguished between two groups of students: (1) students who stopped further practice after feeling as though their comprehension in EMI lessons was sufficient and (2) students who continued listening practice with motivation to improve for future-oriented goals (e.g. using English for personal development and/or interpersonal communication).

4.3. (T3) towards the end of the transition: a pursuit of sustainability

From the midterm towards the end of the term, progress in listening was perceived by many students to slow down, triggering doubts and uncertainties that led to further ‘dropouts’. For those who persisted, however, they adjusted their effort management and material selection, showed high resilience to adversity, and developed a mechanism of reflective, positive feedback for sustainable SRLP.

Different from the midterm, when students generally perceived notable progress in listening, many students at T3 reported that their progress slowed down during the latter half of the term. Students described their progress in listening as ‘very slow and almost unperceivable’ (S28, LP), and despite ‘some improvement’, it was ‘not quite as expected’ (S12, MP). Some students even felt that they reached a plateau where it might be hard to improve any further. As such, another three LP and MP students were discouraged and stopped SRLP.

A key factor distinguishing the ‘continuers’ from the ‘dropouts’ seemed to be their ability to conduct reflective self-diagnosis to gauge their progress and a resilience towards adversity. For the former, ‘progress’ tended to be defined in vague terms, which seemed to easily result in negative feedback discouraging further practice effort. An example of this is S30 (LP), who lamented a lack of perceived progress since the midterm: ‘It’s hard to gauge progress in listening; it’s abstract and has no output. I feel I haven’t found the right method, that is, a method that can keep me going’. In contrast, the ‘continuers’ showed more patience and resilience in the situation, accepting that improving listening ability might be time-consuming: ‘From the start of the term till now, the progress in my listening is like from a seed to the first leaf, but it might take a while for it to blossom’ (S33, LP). They also seemed to be more capable of reflectively diagnosing and confirming progress in their listening in a more concrete manner. An example of this is S16 (MP), who commented on how her ability to ‘listen impromptu’ was improved due to her practice:

I feel my impromptu listening has improved recently. Before, I could understand only when I was prepared, like, when I knew I’d listen about a particular topic. But now, I can better understand when entering a listening scenario unprepared. I can grasp the main message even if I don’t know what the person has been talking about. I think after enough practice, I’m more sensitive to some sentence structures, and I can guess more accurately when I miss certain parts.

Recognizing that ‘improving listening is a long-term process’, the student stressed that she would not feel defeated easily: ‘Sometimes during practice, I still couldn’t understand a lot, but I’d tell myself, that’s okay, and not to feel bad about it’ (S16, MP).

Another important feature identified across the ‘continuers’ narratives was their timely adjustment of practice methods and materials. In the face of perceived deceleration of progress, several students mentioned attempting to balance between their invested effort and long-term returns, adapting the amount of practice per day to trade for higher quality practice and/or persistence over time. S20 (MP), for example, reported compressing practice time from one hour per day to 40 minutes per day in the hope of improved quality of concentration: ‘I’m worried that if I overstretch myself, this could lead to low effectiveness.’ This relatively small change made the student feel the practice was more sustainable long term. Similarly, S31 (LP) stated that ‘I no longer force myself to listen to a word’s pronunciation until I fully master it, because that’s too complicated, and I can’t persist.’ Instead, she reported that she would ‘note the words down, rehearse several times, and try to memorize them’, believing that she could ‘persevere in the practice’ by leaving unmastered content for another day.

The pursuit of a more sustainable way of practicing was also reflected in students’ selection of listening materials. Those materials with high entertainment value (e.g. films, documentaries) replaced materials with ‘serious’ topics, which were popular at the start of the term: ‘I tried to listen to the BBC news, but it’s too detached from our life, and there were many terms I couldn’t understand. Now, I mainly watch English TV series or documentaries, for fun, and for learning’ (S19, MP).

In sum, from the midterm to the end of the term, students seemed to encounter a bottleneck in improving their listening. This perceived bottleneck discouraged some students from SRLP. However, for those who persisted, they agentively adapted practice, reflectively assessed their own progress, and resiliently sought more sustainable ways for directing and regulating their own learning.

5. Discussion and pedagogical implications

This study provides insight into the out-of-class self-regulated learning that Chinese students engaged in over their critical transition term from Chinese-medium high schools to an EMI-TNHE university. To our knowledge, it is the first longitudinal study to use in-depth, qualitative research to explore SRLP in such a context. Based on our synthesis of 102 interviews with 34 students, we will discuss what we have identified as the main takeaways in relation to the theoretical framework and existing literature. We present this section as a meta-narrative of its own that progresses in relative chronological order but with a more thematically oriented discussion than the findings above.

5.1. Appropriate strategies to reach the initial language threshold and beyond

Our findings show that upon arrival at the EMI-TNHE university, students generally reported initial enthusiasm (motivated intention) to engage in SRLP (Zimmerman 2000). However, practice effort varied across participants’ different proficiency levels. LP and MP students appeared to be primarily driven by an outcome expectation to ‘survive’ their English-medium lectures. Accordingly, these students planned industrious SRLP. They relied on intensive listening strategies at T1 that modelled their form-focused, examination-oriented language education in secondary school. Although these strategies could help the students address decoding issues (e.g. segmenting speech) (Field 2019), they would later report that their initial strategies did not seem to prepare them for complex communicative tasks and were unsustainable for long-term development. These findings highlight that students’ transitional learning strategies were heavily shaped by their stabilized knowledge structures about how to learn that were developed during their previous educational experiences. This echoes Macaro et al.’s (2018) view that understanding transition into EMI needs to consider what comes before it. Therefore, for successful transitions to take place without initial panic, students need to move from a compartmentalized view of language as a series of linguistic components to be accumulated to a view of language as a social semiotic that enables meaningful communication (Bowen and Thomas 2022) and disciplinary literacies (Airey 2020).

In contrast, HP students at T1 approached their SRLP in a relaxed, intrinsically motivated manner. They recognized the benefit of SRLP but were mainly motivated to engage in activities they enjoyed, echoing Jiang and Zhang’s (2019) finding that high achievers often reach a certain threshold where pressure from examinations no longer motivates their language learning. Indeed, HP students at T2 reported that without a national examination to prepare for (cf. their Chinese-medium university peers) and a relatively ‘comfortable’ level of comprehension in their classes, they became demotivated for SRLP (including some LP and MP students who perceived themselves as reaching an adequate level). In the absence of clear language admission thresholds (Ali 2013; Aizawa and Rose 2020), students’ self-assessment of their need to engage in SRLP may be problematic for their advancement, as there is little evidence that an EMI context alone promotes significant language development (Lei and Hu 2014; Macaro et al. 2018). In the current study, many students seemed under-prepared for how best to select appropriate strategies to regulate their learning. In a context where the curriculum does not have an official self-regulated learning component (e.g. Zhou et al. 2022), advising should be provided to enhance strategy awareness and expand students’ beliefs about how to learn, guiding them through pathways for improving beyond the initial language threshold when and where appropriate (see Mynard and Shelton-Strong 2022).

5.2. Motivational shifts from the immediate to the future

Most LP and MP students successfully navigated the bifurcation that took place at the midterm – a watershed moment in which some students continued their SRLP while others reduced or abandoned it. Those who persisted were pushed beyond immediate ‘survival’ and into more integrated, future-oriented motivational trajectories, as described in their discussions of their future selves communicating beyond the university, in study abroad contexts, or for future development in general. These findings resonate with a number of studies in which motivation for learning appeared to be more strongly influenced by longer-term, future-oriented goals than by the immediate context (e.g. Lasagabaster 2016; Jiang and Zhang 2019; Yuan et al. 2023), especially with regard to students’ emerging views of English for international use (e.g. Iwaniec and Wang 2022) and in study abroad (e.g. Liu et al. 2021; Boonsuk and Fang 2023).

As mentioned above, initial incentives for language development in EMI settings are often unsustainable (Gao 2008). In the current study, students who persevered in their SRLP often indicated strong orientations towards immediate, extrinsically motivated goals at T1 but later evidenced a shift in motivational orientation towards longer-term goals at T2 and T3. Although still considered extrinsic motivation, these shifts often resembled a progression towards highly integrated and/or intrinsic motivation. Such shifts have long been part of self-determination theory’s conceptualization of human thriving (see Ryan and Deci 2020), Thomas and Rose’s (2019) continuum of strategic behavior, and, more generally, Zimmerman’s (2000) aims for self-regulated learning. All three theoretical perspectives indicate that these shifts are positive and should be seen as educational ideals (Thomas et al. 2021).

5.3. Metacognitive awareness as enabling sustainable self-regulated learning

Although more students considered themselves SRLP ‘dropouts’ at T3, there were certain factors that the ‘continuers’ had in common. Overall, continuers displayed strong metacognitive awareness in monitoring their SRLP with an orientation towards resilience. Their poised reflection enabled them to self-diagnose areas for improvement, even in the face of adversity (e.g. misunderstandings or perceived slowed progress). They used emotion regulation strategies such as positive self-talk to maintain their motivated intention (see Gross 2015), which, in turn, created continued opportunities for them to engage in additional SRLP. Expanding on emerging EMI research on student emotions (see Şahan and Sahan 2023), our finding suggests that emotion regulation such as cognitive change through positive thinking might help students reappraise their progress and sustain self-regulated learning behaviors. Although the continuers’ ability to develop as self-regulated learners on their own is a testament to their forethought, performance, and reflection feedback loops fueling their learning mechanisms (i.e. self-generated feedback as feedforward for learning) (see Zimmerman and Moylan 2009), teacher-led metacognitive training could have enabled more of the participants to continue (see Goh and Vandergrift 2022; Bozorgian and Shamsi 2023).

Finally, it is important to note that the students’ SRLP during transition was, in essence, an exploration of independent learning during a time when many students experience feeling ‘lost’ or ‘confused’ for reasons beyond language proficiency (Ding and Curtis 2021; Wang and Chao 2022). Indeed, transition of any sort can be difficult and often leads to both learning and identity-related transformations. Schneider (2022) considers strategic behavior as acts of identity for students orienting to new contexts. In the current study, general trends in strategy use for the continuers can be mapped onto three strategy categories Schneider (2022) developed: survival (T1), investment (T2), and growth (T3). We interpret growth as not just an attempt to improve listening proficiency but as an expansion of the learners’ mindset towards more sustainable listening practices. However, we believe this growth could not have occurred in the absence of their ability to develop strong metacognitive awareness. Thus, we see clear links between cognitive and identity-oriented perspectives of strategic learning. We believe that in supporting students to develop metacognitive awareness via guided facilitation (see Goh and Vandergrift 2022), educators can better support learners in their self-regulated learning endeavors. This may reduce the number of ‘dropouts’ in future transitions and enable students to thrive in the immediate context and beyond – as both lifelong and lifewide learners across contexts (Reinders et al. 2023). With a self-informed plan for learning, students can better manage their expectations about listening development and practice sustainably: more is not always better.

6. Conclusion

Tapping into the central and controversial issue of students’ language development in EMI contexts, this study analyzed how 34 students engaged in out-of-class SRLP during their first term transitioning into an EMI-TNHE university. Following the analysis of individual narratives, we identified general trends within the group and presented these as a synthesized narrative of the learners’ practices. In doing so, we identified key learner factors that seemed to mediate the students’ SRLP. Findings highlight that an EMI environment provides initial incentives for low and medium proficiency students to engage in self-regulated learning in hopes of achieving an abstract language threshold for ‘surviving’ academic lectures. However, SRLP ‘dropouts’ were common. Those who continued SRLP after reaching a perceived threshold appeared to be driven by longer-term, future-oriented, and often intrinsically motivated communicative goals. The ‘continuers’ also reflected higher metacognitive awareness to monitor and evaluate their learning outcomes, using effective emotion regulation strategies to help them persevere against adversities. Their metacognitive awareness also helped them to adjust their strategic behavior in pursuit of sustainable self-regulated learning to thrive in the EMI-TNHE context and beyond.

Some limitations of the study should be noted. First, data were collected at an EMI-TNHE university that uses English for instructional and administrative purposes more widely than other EMI contexts where multilingualism is usually prevalent (see Zhou et al. 2022). Hence, implications from the study should be considered with findings from contexts where teachers and students may harness other semiotic resources for lecture comprehension and listening development. Second, we sampled social sciences, humanities, and business majors. Future studies may seek to diversify the demographics of the participants to examine whether disparities exist in their self-regulated language learning. Finally, future research can also consider using other research instruments such as learning journals and more objective measures to triangulate the findings. Despite these limitations, the current study is one of the few empirical investigations in L2 self-regulated learning research that utilized a longitudinal qualitative design to explore the under-investigated skill of listening. It is also among the first studies to combine thematic and narrative analysis to unpack the complexities of students’ self-regulated language learning outside of class in an EMI context. We call for more in-depth longitudinal studies to unpack the processes involved in students’ language learning behind the scenes. In doing so, implications can be offered to enable effective support for students to thrive within their EMI study and beyond.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Direct Grant [4058098] at the Chinese University of Hong Kong.

Notes

1 In CEFR levels, A1 and A2 refer to ‘Basic user’, B1 as ‘intermediate’, and B2 as ‘upper-intermediate’. C1 and C2 represent ‘proficiency user’. In our study, we align with the CEFR grouping to define low proficiency group (A1 & A2), and medium proficiency group (B1). However, for high proficiency group, we use B2 as a cut-off point because previous research has suggested it as a commonly adopted threshold for successful academic study at the university level (Harsch 2018).

2 In China, non-English major undergraduate students need to take the College English Test (CET) Band 4, which is a prerequisite for obtaining a bachelor’s degree in many Chinese-medium universities.

Appendix 1.

Interview protocol

1. What listening expectations (or goals), if any, have you set for yourself to improve listening? Why have you set these expectations (or goals)?

2. What practice plans, if any, have you made for achieving the goals?

3. How have you implemented your practice plans? Could you describe in detail how you usually practice listening?

4. How do you feel when practicing listening? What do you usually do about these feeling?

5. What difficulties have you encountered during practice? What do you usually do about these difficulties?

6. How do you think of your listening practice process (e.g. methods, materials, etc.)? Any problems or issues you have noted?

7. How do you think of your progress in listening since our last interview? What do you think are the factors that have affected your progress?

8. What new goals (or plans) for listening practice would you like to set for yourself in the following weeks? Why?