Introduction

The Australian higher education sector provides tertiary education to a large number of international students who speak English as an additional language (EAL). Studies on EAL international students have shown that English language proficiency (ELP) is a critical factor related to their well-being (e.g. Li et al., 2014; Marginson et al., 2010). Our study examines the effects of a personalised autonomous approach on international PhD students’ English self-efficacy, their psychological well-being, and the interrelation between these factors, ultimately influencing their research performance. While research underscores the essential role of ELP in the psychological well-being of international students (e.g., Li et al., 2014), there is scant research that explores how to assist students in bolstering their English communication confidence and its subsequent impact on their well-being.

This paper is an outcome of a collaborative research project between the Graduate Research School (GRS) and the School of Humanities and Languages at the authors’ university.Footnote 1 Since 2017, GRS has been supporting international research students enrolled in a PhD or Master by Research program through a course called Personalised English Language Enhancement (PELE). The PELE course has been proven to be effective with significantly positive impacts on students at all levels of study (undergraduate, postgraduate coursework and postgraduate research) in “a) their confidence with English skills in both academic and everyday contexts, b) self-efficacy for self-regulated English learning and academic and social engagement, and c) their sense of belonging to PELE, faculty and university” (Kim, 2023:615).

However, the prior study has limitations due to its absence of a control group. Thus the present study aims to complement the previous study in the following ways. Firstly, it measures the effects of PELE by contrasting PELE participants with international students who did not undertake PELE (referred to as non-PELE). Secondly, given the significant impacts of PELE on students’ confidence with English skills, self-directed learning skills, behavioural engagement, and sense of belonging already documented in previous research, this study examines the relationships between these variables and student well-being. Finally, it seeks to investigate the underlying principles and the practical pedagogical methods of PELE that contribute to effective student learning, as evidenced by both quantitative and qualitative data analyses.

In the following sections, we will first introduce the PELE approach and present the methodology chosen to investigate the research questions. This will be followed by a detailed presentation of the findings from the survey and focus group data. After that, we will synthesise these findings with theoretical explanations and discuss the broader implications for higher education sector.

PELE Approach

The pedagogy of PELE is a personalised autonomous (PA) model developed by Kim in the context of translator and interpreter education through a series of action research (see Kim, 2014 and Kim & Jing, 2019 for details). The PA model promotes a cyclical exploratory research approach to learning in which students are guided to personalise their own learning in five steps: 1) becoming aware of the context in which they want to communicate better in English; 2) analysing their linguistic needs; 3) designing a personal project with the methods and resources they prefer to use; 4) implementing the project for 5–6 weeks; and 5) reflecting on their progress. This approach is deeply grounded in sociocultural theories of learning (for details see Kim, 2023). This learning cycle was devised to equip students with the ability to independently enhance their language skills, or to acquire any other knowledge or skill they desire, even beyond the course’s duration. Therefore, the framework enables students to exercise their agency, guiding them to develop reflective learning practices that are integral to sustainable language development, but are lacking in existing ELP models.

Throughout the term, PELE students are encouraged to work with the five principles:

  1. 1.

    Work smarter not harder.

  2. 2.

    Make plans but remain flexible for the unexpected.

  3. 3.

    Communicate honestly and respectfully.

  4. 4.

    Have a growth mindset rather than a fixed mindset.

  5. 5.

    Have fun!

The primary principle is to encourage students to explore new and innovative approaches for enhancing their English skills, especially when traditional methods prove ineffective. Thus, in the second stage of the learning cycle, students are introduced to numerous resources and tools (for details see Kim et al., under review, Bedford et al., in press, Li & Kim, under review). The second principle aims to bolster students’ capacity for responding to unforeseen circumstances with flexibility and resilience, such as when their project does not proceed as planned. Students are encouraged to modify their initial project design as needed during implementation to ensure the best possible outcomes. The third principle, significant for any communicative context, involves introducing students to the fundamental aspects of crucial conversations and strategies for managing them, drawing on communication principles from books such as ‘Nonviolent Communication’ (Rosenberg, 2015) and ‘Crucial Conversation Skills’ (Patterson et al., 2002). The fourth principle encourages an enthusiastic attitude towards learning and persistence in achieving goals and the ability to learn from failures and mistakes. The last principle, which is emphasised throughout the course, recognises that genuine enjoyment in learning is indispensable for language acquisition and its practical use in real-world contexts. Thus, students are guided at every stage to identify and pursue what they find enjoyable, be it a project goal, method or resource. Should they find their chosen method no longer enjoyable, they are encouraged to make adjustments.

Additionally, PELE fosters a safe learning environment, largely facilitated by peer mentors who have previously completed the course. These mentors share their learning experiences and resources during classes, and lead social clubs that reflect their interests, such as a book club, singing club, and free talk club. A review of their expression of interest forms, in which they explain their motivations for volunteering to be peer mentors, consistently reveals three main reasons: 1) a desire to further improve their English communication or leadership skills; 2) an intent to contribute back to the PELE community, acknowledging the benefits they have gained; and 3) a wish to maintain their connection with the PELE community. These social clubs offer genuine opportunities for students to practice English and make friends.

Methodology

This study took a mixed method approach, collecting data across two terms (Term 3 2021 and Term 1 2022) through entry and exit surveys, as well as focus groups and individual interviews. Surveys were the primary tool to understand general patterns whereas focus groups and interviews were utilised to attain a more insightful understanding of those patterns.

Survey Design

Entry and exit surveys started with demographic questions including students’ native language, residency status, and current location (onshore or offshore) and include a series of questions to assess students’ self-efficacy in English language skills and self-directed learning, well-being, as well as sense of belonging to the university community. Self-efficacy is people’s feeling about their abilities to deal with specific tasks and proves a reliable variable in predicting students’ performance (Bandura, 1997). There is currently no standardised measure of self-efficacy in language learning. We designed 25 “can do” statements regarding English usage across the four language skills in both academic and everyday contexts (e.g., “I can communicate my thoughts in class or lab in English”). Students indicated their English capabilities by rating their agreement with these statements on a 5-point Likert scale from 1 (I cannot do it at all) to 5 (I can do it easily). We used Bandura’s (2006) framework for building a self-efficacy questionnaire and modelled our questions on a similar survey designed by Wang (2004), which was developed to assess Chinese students’ self-efficacy with everyday English usage in the United States. Social self-efficacy was measured by giving students two “can do” statements about their ability to make friends and work in a group. As the PELE approach emphasises self-directed learning, we also designed four questions assessing students’ self-efficacy regarding each of the personalised learning skills PELE teaches (e.g., “I can identify areas that I could improve in my studies”). After the self-efficacy questions, students were asked to indicate their sense of belonging to the university community (i.e. How much do you feel that you are connected to the overall UNSW community?).

Finally, the Warwick-Edinburgh Mental Well-being Scale (WEMWBS; Tennant et al., 2007) was utilised as a measure of psychological wellbeing. The WEMWBS has been widely used in various contexts with high validity and reliability (Cronbach alpha = 0.91, retest reliability at one week = 0.83). In addition, the items in the scale are worded simply and clearly, and focus entirely on positive psychological functioning, with items that measure positive affect, positive functioning, and positive inter-personal relationships. This makes it well-suited for investigating the well-being of international students.

The same questions were asked of PELE and non-PELE students at entry and exit. Entry surveys also included their perception about support for language learning skills at the university. PELE students were asked a few additional questions in the exit survey regarding how PELE impacted their perceptions of language-learning, their satisfaction with the PELE course, and sense of connectedness to the PELE community.

In addition, supervisors of PELE students were invited to take part in a survey at the end of each term, to indicate whether they noted any improvements in their students’ communication skills after taking the course. Supervisors were asked to rate their agreement, from strongly disagree (1) to strongly agree (5) with how much their students had improved in 10 aspects of communicating with them. They were also asked how likely they were to recommend PELE to future international students.

Survey Participants

Surveys were conducted at the beginning and end of Term 3 2021 for both PELE and non-PELE students. Further PELE data was collected in Term 1 2022 to increase the sample size of pairwise comparisons and strengthen evidence of relationships between confidence and well-being. In both terms, international students were banned from entering Australia due to the COVID-19 pandemic. That is why many participants took the surveys overseas. PELE student surveys were administered online during class by the third author, who was not involved in the teaching of PELE. Surveys were roughly 10–15 min in duration. On average 45 students took the course in each term. For non-PELE students, an invitation was sent out to 1,089 early-stage postgraduate research students (up to 1.25 years) who were not taking PELE that term. Most of them were enrolled in a PhD program and some were in a Master by Research program. All students to complete both entry and exit surveys were entered into a draw for ten $25 vouchers for the university bookshop.

Final samples consisted of 62 PELE students and 35 non-PELE students who completed both entry and exit surveys. They were largely international students but some of them were non-native speakers of English who held domestic residency status through various visa-types. These students can easily fall through the system as they are classified as domestic students but still need English support. In this paper, we use “international students” to refer both groups because most survey participants were international students and there was no distinction between the groups when it comes to English language proficiency. Detailed demographic information is provided in Table 1 below.

Table 1 Participant demographics
Full size table

Participants’ English Support Needs

In the entry surveys students were asked if they felt they needed or if they had been advised to seek English language support. As shown in Fig. 1, most PELE and non-PELE students indicated that they needed support. However, there were many non-PELE students who were unsure, and some who felt they did not need any support. Inspecting the ‘Advised’ graph, we can see that most PELE students had been advised to seek support, whereas most non-PELE students had not. Potentially, being advised to seek out language support may be a significant determinant of whether students feel like they need English language support and choose to enrol in PELE.

Fig. 1
figure 1

(a) Self-perceived needs for English support and (b) Advised to get English support

Full size image

Major Findings from Survey Data

Survey Data Analysis

Raw scores from each item of all the measures were summed to calculate total ELP self-efficacy, self-directed learning (SDL) self-efficacy, social self-efficacy, sense of belonging to the university community, and well-being scores for each student. Mean scores for PELE and non-PELE students at entry and exit were calculated for each measure and presented in Fig. 2. Cohen’s d was also calculated as a measure of effect size for each analysis (denoted by the lowercase ‘d’ in the reported statistics), which indicates how large statistically the differences between the groups were (as opposed to the significance value ‘p’ which reflects the probability of this difference occurring by chance). In this section, we will present results comparing the two groups, PELE and non-PELE, at entry and exit points.

Fig. 2
figure 2

Comparison of mean scores between PELE and non-PELE at entry and exit. Error bars represent standard error of the mean

Full size image

PELE Entry Vs. Exit

Paired samples t-tests were run to compare entry and exit scores for PELE students and statistics are presented in Table 2. As shown in Fig. 2, PELE students significantly improved from entry to exit in ELP self-efficacy, SDL self-efficacy, social self-efficacy, well-being, and sense of university community. Measures of effect size (indexed by cohen’s d column in Table 2) indicate that most of these differences were a medium effect (0.5–0.8). However, the changes in university community showed a large effect (> 0.8), indicating that PELE had a large effect on the feeling of belonging to the university community in the current sample.

Table 2 Paired samples t-tests comparing PELE students’ scores from entry to exit
Full size table

Non-PELE Entry Vs. Exit

Table 3 presents statistics from the paired sample t-tests comparing non-PELE students from entry to exit. As shown in Fig. 2, non-PELE students did not significantly improve in any variable from entry to exit.

Table 3 Paired samples t-tests comparing non-PELE students’ scores from entry to exit
Full size table

Comparison of Improvements

Figure 3 shows that mean improvements from entry to exit were greater for PELE than non-PELE students in all the dependent variables presented. Comparison of these improvements (statistics are presented in Table 4) reveals that PELE students improved significantly more than non-PELE students in ELP self-efficacy and sense of university community, with moderate to large effect sizes. The averages for well-being and SDL self-efficacy were much higher in the PELE group but the differences were not as significant as ELP self-efficacy and sense of university community. The lack of significance can likely be attributed to large variability in the samples.

Fig. 3
figure 3

Comparison of mean improvements between PELE and non-PELE students. Error bars represent standard error of the mean

Full size image
Table 4 Between-subjects t-tests statistics comparing improvements between PELE and non-PELE students
Full size table

To better understand what students believe contributes to their sense of belonging to the university, we analysed more than 200 comments in the entry survey. Those who felt strong connections to the university (rated 4 or 5) mentioned strong support and kindness from various sources, including supervisor, school, faculty and university. The majority of those who felt less connected to the university (rated 3 or lower) mentioned COVID-related restrictions and/or being a new student, which are understandable considering the pandemic restrictions imposed in both terms. However, there are many comments about their self-perceived lack of English skill and unwillingness to engage in university activities from both PELE and non-PELE groups. Some comments indicated that these were related to each other:

I feel like spoken English is a big problem for me, it does impact me to make friends and express my thoughts. … I often worry that native English speakers wouldn't listen to me because I speak slowly and always pause. … Because of the language barrier I find it difficult to integrate into Australian culture, in fact I would like to be integrated here. This made me feel helpless and sad. (non-PELE entry survey)

Interestingly, comments to the same question in the exit survey differed strikingly between the groups. While the non-PELE group was still exclusively concerned about COVID restrictions, there were few COVID-related comments from the PELE group. Rather, most were about PELE’s contribution to their sense of connection to the university:

Definitely, better than before! I became a member of the PELE community, and feel more a member of the UNSW community member.

Well-Being Analysis of Individual Items

We used the 14-item WEMBWS scale (see below) to examine students’ well-being at entry and exit:

  1. 1.

    I've been feeling optimistic about the future.

  2. 2.

    I've been feeling useful.

  3. 3.

    I've been feeling relaxed.

  4. 4.

    I've been dealing with problems well.

  5. 5.

    I've been thinking clearly.

  6. 6.

    I've been feeling close to other people.

  7. 7.

    I've been able to make up my own mind about things.

  8. 8.

    I've been feeling interested in other people

  9. 9.

    I've had energy to spare

  10. 10.

    I've been feeling good about myself

  11. 11.

    I've been feeling confident

  12. 12.

    I've been feeling loved

  13. 13.

    I've been interested in new things

  14. 14.

    I've been feeling cheerful

Figure 4 indicates that the PELE group experienced a greater improvement than the non-PELE group. However, the PELE group showed a statistically significant improvement in confidence (item 11) compared to the non-PELE group.

Fig. 4
figure 4

Breakdown of PhD students average WEMBWS scores

Full size image

Correlations of Well-Being and Other Variables

To assess whether improvements in students’ ELP self-efficacy, social self-efficacy, SDL self-efficacy, and sense of belonging to university community were accompanied by enhancements in their well-being, we ran Pearson’s r correlation tests on the difference scores (exit minus entry) for each dependent variable. Figure 5 shows the correlation matrices with Pearson’s r values and significance (indicated with asterisks) for PELE and non-PELE groups separately. For PELE students, well-being showed significant positive correlations with all other dependent variables, and particularly strong correlations with the self-efficacy variables. The same was true for non-PELE students, although well-being was not correlated with sense of belonging to of university community for these students. Also, there were significant correlations between all the self-efficacy variables in both groups. Interestingly, sense of university community was not correlated with social self-efficacy, but it was positively related to self-directed learning in both groups, and to ELP self-efficacy in PELE group. Not shown in the figure, for the PELE group there was also a significant positive correlation between their sense of PELE community and university community at exit (r = 0.441, p < 0.001).

Fig. 5
figure 5

Correlation matrices presenting Pearson’s r values for correlations between all measures. Single asterisk represents significance at the 0.05 level, double asterisks represent significance at the 0.01 level

Full size image

Additional PELE Exit Questions

In the exit surveys, PELE students were asked additional questions regarding changes in their perceptions of learning, the impact of PELE on their overall university learning experience, their sense of belonging to the PELE community, and their satisfaction with the course and the university. Figure 6 displays frequencies of students’ ratings for each of these questions. Firstly, when asked to rate from 1 (not at all) to 5 (yes, definitely) whether the PELE course changed the way they think about learning, most students indicated a 4 or 5 (‘Learning’ graph in Fig. 5 below) and a significant majority of students indicated that they preferred the personalised learning approach of PELE to a standard curriculum.

Fig. 6
figure 6

Frequencies of student ratings across various questions at exit surveys

Full size image

When asked if the skills they learnt in the PELE course helped them to get the most out of their university experience from 1 (strongly disagree) to 5 (strongly agree), most students rated a 4 or 5 (‘Skills Impact UNSW comm.’ graph). The results show that the impacts of students’ learning go beyond PELE. Most students felt a 4 or 5 (from 1 = not at all connected, to 5 = strongly connected) in their sense of belonging to the overall PELE community.

Many respondents provided additional comments that PELE triggered a change in their thinking, allowing them to extend their learning process and develop as learners. One student commented, "PELE educated me to develop a growth mindset and use self-reflection to adjust my learning activities." Another student shared, "I have discovered that learning could be fun. The supportive and positive energy of PELE was amazing." Many students mentioned that they found (unexpected) joy in improving their language skills, which they attributed to the PELE community and conducive learning environment.

The satisfaction levels of students in PELE were consistently high, with a large positive correlation found between student ratings of PELE and university satisfaction (r = 0.563, p > 0.001). According to one student, "PELE has been the most productive and efficient program that I have ever been involved in." Another student expressed gratitude for the course, saying, "I feel thankful for PELE and the opportunity to learn and improve my English skills in a supportive environment."

Supervisor Data

Out of 95 supervisors who were invited to complete the survey at the end of each term, 16 supervisors took part (T3 = 11, T1 = 5). Figure 7 displays the frequencies of supervisors’ ratings across each variable regarding their perceptions of their students’ communication. As shown, most supervisors agreed that their students improved in confidence, verbal fluency (speaking clearly), willingness to communicate their needs, taking more initiative to communicate, enjoying communication, and oral presentations. Otherwise, supervisors mostly indicated that they neither agreed nor disagreed as it was difficult to see improvement in such a short period of time. Only a very small number disagreed with any of the statements.

Fig. 7
figure 7

Frequencies of supervisors’ ratings for each communication variable

Full size image

When asked to elaborate on their responses qualitatively, some supervisors observed some tangible improvements, while others could not tell immediately:

Spoken English takes a while but I can see good signs

It was very effective and great way for them to improve their English

Apart from confidence and overall improvement in English language skills, [student name] is more self-aware about how he allocates his time. He has also learnt the importance of time management and focus and is actively applying them to his research program.

Finally, when asked about the likelihood of recommending PELE to future international PhD students, nearly all respondents indicated they were either 'highly likely' (12) to recommend it, or 'likely' (2). Two respondents were unsure.

Three supervisors took part in a focus group and said that they noticed improved confidence and communication skills in their students who had taken PELE. One supervisor in particular now recommends PELE to most of his international students as he eye witnessed such positive changes in his PhD students who completed PELE in the past. He emphasised that confident communication skills are significant in STEM fields where students often engage with industry partners through presentations or research collaborations.

Major Findings from Focus Groups and Individual Interviews

The survey data clearly showed that PELE students’ self-efficacy in using English improved significantly after completing the course. However, the data was limited in understanding why it improved, and how it impacted their research and well-being. Thus, focus groups were conducted to gain a deeper understanding about the questions.

Focus Group Questions and Participants

Students were invited to participate in focus groups by ticking a box in the exit survey. In total, 12 students (F = 7, M = 5) took part in one of three PELE groups and one non-PELE group, as outlined in Table 5 below. Pseudonyms have been assigned to ensure confidentiality of responses.

Table 5 Focus group participants
Full size table

All four non-PELE students who took part in the focus group in Term 3 2021 enrolled PELE in Term 1 2022. They were individually invited by the third author for an interview to compare their experiences before and after taking the course and two of them (Xiang and Feng) agreed.

All focus groups and interviews were conducted by the third author via Zoom due to ongoing COVID-19 pandemic restrictions and the dispersed geographic locations of participants. They were transcribed from recordings and data analysis was conducted by the third author using NVivo12, following the six-phase approach to Thematic Analysis (TA) developed by Braun and Clarke (2012). A predominantly inductive approach to coding and analysis was employed, whereby codes and themes were generated from the content of the transcripts.

Findings from Non-PELE Focus Group

Non-PELE participants were asked about their perceptions about their English language skills and confidence using English, as well as their well-being. It was strikingly noticeable that they constructed meanings around their English language skills using a deficit discourse, mainly focusing on their shortcomings or insufficiencies. While students reported using various tools and techniques to enhance their English proficiency, they still faced ongoing difficulties with communication. Their perceived shortcomings in English language skills appeared to have an impact on their research and overall well-being.

Deficit discourse of framing language ability

Non-PELE focus group participants unanimously perceived their English language skills to be below standard. They reported particular difficulty with verbal self-expression, and how this impacted their social life:

I can't express myself smoothly by English. … it is very difficult, for me. (Feng)

… if when I need some professional help, it does matter … I don't know how to talk about my back pain, or something to the doctor in a very detailed way” (Xiang)

This sense of deficiency appeared quite deep-rooted. While Moni seemed to have internalised and extended this to her entire nation, Juno’s perceived lack of language translated into his ideas seeming less worthy than those of his native-speaker colleague:

My English is still not quite good because Bangladeshi people are not good at English. (Moni)

… in case of my colleague who is … always … his idea is [more] attractive than me [mine]. … He … when supervisor ask certain question, he never nervous about the question and then, very fluent about the answer. (Juno)

Impact of English Language Skills on Research and Wellbeing

Non-PELE students shared various strategies they had used to improve their English language skills. These included engaging with UNSW English support services, using online and digital technologies such as BBC Learning English and online oral tutors, as well as seeking to improve their English skills through social interactions and participation in events. Despite employing a range of strategies, students commonly reported that their limited English language proficiency had a detrimental effect on their research and supervision experiences, which were intricate and multifaceted:

I always prepare my presentation for the meeting, but when I have a meeting with my supervisor, I am nervous and then sometimes I forgot the word I need to speak for describing my idea. (Juno)

…the most difficult part for me in this research is to discuss with my supervisor and my classmates. You know, especially when the discussion is long when the time period becomes long, my English level will drop down. (Xiang)

Participants also viewed their inability to communicate effectively in English as a major stressor that impacted their wellbeing:

…after arriving [in] Australia, I found I can't understand what people [are] saying … [I] lost my confidence in the first year... So, so, it makes me feel very … lost and very …shamed for myself (Feng)

However, students’ awareness of what they lack appeared to be a motivator for their desire to improve their English language skills:

I become frustrated about [my poor English] after the meeting …but I think it is one of my press [motivations] to improve my English. (Juno)

I really want to improve those problem and during my PhD study. … I have to make myself the language level and it's at the same level as the native speaker because I need to find a job in Australia, so that that's a very important part of my study life. (Feng)

In summary, non-PELE participants reported utilising various tools and techniques to enhance their English language skills, which resulted in some improvements. Yet despite strong willingness and motivation to keep improving, most still faced challenges with communication at the time of the focus group and felt that it had consequential effects on their research and overall well-being.

Findings from Interviews

Two non-PELE focus group participants, Xiang and Feng, who took PELE in the following term were interviewed individually. Both said that they felt that their English skills and well-being had significantly improved after PELE.

Feng had specifically chosen a reading project, explaining that where previously it had taken her a full day to digest an article, she was now “very happy” with her reading speed:

I can easily find information in articles and from several article at same time I can find what I want to read.

Xiang, who appeared visibly happier, experienced improvements in multiple domains. He said that he improved his speaking skills through his speaking project but he also noticed improvements in his reading skills thanks to the supportive PELE book club:

I don't have a confidence to finish a fiction book before. I tried but it's really hard. … I have more confidence to, to complete by myself now.

Both students were asked whether they could have achieved the same improvements without taking PELE. Feng said that she could have made similar improvements on her own but the process would have been slower:

… the process will be much slower … if I just depend myself, depend on myself, it's time consuming to find the useful tools

In response to the question, Xiang noted that the most significant impact of the course on him was a boost in confidence, stemming from a conscious change in self-perception during PELE. He began to see himself not just as a 'language learner' but as an 'English speaker’:

It gives me a huge confidence now. And I will not stop speaking. [laughs]

The responses to these individual interviews strongly resonated with the themes from the PELE focus groups, as will be discussed further in the following section.

Findings from PELE Focus Groups

In a marked contrast, participants from the PELE focus groups displayed significantly higher levels of happiness and confidence compared to their non-PELE counterparts, who displayed feelings of futility. They attributed their increased happiness and confidence to tangible improvements in their English abilities, which were not only self-recognised but also remarked upon by their supervisors. Echoing Xiang’s experience presented above, students often found that their efforts to enhance one particular skill, like reading or writing, led to progress in other areas, too. This holistic improvement in their language skills fuelled students’ confidence, enabling more effective communication with their supervisors and discussions about their research, thereby positively influencing their overall well-being.

For instance, Amaya focused on improving her writing skills in her personal project and felt satisfied with her progress as she received less negative feedback from her supervisors. Similarly, John felt motivated with his supervisor's feedback on his writing:

I felt like I'm getting less comments from the supervisors as well. … it's really good, because we can reduce the review cycles and hope we will get less comments from the reviewers as well in the future. (Amaya)

I had a meeting with [my supervisor] yesterday …and his feedback to me was, here, I actually noticed an improvement in your English writing so that was like motivational for me. And encouragement as well. (John)

Many shared that the increased confidence led to them to take initiatives in communicating with their supervisors and peers. Jing was keen to share how her emotions and attitude about her supervisory meetings had transformed after completing her speaking project. Where she once felt fear, she now takes the initiative in meetings and has a better understanding of her supervisor:

Let me tell you. Before I'm really, I'm really scary [scared when] my Professor asked me to have a meeting … I maybe need to write down the question, even the answers … But …[recently] I’m asking my professor to have a meeting with me: ‘Can you please have a meeting with me, maybe next week, can update’ … I'm like not scared to have meeting anymore … I can convey my idea, communicate more much more fluent than before … I can understand more what he is talking about. (Jing)

While Jing’s “I can” statements indicate increased communicative self-efficacy, other students articulated an increase in confidence as the biggest benefit of PELE:

… my writing before PELE was just terrible. … Because I, yeah I lacked confidence to be honest, I lacked confidence in writing but PELE like gave me the tool, the learning process, and now I’m more confident. (John)

This virtuous cycle of English language enhancement followed by increased confidence and then active engagement in communication is what most PELE focus group participants, including interviewees Feng and Xiang, experienced. When asked about which aspects of the PELE course contributed to these positive outcomes, participants pointed out three factors: the PA model, course resources and tools, and the PELE community. We will explore them one by one in the following section.

The PA Model

Participants unanimously agreed that the PA model was a significant contributor to their perceived increase in English language skills. In the PELE exit survey the majority of the students indicated that they prefer the personalised learning approach to a standard curriculum. Focus group participants appreciated its interactive nature and the individual attention; the freedom to focus on what they wanted to improve and at the level of their individual ability.

…the personalised approach was really good for me, because I was able to improve what I thought I wanted to improve on, it's not a curriculum, you know, like a structured one when I have to focus on reading first writing next and this one next and this one next. But then it was like, ‘okay, I wanted to improve on this; let's see what the diagnostics say’ and then ‘Oh, really? This is what I need to improve so I’m going to work on that’. (Ira)

Most importantly students said that they learned how to learn.

… more than English learning in the PELE course, I found out how to learn by ourselves … in a way which is comfortable for us … rather than teaching like spoon-feeding everything by the teacher … And the thing is this can be applicable for any other any other studies, not only for the English (Isuri)

Participants found that the learning skills learned through conducting their personal projects carried over into their research:

… the main thing I learned was to plan. … like my writing was so much easier than before, because I know where to go, what to look for, and I had a clear idea of what I was doing … then I used it also in my research, to write my literature review. So, that really helped me (Ira)

However, participants hesitated before fully committing to the course for various reasons. Some students shared their initial doubts when they discovered that the course would not focus on teaching English skills like grammar and vocabulary in the way a more traditional curriculum might.

…Most of them [previous English courses] have very regular agenda, every time, and we discuss or learn the same vocabulary together, every time. Some of my friends in PELE said to me, they really learn nothing about new vocabularies in [Convenor]’s lecture. … and, at the beginning, they said they don’t know if they should continue the course. (Xiang)

For some, previous English learning experience was a barrier for them to enrol in PELE:

There was also this English course as part of our introductory to the Master's program which was good, but um, there was not a lot that I learned from that experience .... And so that was the reason behind my uncertainty enrolling and but then man, having gone through the program I'm just overwhelmed. … It was beyond my expectations. (John)

Others mentioned their hesitations about enrolling or continuing in the course due to concerns about PhD workload and required time commitment. However, ultimately students found the time investment to be worth it and came to enjoy the new learning approach.

Tools and Resources and ‘Sharing’

Numerous participants highlighted the instrumental role of the diverse tools and resources introduced through PELE in enhancing their English language proficiency. These resources range from academic and non-academic references, such as Rosenberg's 'Nonviolent Communication', to digital tools like the diagnostic questionnaire, the Academic Phrase Bank, OneNote, and various AI-based applications. These were perceived as useful not just for their English learning but also for broader academic and professional advancement:

PELE like, it gives you the set of tools that you could use not only to improve English language but … that is transferable to other aspects of our academic career and not only academic, they can also be used in a professional life as well (John)

Isuri, a self-proclaimed introvert, practiced her speaking mainly using AI tools in PELE and gained confidence in communicating with humans as well. She viewed her participation in the focus group itself is evidence of her increased confidence:

Now I can share my experience with you because I am able to talk with you...So, I think this is my improvement.

However, while students found AI tools useful, when questioned whether they could have achieved the same improvements using AI on their own, it was clear that the PELE community provided crucial human support and accountability:

…it made a world of difference having the PELE community support […] without the community support I think of my results would be different. (John)

I don't think I will be motivated to do that things … I don't think we will be practice like in a daily basis, so we may, to some level, but not in a daily basis. (Amaya)

Feng credited the PELE community for her expedited discovery of beneficial resources, as recounted in her interview. The PELE teachers' expertise in resource selection, along with a regularly updated compilation of AI tools, was highly valued by the students. The act of resource sharing fostered a strong sense of community, with 'sharing' being a recurrent theme in discussions. This culture of exchange not only encouraged students to explore innovative tools but also established the sharing of practical advice and personal narratives as a treasured component of their collective experience:

you get some takeaway messages from others’ experiences as well because we talk a lot about that. (Ira)

Sharing information via PELE was especially important during the pandemic when physical resources were limited. Peer recommendations played a direct role in students trying out new tools, and even in the uptake of English language support services at the university more broadly.

PELE Community

As seen above, the PELE community of students, teachers and mentors functioned as an effective information-sharing network. However, its importance is far greater than this. Exit survey results indicated that most students felt either connected or strongly connected to the PELE community and this was mirrored in focus groups where students found the PELE community to be a source of peer support, providing a safe environment for speaking up and sharing their experiences. This even extended to the OpenLearning platform, which was seen as a better alternative to Moodle, as it facilitated a greater sense of community and allowed students to share their lives with each other.

The PELE community was also experienced as a safe space for exposure for other cultures. The value of such a secure and supportive environment was repeatedly emphasised in student responses as pivotal in enhancing their confidence. Ira’s experience, in particular, demonstrates how engagement with the PELE community significantly improved her confidence in speaking English and, as a result, positively affected her overall well-being:

… when everyone share their experiences, … it was huge knowledge for me … it definitely improved my confidence speaking with different people from different faculties, different countries, you know. That really encouraged me, and I think I made a couple of really good friends from PELE, yeah that’s that’s really wonderful. (Ira)

The above experiences highlight the benefits of the PELE community, allowing students to interact and develop friendships across discipline boundaries and degree levels. Students not only found this enriching but remarked that it was their only opportunity to do so. Cross-discipline interaction had a positive impact on both well-being and research mindset:

…by taking part in PELE, I feel that I'm still at this at the campus, I'm still at the School, I'm still learning…. So every time we joined together to say something, even, even if I cannot, I'm not speaking, … we are still together in a big screen and yeah, we can see each other that we can hear each other, so I can feel very strong, connected feeling yeah. (Sheng)

The PELE community’s role as a safe space for cultural exposure and as a crucial source of peer support and English practice was particularly underscored during the pandemic. Yet, the sense of isolation commonly experienced while pursuing a PhD was highlighted by Yang, who noted that such solitude can significantly impact well-being, even without a pandemic:

…even before pandemic, I also always feel very stressed, because I didn't know so many people doing research, except those in my office or some of them in my school. But I couldn't find any chance to meet the people, to be social with other students in different subjects.

He specifically emphasised the significance of this for international students, who often face additional challenges in finding opportunities to practice and enhance their English proficiency:

I do feel that when we go to HDR [Higher Degree Research] this stage, we have fewer people to talk with, especially for international students. So, sometimes we are just stay at the office. The only chance for us to talk talking English maybe just for supervisors, or just for colleagues, maybe just maybe no more than 10 people… I think even sometimes, we, we very require or just very urge to want to talk English, but we don't have this chance.

Students highlighted the freedom to share experiences and communicate openly as a significant benefit. The supportive environment of PELE was instrumental in enabling students to forge diverse friendships, contributing positively to their well-being. This was especially crucial as discipline-specific English Language Proficiency (ELP) models may not facilitate cross-faculty or cross-level interactions. Such interdisciplinary engagements offered students broader perspectives, deepened their research reflections, and reinforced their sense of university belonging, even in the context of fully online coursework.

Discussion

Our study examined the effects of the PELE approach on international PhD students’ ELP self-efficacy, well-being, and the interrelationship between these factors and their research. Survey data revealed significant improvements for the PELE group across all variables, including ELP self-efficacy, well-being, SDL self-efficacy, social self-efficacy, and a sense of belonging to the university community, unlike the non-PELE group which did not exhibit significant improvement in any variable. Between-subjects t-tests comparing improvements between PELE and non-PELE students (Table 4 above) indicated that PELE students improved significantly more in ELP self-efficacy, well-being, SDL self-efficacy, and sense of belonging to the university community, with particularly notable differences in ELP self-efficacy and sense of belonging to the university community, with moderate to large effect sizes.

However, given the low response rate among non-PELE students, generalizing the findings from this group is not feasible. There might be some ceiling effect in the non-PELE group as participants who completed the entry surveys exhibited higher self-efficacy across all measured variables. The observed differences in improvement between the groups could potentially have been more pronounced if the non-PELE group participants had self-efficacy levels similar to, or lower than, those in the PELE group. The degree to which these participants accurately represent the broader international EAL PhD student population remains uncertain. A further study will be required to measure the effects of PELE in comparison to non-PELE students with different levels of self-efficacy.

Focus group and interview data revealed three key factors that contributed to significant enhancements in students’ confidence in English: the personalised autonomous pedagogy, PELE tools and resources, and the PELE community. Students' bolstered confidence led to increased agency in taking initiatives, such as arranging supervisory meetings, receiving feedback positively, and making inquiries when necessary. These attitudinal changes had a profound impact on their research performance, relationships with supervisors, and overall well-being. In the following section, we will highlight the relationship between confidence and wellbeing further and explore the three key factors of PELE with potential theoretical explanations.

The virtuous cycle experienced by PELE students appears to start from increased self-confidence, which is individuals’ feeling about their capability to deal effectively with various tasks. In our study, students’ confidence was measured using self-efficacy questions about their abilities to perform specific tasks in the surveys. However, focus group participants used confidence to refer to both self-confidence and self-efficacy. Therefore the two terms are used as interchangeable concepts in our study. In spite of the technical difference, the notable fact in the concepts is they are people’s feelings and emotions. The importance of emotions in educational settings has been recognised in educational psychology as a fundamental source of motivational and physiological energy (Pekrun et al., 2002) and as one of the most influential motivators that influence people’s behaviours (Bandura, 1986). Self-efficacy proves a reliable variable in predicting students’ performance (Bandura, 1997). It is also confirmed that self-efficacy is “a strong predicator of performance in different language skills and tasks” in a study that reviewed 32 empirical studies in the second language learning field (Raoofi et al., 2012:60). The socio-contextual model of second language (L2) learning (Clément, 1980) postulates that:

… L2 confidence is the key construct underlying L2 behaviour as well as its social psychological correlates. L2 confidence corresponds to a relative lack of anxiety when using an L2 coupled with the belief in being able to cope linguistically with the L2 situation at hand. It is specifically derived from relatively frequent and pleasant contacts with the members of the L2 community whether through face-to-face interactions or L2 media. It has been shown to sustain motivation to learn an L2, to promote identification to the L2 group, to be related to better production of the L2, and finally a better adaptation among those living in a bilingual context … (Rubenfeld et al., 2006:610, our emphasis).

Technically, in the PELE context, English is an additional foreign language rather than L2 but the L2 model’s propositions are manifested as students learn English in authentic English contexts.

Therefore it is not surprising that our data showed that PELE students received positive feedback from their supervisors about their enhanced English language proficiency, as shared in the PELE focus groups as well as in the supervisor focus group as presented above. Increased confidence in English language proficiency led to tangible outcomes like a successful confirmation review, paper acceptance and an award. Kissa was presented a best paper award at a conference, which makes her even more confident:

…working on writing it was first was like pushing and now it is making me feel comfortable in this type of pushing… so I started writing more and, recently, I guess, I got this this paper awards in the conference, so I became more confident (Kissa)

The significant role of confidence was also confirmed in the well-being data. PELE students’ well-being improvement was greater than their counterparts’ in 13 items out of 14, with significant difference in the confidence question. This may be interpreted that increased confidence might have had ripple effects. For instance, the supervisory relationship is a significant factor that influences PhD students’ well-being as they are the main source of support (Cotterall, 2013). As discussed above, PELE students’ emotions and attitudes towards their supervisors and supervisory meetings changed. As they feel more confident about themselves, they become less defensive when they receive critical feedback from their supervisors but more inquisitory, asking for clarification or elaboration, or explaining why they disagree.

I think I will ask more questions than before. Like before I just know, oh what you're talking about, okay that's your story, but now I, I want to ask why we think in that way, how, why and how and more yeah more in detail because maybe I improved my confidence [laughs] (Jing)

These open dialogues lead to enhanced interpersonal relationships and strengthen students' agency to communicate with their supervisors, for instance initiating supervisory meetings and reducing negative emotions (e.g. not feeling “scared” anymore). This can be a great source of happiness and is significant for PhD students because supervision meetings are a primary context where they communicate in English and can be a major source of anxiety and frustration if not handled well.

Now let us look into the three key elements that contribute to enhancing students’ confidence. A powerful theoretical explanation can be found in self-determination theory (SDT). SDT suggests that all humans are inherently driven by three basic psychological needs: autonomy, competence and relatedness (Ryan & Deci, 2017). In PELE, students are given autonomy to choose what they want to improve and how to improve it; their need for competence is satisfied by their improving English language skills; and the sense of community fulfills their need for relatedness, which is particularly significant for international students who pursue their aspirations far away from their family and friends.

In addition, many students recognise the significance of improving English language skills and attempt to do so. However, they lack effective methods and resources. PELE's introduction of innovative language learning strategies, drawn from literature on language learning strategies, second and foreign language teaching, and applicable linguistics, has helped to galvanise students' efforts to improve their English skills (for more details see Kim et al., under review). With the rapid development of AI tools, it is essential to provide explicit guidance about how to ethically use them as effective learning tools.

The exit survey data showed a significant positive correlation between the sense of PELE community and university community suggesting that PELE improved students’ overall connection to the university community. Interestingly, Pearson's r correlation tests on the difference scores showed that for only PELE students, the sense of university community was significantly and positively correlated with well-being. Being a part of the PELE community enabled students to build the confidence to start communicating in English. The theme of 'sharing' emerged as crucial. This idea of sharing became a lifeline for research students, who may not have a ready-made community.

Last but not least, PELE guides students to learn self-directed learning skills and become autonomous learners. This is intentional because the process of improving communication skills for advanced English learners like international PhD students demands life-long learning. In fact, this applies to many professional English speakers regardless of whether or not English is their mother tongue. However, it is a sad reality that many educational systems dating back to the industrial revolution have prioritised training compliant citizens rather than promoting deep learning and individual creativity (Jiménez Raya & Vieira, 2015). As a result, self-directed learning is often neglected in school education, and students are not taught the necessary skills to become autonomous learners. The transformational changes impact on their lives beyond PELE and greatly improve their well-being. Survey and focus group data clearly showed that many students appreciate PELE for the life skills learned in addition to language and academic skills.

Conclusion

It is a common and expected phenomenon that international PhD students arrive with robust English language skills, a foundational requirement for their admission into doctoral programs. Despite a high level of their initial proficiency, they often require additional time to refine these skills within their new research environment, where English is not merely a medium of communication but also a vital instrument for academic success. A study surveying 908 environmental science researchers uncovered a notable disparity: non-native English speakers need 46.6% to 90.8% more time to comprehend academic papers in English compared to their native-speaking counterparts (Amano et al., 2023: 3). It is, therefore, crucial to offer these students comprehensive support, including access to strategic resources and development of self-directed learning skills for autonomy and a nurturing and inclusive community. This support is comparable to providing seeds with sunshine, water, and fertile soil, necessities for their growth and thriving. With the right support, these international scholars can reclaim the confidence they once held as high achievers in their respective professional or academic fields before embarking on their PhD journey. Restoring this confidence is crucial, as its loss can have a deep and widespread psychological impact on this highly skilled and aspirational group.

We hope that the PELE approach can serve as a new innovative model for higher education institutions striving to support culturally and linguistically diverse PhD students, both from international and local backgrounds, enabling them not only to survive in their academic endeavours but also flourish and cultivate a positive researcher identity. It is imperative to attentively listen to our students' voiced needs and actively pursue holistic, and sustainable methods of support, moving away from outdated practices that may no longer be effective.