From a Doctoral
Student to a Doctoral Supervisor: A Longitudinal Autoethnographic Study
Junjie Gavin Wu[1]*
|
*Correspondence: gavinjunjiewu@gmail.com |
|
1
Macao Polytechnic University, Macao, China |
Abstract
The
final year of a Ph.D. program is a crucial transition from doctoral student to
independent academic and social actor. Within years, new graduates often become
supervisors themselves, navigating an academic ecosystem increasingly
transformed by generative artificial intelligence (GenAI). Despite its
importance, research tracking individuals across both stages remains scarce. Grounded in ecological systems theory, this
autoethnographic study examines critical incidents and emotional encounters
across two phases of my academic journey: completing my doctorate in Hong Kong
(2019–2020) and serving as a research supervisor at a Macau university (2023–2025).
Drawing on reflective journals, I document my thesis completion, the shift to
online teaching, job searching, academic recognition, and the emergent
challenges of guiding students who utilize GenAI tools. Findings reveal that
both transitions were profoundly social, psychological, and emotional. Feelings
of empowerment and disempowerment fluctuated based on interactions across
ecological systems. Notably, exosystemic and macrosystemic constraints
encountered as a student did not vanish upon becoming a supervisor; instead,
they resurfaced as structural pressures I had to either reproduce or interrupt.
The paper concludes with recommendations to better resource final-year students
and novice supervisors through these critical transitions in the GenAI era.
Keywords:
transitional
development; ecological systems theory; autoethnography; doctoral supervision;
generative artificial intelligence; doctoral education
1. Introduction
In the past decades, universities across the globe
have witnessed the academisation and marketisation of higher education.
Doctoral education in particular has captured educators’ attention, as societal changes have challenged
traditional provision with new types of programme on offer (e.g., Doctor of
Education, Doctor of Business Administration). Additionally, in the
post-pandemic era, with reduced mobility, an economic downturn, and the
unavoidable involution (an anthropological concept that population growth leads
to contests for limited resources without an increase in productivity),
competition among university graduates has been exacerbated. This
intensification has been compounded by the arrival of generative artificial
intelligence (GenAI). As drafting and literature-synthesis costs fall, research
output rises across the sector, and candidates find that the markers by which
they once demonstrated distinction, a manuscript in press, a polished job talk,
a rapid turnaround on revisions, are more easily matched by their competitors (Jensen et
al., 2025; Lai et al., 2025). Survival in doctoral
education has therefore become less a matter of productivity than of navigating
an ecology in which productivity itself is inflated, contested, and unevenly
recognised across academic cultures (Bao, 2025). It is to a certain extent
a fait accompli that a doctoral degree is now essential for entering the higher
education job market, especially given the importance of university rankings
and their emphasis on quantifying staff with Ph.D.-level education (Ek et
al., 2013).
This landscape has driven
students, supervisors, and managers to pay unprecedented attention to the
process of doctoral education (Park, 2005). The process includes not
only a well-crafted thesis from which graduates can publish, but also the
training of students as future competent researchers, qualified teachers, and
mature individuals ready to contribute to society. Carter et al. (2021) describe the doctoral
journey as “doctoral borderlands…marked by uncertainty, hope,
risk, and identity transition” (p. 283). The doctoral
voyage is thus much more than writing a thesis; it can be rewarding, inspiring,
complex, and arduous, involving socialisation from the academic to the social
world where students juggle cognitive, affective, and social challenges (McGloin,
2021).
Previous studies have
examined various facets of doctoral education, including the progression of
study (McGloin, 2021), the enhancement of
academic teaching skills (Rao et al., 2021), the provision and uptake
of thesis feedback (Han & Xu, 2021), and the emotional
engagement that sustains or erodes candidature (Alves et al.,
2024). The identity construction of doctoral students has
been widely investigated: Teng (2020) reported on a Mainland
Chinese student developing his researcher identity, finding that he struggled
because of prior learning experience and external restraints, yet that agency
was the most essential variable in identity formation. Working with a larger
group, Cong-Lem et al. (2025) traced the identity
transitions of Asian international TESOL doctoral students through
collaborative autoethnography and likewise found agency to be exercised
unevenly, always in negotiation with institutional and cultural structures
rather than in isolation from them. Although these studies illuminate different
stages of doctoral education, little is known about the final-year transition.
The final year matters
particularly because it is when students begin to move from academic learning
to considerations of their future careers. It involves not only the pressure to
submit and defend a thesis; a student also takes on new identities as a job
applicant, a novice scholar, a breadwinner, and a novice teacher. The
socialisation of final-year students therefore requires a holistic, ecological
understanding that goes beyond the knowledge of completing theses (Horta,
2018).
Less well understood still is
what happens next. Within a few years of graduating, many early career
academics are entrusted with supervising research students of their own,
frequently without formal preparation and often in institutional systems quite
different from the one in which they were trained. Lundberg (2025), in one of few
autoethnographic accounts of this shift, documents how developing a supervisory
identity requires adaptations that doctoral training does not supply. In the
Chinese context, where recruitment, promotion, and publication recognition
operate through distinctive structures (Liu & Li,
2024; Wang et al., 2025), and where GenAI has
rapidly become part of how students read, write, and seek feedback (Lai et
al., 2025), the novice supervisor occupies a doubly unfamiliar
position: still adjusting to the local academic culture as a newcomer, while
already expected to induct others into it. Examining the student-to-supervisor
trajectory as one continuous process, rather than as two unrelated literatures,
allows us to ask whether the ecological constraints experienced as a supervisee
are reproduced, resisted, or simply inherited once one holds institutional
authority.
The present paper therefore
follows my learning, working, and supervising trajectory across two phases: the
final year of doctoral study in Hong Kong (2019–2020) and the assumption of a supervisory role at
a Macao university (2023 to 2025). Guided by ecological systems theory, it asks
how I behaved and perceived myself at different stages, and how the interacting
ecological systems empowered or disempowered me at each.
2. Ecological Systems Theory
Ecological systems theory (Bronfenbrenner, 1979) is a widely recognised
framework for accounting for human growth into a competent member of society
from a developmental psychology perspective (Bronfenbrenner,
1992). The ecological environment, historical, cultural,
relational, and fiscal, is central to understanding an individual’s contextual development, because relationships
between person and environment are reciprocal and bi-directional (Bronfenbrenner & Morris, 2006). An individual is an
agentive person who exerts impacts on the environment, while that environment,
comprising the microsystem, mesosystem, exosystem,
macrosystem, and chronosystem, influences the person to “adapt to its conditions and restriction” (Härkönen, 2007, p. 7).
In doctoral education these five systems capture a
learner’s interactions at different
distances from the self. The microsystem describes the immediate environment,
university, family, friends, whose interactions are held to be most powerful
for a student’s growth; a supervisor can
directly modify a student’s beliefs by encouraging or
reproaching. The mesosystem concerns interactions between microsystems, such as
family emergencies colliding with course assignments; Härkönen (2007) argued that the extent to
which microsystems support or hinder each other shapes belief and behaviour.
The exosystem addresses linkages between immediate
and non-immediate environments, as when a supervisor becomes less attentive
because of a lockdown affecting the supervisor’s own household. The macrosystem refers to wider
social and cultural environment, cultural values, national laws, economic
conditions, political development, and changes here tend to influence all other
systems (Cross & Hong, 2012). The chronosystem directs
attention to milestones and significant historical events, which contribute to
a person’s development.
Two features of the framework suit the extended,
two-phase account offered here. First, the chronosystem is not merely a
container for events but an analytic resource for tracing how the same person
is repositioned over time: the pandemic, graduation, relocation, and the
arrival of GenAI are all chronosystemic ruptures whose effects can only be read
across, not within, a single moment. Second, because the theory treats person
and environment as mutually constitutive, it accommodates the change in structural
position that occurs when a former supervisee becomes a supervisor. What was,
in Phase 1, an exosystemic force acting upon me, like departmental workload
allocation, institutional publication metrics, national recognition of indexing
systems, becomes, in Phase 2, part of the microsystem I myself constitute for
my students. The framework thus permits examination of how ecological
conditions are reproduced across academic generations, an issue that studies
confined to a single career stage cannot address (Bronfenbrenner
& Ceci, 2005).
3. The Present Study
The study uses autoethnography to scrutinise my
educational, professional, and personal challenges, and the emotional
rollercoaster of the final doctoral year and the subsequent years of
supervisory practice. Autoethnography is both a product and a process (Ellis et
al., 2011), allowing a researcher to become a participant
observer and to develop a systematic, emic understanding of cultural experience
from a personal, accessible, and evocative perspective. In contrast to the
positivist paradigm, it “acknowledges and accommodates
subjectivity, emotionality, and the researcher’s influence on research” (p. 274). Epiphanies or critical moments,
retained by memory over time, help both researcher and readers relive and
analyse that experience; the first-person pronoun is therefore retained
throughout. The approach has recently been used at both ends of the trajectory
examined here by Cong-Lem et al. (2025) to trace doctoral identity
transitions, and by Lundberg (2025) to examine the move into
supervision.
A common critique of autoethnography concerns the
accuracy of data and the extent to which the researcher offers an unbiased
account. Yet the aim is not to uncover absolute truth but to promote social
justice and improve the living world (Ellis et al.,
2011). Special attention was therefore given to genuine
thoughts and experiences, and events that mattered to me were recorded at
personal, university, and societal levels as they occurred. Though the findings
may not be generalisable, the longitudinal design affords an in-depth,
contextual perspective on the move from final-year student to research
supervisor.
3.1
Research Context and Participant
Higher education in Hong Kong is acknowledged for
its excellence in research and professional education, and its use of English
as the medium of instruction has attracted many international doctoral
students; according to Shin et al. (2018), more than 70% of doctoral
students registered in Hong Kong universities were not natives of the city. The
Phase 1 research site is a top public university in the Asia-Pacific region,
where the quality of research and teaching is underscored among teachers and
doctoral students. From the outbreak of COVID-19 in February 2020 the
university adopted Zoom-based online teaching, which affected both my learning
and my teaching, as I performed research and teaching duties in the department.
Some personal background will assist readers. I
hold a degree in second language education and had taught English at a Mainland
Chinese university before moving to Hong Kong, where I coordinated several core
courses as a novice teacher. I enjoyed university teaching and also aspired to
do research, but was aware that a Ph.D. would improve my prospects; after two
years I resigned from a contract-based post to begin doctoral study. The Hong
Kong university required me not only to complete research but also to teach
undergraduate courses independently and to support departmental teaching
matters such as exam invigilation and course coordination. Over the first two
years I published in well-regarded journals, edited a volume with a mainstream
publisher, and received several international awards. This matters because it
explains the sense of failure I experienced in my final year. By the end of my
study I decided to return to Mainland China to begin my research career.
Phase 2 is set in Macau. Since 2023 I have held a
research and teaching post at a teaching-oriented university in Macau, where my
duties include supervising postgraduate research students in educational
technology; at the time of writing I am supervising over ten doctoral students.
The institutional environment differs from my doctoral training site in ways
that matter analytically: publication is assessed against its own indexing
systems, contract renewal is tied to quantified outputs, and supervisory capacity
is allocated administratively rather than negotiated between individuals. Two
further conditions distinguish this phase. My students entered postgraduate
study after GenAI tools became widely available, and institutional guidance on
their use in thesis work was not clear. I had received no formal training on
the use of GenAI in research supervision before taking on supervisees.
3.2
Data Collection and Analysis
Phase 1: The
final doctoral year (2019–2020)
Two methods tracked my thoughts, feelings, and
life stories over one year. First, a reflective journal was kept from September
2019 to July 2020, allowing me to record emic, insider thoughts and feelings
that external researchers find hard to obtain over a prolonged period (Mackey
& Gass, 2016). The journal contained 11,946 Chinese
characters, analysed directly for their originality and accuracy following
Braun and Clarke’s (2006) procedure: entries were
read repeatedly to establish familiarity, initial codes were assigned and
candidate themes sought, themes were reviewed to generate a thematic map, and
themes were finalised with representative excerpts selected for clarity and conciseness.
To improve the trustworthiness of the coding, the second author coded the data
set, achieving inter-rater reliability above 93%; disagreements were resolved
through discussion. Examples of coding are presented in Appendix 1.
Second, on the basis of the initial journal
findings, a senior researcher, my doctoral supervisor, conducted a one-hour
interview with me. This served to elaborate crucial moments (e.g., feelings
when teaching online during quarantine) and to clarify unclear or missing
points (e.g., rationales for applying to Mainland Chinese universities). The
discussion helped me develop an etic, outsider
perspective on my own experience, so the paper goes beyond a conventional
analytic approach by inviting a senior researcher to help interpret
autoethnographic findings. The interview was conducted in English and analysed
with the same steps (see Appendix 2) to triangulate the journal. Course
materials and the department website were also collected as supplementary
sources.
Phase 2: Becoming a research supervisor (2023 to 2025)
Phase 2 draws on a second reflective journal kept
from September 2023 to May 2025 following the same protocol, in which I
recorded critical incidents in my supervisory practice and the emotions
accompanying them at the points at which they occurred. It contains 8,420
Chinese characters across 47 dated entries. Two supplementary sources were
added: 63 sets of my written feedback on student drafts, which document the
substance of supervisory interaction rather than my recollection of it, and 14
sets of notes from institutional meetings concerning research training,
supervision allocation, and the use of GenAI in postgraduate work. Analysis
followed Braun and Clarke (2006) as in Phase 1. Because
Phase 2 remains ongoing, it is presented as an open rather than a completed
stage, and the analysis attends particularly to moments in which my earlier
experience as a supervisee became visible in my conduct as a supervisor.
Because Phase 2 concerns my own supervisory
conduct, no student is named, quoted, or otherwise identifiable; where a
supervisory episode is discussed, identifying details have been altered or
omitted and the analytic focus remains on my own responses rather than on any
student’s performance.
4. Findings
This section presents five stages. The first four
cover the final doctoral year in chronological order (i.e., the chronosystem),
assisting readers in living through the critical moments; the fifth turns to my
current position as a research supervisor. The Discussion then applies
ecological systems theory across all five.
4.1
Stage 1 (September 2019 to January 2020): “I feel confident that I can land a
good job”
My final-year journey commenced in September 2019
with a scholarly visit to a world-renowned university and an international
award from an esteemed organisation, which together offered a strong sense of
belonging within the academic community. In my journal, I wrote:
Though I only stayed in
Europe for a few weeks, it was a great opportunity to visit a leading
university and established some connections. I felt more closely associated
with the international community. I really enjoyed my visit and appreciated
that the international institute acknowledged my intellectual work. (Journal entry, 26 September
2019)
Successful interactions with
international colleagues and recognition from my research area were sources of
positive feeling and facilitated an active scholar identity. Yet this was not
always the case, as we will see when I attempted to establish my scholarship in
Chinese academia. In November, a lower-ranked Mainland Chinese university
invited me for interview and offered me a lecturer position. Though not my
priority choice, I was strongly positive about my prospects:
The Head was quite friendly
and sincere. The department covered all my transportation and accommodation
fees and offered me a position after my presentation. I was told that I was the
only attendee among others with a few SSCI publications. I feel confident that
I can land a good job as long as I finish my study. Now I only need to complete
my thesis on time. (Journal entry, 6 November 2019)
The sincerity shown by the
Head and the allowances covered by the university made the experience
authentic, interactive, and positive. As a result, I felt appreciated and
acknowledged for my achievements, which raised my expectations and
self-efficacy and left me more committed to writing the thesis.
I negotiated with my
supervisor a quite detailed timeline of my final Ph.D. stage, including when I
should finish the first, the revised, and the final drafts. With the timeline,
I am clear that my supervisor is supportive and I will be able to graduate without
extensions. (Journal entry, 17 December 2019)
A well-defined timeline
instilled confidence that I was on the right track. Alongside the thesis I was
working on journal articles and book chapters, and despite the workload I
enjoyed this stage of Ph.D. life:
Also, I am writing papers
with my supervisor. I enjoy working with him as I could constantly learn from
him about academic writing. It’s worthwhile to have a busy
academic life. (Journal entry, 29 December 2019)
I was also invited to give a
presentation at an international conference outside Hong Kong in early 2020,
and felt more positive about my future career as a result:
I delivered a specially
invited talk at an international conference and received some positive
feedback. I shared the news with my friends and teachers and they felt quite
proud of my achievements. (Journal entry, 23 January 2020)
In summary, during this first
stage I felt optimistic about my prospects, had received international and
domestic recognition, and enjoyed the multitasking life of writing both a
thesis and journal papers.
4.2
Stage 2 (February to March 2020): “I was facing multiple challenges”
As part of my duties I was required by my home
department to teach a General Education course to undergraduates. The pandemic
then interrupted my doctoral journey: from February 2020 Hong Kong universities
responded to the outbreak with an abrupt shift to online teaching while I was
quarantined. The sudden change brought challenges:
I am caught off guard when it
becomes mandatory to teach via Zoom. I’ve never used it before and now
I have to adapt my teaching to a new form of instruction. Though my research is
about education, I am not sure how I could make quick adjustments to online
teaching and what ways will be useful. Meanwhile, the syllabus is designed
based on face-to-face teaching. I wish the course team would adjust the
assessment. (Journal entry, 28 February 2020)
Novice teachers such as
myself faced significant pedagogical challenges during the pandemic (Miller
& Wu, 2022). Without prior experience of Zoom I struggled to
identify effective approaches, and the absence of pedagogical guidance left me
unsupported, especially as the syllabus had been designed for traditional
instruction. The extra time required burdened me as I neared submission:
I am also overwhelmed with my
thesis writing and editing. As the deadline for my thesis submission is
approaching, I am under huge pressure. Though my supervisor chats with me via
FaceTime constantly, it is still stressful, particularly with the pandemic. I
wish our department could have exempted final-year PhDs from their teaching
duties and let us focus on writing! (Journal entry, 6 March 2020)
Although academic and
emotional care from my advisor was provided, I felt unsupported by the
department, still struggling with teaching and research duties at the final
stage of thesis writing. The situation worsened when I had to quarantine in a
compact room:
I had to stay at home 24/7 as
everyone was scared of the pandemic. I could not get new facial masks anywhere
so I had to keep using one mask… I was facing multiple challenges, including the
pressure of finding appropriate ways of teaching online, the pressure of
finishing up my thesis, and most importantly the psychological stress of being
locked down. Things had gone out of hand and I started to have trouble
sleeping. (Interview excerpt)
During this second stage I
experienced a difficult time for want of sufficient professional, collegial,
and institutional support. Holding the dual role of instructor and research
student, I underwent substantial pressure and had to cope with change pedagogically,
physically, and psychologically.
4.3
Stage
3 (April to May 2020): “I was stuck in the middle”
Having submitted my thesis for examination in
April, I resumed the search for a job and began applying for teaching posts in
Mainland China. Though earlier job-hunting had been positive, at this stage I
encountered a serious confidence and identity crisis:
I have applied for a few
positions in Mainland China but I haven’t got any interviews yet. I
am starting to worry whether I am not a strong candidate. I do believe I have
an ‘okay’ package as I have full-time
tertiary teaching experience and some international publications. I have been
waiting for quite some time since the submission of my job applications and I
am not sure when I will be interviewed. (Journal entry, 12 April
2020)
COVID was in full swing, and
I was concerned about its effect on the job market and my own competitiveness:
My colleagues, working at
Chinese universities, told me that many Westerners had applied for teaching
jobs in their Chinese universities partly because the pandemic reduced the
number of job vacancies in the West. They are native speakers of English. I don’t believe I can compete with
them, though they may not have doctoral degrees. (Journal entry, 25 May 2020)
My identity as a non-native
English speaker undermined my confidence, since I believed a doctoral degree
might be less attractive than the identity of a native speaker. My
self-confidence thus changed dramatically: initially positive about securing a
position, I was gradually worn down by the reality of the pandemic job market.
The loss of self-esteem intensified when I discovered that:
I was stuck in the middle. I
was told that Scopus and ESCI are not recognized as proper indexing systems
among most Chinese universities. In that case, most of my publications were
void and they could not be counted towards my academic achievements. I don’t think that’s fair as they were
peer-reviewed and took me a lot of time to get them revised and published. (Interview excerpt)
The cultural disparities
between the international and Chinese academic communities shocked me; it had
never occurred to me that my doctoral efforts would be undervalued within
Chinese academic culture, and I was unprepared for the shock. Despite
recognition from the international community and a developing researcher
identity, my identity was largely threatened within the domestic community
because my work did not conform to the local research culture, causing a marked
decline in self-efficacy. This misrecognition accords with evidence that
overseas-trained candidates cannot straightforwardly convert international
credentials and networks into domestic academic value (Wang et al.,
2025). Financial and living difficulties compounded matters,
as I had decided to move back to Mainland China but could not retrieve my
savings from Hong Kong under quarantine policies:
I underwent a phase of
depression. Nothing worked out well at that stage. I was unhappy most of the
time. Luckily my Ph.D. peers and supervisor were always there for me. We shared
our concerns and encouraged each other. That absolutely helped me overcome the
negative moments. But I still wished that our university could have offered
more support to final-year students. (Interview excerpt)
Despite a hard time in career
and life, I remained grateful for the support of peers and teachers. No support
at departmental or university level was reported at this stage either.
4.4
Stage 4 (June to July 2020):
“I am willing to take a chance”
Near the end of July I officially finished my
study and regained self-confidence through two critical incidents: passing my
online viva and receiving job offers. The pandemic left me no choice but to
accept an online viva conducted over Zoom, which proved successful:
It was a relaxing viva and we
exchanged opinions and thoughts about my thesis. I enjoyed the process and I am
honoured that my work is recognised by international scholars. (Journal entry, 10 June 2020)
In multiple entries I
referenced the term renke (认可, recognition or
acknowledgement). Being recognised by the academic community evidently had an
irreplaceable impact on my identity construction as a successful Ph.D. student,
and similar comments appeared when I was offered positions at Chinese universities:
At last! I received a few
offers and honestly, they should have happened months before. (Journal entry, 18 July 2020)
With offers at hand I rebuilt
my confidence and displayed a confident identity, believing the offers were
overdue. Yet a thorny issue followed, i.e., making a final decision:
I am not sure which
university I should choose. Teachers in Mainland China are underpaid with
demanding research and teaching requirements. I come from a working-class
family so I have to take the financial factor into serious consideration.
Meanwhile, due to the low salary, I am not sure whether I can focus on doing
research and teaching. I have to establish a personal life, right? (Journal entry, 23 July 2020)
I was hesitant because
leading universities in China may not offer competitive salaries, yet I was
equally unsure about trading career development for pay. This is a
disappointing but realistic situation for fresh graduates, as university
teachers earn much less (on average around GBP 1,100 per month for a lecturer)
than other professions. The tension is not idiosyncratic: in Liu and Li’s (2024) study of Ph.D. graduates
entering universities in the Greater Bay Area, financial concern and risk
mitigation operated as two of five competing strategic logics shaping career
decisions. Other obstacles for novice university teachers in China also weighed
on me:
Many Chinese universities
have adopted a mixed-mode of university systems from everywhere (e.g., UK, USA,
Germany…). Novice teachers are under enormous pressure
just by looking at the contractual and promotion requirements. I have to figure
out which job offer is more reasonable. (Journal entry, 25 July 2020)
Family also stood out in my
decision-making. My parents would have liked me to work near my hometown,
whereas I intended to stay in a city far from it. As the only child, I
commented:
In China, we have a saying ‘While one’s parents are alive, one
should not travel to distant places.’ I know as the only child in
my family, I need to take care of my parents at some point. Yet, I would like
to stay in Guangdong for a while. (Interview excerpt)
This last stage was
bittersweet: I regained a sense of pride but was caught in dilemmas when making
career decisions. After weighing the various factors, I decided to take a
chance at a teaching-oriented university far from my hometown.
4.5
Stage 5 (2023 to 2025): “Am I the one they are waiting on?”
From 2020 to 2023, I worked at a Mainland Chinese
university, where teaching was prioritized over research, although the position
was well paid. Seeking a more international academic environment, I decided to
move to Macau.
In Stage 5 I was allocated my
first research supervisees. The transition was administrative rather than
ceremonial: no training preceded it and no conversation marked it. What changed
was the direction in which expectation flowed. Where I had waited on a
supervisor’s comments, a department’s decisions, and a job market’s verdict, I was now the person others waited
upon. Four themes trace how that reversal was experienced.
4.5.1
Inheriting a role without a script
My first sustained difficulty was relational
rather than intellectual. I knew the content of educational technology and had
learned to write for international journals, but I had no repertoire for
supervision itself: how much direction to give, when to withhold a solution so
a student might find it, how to say a draft was not yet adequate without
extinguishing the confidence needed to revise it. The only available model was
the one I had lived through:
What language style should I
adopt when providing revision comments to my students? My own supervisor was
democratic and open-minded, offering general guidance while expecting me to
take intellectual initiative rather than wait for direction. In my current
context, however, all my students are self-financed and some hold jobs
alongside their studies. This leaves me uncertain: would they prefer a
traditional top-down Chinese supervisory style, or a more collaborative Western
co-negotiated approach? … I think I will start by applying my usual
approach with students and then see how they respond. (Journal entry, 14 October
2023)
The entry shows a supervisory
repertoire being improvised in real time, with my own supervisor’s practice as the sole reference point. This
echoes Lundberg’s (2025) account, in which
supervisory identity had to be constructed largely from personal experience of
being supervised rather than from formal preparation. The absence of
institutional induction meant that the microsystem I built for my students was,
in effect, a copy of the one I had inhabited - a mode of transmission that
carries forward good practice and unexamined assumptions with equal fidelity.
4.5.2 The
metrics return, from the other side
The indexing systems that had invalidated much of
my publication record in Stage 3 did not become less consequential once I was
employed; they became the terms in which I was expected to advise. My students’ progression, my own contract renewal, and the
resources available to my supervision were all calibrated against recognised
output:
Today, I convinced my student
to drop the classroom ethnography in favour of a questionnaire. I cited
methodological reasons and they were sound. But honestly? Ethnography would
take two years, and our faculty tallies recognised publications every semester.
Students need Scopus-indexed papers to graduate. Three years ago, my Scopus
papers were deemed irrelevant. Now I fear my students will suffer the same fate
when they go on the job market. (Journal entry, 9 May 2024)
What had been, in Stage 3, an
exosystemic constraint imposed upon me from above became, in Stage 5, a rule I
was now compelled to actively pass on to my own students. This is where the
ecological framework does analytic work a single-stage study could not: the
constraint was not removed by my change of position, only relocated from the
environment acting upon me to the environment I constitute for others (Bronfenbrenner & Morris, 2006). Wang et al. (2025) show how deeply the
recognition of credentials and networks is embedded in Chinese academic
recruitment; my findings suggest that this embedding is sustained partly
through the ordinary, well-intentioned advice of supervisors who were
themselves disadvantaged by it.
4.5.3
Supervising alongside generative AI
The most substantial discontinuity between my own
candidature and my students’ is that theirs is conducted
with GenAI as a constant presence. Drafts arrive fluent; literature reviews
arrive comprehensive in coverage and thin in judgement. The diagnostic signals
I had been trained to read, such as awkward phrasing indicating unresolved
thinking, a missing reference indicating an unread field, no longer indicate
what they once did:
The literature reviews were
better written than anything I produced in my doctoral years. Every sentence
was smooth and every citation existed. But when I asked him why he referred to
this theory, he could not answer. He had not made that decision; AI had made it
for him. I asked him to talk me through the arguments… With AI handling the routine
supervisory work, perhaps I can now focus on higher-level guidance: nurturing
independent thinkers, compassionate educators, and real humans, not paper
machines. (Journal entry, 22 July 2024)
These experiences correspond
to the tensions documented by Lai et al. (2025), who found that GenAI
adoption generated conflict not only between affordance and limitation but
between productivity-oriented practice and established academic norms,
surfacing interpersonally as divergent expectations among students,
supervisors, and peers. They also illuminate what is at stake in Jensen et al.’s (2025) finding that chatbot
feedback is task-focused and agency-enhancing whereas supervisory feedback is
relational, contextual, and developmental. From the supervisor’s side that difference is experienced as a
redistribution of function: the tasks by which I formerly demonstrated my
usefulness, including line-level correction, pointing to overlooked sources,
are now performed adequately without me, which obliges me to make the
relational and developmental dimensions of supervision explicit rather than
incidental. In ecological terms, GenAI entered as a macrosystemic and
chronosystemic change and rapidly reorganised the supervisory microsystem
itself.
4.5.4
Choosing what not to reproduce
Against these reproductions stands what I have
declined to pass on. The most acute deprivation of Stages 2 and 3 was not any
single hardship but the silence of the department: teaching duties that were
not lifted, guidance that was not offered, wellbeing that was not asked after.
Having named that absence in my own account, I found it difficult to replicate:
Students spoke about both
academic and personal hurdles, including no library access to journals, no
funding for conferences, and unhealthy competition among supervisors’ groups. I’ve tried to treat them as whole
people who get sick, feel down, or just need a break. Yet the underlying
tension remains: if they do not keep writing and submitting, their annual
report to the faculty will reflect poorly on them. (Journal entry, 6 January
2025)
The practices I have adopted most consciously are corrective ones: asking about circumstances before asking about progress, protecting writing time from administrative encroachment where I have discretion, and stating plainly that the recognition economy my students are entering is arbitrary in ways that are not their fault. Yet my capacity to act remains bounded. As a junior faculty member subject to the same output requirements as my students, I hold limited authority over workload allocation, admissions, or policy. The agency I exercise is real but narrow, operating in the interstices of structures I did not design and cannot alter, close to what Cong-Lem et al. (2025) describe as agency negotiated within, rather than against, structural constraint. Stage 5 therefore does not resolve the disempowerment of Stages 2 and 3 so much as recast it: I moved from lacking power over my own conditions to lacking power over conditions for which I am now held responsible. Becoming a supervisor did not complete the transition begun in Stage 1; it opened a further one, in which the ecological constraints of my own candidature reappeared as material I had to handle on behalf of others, under a technological condition that had no counterpart in my own training.
5. Discussion
The study illustrates that both as a doctoral
student and as a novice supervisor I experienced innumerable critical moments.
The last year of doctoral education was not merely a matter of writing up a
thesis but involved multiple tasks and a psychological journey shaped by
interactions across a wide range of social contexts; the extension into Stage 5
suggests that graduation did not conclude that journey. Below I interpret both
phases through ecological systems theory.
First, the transition from final-year student to
scholar and social being is inherently continuous and social. In alignment with
Gardner (2008), final-year students often
encounter “a drastic transition” (p. 344) from school to workplace. Strong
self-confidence marked Stage 1, yet when I turned seriously to job hunting in
Stage 3 I lost that confidence as severe challenges were met, and recovered it
only on receiving offers. As Gravett (2021) argues, completing a thesis
does not in itself signify transitional development; it is a dynamic and fluid
process of emotional and psychological instability arising from interactions
among ecosystems, for which most students are unprepared. That instability is
structurally patterned rather than merely personal: Liu and Li (2024) identify five competing
strategic logics through which Ph.D. graduates in the Greater Bay Area
negotiate career decisions, and Wang et al. (2025) show that the social
capital facilitating an academic job search in China is unevenly distributed
between domestically trained and returnee candidates. The confidence I lost in
Stage 3 was thus less a failure of self-belief than an accurate reading of a
market whose terms I had not been taught.
My narrative also highlights a progressive
trajectory of becoming a scholar. Recognition of my work prompted greater
emotional and physical investment in academic activity, which in turn furthered
recognition of my academic status, a bi-directional relationship between
doctoral engagement and academic development (Baker &
Pifer, 2011). Through interaction within research
environments and the development of research skills, students can re-negotiate
their researcher identity, exercise agency, and develop belonging (Cong-Lem
et al., 2025; Teng, 2020). Yet my agency was largely
subject to contextual factors in the meso-, exo-, and
macrosystems (Mantai, 2019). The external factors named
across Stages 1 to 4, economic, political, cultural, re-positioned me through
an emerging (Stage 1), reducing (Stage 3), and re-emerging (Stage 4) sense of
being a competitive applicant, while I had limited agency and scarce resources
with which to influence the workplace.
Second, renke, a keyword
repeatedly stressed in my journal, indicates the significance of recognition
and validation from supervisors, peers, academic organisations, and the
workplace. In accordance with Mantai (2017), “validation (recognition by others) influences
internal validation (self-belief)” (pp. 647–648): external validation
supports a positive orientation towards oneself and motivates students to
resolve difficulties. Mantai reported three routes to
such validation, closely tied to the micro- and mesosystems, including
publishing, conducting projects, and disseminating findings. My findings
complicate this. Although I held publications and international recognition, I
met disappointment in the Chinese job market, because Hong Kong, Macau, and
Mainland Chinese universities have distinct learning, teaching, and management
cultures, and most Mainland Chinese universities do not recognise
Scopus-indexed papers as proper publications. In addition to Mantai’s three microsystemic constructs, self-belief and
emotion are shaped by the wider academic culture and by social rules that
operate across ecosystems. Wang et al. (2025) supply a mechanism: where
recruitment operates through renqing and guanxi as
well as formal criteria, advisor-related ties and institutional prestige
accumulated in one system do not transfer intact into another.
Third, doctoral learning is an emotional practice.
Emotions trigger physical and psychological change, shaping and being shaped by
the learning process (Naude et al., 2014), yet emotional processes
remain comparatively underexplored relative to cognitive ones (Trigwell et al.,
2012). Alves et al. (2024) show that doctoral
emotional engagement is fostered by autonomy-focused supervision, perceived
project relevance, personal development motivation, and an inclusive academic
context - precisely the conditions whose presence and absence structure my account.
Novice graduates often experience anxiety when expectations go unrealised (Teng,
2017), and Rabe et al. (2021) found supervisors sometimes
overlook students’ “emotional rollercoaster journeys” (p. 318). Differing from Rabe et al., I received
abundant support from peers and supervisor, but complained of the vulnerability
of studenthood and the absence of support at
departmental and university levels, since as a final-year student in quarantine
I still had to teach. That absence, with my troubled job search, produced the
depression and despair of Stages 2 and 3. The results suggest that students
find it hard to regulate intrinsic emotions “through seeking extrinsic emotion regulation from
others” (Han & Xu,
2021, p. 2), because they lack power to influence
departmental decisions or pandemic policy. Power mediates emotional
self-regulation: the micro- and mesosystems contributed positively to my
emotional development, while the wider exo- and
macrosystems posed insurmountable challenges.
Fourth, the extended timeframe shows most
distinctively that these constraints did not dissolve with my change in
structural position. Stage 5 shows the same exosystemic and macrosystemic
forces reappearing on the other side of the supervisory relationship, where I
was no longer their object but their conduit. The indexing regime that
invalidated my publications in Stage 3 became the criterion by which I advised
my students; the absence of supervisory preparation documented by Lundberg (2025) meant the microsystem I
built for them was largely a copy of the one I had inhabited. Two implications
follow. Transmission in doctoral education is not confined to knowledge and
skills: ecological conditions are themselves inherited, and most faithfully
where no institutional structure makes them visible and open to revision. And
the agency of a novice supervisor is a poor instrument for interrupting them;
my capacity to shield students was confined to matters of personal discretion,
while workload allocation, output requirements, and recognition criteria
remained where they had always been. If the disempowerment of Stages 2 and 3 is
to be addressed, it will not be addressed by waiting for wounded students to
become better supervisors.
The technological condition of Stage 5 sharpens
the point. GenAI entered doctoral education as a macrosystemic change and
reorganised the supervisory microsystem with unusual speed, redistributing the
functions through which supervision has conventionally demonstrated its value (Jensen et
al., 2025) and generating interpersonal tension over legitimate
use and appropriate expectation (Lai et al., 2025). In my experience
institutional policy followed adoption rather than preceding it, leaving
individual supervisors, often, like me, junior and untrained, to negotiate a
novel and consequential boundary without a shared framework. This is an
exosystemic failure of provision of the same kind described in Stage 2,
recurring in one account for the same structural reason: the burden of adapting
to abrupt environmental change was devolved to the individual least equipped to
bear it (Teng & Wu, 2021). In short, landmark
learning and life moments can facilitate or hinder transitional development for
final-year students and for those who go on to supervise. Bronfenbrenner (2005) argued that special care is
needed to support a person through life changes; care required, on this
evidence, not only at graduation but at the point of assuming responsibility
for others.
6. Conclusion and Implications
The paper offers evidence of how a wide range of
ecological factors influenced a doctoral student’s transitional development, and how those same
factors persisted into the early years of a supervisory career. Final-year
doctoral study can be severely challenging, with demands ranging from academic
requirements to online teaching, job searching, and personal life management
amidst a pandemic; and the transition into supervision brings a further,
largely unsupported adjustment, now complicated by GenAI. Five implications
follow.
First, effective training and
support are critical to help final-year doctoral students move smoothly from
university to the job market. Although a major task of doctoral education is to
prepare students to conduct research independently, research competence alone
does not suffice in today’s market. Programmes might
also help final-year students develop a repertoire of strategies for meeting
career expectations and adjusting to job markets in different cultural
contexts, such as familiarising them with local academic cultures, recognition criteria,
and recruitment practices, and preparing them for job talks. Given that social
capital operates differently for domestically trained and returnee candidates (Wang et
al., 2025), such preparation should be explicit about the
informal as well as the formal terms on which appointments are made.
Second, because of cultural,
economic, and political differences, those who learn transregionally or
transnationally may struggle to adjust to different ecological cultures at
work. Proactive agency in responding to these challenges is called for, but joint
support across ecological systems is necessary for that agency to be exercised,
for instance, Chinese universities recognising Scopus-indexed journals as
proper venues, and university management developing a healthier perspective on
rankings.
Third, emotional flux is an
important yet overlooked aspect of doctoral education. Emotions, embedded in
sociocultural context, play an invisible but key role in shaping future
commitment. Support from supervisors and peers was constructive here in combating
negative feelings during the pandemic, but the lack of care at departmental
level caused distress, indicating that departmental support is essential in
regulating final-year students’ emotional states. The
predictors identified by Alves et al. (2024), including autonomy-focused
supervision, project relevance, and an inclusive academic context, offer
concrete points of intervention rather than general exhortation.
Fourth, the transition into
supervision warrants the same institutional attention as the transition out of
doctoral study. Consistent with Lundberg (2025), novice supervisors
improvise from their own experience of being supervised, transmitting both its
strengths and its deficits. Structured supervisor development, collegial
mentoring, protected early-career time, and non-punitive discussion forums
would render inherited supervisory practices visible and open to revision. This
is both an equity and a quality concern: without such provisions, it is the
students of the least supported supervisors who pay the price.
Fifth, institutions need to
move ahead of GenAI adoption rather than behind it. Supervisors are currently
negotiating questions of legitimate use, authorship, and the assessment of
understanding case by case and without shared criteria, which generates avoidable
friction in the supervisory relationship (Lai et al.,
2025). Because GenAI most readily substitutes for the
task-focused components of feedback while leaving the relational and
developmental components untouched (Jensen et al.,
2025), policy and training should aim not at prohibition but
at helping supervisors and students identify what supervision is uniquely for.
Critical GenAI literacy might usefully be treated as a shared curriculum for
candidates and supervisors alike.
The study is not without
flaws. Readers should interpret the findings with caution, as this is a single
person’s journey; the experience
reported was distinctive in that I acquired academic rules in Hong Kong while
pursuing a career in Mainland China, and may not transfer to other contexts. A
larger study with more participants would yield more generalisable results. The
two phases are also not methodologically equivalent: Phase 1 was documented
contemporaneously across a bounded year, whereas Phase 2 concerns an ongoing
period and therefore reports a transition still in progress rather than one
whose outcome is known. Finally, interviews with supervisors, department heads,
students, and parents could help triangulate a single narrator’s account.
Declaration
of GenAI Use in the Writing Process
The authors used ChatGPT to improve language,
clarity, and readability. The authors reviewed and edited all AI-generated
content as necessary and take full responsibility for the content of the
published article.
Funding
This research received no external funding.
Conflict
of Interest
The authors declare that they have no competing
interests.
References
Alves, P., Lopes, A., Cruz-Correia, R., & Menezes,
I. (2024). Predictors
of emotional engagement in doctoral education. Studies in Higher Education,
50(6),
1–16. https://doi.org/10.1080/03075079.2024.2369704
Baker,
V. L., & Pifer, M. J. (2011). The role of relationships in the transition
from doctoral student to independent scholar. Studies in Continuing
Education, 33(1), 5–17. https://doi.org/10.1080/0158037X.2010.515569
Bao,
J. (2025). Claiming legitimacy in the academic field: Chinese university TESOL
lecturers’ investment in doctoral learning. International Journal of TESOL Studies, 7(4), 33–52. https://doi.org/10.58304/ijts.250627
Braun,
V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in
Psychology, 3(2), 77–101. https://doi.org/10.1191/1478088706qp063oa
Bronfenbrenner,
U. (1979). The ecology of human development: Experiments by nature
and design. Harvard University Press.
Bronfenbrenner,
U. (1992). Ecological systems theory. In R. Vasta (Ed.), Six
theories of child development: Revised formulations and current issues (pp. 187–249). Jessica Kingsley
Publishers.
Bronfenbrenner,
U. (2005). Making human beings human: Bioecological perspectives on
human development. Sage.
Bronfenbrenner, U., & Ceci, S. J. (2005). Heredity,
environment, and the question ‘how’: A first approximation. In U.
Bronfenbrenner (Ed.), Making human beings human: Biological perspectives on
human development (pp. 174–184).
Sage.
Bronfenbrenner,
U., & Morris, P. A. (2006). The bioecological model of human development.
In W. Damon & R. M. Lerner (Eds.), Handbook of child psychology (pp. 793–828). Wiley.
Carter,
S., Smith, K., & Harrison, N. (2021). Working in the borderlands: Critical
perspectives on doctoral education. Teaching in Higher Education, 26(3), 283–292. https://doi.org/10.1080/13562517.2021.1911098
Cong-Lem,
N., Truong, K. D., Zhu, B., & Long, Z. (2025). Identity transitions,
experiences, and agency: A collaborative autoethnography of Asian international
TESOL doctoral students. The Language Learning Journal, 53(2), 143–158. https://doi.org/10.1080/09571736.2024.2349052
Cross,
D. I., & Hong, J. Y. (2012). An ecological examination of teachers’
emotions in the school context. Teaching and Teacher Education, 28(7), 957–967. https://doi.org/10.1016/j.tate.2012.05.001
Ek,
A. C., Ideland, M., Jönsson, S., & Malmberg, C. (2013). The
tension between marketisation and academisation in
higher education. Studies in Higher Education, 38(9), 1305–1318. https://doi.org/10.1080/03075079.2011.619656
Ellis,
C., Adams, T., & Bochner, A. (2011). Autoethnography: An overview. Historical Social Research,
36(4),
273–290.
Gardner,
S. K. (2008). ‘What’s too much and what’s too little?’: The process of becoming
an independent researcher in doctoral education. The Journal of Higher
Education, 79(3), 326–350. https://doi.org/10.1080/00221546.2008.11772101
Gravett,
K. (2021). Disrupting the doctoral journey: Re-imagining doctoral pedagogies
and temporal practices in higher education. Teaching in Higher Education, 26(3), 293–305. https://doi.org/10.1080/13562517.2020.1853694
Han,
Y., & Xu, Y. (2021). Unpacking the emotional dimension of doctoral
supervision: Supervisors’ emotions and emotion regulation strategies. Frontiers in Psychology, 12, 651859. https://doi.org/10.3389/fpsyg.2021.651859
Härkönen,
U. (2007). The Bronfenbrenner ecological systems theory of human development. 1–19. http://wanda.uef.fi/~uharkone/tuotoksia/Bronfenbrenner_in_English_07_sent.pdf
Horta,
H. (2018). PhD students’ self-perception of skills and career plans while in
doctoral programs: Are they associated? Asia Pacific Education Review, 19, 211–228. https://doi.org/10.1007/s12564-018-9532-y
Jensen,
L. X., Bearman, M., Boud, D., & Konradsen, F. (2025). Feedback encounters
in doctoral supervision: The role of generative AI chatbots. Assessment & Evaluation
in Higher Education. Advance online publication. https://doi.org/10.1080/02602938.2025.2478155
Lai, S., Liu, S.,
Dai, Y., Lim, C. P., & Liu,
A. (2025). The
impacts and tensions of generative AI on doctoral students’ supervisory and
peer dynamics: An activity theory analysis. Australasian Journal of
Educational Technology, 41(5), 1–17. https://doi.org/10.14742/ajet.9916
Liu,
Y., & Li, D. (2024). Unsettling academia: Academic career choices and
pathways of PhD graduates in the Greater Bay Area in China. Higher Education Quarterly,
78(2),
388–403. https://doi.org/10.1111/hequ.12460
Lundberg,
A. (2025). Transitioning from doctoral student to doctoral supervisor. Research in Post-Compulsory
Education, 30(3), 517–535. https://doi.org/10.1080/13596748.2025.2522549
Mackey,
A., & Gass, S. (2016). Second language research: Methodology and design (2nd ed.). Routledge.
Mantai, L. (2017). Feeling like a
researcher: Experiences of early doctoral students in Australia. Studies in Higher Education,
42(4),
636–650. https://doi.org/10.1080/03075079.2015.1067603
Mantai, L. (2019). ‘Feeling more
academic now’: Doctoral stories of becoming an academic. Australian Educational
Researcher, 46(1), 137–153. https://doi.org/10.1007/s13384-018-0283-x
McGloin,
R. S. (2021). A new mobilities approach to re-examining the doctoral journey:
Mobility and fixity in the borderlands space. Teaching in Higher Education,
26(3),
370–386. https://doi.org/10.1080/13562517.2021.1898364
Miller,
L., & Wu, J. G. (2022). Re-learning, reacting, and reflecting on sudden
pedagogical adaptations. In Pun, J., Curle, S., & Yuksel, D. (Eds.), The use of technology in
English medium education (pp. 145-158). Springer.
Naude,
L., van den Bergh, T. J., & Kruger, I. S. (2014). ‘Learning to like
learning’: An appreciative inquiry into emotions in education. Social Psychology of
Education, 17(2), 211–228. https://doi.org/10.1007/s11218-014-9247-9
Park,
C. (2005). New variant PhD: The changing nature of the doctorate in the UK. Journal of Higher Education
Policy and Management, 27(2), 189–207. https://doi.org/10.1080/13600800500120068
Rabe,
M., Agboola, C., Kumswa, S., Linonge-Fontebo, H., & Mathe, L. (2021).
Like a bridge over troubled landscapes: African pathways to doctorateness.
Teaching
in Higher Education, 26(3), 306–320. https://doi.org/10.1080/13562517.2021.1896490
Rao,
N., Hosein, A., & Raaper, R. (2021). Doctoral students navigating the
borderlands of academic teaching in an era of precarity. Teaching in Higher Education,
26(3),
454–470. https://doi.org/10.1080/13562517.2021.1892058
Shin,
J. C., Postiglione, G. A., & Ho, K. C. (2018). Challenges for doctoral
education in East Asia: A global and comparative perspective. Asia Pacific Education
Review, 19(2), 141–155. https://doi.org/10.1007/s12564-018-9527-8
Teng,
M. F. (2017). Emotional development and construction of teacher identity:
Narrative interactions about the pre-service teachers’ practicum experiences. Australian Journal of Teacher
Education, 42(11), 117–134. https://doi.org/10.14221/ajte.2017v42n11.8
Teng,
M. F. (2020). A narrative inquiry of identity construction in academic
communities of practice: Voices from a Chinese doctoral student in Hong Kong. Pedagogies: An International
Journal, 15(1), 40–59. https://doi.org/10.1080/1554480X.2019.1673164
Teng,
M. F., & Wu, J. G. (2021). Tea or tears: online teaching during the
COVID-19 pandemic. Journal of Education for Teaching, 47(2), 290-292. https://doi.org/10.1080/02607476.2021.1886834
Trigwell, K., Ellis, R. A., &
Han, F. (2012). Relations between students’ approaches to learning, experienced
emotions and outcomes of learning. Studies in Higher Education, 37(7), 811–824. https://doi.org/10.1080/03075079.2010.549220
Wang,
Y., Xie, Y., & Jing, X. (2025). The multi-dimensional, multi-level impact
of social capital on academic job search in China: A comparison between
domestic PhDs and PhD returnees in the social sciences. Higher Education. Advance online publication. https://doi.org/10.1007/s10734-025-01532-x
Author biodata
Junjie Gavin Wu is an Assistant
Professor and Ph.D. Supervisor at Macao Polytechnic University, as well as the
Vice President of PacCALL and the Vice President of GloDAL. He is Editor-in-Chief of SN Social Sciences
(Scopus, Springer), Editor-in-Chief of Applied Language Sciences, an Executive
Associate Editor of Computers & Education: X Reality (ESCI), and an
Executive Associate Editor of Artificial Intelligence in Language Education (Castledown). Gavin has authored around 80 English
publications, with over 35 papers appearing in SSCI/SCI journals. He has
published books with Cambridge University Press, Routledge, and Springer, and
spearheaded over 10 special issues. Gavin serves on the editorial boards of
several leading journals, such as the British Journal of Educational
Technology, Interactive Learning Environments, Language Learning &
Technology, Computers and Education: Artificial Intelligence, IJCALLT,
Multimedia-Assisted Language Learning, and so on. He is also an award-winning
teacher in China.
E-mail:
gavinjunjiewu@gmail.com
ORCID: 0000-0003-4937-4401