Enhancing factors for doctoral students’ writing self-efficacy: A narrative approach

Self-efficacy is an important construct doctoral students need to work on in their academic writing. Elevating academic writing self-efficacy to achieve the international standards for EFL doctoral students is a pursuit of success. Previous studies seem to focus on anxiety commonly experienced by the students rather than on EFL their positive self-concept and high persistence, parts of self-efficacy. Thus, this study aims to investigate factors enhancing the academic writing self-efficacy of doctoral students studying overseas (e.g., Australia, Germany, Korea, Thailand, and the U.K). For this narrative study, a set of narrative frames and interview guides were used to collect data from eight volunteer participants who were purposively selected on these criteria: coming from the Eastern part of Indonesia, under competitive scholarship awards, having at least 10-year EFL teaching background. Thematic analysis was used to analyze the collected data. The findings indicate that the contributing factors include good self-concept, self-belief in academic writing, doctoral study (supervisory atmosphere, the supportive school system, and scholar community), self-efficacy in writing (logic, argumentation, near-native style, integrating references). Through this study, the narrative frames proved to be practicals tool for exploring the elusive construct of self-efficacy in EFL writing. The study suggests training on writing self-efficacy for graduate schools in EFL contexts along with establishing a support system, and this aims at achieving international academic writing standards for the prospects of publication.


INTRODUCTION
Academic writing has perturbed many English as a Foreign Language (EFL) doctoral students having the writing hurdles to overcome.Perhaps the concern comes from the intricacy of academic writing that many struggle with not only technical writing abilities but also the feelings triggered by the difficulty and pitfalls of the writing process, especially for those just starting their doctoral study (Wilson & Cutri, 2019).Nevertheless, special retreat addressing doctoral students' academic writing has been scarce (Vincent et al., 2023).Academic writing requires many aspects.The English-speaking community's thinking and writing styles can sometimes force EFL writers to reframe their academic writing for a different audience or a greater focus on theories, which can help them focus more sharply on problems when they arise (Kim & Saenkhum, 2019).The disciplinary community also requires the EFL writers' critical thinking and position (logic, objective and updated to the recent development).Indonesian doctoral students in an English Language Education program did not demonstrate contextualization of research in the relevant body of knowledge within their dissertation, in addition to two other issues of lacking argumentative load and the absence of research gap identification (Basthomi et al., 2021).
Indonesian academic writers tend to have potential problems with writing research articles in English rhetorical styles, which are different from Indonesian rhetorical styles (Arsyad, 2018).He further explains that the English rhetorical style covers logical reasoning and topic justification, and if they miss these, the writing does not pose the potential for publication.International students pursuing study in Malaysia encounter difficulty concerning second language (L2) writing, experiencing the effects of the first language (L1) to L2 (Jomaa & Bidin, 2017).EFL postgraduate students of TESOL programs in Australia who originally come from China, Korea, Oman, and Taiwan, also share similar concerns regarding coherence, cohesion, articulating voice, relevant topics, referencing, and citation (Al-Badi, 2015).Due to a lack of knowledge and awareness, Arab EFL students struggle with citations of various functions, which has pedagogical repercussions for reasoning (Jomaa & Bidin, 2017).
These similar issues continue to perplex EFL doctoral students globally, including Indonesians, in working to the improvement of writing aspirations.Writing, a complex task requiring simultaneous cognition management, motivation and linguistics processes, increases its complexity in L2 context (Mendoza et al., 2022).As academic writing is an iterative process, various circumstances influence the writer internally and externally (Vincent et al., 2023).Internal factor writing self-efficacy is extremely pertinent to doctoral students' challenges (Huerta et al., 2017).While externally, selfassessment of ability is also influenced by the judgments of significant others (Bandura, 1997).Regarding this, Purwanto et al. (2020) identifies some problems that prohibit publication, namely the absence of EFL writers' confidence in writing quality, lack of writing investment of preparation, and high dependence on external rewards upon publication.
The causes doctoral students encounter barriers to publishing in the high-impact journal come from negative review results, financial publication issues, long-response time, low English proficiency, preparation and submission time constraint, limited access to journals and references, and technology (Purwanto et al., 2020).The facts may worsen the spirit of aspiring for publication among EFL writers, and the concern started to crystallize, i.e., when EFL doctoral students are confronted with the obligation to write academic papers from their studies.This implies the need for writing self-efficacy.

Self-Efficacy to Promote International Academic Writing Practice
Self-efficacy is context-specific and directly related to the surrounding situation and could be contributed by the enactive experience of the student-writer (Schunk & Zimmerman, 2007).Furthermore, writing self-efficacy is a theoretical notion that incorporates everything inherent in a writer's belief in their ability to write, such talents requiring many skills, methods, and knowledge within certain situations (Mitchell et al., 2017).Interestingly, while there is ample literature suggesting the correlation between self-efficacy and writing skills, such as Bai et al. (2020), Bingöl et al. (2019), Fathi et al. (2019), Fatima et al. (2020), Feizi (2019), Ho (2016), Lichtinger (2018), Litson et al. (2021), Sun andWang (2020), andSu et al. (2018), other studies show the relation between selfefficacy in writing to writing competence and achievement (Khojasteh et al., 2016;Sabti et al., 2019;Tai, 2016).
Since self-efficacy in L2 writing is operationalized as transferable and interactional per se, according to Bai and Guo (2018), there is a significant correlation between the use of selfmonitoring and the organization.However, the strongest correlation occurs between revising and self-initiating.Ho (2016) concludes that the greater the complexity and higher frequency of revision in L2 writing, the higher one's self-efficacy belief in engaging in challenging activity.Strong beliefs about competence enable one to approach challenging tasks (Schunk & Meece, 2010).In contrast to individuals who doubt their abilities, self-efficacious people are more eager to engage, work harder, and stick with projects for extended periods.They also respond less negatively when difficulties arise (Jalaluddin et al., 2013).Bai et al. (2018) indicate that high achievers or skilled writers apply more planning, self-monitoring, textgenerating, and revising strategies than low achievers or unskilled writers.The writers' selfefficacy in writing with a certain level of confidence enables them to perform writing tasks and achieve desired outcomes (Bandura, 1982;1997) and its effects on task choice, efforts, perseverance, and success (Britner & Pajares, 2006).Bandura (1997) lays the groundwork for writing self-efficacy at the outset by stating that self-efficacy in writing expands within someone by their conscious or subconscious receiving, processing, and acting on self-efficacy in writing.All influence sources including personal successes (mastery), peer success (role model), verbal-received persuasion, and physiologically state enhancement.
In particular, three characteristics are associated with writing self-efficacy based on Bandura's concept as described in categories by Wang and Jape (2007), and the present study modifies the description to suit its context.

Categories Description 1. Academic writing persistence
Longer persistence in writing owing to solid belief in the work indicates high self-efficacy.

English proficiency self-awareness
Perceived competence refers to being conscious of one's English proficiency.Once a student demonstrates competence (e.g., vocabulary or tenses, awareness of coherence in writing), s/he is said to be very self-efficacious.

Language learning engagement
Students are enthusiastic about participating in writing-related activities.Narrative frameworks may provide evidence for the association between students' writing self-efficacy and their willingness to participate in professional development (PD) programs for academic research writing.Wang and Jape (2007) indicate that students' interests, teachers' roles, the complexity of needed tasks, learners' performances, comparison to other learners, efforts put into the task, and metacognition awareness have all influenced the categories of selfefficacy.
There are several psychological variables concerning self-efficacy and academic writing.Fathi et al. (2019) accentuate a positive correlation between self-efficacy in writing and students' interests, sustained efforts, self-regulatory capacity, writing concept, goal achievement, and successful writing performance.Psychological resilience and positivity were significant predictors of self-efficacy in writing (Bingöl, et al., 2019).Writing selfefficacy is also a generative capability, and that requires the effective coordination of cognitive, motivational, emotional, and behavioral skills, and hence that capability is to properly fulfill various purposes (Teng et al., 2018).Besides intellectual, emotional challenges accompany writing (Teng et al., 2018).All of those challenges have a significant impact on the writing process overall.Litson et al. (2021) disclose that early preparation programs for doctoral students will develop self-efficacy in writing for a successful doctoral journey.Before the degree enrollment, the students often have an insufficient apt level of structure and strategic communicative skills with supervisors, administration and colleagues (Feizi, 2019).According to Schunk and Pajares (2002), there may be a developmental component to the formation and growth of self-efficacy and the predictive and mediating roles of self-efficacy depending on study level and years of schooling.The role of self-efficacy in writing is neither wholly stable nor homogenous across subgroups or levels of doctoral study (Litson et al., 2021).Steady selfefficacy is essential to maintain while pursuing the degree.For example, they stayed steady after receiving down-breaking feedback from reviewers.On the other hand, Ho (2016) attempts to see the possible challenges that may have to do with academic writing for publication in the doctoral requirement.He recommends research to continue study on doctoral students' development of writing self-efficacy, particularly regarding writing productivity and quality.Therefore, doctoral students need to invest efforts by making the most out of all related-supported writing activities, aiming at consistency all through the stages of their doctoral study.As accentuated by Bingöl et al. (2019) that investing in self-efficacy while maintaining a fundamental idea of identifying and achieving personal life goals will primarily contribute to people living longer while expanding their knowledge of themselves through access to references for enriched-compacted writing content.
Yet, self-efficacy is an elusive construct, and the appropriate level of specificity in measuring self-efficacy has been challenging to discern (Tschannen-Moran & Hoy, 2001).Researchers, nonetheless, up to today, have ever since had difficulty developing a measurable tool for capturing it.Bruning et al. (2013) acknowledge their restricted self-efficacy writing scale to one factor and three-factor proposed model to precisely depict the operationalization of the self-efficacy construct; hence other models are welcomed.Macalister (2012) entails that questionnaires and surveys could not facilitate the depth study.Hence, we pinpoint the multi-dimension of writing self-efficacy through the present narrative study.
The study aims to investigate the writing selfefficacy of EFL doctoral students during their study overseas.The research question to answer is "What are the succeeding factors of EFL doctoral students' writing self-efficacy while they are studying overseas?"Thus, the central to our views is simply that all writing self-efficacy in EFL context of doctoral students is examined from the perspectives of the persons involved (Creswell & Creswell, 2018) to apprehend the elusive construct.Therefore, much is still not well informed regarding incorporating narrative frames in combination with a narrative interview as instruments of collecting data on underlying self-efficacy in doctoral academic writing.More studies need to illuminate these issues from the context of EFL.Barkhuizen et al. (2014) identify narrative frames in researching language instructors' experiences as potentially rich sources of information in terms of practicality in narrative inquiry.Inferring from Marschall and Watson's (2022) study, we conclude that narrative frames play a role in self-schema adaptation which is a model of self that is dynamic and responsive to experiences, memories, social work, and future possibilities.We recognize the importance of depth examination on writing self-efficacy, yet the study about this is scant.That implies that utilizing narrative inquiry study would unravel the elusive construct of the efficacy which is far beyond the numbers and calculation.

METHOD
This study applied a narrative inquiry study to determine Indonesian doctoral students' self-efficacy in academic writing.Narrative inquiry (Barkhuizen et al., 2014), also known as personal experience (Clandinin & Connelly, 2000), allows people to participate in the meaning-making of others.This method, noted for its small number of research participants, may answer the question of reality by re-constructing experiences gleaned from societal and personal stories after getting momentum from human intricacies (Webster & Mertova, 2007).The narrative inquiry study employs thematic analysis of essential themes (Clandinin & Connelly, 2000), of the participants' shared real stories.

Respondents
The idea behind the narrative inquiry study is to choose participants who best assist the researchers on the issues raised in research questions of their shared experiences (Creswell & Creswell, 2018).The merits of the participant selection of the study include individuals from the eastern part of Indonesia who have been awarded a competitive scholarship and had at least a 10-year EFL teaching background.Furthermore, the reasons for demographic selection indicate that participants from underdeveloped areas have resilience to the challenges of studying at prestigious universities overseas.In total, eight participants (four females and four males) voluntarily involved, as shown in the profile below (pseudonyms are used*).

Instruments
The study collects data utilizing two major instruments, narrative frame and interview.Both instruments allow us to reveal the participants' elusive construct of self-efficacy, manifested in their academic writing during their doctoral study.From Lam's (2020) qualitative study, narrative frames collect wealthy participants' data from written narratives.That said, the narrative frames have sentence starters to complete in uncovering the participants' self-efficacy to handle the academic writing complexity.Seven frames were provided to compile data on self-efficacy and self-regulation, while the present study is mainly focusing on the first construct, i.e., self-efficacy (1st phase of dissertation research).In addition, the narrative interview was used to attend to the flow of the story told and expressed by participants without interrupting (Clandinin & Connelly, 2000).The interview shares advantages, namely practicality when direct observation is unavailable; providing past information of participants; researcher's control over the line of questioning (Creswell & Creswell, 2018).Both Creswell enlist essential interview components, consisting of content questions with probes and closing instructions, yet the questions should go naturally with the investigation.Thus, aspects of self-efficacy writing are accommodated through the instrument.

Procedures
First, we prepared narrative frames referring to some of Clandinin and Connelly's (2000) elements and interview guides referring to Creswell and Creswell's (2018) components.Next, the frames and the guides, along with the consent form, were administered to participants (overseas) through email.However, during data collection, we met in person with participants staying in Indonesia.Within three days, the consent form and the frames were returned.Any raising issues from the frames were noted for clarification in the interviews which had been conducted twice for approximately 1 to 1.5 hours.Depending on the participants' location during data collection (in the country or overseas), we conducted the on-site or online interview (by zoom).The interviews are semi-structured and provided with an interview guide and probes.
During the interview, attending was employed but not interrupting (Barkhuizen et al., 2014).The data saturation technique was applied by having followup interviews.
Then, a professional helped transcribing the interview before the participants checked for data validation.Collaboration with participants as storytellers in all phases of the research using the retrospective technique (Birt et al., 2016) is required.The checking aims at achieving rigor and trustworthiness because the result credibility is the bedrock of high-quality qualitative research (Clandinin, 2006).In other words, it is called participant validation to explore the result's credibility.After that, we cleared data by dropping unnecessary parts for further analysis.

Data analysis
The present study's data of the two instruments are under thematic analysis by determining key findings.Employing NVivo 12 Plus, the analysis operation focuses on matrix coding query and project mapping as measured by several coding references.A term or short phrase that symbolically assigns a summative, salient, essence-capturing, and/or evocative attribute to a portion of languagebased or visual data is known as a code in qualitative research (Saldana, 2013).Then the theoretical lenses guide the interpretation (Pahlevi, 2020) of the study's results.

FINDINGS AND DISCUSSION
Of the data operationalization, four main nodes appear.Under each main node, there are several sub-nodes namely self-concept (being an awardee of competitive scholarship, a scholar, a non-native writer, a PhD student), self-belief (contribution to others, writing competence, efficacy, study goal), doctoral study (supporting system, scholar community, facilities and accesses, school program, supervisory atmosphere), writing self-efficacy (writing skill, writing investment, high self-efficacy, anxiety).In general sense, nodes are equal to factors, whereas sub-nodes mean sub-factors.The skills needed for academic writing (sub-nodes) include paraphrasing, synthesis, coherence, rhetorical expressions, reference integration, meaningfulness, academic vocabulary, organizing, structure, near-native-like styles, advanced tenses, grammar, logic, and arguments.Whereas, writing investment (sub-node) is in the forms of joining professional development program, reading on references, proofreading, draft-practice-revision.In addition, there are several elements under high selfefficacy (sub-nodes), such as creativity, scholarly sharing, self-reflection, openness to feedback, persistence, and self-encouragement.

Succeeding Factors Leading to EFL Writing Selfefficacy Self-Concept Factors
The data on thematic analysis reveals that the good self-concept of the doctoral students comes from four concurrent roles attached to them (Figure 1) either as a PhD student (78 coding references count), a non-native writer (38), a scholar (23), and a competitive scholarship awardee (19).Students' self-concept is accumulated from their personal experiences and responses to the surrounding situational settings (Pajares & Schunk, 2005).In turn, the external frame of reference infers one's skill in comparison to that of others (Mendoza et al., 2022), whether they are native or non-native writers.

Figure 1 Self-concept mapping
Throughout the doctoral study, the self-concept develops along with enriched experiences.According to Pajares and Schunk (2005), when experiences increase, self-concept can have both a constant and evolving component.In her selfinitiated follow-up interview, one of the participants, Armi, said converting her study stream from a Doctor of Education (EdD) to a PhD resulted from her excellent seminar paper at the start of the third semester, which received outstanding comments from the examination panel.That indicates a good self-concept empowers self-belief, contributing to succeeding factors towards writing self-efficacy.

Self-Belief Factors
The EFL doctoral students' self-belief comes from setting their study goal (prior to and during the study): having standard writing competence, an expectation to contribute to others, which is counted in 59 references coding, 32, 29, and 22, respectively.The second factor is illustrated next in Figure 2.

Figure 2 Self-belief mapping
In the realm of doctoral study goals, the participants have varied objectives either short or long terms such as keeping up dissertation writing to timeline, having supervisors' satisfactory remarks on confirmation exam, passing dissertation exam without conditions, writing for international reviewed journals.Interestingly, Adaline set high goals which are passing on her dissertation defense from well-respected linguists and expecting others to read her works and cite.According to Bandura (1982), people assess their abilities which is resulting in sense of efficacy leading to their conduct.Furthermore, Bandura mentions that higher levels of perceived self-efficacy are associated with higher performance levels.The eight participants' higher performance attainment was attested by the recent four roles and being awarded competitive scholarship, and hence greatly influence self-belief on writing competence, contribution to others and efficacy in life.

Doctoral Study Factor
The third succeeding factor, doctoral study, highlights the necessity of EFL doctoral students' overseas writing self-efficacy (Figure 3).Interestingly, the supervisory atmosphere positions relatively high with 135 coding reference counts.These four sub-nodes under the doctoral study node significantly reveal an exciting issue to discuss.

Figure 3 Doctoral study factor (coding reference count)
First, the findings imply that the supervisory atmosphere is influenced by the supervisor's rapport with doctoral students, equality principle, maintaining supervisees' mental health, nurturing flexibility of consultation, intensive writing assistance, and helpful feedback.Through the equality principle shared between the supervisor and the supervisee, both sides take care of the responsibility of dissertation research.Hugo states that: "When I came to Germany, my professor said that was my time to do my research.She said I have my proposals, read them deeply, and then start writing everything about my research.She told me she was working to guide me in whatever I wanted.She reminded me to manage my own time whether I can finish in 4 years, ten years, or 20 years, and she is available for that."Kornharber et al. (2016) uphold their claim for academic writing retreat by establishing a structure and setting specific goals while developing selfmotivation and self-confidence.Hugo's supervisor exemplifies to her supervisee the importance of goal setting prior to the study.Further about doctoral study, Rowe and Fitness (2019) address the need to maintain doctoral students' well-being during study overseas.Feizi (2019) underlines that inadequate support leads to a substantial source of stress for doctoral students.High caution about maintaining the supportive relationship between supervisors and supervisee might prevent unnecessary problems.Armi recalls that her supervisors endeavor to maintain her mental health realizing doctoral study overseas has its challenges: "I once had an experience she visited me in person.Like last time I got a terrible flu.So, I did not come, and she was so worried.My friend in the office told her about my condition.Then I sent her an email to say sorry.She suggested I take some herbs.On my articulation exam, the other supervisor gave me a pink hot water bottle after knowing how terrible my flu was" (while showing the bottle in the zoom interview to the interviewer).
The participants' supervisors highlighted partnership for dissertation research, including equal contribution and role of the students and the panel of supervisors.Thus, they open the flexibility of communication.Adaline, in a high efficacy manner, admits that "I am the one who decides my supervisory team from the names the school enlisted by providing reasons for my selection to the school.My supervisors treat me like a colleague.We PhD students, look like staff, so every announcement or post they send to us through mail or email.Everyone in the office gets information quick.After receiving feedback from my supervisors, I chose whether I needed to meet them and would ask them if I needed a meeting for discussion.I think it is more informal and open in Australia compared to Indonesia.You can disagree with your supervisor.They will not take it personally." Besides the rapport, the school programs designed at least a few considerations for advanced doctoral study (82 ref coding count), including available working space (office, laboratory, cubicle), meeting schedule, approval on joining relevant training for research projects, and study progress monitoring.In similar aspiration, though it is different from one country to the other, mostly the participant's host university gives access to facilities and rich resources (18 ref coding count), e.g., PhD computer, journal portals, grammar checker, programmed PhD community of reading and writing, dissertation editing, and financial assistance.Additionally, the supporting system of the university (recorded at 15 ref coding count) intensively leads the students working on a dissertation by facilitating research and academic writing through different areas.For example, from four Australian universities (see Table 2): cohort FB group (for dissertation discussion), biweekly 'shut up and write' group (to 2.5-hour-individual reading and writing session with free dinner), and various linguistic reading groups.In brief, the school's community providence validates Kornhaber et al.'s (2016) writing retreats, i.e., to legitimize the acts of writing in a comfortable working zone without interruption in a fostered supportive community.In another instance from Korea, Jack narrates that his school has urged the students to research and writing collaboration, whereby his supervisors had expanded a network in Jack's interests.Proudly speaking, he managed to present his writing not only in South Korea but also in the U.S. and publish seven academic papers (including a dissertation and two conference papers) during his study.

Writing Self-Efficacy Factors
The fourth factor (Figure 4), writing self-efficacy, has been coded with four sub-nodes: high selfefficacy, writing investment, writing skills, anxiety.The figure implies that 'writing' (main theme) has close association with 'progress' to indicate that academic writing progressing throughout the participants' study, and mostly the 'feedback' from 'supervisors' comments' are 'sometimes' felt 'painful', yet the doctoral students 'understand' and continuously 'work' to improve the 'draft' and achieve the international 'standards' of academic writing.

Figure 4 Writing self-efficacy coding query
Supported by Wang and Jape's (2007) categories of writing self-efficacy, the eight participants have achieved writing self-efficacy, justified by persistence, awareness, and study engagement.Needless to say, although anxiety is inseparable from writing self-efficacy development, it lasted quite short due to right self-concept (Figure 1), self-belief (Figure 2), and other related subfactors such as high self-efficacy (with the occurrence of 6 elements in references coding), writing investment (4), and writing skills (11).In contrast, anxiety, which is a part of the writing selfefficacy development, is at the 4th place, with the least references coding count (33) from 16 files, out of total 44 data files.Confirmed in the interview, none of the participants take personally any supervisors' hard feedback, instead perceived them professionally, although most of them need to step back for a while to refresh before proceeding.The next diagram relates to this issue.
Figure 5 presents those giving initial impetus to high self-efficacy (sub-factor) namely selfmotivation, persistence, openness to feedback, selfreflection, academic sharing, and innovation.The study opposes Huerta et al's (2017) findings that inconclusively reported that English-native students have higher EQ than those of non-native students with lacking cultural variability within the measurement used and data collection prior to the writing process.The present study, however, unravels that these students' writing self-efficacy evolves, and they are emotionally mature in managing writing tasks despite various hurdles encountered.

Figure 5 High self-efficacy sub-factors (coding ref count)
Figure 5 shows self-encouragement at the highest peak (54 counts), followed by academic writing persistence with a minor difference at 52.This explains why anxiety accounts (see Appendix) for only 33 counts of reference coding, and the gathered narratives show that it is quite low and fades quickly.As expected, Huerta et al. (2017) mention that these students' ability is to positively control their emotions to relieve stress, communicate effectively, and overcome problems -those with higher EQ outperform those with lower EQ.
In regard to the writing investment (Figure 6) towards writing self-efficacy, the participants have paid off hard work in different areas such as outlining, drafting, practice up to revision; peer review and proofread by professionals; browsing and reading relevant references; joining any professional development (PD) programs to solve each part of the dissertation writing and analysis.

Figure 6 Writing investment sub-factors (coding ref count)
More specifically to writing skill (Figure 7), the participants have been aware of elements of advanced writing with 11 the most essential ones, described as follows.
Logic and argumentation succinctly cover all the salient points within academic writing.In relation to integration of references into the writing, the students have embarked on their investigative area since 1st semester and progressed.For example, Moses was aware the need of enriching his insights before developing arguments and making argumentation smooth flow (1.49%); Armi believed she can connect and contrast theories to the factual phenomena (1.21% coverage on the coding); Izril learnt to elaborate ideas on common ways of academic writing while trying to link the relevancy of one part to the others (1.21%); Hugo limited his own assumptions and subjectivity among the argumentation (1.17%); Lia equipped her arguments with scientific evidence in support (1.12%); Jack adapted to EFL context for argument building (0.89%); Nita was keen to find infeasible answers to many research questions in her minds (0.80%); Adaline made counter-arguments for quality paper (0.67%).Swales (2009) underlines the incorporation of any insights obtained into writers' subsequent writing since short-term enthusiasm does not automatically mean long-term improvement, irrespective of highly positive response.He adds that summarizing references, critical review, and a topic discussion have been designed to demonstrate disciplinary differences within a single genre, but as the number of texts required increases, so do the time and effort required to search relevant literature, while the chances of a serendipitous encounter decrease.In this sense, maintaining both self-belief and self-efficacy in writing should be made continuously.
As EFL writers, the eight participants realized that peer review and proofreading cannot always happen due to tight schedules of either the participants or their peers.Instead, they employed applications to the writing besides their EFL intuition: Grammarly (grammar checker, sentence construction and plagiarism detector), Quillbot (paraphrasing), Ludwig (word similarity of different contexts).Not only can advanced tenses and grammar draw the attention of the doctoral students, but the other challenging issues such as near nativelike style, being logic and critical, can also draw their attention to targetting the readers' perceived meanings.Mostly the students aimed high to achieve near native-like style of writing, arguing that they are communicating written thoughts to international audience.Jack during his first semester was reminded gently by his supervisors about 'salad bowl' writing where he put everything like the mixture of all 'ingredients' (information) but then he exercised his logic to sort it out and insert it in.Adaline attended two tutorial classes before starting her first year, in order to be able to evaluate critically.She believes being logic similar to English native speakers is her issue, and hence, she needed to strengthen her arguments justifying that the sound arguments have clear and structured ideas.Similar thoughts were shared by Armi, in regard to logic and argumentation, which according to her that clarity of ideas should be performed through sentences.She added the logic behind those sentences are important factors to polish her writing to look sophisticated and smooth flow.In fact, Armi's writing self-efficacy is significantly improved in the 2nd year in which she had strong theoretical framework proposing a new study out of unavailable literature on Indonesian translanguaging in primary schools.On the other hand, Hugo, being home-context oriented, seemed to load subjectivity in his research paper, to address a real situation within his academic writing.Many EFL writers tend to fall into this temptation though.Then, Hugo learnt to re-write near native-like style by being objective to deliver his voice.Relating to objectivity versus subjectivity within writing, Nita argues, "I think I tried to make it more objective and not to influence the readers.Yet, at the same time, of course, there are a lot of papers for the topic that you are investigating.You need to be able to pick which ones are suitable to use, to drive your logic, and to support or even to argue.Maybe you could use them to contradict with what you have done." From different aspects of academic writing, one might also need vocabulary and selective phrases to build arguments.Having a common belief as an EFL writer, Izril shares his concern in addressing comprehensive writing to his Thailand university context and international audiences by claiming that, "The most difficult part is lacking strong keywords." In line with Kweldju 's (2003) study, students' vocabulary learning is activated through constant mindful raising to make correction useful since near native-stylistic problems were repaired by reformulation.Although sharing similar perspective as Izril, however, Nita came to a term that she needs to revisit her previous view by reflecting that, "Professionals editor just noticed my piece of writing and simplified my advanced English words.This is not like what I learned, so I asked the editor about this.What surprised me more was that he said that academic writing doesn't need complex or high vocabulary but saying straightforward.I assume most academic papers nowadays have shifted to simpler word use so that readers can easily follow the arguments." We presume that the editor highlighted the essential idea of the paper's content while leaving alone the 'wrapper' (to use Basthomi's term) or rhetoric.Turning to the English writer community, Basthomi (2006) raises the possibility of English rhetorical expectations in academic writings for international publication.Basthomi's cutting-edge rhetorical journey was based on his EFL writer background.It included his sensitization to English academic writing rhetoric as it is normalized in the English-speaking community.According to him, Indonesian editors of English journals tend not to share the importance of rhetoric, unfortunately.
In Figure 7, the reference coverage count was relatively small, with 33 from seven files (narrative frames and interview transcriptions).Looking from Holliday's (2005) school of thoughts many have turned from promoting native-speakerism over EFL writing works, as this may distort itself arguing on possible disadvantages against non-native writer's contribution.
The remaining abilities leading to the succeeding sub-factors of writing self-efficacy are meaningfulness, coherence, synthesizing, and paraphrasing, which have become a regular practice in academic writing of the doctoral students.

CONCLUSION
To the authors' best knowledge, previous studies partially unwrap self-efficacy factors associated with doctoral students' writing complexity, majorly focusing on anxiety, whereas the current study has served to unravel positive self-concept and high persistence of the EFL students.
As personal narratives are unique in their essence, the present study broadens ELT perspectives on self-efficacy for advancing academic writing.The study reveals that narrative frames have served as practical tools disclosing the elusive construct of self-efficacy in academic writing.Despite the limitation, however, the frame's benefits surpass the limitation.Futher research needs to see how writing self-efficacy training for graduate schools in EFL contexts, as well as the establishment of a support system, can meet the goal of international academic writing standards.
A writing self-efficacy program by graduate schools for EFL doctoral students is highly recommended aiming at speed academic writing development and high publication potentiality.

Figure 7
Figure 7Writing skills sub-factors(coding ref count)

Table 2
Participant profile