How supervisors provide and students react to EAL thesis supervision: Voices from Sweden and Indonesia

Introduction Thesis supervision is a critical part of students’ academic literacy development. Previous research has shown different dimensions of this development with limited attention to cross-cultural aspects. In particular, there has been little research on how students and supervisors negotiate supervision practices in non-anglophone contexts. This study aimed to explore students’ and supervisors’ reported priorities and experiences regarding the provision and reception of feedback in English as an Additional Language thesis supervision. Method We conducted a qualitative case study to illuminate supervisor’s and students’ experiences of supervision in Sweden and Indonesia. It involved 39 participants (14 supervisors and 25 students) from one Swedish and three Indonesian universities. One-on-one semi-structured interviews were conducted and analyzed thematically using Biesta’s functions of education, Habermas’ communicative action theory, and perspectives on academic literacy. Findings Firstly, we found that Swedish and Indonesian supervisors had different feedback provision priorities. Swedish supervisors described prioritizing content-focused feedback to facilitate students’ socialization into academic writing. Conversely, most Indonesian supervisors expressed balancing content- and form-focused feedback with a greater emphasis on qualifying as English teachers. Despite these differences, supervisors in both contexts tended to isolate academic language use from discipline-specific values and practices. Secondly, students in both contexts largely expressed an instrumental orientation to achieving their goals and were frustrated by supervisors phrasing feedback as questions. Many students expressed unfamiliarity with necessary methodologies and theoretical frameworks, which made supervisors’ feedback difficult to decode. Discussion Since only a few of the students viewed the feedback as a support for their process of learning, this study calls for a clear communication about the academic socialization intention through supervision. However, academic socialization cannot solely be the responsibility of supervisors but must be embedded in the curriculum courses.


Introduction
This article is a part of a PhD project that focuses on thesis supervision in English as an Additional Language (EAL) in Swedish and Indonesian universities (see also Nangimah and Walldén, in press). The Swedish and Indonesian contexts were chosen to contribute to the identified need of insight into academic literacy development in non-anglophone EAL writing contexts (e.g., Canagarajah, 2022). Furthermore, in the context of Northern-Southern perspectives, Sweden is associated with the privileged Northern context whereas Indonesia is associated with relatively marginalized Southern context (Pennycook and Makoni, 2020). Hence, we studied Swedish and Indonesian thesis supervision to illuminate supervisory practice in both privileged and marginalized contexts.
The overall goal of the project is to offer understanding how students' academic literacy is developed and socialized through thesis supervision. Our point of departure is that students' communication within academic discourse communities is frequently shown through academic writing, where students need pedagogical support to develop research expertise and communication skills (see Hands and Tucker, 2022) so they can "learn how to think and act" like members of an academic community (Golde, 2010, p. 82). In the present article, we use interview data to shed light on the priorities and experiences students and supervisors express regarding the provision and reception of feedback.
As a form of pedagogical writing support, thesis supervision is expected to help students gradually develop their academic acculturation by providing contextualized instructions related to disciplinary areas (Strauss and Mooney, 2011) and usable feedback that is, for instance, "comprehensible, process-oriented, dialogic" (Vattøy et al., 2021(Vattøy et al., , p. 2332 and "detailed, considerate of affect, and personalized to the student's own work" (Dawson et al., 2018, p. 25). Further, students need knowledge of different discourses for different purposes and competence to communicate ideas in certain social activities and situations (Gee, 2014). Hence, "authorial voice, power relations, identity construction, as well as cross-cultural and crosslinguistic features" (Flowerdew, 2020, p. 588) need to be considered to foster students' socialization. This is particularly visible in supervision situations.
However, students and supervisors may have different priorities and assign different functions to thesis supervision. For instance, they may differ regarding the desired focus of the feedback and expectations of students' independence (Zacharias, 2007;McMartin-Miller, 2014;Dawson et al., 2018). They may also have different perspectives and attitudes toward thesis supervision. Supervisors may see effective supervision as a purposeful coaching based on its various intentions, such as functional project management; enculturation to the academic community; and the development of critical thinking, emancipation (Lee, 2011), students' agency, and problem-solving skills (Wilson and James, 2022). On the other hand, students may have positive and negative perceptions towards supervisor' actions. Misunderstanding might also occur, which may lead to the need for clear guidance and a shared agenda between supervisors and students (Agricola et al., 2021). In certain circumstances, thesis supervision can also make little affordance for students' academic discourse socialization, for example, when it discourages students due to critique, non-verbal comments, lack of participatory and collaborative practices, supervisors' time constraints, and supervisors' lack of priority to develop students' academic literacy (Bastola, 2021).

The existing perspectives on academic literacy
The perspectives on supervision explored in the present study builds on the academic literacies (see Street, 1998, 2006). It expands on other salient perspectives in academic writing: the study skills perspective, which focuses on formal features while paying little attention to context, and the academic socialization perspective, which takes the disciplinary context into account but focuses on teaching specific genres and linguistic patterns, according to an English for Academic Purposes (EAP) approach, rather than on power relations and individual factors among students (also discussed in Wingate and Tribble, 2012). The similar notable isolation of study skills from academic socialization is also associated with the EAL context. Research on EAL students as newcomers in the discourse community mostly relied on a language-based approach despite the sociocultural factors, including identity conflict and transformation, and challenges on authorizing "dialogical selves and voices" (Choi, 2021, p. 538). A study skills approach to researching academic writing is evident in studies focusing on finding effective strategies to develop students' writing competence in the area of coherence and cohesion, lexicon, grammatical range and accuracy (Ebadi and Rahimi, 2019), task response, and lexical resources improvement (Hoang and Hoang, 2022). In relation to Li's (2022) systematic review, 1 these studies call for moving beyond a pure language-based approach to academic literacy in favor of approaches that consider disciplinary and sociocultural dimensions. According to Street (1998, 2006), these dimensions are intertwined in the development of academic literacies.
In the language-based approach, EAL students' struggle with aspects of academic discourse has commonly been misrecognized as language shortcomings regardless of their lack of academic literacy. It has also frequently been seen as a deficit that needs remedy rather than as an academic development phase. The attempt to "fix the error" frequently takes the form of a writing program provision that is separate from the curriculum. Examples include pre-sessional courses for international students in English speaking countries (Hajar, 2020) and out-of-class interaction support (i.e., using network ties to get peer readers and community to share writing difficulties and strategies) for domestic EAL students (Bankier, 2022). While we understand the impulse to overcome students' academic discourse challenges, we agree with Donovan and Erskine-Shaw's (2020) suggestion to move from the deficit-repairment model. Instead, we recommend academic literacy support to create a shared academic literacy and develop a sense of belonging to the community because academic literacy needs academic identity adaptation, which involves affective and intellectual efforts.
From the perspective of academic socialization, students' engagement with academic discourse communities is a complex process of engaging in socially and culturally situated practices. Bailey (2018) points out that academic literacy development does not occur instinctively but frequently needs explicit instruction and intervention to meet situated learning needs. In this process, students need guidance and encouragement. Lillis (2003) recalls the need for a pedagogical practice that facilitates dialogic knowledge production by involving negotiation among students, teachers, and academic institutions. In alignment with Lillis (2003), who demands emphasizing students' awareness of meaning-making among students and teachers/institutions, Hathaway (2015) further underlines the need to address students' writing socialization as a part of the academic development process. Moreover, students' academic socialization needs to be seen as what Li (2022) refers to as "agentive scaffolded learning activities" rather than mere textual problems (p. 20). It requires effective instruction and approaches that consider students' different languages and cultural backgrounds to create meaningful and culturally relevant learning practices. According to Li, academic socialization constitutes the integration of language development and content learning to achieve knowledge production, effective communication, and social transformation.
To clarify the notion of academic literacy used in this study, we refer to Gee's (2014) and Wingate's (2018) academic literacy concept. It follows that we view literacy as a multi-faceted social practice and cognitive activities that are shaped by context, meaningmaking, authority, identity, power relations, and institutional practices. Accordingly, we address academic literacy as the ability to competently communicate ideas in an academic discourse community, which encompasses the skills to read, evaluate, present, debate, and create knowledge through speaking and writing. It covers the ability to apply these skills to relate, interpret, and understand both spoken and written information and communicate it effectively. In the context of thesis supervision, a simple example of academic literacy is the ability to understand the verbal and written information gained from thesis supervision and individual search, create knowledge based on the gained information as a result of the meaning making process, and respond to the feedback by expressing it verbally during the supervision and textually in the form of written draft revision.

Students' academic literacy development in the EAL thesis supervision
Although different socio-cultural situations can bring complexity into the academic socialization process, how students in EAL contexts develop their academic literacy to engage with their academic community through thesis supervision is comparatively underexplored in research. Previous research frequently focuses on international students as newcomers to the academic society in English speaking countries (i.e., the USA, Australia, and Ireland) (Sheridan, 2011;Choi, 2021;Creely, et al., 2021). Ma (2021), who studied different EAL doctoral students' academic literacy in Australia, found that students rely much on L1-L2 translation, lack local learning experience, have different learning expectations, and experience social isolation. Similarly, Elliot, et al. (2016) highlight the importance of "a third space" (an informal space for learning and enjoyment) as a coping mechanism. Whether EAL students who study in home countries and in lower levels of education (undergraduate and postgraduate) experience similar linguistic and socio-cultural challenges is unclear. Furthermore, the cross-cultural perspective has been given little attention regardless of thesis supervision's complex dimensions. Thesis supervision deals with students' challenges of mastering scientific writing (Denis, et al., 2019) and issues such as time constraints, diverse perceptions, technological use, and academic collaboration problems (Zaheer and Munir, 2020). Hence, it needs more than a language-based approach. Moreover, research on how students and supervisors experience EAL thesis supervision as part of students' dynamic academic socialization is scarce despite the identified need for new strategies for academic literacy development to manage multicultural assets with "diverse rhetorical, disciplinary, and communicative contexts" in EAL academic writing and publication (Canagarajah, 2022, p. 18).
The present study contributes to the existing EAL supervision research by focusing on experiences of thesis supervision in English-Medium Programs at Swedish and Indonesian universities. With regard to the aforementioned complexity of thesis supervision, we shed light on how Swedish and Indonesian students and supervisors voice their experiences of thesis supervision as the iterative development of academic literacy and academic socialization. As Yang and Carless (2013) argue, supervision with dialogic feedback provision needs to interplay cognitive scaffolding, affective support, and organizational elements (i.e., feedback timing, sequencing, and modes). Yang and Carless also claim that this supervisory model can enhance supervisors' responsibilities, student-supervisor relationships, and students' engagement and self-regulation. This type of supervision is assumed to be able to develop students' feedback literacy 2 and enable them to engage more with their project, acquire and develop academic literacy, and prepare them to join the academic and professional community. On the contrary, Castanheira et al. (2015), who draw upon the dissimilarity of academic registers across disciplines, question how far thesis supervision enables students to recognize, reflect on, and acquire different genres to develop their academic literacy. Supervision calls attention to the challenge for all students, regardless of linguistic background, to acquire the secondary discourses of higher education and take on new social roles associated with research and academic writing (Gee, 2014). As discussed by Lea and Street (2006) and Wingate and Tribble (2012), this academic literacy development may be restricted by decontextualized and formfocused views of language capabilities.
The present article develops on a prior publication based on the partly the same material (Nangimah and Walldén, in press We employed an appraisal-based discourse analysis to show that thesis supervisors in Sweden and Indonesia described experiencing tensions related to unbalanced relationships with students and colleagues, dealing with various roles, and having feedback priority management with regard to students' instrumental goals and the desired intellectual development. Building further on the interview data and using a different conceptual framework, the present study instead focuses on both supervisors and students' experiences in giving and receiving feedback. Employing an analytical framework combining an academic literacies perspective with Biesta's pedagogical functions and Habermas communication theory (see below), we offer critical insight into thesis supervision as part of students' learning and academic literacy development through the students' voices.

Aim and research questions
We strive to nuance the understanding of supervision as an educational practice based on supervisors' and students' voices with an emphasis on the latter. Our aim is to contribute knowledge of EAL supervisors' and students' priorities and experiences regarding the provision and reception of feedback in two different contexts. Specifically, we aim to investigate the following research questions: 1. What feedback priorities do Swedish and Indonesian supervisors and students describe in thesis supervision? 2. How do Swedish and Indonesian students express tackling the feedback received in thesis supervision?

Conceptual framework
We frame this research in Biesta's (2009) functions of education and Habermas ' (1984) communicative action to understand how thesis supervision as academic discourse socialization is described by students and supervisors. We consider that supervisors and students may have different priorities and strategies to achieve their goals and deal with challenges within situated circumstances. As Gee (2002, pp. 167-168) points out, discrepancies in discourse models can lead to problems, for example, when supervisors and students carry conflicting assumptions about the process of supervision and thesis writing. In addition, thesis supervision does not cover only cognitive but also affective elements that require reciprocal negotiation and knowledge production between both active agents (students and supervisors). Hence, we try to understand how supervision encourages or discourages students' academic literacy development based on their learning orientation.
Through Biesta's (2009) perspective, we see supervision as a form of education with three potential functions: qualification, socialization, and subjectification. Qualification concerns the students' acquisition of skills, knowledge, and dispositions that enable them to do what is expected of them after completing the education. For most of the students participating in the study, this entails teaching English. Socialization involves the students being inserted in established ways of being and doing things. It is considered an unavoidable part of education but can be conducted in more or less explicit ways. In the present study, the "ways of being and doing" mainly relate to conducting research and writing theses on an undergraduate level. Supervision practices constitute a clear example of relative novices in the field learning from a disciplinary expert, which makes the socialization aspect salient. Opposite to socialization, subjectification entails acting in independent ways not restricted by an established order. These concepts are used to explore both the students' and the supervisors' different priorities regarding providing and receiving feedback. The academic literacy perspective foregrounds the latter function since it encourages individual perspectives and critical interrogation of dominant practices, while EAP aligns more with socialization since it emphasizes learning the specific (see Lea and Street, 2006;Wingate and Tribble, 2012). Biesta's concepts have been similarly operationalized in a previous study of literacy practices in a linguistically and culturally diverse contexts (Walldén, 2022).
Habermas ' (1984) communicative action theory foregrounds the important connection between the objective, the social, and the personal in which knowledge is shaped socially and comprehended through communication. Based on Habermas' thinking, an ideal thesis supervision requires deliberative democracy to achieve mutual understanding without coercion. Hence, thesis supervisors are expected to not fully control the students' learning process through supervision, giving students agency to regulate their learning. Habermas fundamentally distinguishes communicative action (having interaction to freely agree on mutual understanding) from strategic action (getting things done). The simple pragmatic manifestation of this distinction can be seen through supervisors' feedback provision priorities and students' reactions to supervision. The strategic aspect of completing educational goals and thus achieving qualification is always present in formal education. However, studies exploring writing practices in educational settings have shown that it may be fruitful to distinguish between literacy practices oriented to completing goals and literacy practices that put a greater emphasis on the processes of seeking understanding and developing new knowledge (e.g., Berge, 1988;Lindh, 2019;Walldén and Lindh, 2021). We use Habermas' communicative action concept to see how students and supervisors describe the feedback provision, which is supposed to be a medium to allow a meaning-making process through negotiation. We acknowledge that mutual understanding of supervision may require not only linguistic competence to clearly articulate ideas but also shared feedback literacies and effective feedback and supervision criteria between students and supervisors. Different perspectives between students and supervisors can create tensions and dissatisfaction. Gee (2014) describes such discrepancies between students and supervisors in terms of contrasting figured worlds or discourse models. Furthermore, we use the strategic action concept to analyze supervisors' priorities and students' reactions in thesis supervision as they are expressed through interviews.
Finally, we use constructs for the academic literacies perspective to explore the supervisors' and students' view of supervision and undergraduate thesis writing as part of academic literacy development. The two constructs education function and communicative action are used to interpret priorities and experiences relating to feedback, as expressed by both students and supervisors. We take an interest in how participants relate to the different views of academic literacy development, involving study skills, academic socialization, and academic literacies. Thus, we are able to consider experiences of supervision in light of both dimensions of academic literacy development and general educational goals.

Research design
This is a qualitative multi-case study design (Yin, 2018) that focuses on supervisors' and students' voiced experiences of feedback provision priorities and reactions in EAL thesis supervision. The first research question (What feedback priorities do Swedish and Indonesian supervisors and students describe in thesis supervision?) focused on supervisors and students as a case; the second research question (How do Swedish and Indonesian students express tackling the feedback received in thesis supervision?) focused on students as a case. The multi-case study approach was useful to shed light on thesis supervision in two different cultural contexts by answering the how and why questions and doing recursive and contextual data analysis (Harland, 2014). This study complies with the European Code of Conduct for Research Integrity 3 and GDPR 4 requirements.

Contexts
The study was conducted at two English-Medium study programs: English Studies and English Language Education in Swedish and Indonesian universities. It investigated the undergraduate thesis supervision that was offered to help final-year students write a thesis in English as their Additional Language. Students wrote their theses individually (all Indonesian students and some Swedish students) or in pairs (Swedish students). Publishing the thesis into peer reviewed journals was expected of the Indonesian students (Kementrian Pendidikan dan Kebudayaan Direktorat Jenderal Pendidikan Tinggi, 2012), but not of the Swedish students (The Swedish Higher Education Act, 1992:1434, n.d., section 6). Further differences are highlighted in the results section.

Recruitment and participants
Participants were recruited by email through faculty research managers. They were obtained by using a convenient sampling (Robinson, 2014) that matched the cases: participants had personal experience in their respective contexts of undergraduate EAL thesis supervision. No incentive was provided for the participants, thus their involvement in this project could be ensured as fully self-regulated and free from influence (Creswell and Creswell, 2018). Prior to the data collection, we gained Indonesian universities' research approval (not needed in the Swedish context) and the participants' written consents (for both Swedish and Indonesian contexts). Supervisors and students were informed that this research investigated their supervisory experience, particularly in giving and receiving feedback. Thirty-nine participants (14 supervisors, five Swedish and nine Indonesian; and 25 students, 10 Swedish and 15 Indonesian) agreed to participate in this research. We refrained from collecting personal information about participant's age. However, the supervisor's professional experience in that role ranged between 2 and 22 years (see the Appendix 1). Regarding the students' age, it can generally be stated that Indonesian students are in their early twenties, since the system require early entry, whereas Swedish student enter tertiary education late in comparison to other countries 5 .
This study involved imbalanced number of participants from one Swedish university and three Indonesian universities due to some participants' withdrawal from our research. Despite the uneven comparison number of participants, this study offered explorative supervisory insights from Northern-Southern contexts without making generalization according to the nature of case study (Yin, 2018). Further details about the participants are presented in Appendix 1, where pseudonyms were assigned to all participants to maintain confidentiality and anonymity. For instance, SU refers to Swedish supervisors, IPTU addresses Indonesian supervisors without a co-supervisory system, and IPCU stands for Indonesian supervisors within a co-supervisory system. Meanwhile, students are represented by using letter A for Swedish context and B for Indonesian context. The number following those letters indicates the order of the interview where it followed first-agreed-first-interviewed system.

Semi-structured interviews
One-on-one recorded semi-structured interviews lasting for 1 hour were conducted by the first author through videoconferencing platforms due to COVID-19 restrictions. Most interviews were conducted in English, but some of them were carried out in Indonesian. The first author conducted interviews in accordance with Galletta's (2013) interview protocol (see Appendix 2). Each interview was initiated by explaining the research purposes, interview procedures, freedom of withdrawal, and participants' confidentiality and anonymity to ensure the participants' full understanding of their research contribution as suggested by Pietrzykowski and Smilowska (2021). Basic questions related to the participants' latest activities were then asked to ease the participants into the interview setting. Guided by the main research questions, the following part of the interview focused on participants' supervisory experience related to their feedback provision priorities (for supervisors and students) and their reactions toward the feedback (for students). During the interviews, the first author acknowledged the participants' agency (Latour, 2005) and the possibility of knowledge reconstruction due to reciprocal influences between interviewer and interviewees in a social interaction (Brinkmann and Kvale, 2018). Thus, the first author strove to adapt flexibly to the participants in the interviews to avoid the Hawthorne effect, where participants modify their responses or behavior because of their awareness of being observed (Franz, 2018). Probing information was also used to clarify the answers given as suggested by Brinkmann and Kvale (2018). Before ending the interviews, the first author asked the participants whether they had any additional information and questions related Frontiers in Education 06 frontiersin.org to the interview. Follow-up questions were sent through email or chat applications (whichever the participants preferred) if confirmation(s) were required after the interviews. Despite the use of interview as a sole research tool, we ensure the data saturation by involving 39 participants (more than the number that Hagaman and Wutich, 2017 suggested to have 6-16 participants), conducting one-hour length interview with freedom for participants to add information in the end of interview for data richness. Furthermore, we used sampling technique with specific criteria rather than random sampling to obtain relevant research participants and intended data as discussed by Mwita (2022).

Data analysis
Following Brinkmann and Kvale (2018), the first author carried out verbatim transcription (with omission of superfluous words such as "eh" and "mmm") to contextualize the conversations, create meaning from the participants' stories, and facilitate the interpretative construction. We then analyzed the interview transcriptions through a thematic approach by following Braun and Clarke's (2006) procedures. The themes were analyzed by drawing inferences based on emerging themes as evidence, and we iteratively double-checked the inferences using the theoretical framework as recommended by Timmermans and Tavory (2012). Based on the emerging themes, the specific part of the excerpts that match the themes gained from Indonesian interviews were then translated into English by the first author, as native Indonesian, for data analysis purposes. To ensure the translation validity, both authors discussed the translated parts, also with reference to contextual information, to arrive at translations accurately expressing the original meaning. We intensively discussed how we approached and interpreted the themes by integrating the expressed stories, contextual background, and theoretical framework to achieve inter-coder agreement and ensure saturated data as suggested by Braun and Clarke (2021). Since this research involved participants from different cultural backgrounds, we paid attention to intercultural elements to avoid misunderstanding. This was facilitated by the authors' backgrounds in Indonesia (Nangimah) and Sweden (Walldén).
For instance, when students' responses showed that they experience challenges in dealing with disciplinary genres, we matched them with theories on academic literacy and relevant previous research. In addition, we employed concepts from Biesta and Habermas (see above) to bring additional insight to the themes we developed regarding supervisors' and students' described experiences. In particular, we considered their priorities and experiences in light of different aspects of academic literacy development. For example, when the participants echoed a study skills approach to developing language capabilities, in which language use was treated quite separately from the development of ideas and context, or an academic socialization perspective, in which the language use and participation in academic literacy practices were viewed as intertwined, we had themed these examples accordingly.
To answer the first research question, concerning feedback priorities, we combined the academic literacy perspective with Biesta's (2009) functions of education (qualification, socialization, and subjectification). We used these concepts to discuss the findings in relation to the more general functions of the supervision and thesis writing as part of an educational practice. As Biesta (2009) himself points out, the three functions are interrelated. For example, independent thinking is not only a matter of subjectification because it is commonly viewed as a desirable educational outcome in higher education (i.e., qualification). It follows that independence is likely to be recognized as valuable and desirable if it aligned with the supervisors' beliefs and the ideals held by the existing order in the community (i.e., socialization). In accordance with the academic literacy perspectives and Gee's (2014) concept of discourse models, supervisors and students in different academic contexts may express different experiences, beliefs, and actions in relation to these functions. Since the present study focuses on supervisors' and students' perspectives, it was important to consider the different agendas the supervisors and students might have had in relation to these functions.
To answer the second research question, concerning students' reaction to the feedback, we applied Habermas' (1984) concepts of strategic and communicative actions and Biesta's (2009) concepts. We used communicative action to understand the meaning-making processes during the feedback provision based on students' responses. Strategic action was used to analyze the students' expressed reactions toward the given feedback, including whether they used the feedback to revise their drafts and how they strategically dealt with the feedback. Biesta's (2009) concepts were used to examine students' reactions to the feedback that indicated their involvement in the educational process. For example, our previous study, which focused on supervisors' beliefs about their roles, showed that supervisors complained over students who emphasized instrumental goals over the possibility of academic growth and intellectual development (Nangimah and Walldén, in press). In other words, supervisors' possibilities to foster students' academic socialization and subjectification seem obstructed by students' orientation to achieve minimum requirements for qualification. Students could be presumed to set their own goals and act strategically in the process of supervision to reach them. Besides, students might not directly emphasize their socialization or subjectification as a priority (RQ1), but their experiences of dealing with feedback (RQ2) might still indicate that they take part in the educational processes.

Results
Based on the interviews with supervisors and students about their priorities and experiences regarding the provision and reception of feedback, we found two main themes: (1) content-versus formfocused feedback provision and (2) the students' reactions to the feedback. The latter theme contained five sub-themes: (1) reacting emotionally and fearing losing face, (2) resisting and avoiding feedback, (3) seeking clarification and learning from questions, (4) defending and arguing for choices, and (5) engaging with unfamiliar methodologies and theories.

Content-versus form-focused feedback provision
In different ways, Swedish and Indonesian supervisors explained that their feedback provision needed to balance a focus on content and a focus on form. The supervisors in both contexts differed as to the Frontiers in Education 07 frontiersin.org degree of responsibility taken for fostering the students' command of written academic language. Their priorities in feedback provision can be seen below.

Supervisors' priorities regarding content/ form-focused feedback provision
Most Swedish supervisors shared the perspective of not foregrounding form-focused feedback provision and giving students responsibility regarding language use. They commented on linguistic correctness and other formal features of the writing generally and selectively, preferably late in the process.
1. On the early draft, I basically talk about the content, the ideas, and the arguments.… I can just sit and think about what actually is written.… So, I try to kind of be a reader, not as a supervisor in that session… in the last draft or towards the end, I kind of start pointing out, "Well, listen. You consistently make subject verb agreement mistakes, and I am not correcting them. " (SU2) The quote is representative of the supervisors in both contexts and demonstrates an overarching wish to emphasize the projects' content and ideas and the students' intellectual development. In the Swedish context, SU5 distinguished themself by admitting that they discussed punctuation rules to facilitate correct usage. However, the supervisor expected the student to use this knowledge independently.
2. I point out problems. I show them something that they do, and then I tell them what the rule is. Take, for instance, the difference between hyphens versus en dashes and em dashes. These are three different punctuation marks, but very many students use them interchangeably. So, I point out that this is not how you are supposed to write in English. (SU5) The excerpts 1 and 2 illustrate that the Swedish supervisors, despite their wish to emphasize the content, had specific expectations of students using grammar and punctuation correctly. While Swedish supervisors expressed concordant expectations regarding students' independence in using correct language, Indonesian supervisors voiced varied focus priorities. Some Indonesian supervisors described focusing on the content, stating that they only provided general, corrective form-focused feedback late in the writing process. This was partly because of the clear division of labor between the main supervisors and co-supervisors. IPCU1 and IPCU2, who worked with co-supervisors, explained that the co-supervisors were responsible for checking the students' use of language and punctuation. Hence, the main supervisors focused only on the content. They "do not discuss the language" or "rarely check the grammatical mistakes. " They "only give general comments on grammar" and "ask students to find the problem and solve it by themselves" (IPCU2). IPCU1 stated that the language check was necessitated only by the perception among the main supervisors that the co-supervisors were not doing their job properly.
Other Indonesian supervisors who worked without co-supervisors (IPTU1 and IPTU3) expressed that they "comment on all drafts in detail, except grammar, and ignore it in the beginning to encourage students to develop their writing" (IPTU1). This aligns with the view of the Swedish supervisors. IPTU3 expressed that it has been necessary to "shift from grammar-focus to knowledge generation-focus" in order to "prepare students to have research skills to adapt the Ministry of Education's new policy regarding research publication. " This brought more focus on "whether the research idea is researchable or not" (IPTU3).
Meanwhile, several Indonesian supervisors who also worked without co-supervisors (IPTU2, IPTU4, and IPTU5) explained that they gave both content-and form-focused feedback. During the interview, they acknowledged that combining both types of feedback in the beginning of the thesis project would make students "feel overwhelmed and devastated due to the great number of mistakes pointed out" (IPTU2). However, they chose to balance both contentand form-focused feedback due to the students' role as English language learners who were expected to be English teachers. This concern was not pointed out by the Swedish supervisors.
3. First, the English Education study program focuses on learning language. Of course, the most important is the language first. Their language should be understandable. But, we cannot separate it from content. It focuses on the language, idea organization, and content. If the mistakes are not too bad, I ignore the language. We can focus on it later. I think I pay attention to both content and language, though the comments on language come later… I usually point out the language just to make them realize that they make mistakes from the beginning… so they can avoid it in the future, as a learning process. (IPTU2) 4. I will focus on the research area … the basic linguistic issue: the language, grammar, word choice, sentence structure, and so on. I think the content would be more valuable. The message that a student has to argue and share. But since most of them are expected to become teachers or lecturers of English, so certainly, they cannot ignore it. The language should be clear and semantically [sic, grammatically] correct, or sometime follows the guideline, be called minimum requirement, subject verb agreement, etc. (IPTU4) 5. I expect students to have good writing skills. It's already at the end of their educational stage. So I can focus on the content. But, it turns out that there are students whose English is not good enough or the arrangement of ideas is still messy. I give feedback on both content and language. (IPTU5) The excerpts 3-5 indicate that the Indonesian supervisors stressed the students' language capabilities while connecting these to their study program (learning English) and future role as teachers. Thus, they viewed the acquisition of basic linguistic skills as a necessary qualification. IPTU4 made a clear division between the "basic linguistic issue" and "content … the message that a student has to argue and share. " IPTU5 also made this division and implied that these aspects were intertwined by mentioning messy arrangement of ideas in conjunction with the students' English not being good enough.
Overall, the tendency to divide between "form" and "content" and to accentuate students' use of grammar and punctuation in relation to the form reflects a study skills view, in which language capabilities are viewed in relative isolation of specific contextual and discursive demands. However, the supervisors in the two contexts emphasized different priorities regarding the goals of the supervision and thesis writing. Some Indonesian supervisors foregrounded the need to Frontiers in Education 08 frontiersin.org consider the students' position as language learners and prepare them for their expected role as English teachers-something which Swedish supervisors did not mention, despite the fact that most of them work in teacher education. It seems that Indonesian thesis supervision puts more emphasis on what Biesta (2009) calls qualification since they stressed the importance of students being prepared for their profession. The Swedish supervisors' greater reluctance to correct the students' language and instead focus on content and ideas indicates that they privilege students' independence. This was connected to both subjectification and socialization. However, as expressed by the supervisors, these goals are relatively disconnected from the students' use of language.
Students' perspectives of content/form-focused feedback provision In terms of content/form-focused feedback, students in both contexts gave similar responses. They were expected to independently figure out the language use, but some of them still expressed receiving form-focused feedback with different degrees of linguistic-focused help. Most Swedish students confirmed that they were expected to take responsibility regarding grammatical matters. Their supervisors focused more on "the ideas clarity and the content development" (A2, A3, and A4). A10 mentioned that supervisors were "very good at pointing out that you need to focus on language yourself. " However, some students expressed that their supervisors still paid much attention to the form. A9 described having "a very long discussion once about whether the comma should be inside a quotation or outside of it [laugh]. " On the contrary, A1, who came up as the outlier of content-and form-focused feedback balance, explained the great involvement of their supervisor in structuring the draft, as exemplified in excerpt 6. Even though the Swedish supervisors, unlike their Indonesian counterparts, barely discussed the students as language learners, the students themselves sometimes stressed this aspect. However, as illustrated by excerpt 6, the students expressed this as a linguistic disadvantage in the process of thesis writing rather than as something significant for their qualification as English teachers. Furthermore, the Swedish students' responses indicated that the supervisors did not necessarily refrain from telling students what to do and showing them how to do it regarding the use of language in students' writing despite their stated content focus. The support A1 expressed in reformulating and restructuring the draft and the reference to "ideas" of improvement rather than corrections transcend the study skills approach to academic language, since these were more associated with the genres and linguistic patterns of academic discourse than the purely formal aspects mentioned by A9. Thus, more functional and contextual aspects of language use might play a part in the academic socialization of the students even though this was not explicitly addressed by the students and supervisors. However, the supervisor's "cluttering" of the students' draft, as A1 mentioned, shows that the supervisor became the students' editor despite their stated resistance to assuming that role.
Unlike the other students in both contexts, two Indonesian students (B14 and B15) were supervised by a pair of supervisors, one main and one co-supervisor. They mentioned that they got sequential feedback with different focuses, as exemplified below.
7. Firstly, we discuss research ideas, content of the writing, and research development with the main supervisor… After it finishes, students will be sent to the co-supervisors to get feedback on the writing, grammar, punctuation, word choices, and thesis structure based on the faculty requirement. (B14) The different supervisory functions for addressing ideas and content, on the one hand, and formal aspects (including following the faculty template for thesis structure), on the other, strongly suggests the study skills approach to language use. Students who were supervised by a single supervisor expressed different responses regarding the feedback content/form focus. For instance, B8 and B10 pointed out that the supervisors focused both on content and form at different writing stages. In the beginning of the writing process, "the supervisor has never problematized language. The important thing is finishing it and focusing on the language later" (B10). This reflected the supervisors' overall stated desire to mainly focus on content early in the project (see previous section). Some students mentioned that they got feedback on the content and had responsibility toward grammatical and punctuation use. Namely, the supervisors only gave straightforward and brief comments on the form, such as "correct the grammar, " and they highlighted the required grammatical revision parts (B5, B11, and B12). This responsibility is also shown in excerpt 8.

The supervisor highlights the draft and comments only
"grammar" without any explanation. It's confusing, but we know that it means grammatical mistakes. (B6) In other words, the students were expected to have the necessary study skills to understand the nature of their mistakes and correct them. In contrast, B7's supervisor seemed to give "the correct grammatical use, " as seen in excerpt 9. 9. The supervisor usually points out the wrong part in the draft like "correct sentence structures. " [Supervisor's name] shows the correct structure and writes it down. (B7) In accordance with the priorities expressed by the supervisors, Swedish students experienced a greater emphasis on content and expectations of taking a greater individual responsibility for language use, whereas the Indonesian students received feedback on both content and form. However, the experiences of the Indonesian students differed with regards to the explicit guidance provided by the supervisors. Overall, the students' responses expressed an awareness of content and form as two quite separate dimensions of feedback. Thus, they confirmed the content/form dichotomy and study skills approach also articulated by the supervisors in both contexts. The division seemed particularly clear to the Indonesian public university students, who were asked to go to different supervisors to get support with content and language. Although the Indonesian supervisors stressed the qualification aspect of their students developing the necessary language capabilities for teaching Frontiers in Education 09 frontiersin.org English, neither the Swedish nor the Indonesian students connected the thesis writing or supervision to their future academic or professional life. Based on the experiences they expressed, the students did not seem to perceive learning new ways of using language and writing as part of a learning process of academic socialization and subjectification.

Students' reactions to the feedback
Both Swedish and Indonesian students expressed difficulties in tackling supervisors' feedback in the form of comments and questions during the supervision. The following sections will show the different concerns and approaches described by the students.

Reacting emotionally and having a fear of losing face
The Swedish students who were allowed to do their degree project in pairs and had individual or in-pair supervision barely mentioned the tension or anxiety of losing face by making mistakes in front of their writing partners or supervisors. However, they expressed concern over comments that were considered harsh and even made them cry. One such instance was when a pair of students received the comment that they were "not being serious" (A3 and A4). Furthermore, A10 referred to a supervisor as "a very good teacher, but not always very good at [sigh] expressing himself " in relation to how the supervisor sometimes phrased feedback in a way that distressed the students. Meanwhile, Indonesian students expressed concern over their question-and-answer sessions during a group supervision that sometimes led to embarrassment. It seemed that a question-andanswer session-which, according to the supervisors, was intended to "allow students to learn from each other" (IPTU1 and IPTU3) and "give feedback to their peers to raise awareness of their own writing" (IPTU4)-became a pressure for students. B2 mentioned that the supervision put pressure on them to "think harder" in order to avoid feeling embarrassment. The experience of pressure and fear of being embarrassed was also indicated by other students: 10. The supervisor usually asks students to comment on others' drafts whether the grammar is correct or not. It is a kind of grammar quiz. (B3) 11. My supervisor asks a lot of questions during supervision. I am afraid of being an object of laughter. There are three students discussing code-switching, including me. I am afraid they will say "How come you do not know this?" It's pressure. (B4) 12. I think I know the answer, but I am afraid that my answer is unacceptable, not really convincing, or my reason is not profound enough. For example, [supervisor's name] asks "Why do you choose this participant?" The honest answer is I have a friend as a pronunciation tutor there who can help me to reach out to my participants, but I cannot give the honest answer. "Is that a simple reason, really?" (B7) 13. During the supervision, it's actually a pressure because we have group supervision… "Do I look dumb in front of my friends?" (B9) Based on excerpts 10-13, supervision became a face-threatening activity due to perceived intimidating questions and comments.
Students regarded questions as a form of evaluation rather than a means to guide the discussion to get further information and help them to develop their drafts. B3 even considered the questions as a "grammar quiz. " Importantly, the stakes during supervision were higher for the Indonesian students since they were graded by their supervisors based on both the process and the product. This might entail a competition between the students in group supervision that fueled the fear of losing face. It was apparent that the anxiety was tied to both the correct use of English language and aspects related to research practice, such as understanding theoretical concepts and proper descriptions of methodology. In summary, the group supervision made the qualification aspect salient for the Indonesian students, regarding both their command of the English language and their academic literacy.

Resisting and avoiding feedback
Both the Swedish and the Indonesian students expressed instances of resisting and avoiding feedback in the form of questions. When they were explicitly asked whether they ignored the feedback, most students gave similar answers: sometimes they ignored it. This feedback avoidance mostly occurred with unclear comments or feedback in the form of questions. Some students decided to "make [the commented part] more concise" or "remove it to avoid more questions" (A9 and B5). Meanwhile, B9 decided to see whether there was something important to add or not and made [the draft] as safe as possible, to prevent "further frightening questions. " This statement related to the student adjusting their view on the notion of "discrimination" in the "analyzed movie for their project" to a mainstream perspective instead of maintaining a contradictory stance. This was a clear example of a student expressing adjusting to the expectations of being socialized into an existing order. In contrast, one of the Swedish students expressed disregarding feedback as a way to maintain their own process and aim for the writing.
14. Usually, the questions were not very helpful to me… my writing process is just barfing up a lot of words and making little notes and making a little list, and I would receive comments on those that did not really correspond with my intention with them because they were not for anyone else to read. They were just placeholders for me. I got a lot of comments on those, and those are essentially deleted right away because this is not relevant to the paper. (A1) As shown in excerpt 14, the student was confident in their agency and writing agenda. When the supervisor gave them feedback that was not in line with the writing purpose, the student decided not to use the feedback and delete the commented parts instead. Thus, the avoidance of further feedback appeared a part of the students' subjectification. Self-confidence in writing was articulated differently by B11. The excerpts 14 and 15 demonstrate the students expressing independence in decision making and presenting themselves as confident writers in an academic context. However, in comparison with the Swedish student A1, B11 seemed more concerned with achieving the qualification by meeting the graduation target at the same time as their friends. In contrast, excerpt 16 shows a wish to rely on the supervisors rather than taking individual responsibility. The student (B12) questioned the value of feedback by requesting specific feedback rather than challenging questions requiring analytical responses and re-consideration of choices in the text. Excerpt 16 also illustrates that compared to the students in 14 and 15, B12 relied on supervisors' spoon-feeding, indicating a mismatch in discourse model between the students' strategic orientation of "getting things done" (Habermas, 1984) and the supervisors' wish for students to engage in dialogue and critical thinking about the projects. Thus, the students seemed to resist being socialized into taking individual responsibility for the research and participating in potential communicative actions seeking a mutual understanding with the supervisor. Even though the writing process was inseparable from "the need of deep thought" due to the nature of knowledge transformation, B12 seemed reluctant to engage in the iterative process of dealing with the content and rhetoric problems characteristic of academic writing (Weigle, 2005).

Seeking clarification and learning from questions
The students in both the Swedish and Indonesian contexts exhibited similar attitudes regarding receiving feedback in the form of questions. In such cases, most students described that they asked for supervisors' clarification before revising their drafts by asking, "Is this what you meant? If that is the case, here is my answer" (A3 and B11); "what do you mean by this?" (B6 and B13); "Why do you want us to change it?" (A5); or "Is it okay to go this way to answer this question?" (B9 and B14). A7 mentioned "ask[ing] more questions" to clarify what the supervisor meant. Apart from asking for clarification, some students' responses illustrated more dependency on supervisors (similar to the student in excerpt 16) by asking the supervisor "How do I do this?" (A1) and pursuing answers from the supervisor: 17. My supervisor gives me so many comments and questions, so I ask "What should I do?" I knew from my seniors that I cannot ask such a question.… My supervisor replies "You're the one who does the research. By mainly asking questions to get further suggestions for revisions, the students seemed to focus more on the end-product than potential learning chances. While this stance echoed that of B12 complaining about the supervisors not providing enough guidance (excerpt 16), B15 expressed more awareness of the quality of independence as part of the discourse model and academic socialization of supervision. Furthermore, the student expressed acting strategically to resist it. The students' emphasis on strategic, rather than communicative, action was clear. Students' responses showing appreciation for being challenged in supervision sessions were rare in both investigated contexts, indicating that the students' discourse model for supervision misaligned with the supervisors' model. Among the students, A2 was the only one who expressed perceiving questions as a means to develop logical reasoning. 18.
[R]eflection questions or food for thought… They may be general or abstract sometimes, but that also brings the opportunity to see it in different angles… a question to generate ideas or to generate a line of reasoning. It can also be used as an inspiration. "OK, there was something that we did not actually think about, that we forgot to mention, that we forgot to add in order to make our… introduction stronger. " So, it could be used to elicit a new concept (A2) Unlike the other students, A2 expressed using the questions from the supervisors to acquire new perspectives and generate ideas about how to improve the text. This shows an orientation to communicative actions not immediately concerned with strategic goals of completing the education.

Defending and arguing for choices
The Swedish and Indonesian students experienced problems in answering questions that set expectations for their academic literacy without always realizing the significance of this difficulty as part of their academic socialization. While A2 was the only student viewing questions from the supervisor as a resource for developing academic literacy, other students recognized the role of the questions in preparing them for the opposition. This illustrates how several of the students expressed that the supervision facilitated the practice of answering analytical questions (A3, A8, B8, B14, and B15). In this context, the students were able to interpret the questions as demands for clarification (A4) or rejection of ideas (A8) as part of their academic socialization. However, many of them voiced having difficulties in articulating their reasons for doing certain things in their research: 20. When you write, you know what you write, "Oh, this is this… this…and this…. " When my supervisor asks, "Why do you use this method?" [sigh] It's hard to think logically and answer the question systematically. (B7) Some questions that created difficulty among students to argue for their choices included the following: "Why do you say that?" (A6); "Where do you want to go with this?" (A8); "How do you use this method?" (B2); "Why do you choose this participant?" (B7); "Why choosing this topic, how to relate da da da, why this, why that" (B8); and "How can that be interpreted? What do the authors from different articles define it as?" (A2). None of the students expressed any thoughts about why these questions were important to answer which Frontiers in Education 11 frontiersin.org indicates that they do not perceive these questions as part of their process of learning and academic socialization.

Engaging with unfamiliar methodologies and theories
The students in both contexts stated that they experienced dealing with unfamiliar research methodologies and theories. This was true both for the students who proposed their own research topic and those who got it from their supervisors. A great frustration was shown in the response by A1, who did research as a part of the supervisor's project: 21. My supervisor gave us the topic to research… [supervisor's pronoun] provided lots of sources, lots of material for us, which was very much needed… [laugh]. We had no idea. We were so lost. It took me weeks to figure out what the hell I was doing. I was so confused. I was so frustrated, but that's I think it ties back to what I was saying previously about not knowing anything going in as opposed to already hav[ing] a foundation to stand on and doing a smaller project. I mean, it's easier, but is it more fun? It is more rewarding [to do research as a part of a supervisor's project]. Perhaps not. I do not know. I cannot say. (A1) As A1's irritation was expressed by cursing, questioning whether their project was more rewarding or not, and mentioning "not knowing anything, " other students experienced confusion due to not knowing how to do a statistical data analysis and how to "find an example on how to do it" (B7). The students also expressed confusion related to the supervisors' expectations of adjusting their research methods to specific research questions and expectations of research novelty. This is evident in the excerpts below (22 and 23). In excerpt 22, the student was not aware that their research was expected to follow the methodology of the specific researcher, Sato, due to similar research questions and replicatory nature of the research. While in 23, the student struggled with the expectation of contributing novelty and filling the existing research gap in the field. These excerpts pointed to a process of academic socialization into an established way of doing things that seemed challenging and perhaps not entirely visible to the students. The difficulties likely pertained to both general knowledge about doing research (filling the gap and conforming to using expected methods) and specific knowledge about the current state of the relevant field. Apart from experiencing unfamiliar research methodology and expectations of novelty, students also described having problems in applying theories relevant to their projects (see excerpts [24][25][26] The above excerpts show that students were unfamiliar with the relevant theories for their research. B6's confusion and anxiety occurred due to discrepancies between the supervisor's and the examiner's expertise, whereas A8 experienced confusion due to the unexpected and overlooked disciplinary concepts. In other words, the students seemed to expect more support from their supervisors in developing their academic literacy and dealing with relevant methods and theoretical perspectives. In excerpt 26, the student struggled with understanding the meaning of a concept the supervisor proposed for the project: CLIL. This showed how the students shared research challenges due to unfamiliar research methodologies, lack of theoretical understanding, and lack of disciplinary literacy. Implicitly, they pointed to obstacles in their socialization into established research procedures and perspectives and a lack of mutual understanding with their supervisors.

Discussion
Feedback provision priorities according to supervisors and students While previous research on EAL students' academic writing acculturation has focused on their linguistic and socio-cultural challenges in Anglophone contexts as their educational host countries Frontiers in Education 12 frontiersin.org (Elliot, et al., 2016;Ma, 2021), this study has explored experiences of thesis supervision as a pedagogical practice for students studying English as their additional language in their home countries. The first research question of this study investigated the feedback priorities in thesis supervision reported by Swedish and Indonesian supervisors and students. The stated feedback priorities show different foci. The Swedish supervisors expressed greater emphasis on the socialization and subjectification processes by prioritizing content-focused feedback provision and, to some extent, disconnecting students' language use from the socialization process. On the contrary, Indonesian supervisors mostly prioritized giving both content-and form-focused feedback during thesis supervision, with greater emphasis on facilitating students' qualification as capable teachers of English. Only a few Indonesian supervisors expressed explicitly viewing thesis supervision as a socialization, where they perceive supervision as a means to help students be published members of the research community. The different priorities expressed by supervisors might relate to students' English proficiency and supervisors' beliefs of their roles. Swedish supervisors may benefit from students' greater English proficiency, compared to Indonesian students, 6 and the availability of a writing center to help students with their use of language. Therefore, this may have influenced Swedish supervisors to give less formfocused feedback. Apart from students' language proficiency and writing support availability, supervisors' feedback priorities seem driven by how they view their roles. Our previous research indicates that Swedish supervisors firmly resist becoming students' editors, whereas some of the Indonesian supervisors view their roles as service providers and pseudo-parents (see Nangimah and Walldén, in press). It follows that most of the Indonesian supervisors in this study provide both content-and form-focused feedback and target qualification functions to prepare students for their future careers; in contrast, the Swedish supervisors focus comparatively more on students' thesis writing development than their engagement with the academic writing community. Despite the integration needs of language, disciplinary-specific values and practices, and sociocultural elements in academic writing (Weigle, 2005), the students' use of language (grammar, punctuation, and diction) in both contexts is rather seen based on study skills perspectives (see Lea and Street, 2006;Wingate and Tribble, 2012), where their language competence is relatively distant from their use of language in contextual and genre-specific ways. The dichotomous content/form-focused feedback priorities are also expressed by students in both contexts. These feedback priorities echo Li's (2022) operationalization of academic literacy according to a language-based approach rather than a sociocultural approach, which views the writing as shaped by social and contextual constraints. Based on the expectation that students master the language competence by themselves, the thesis supervision focuses more on the idea development and disciplinary values. 6 The 2021 English Proficiency Index shows that out of 112 countries examined by Education First, Sweden ranked 8th in English proficiency, whereas Indonesia is in 80th place (Education First, 2021).

Students' reactions to the feedback
Our second research question explored how Swedish and Indonesian students express tackling the feedback received during thesis supervision. The findings show that thesis writing seems to be an overwhelming activity for students in both contexts, especially when they receive feedback in the form of questions. While Yang and Carless (2013) suggest a dialogic provision of feedback that focuses on cognitive scaffolding, affective support, and organizational elements, the present study illustrates that the intended dialogic supervision is not necessarily perceived as helpful by students. Regarding group supervision, Indonesian students seemed unaware that this format for supervision was meant to support their development of academic literacy through learning from peers, evaluating information, understanding (potentially) similar problems, and contributing to providing solutions. Instead, they expressed fearing losing face and experiencing more pressure due to peers' progress. Conversely, saving face was never the Swedish students' concern during the interviews. This may be because the Swedish supervisory system-which allows students to be supervised individually or in-pairs (depending on their project)-does not put the Swedish students at risk of losing face in front of their peers. The different reactions among students in both contexts might also relate to features of collective (Indonesian) versus individualistic (Swedish) culture. 7 Therefore, Swedish students can be presumed to focus on completing their own projects with less concern for how they measure against their peers.
The present study adds to the body of research showing discrepancies between supervisors' and students' discourse models during thesis supervision. While previous studies highlight different perspectives on effective supervision, desired feedback, thesis supervisory expectation, and attitudes (Zacharias, 2007;McMartin-Miller, 2014;Dawson et al., 2018;Wilson and James, 2022), this study specifically points out different perspectives on thesis supervision as part of a process of learning. In our previous study, the Swedish and Indonesian supervisors foregrounded the learning purposes of supervision (Nangimah and Walldén, in press). In contrast, this study indicates that most students in both contexts are product-oriented and focus more on strategic action (Habermas, 1984) to complete their theses and graduate. Instead of using feedback in the shape of critical questions as a form of meaning-making that leads to negotiation to reach mutual understanding with supervisors, most students in both contexts strategically resist and avoid the feedback and disengage from the academic socialization process. Only a few of them consider feedback in the form of questions as fruitful and constructive to contextualizing and developing their writing.
Furthermore, in line with Zacharias' findings (2007), few students consider the provided feedback to contradict their ideas and drive away their writing intention. It seems that feedback in the form of questions that are supposed to be process-oriented and dialogic (Vattøy et al., 2021) brings another challenge for students. This challenge seems to be rooted in a lack of feedback literacy and disciplinary literacy. Moreover, the students' preference for corrective 7 For example, Hofstede's (n.d.) influential cross-cultural framework rates Sweden much higher on individualism than Indonesia (14). https://www. hofstede-insights.com/country-comparison/indonesia,sweden/.

Frontiers in Education
13 frontiersin.org feedback and suggestions to questions and their strategic avoidance of getting further questions, either by deleting the commented part or adjusting views related to the tricky concept (i.e., the notion of discrimination in the evaluated movie), hint at students' lack of skills or motivation to evaluate information, debate; create knowledge; relate to, interpret, and understand the feedback; and effectively communicate their stance. Instead of engaging in critical dialogue, they choose to avoid it. This finding indicates that Flowerdew's (2020) suggestion to develop students' "authorial voice and identity construction" (p. 588) needs to be integrated with students' feedback literacy development. Students' independent authorial voice and constructed identity as an independent writer (e.g., A1) and a selfaware writer (e.g., B11) are not necessarily used to help them to engage with the expected academic socialization. Instead, they maintain a strategic orientation.
The findings relating to students defending their choices and being unfamiliar with methodologies and theories indicate that thesis supervision calls for a specific approach that accommodates specific needs. Some students appreciate questioned-feedback's function to debate and create knowledge (to prepare them for doing thesis opposition), whereas others have difficulties in understanding written language and applying skills and concepts (either methodological or theoretical). The supervisors might expect that the students in the final stages of their education have sufficient research skills, are used to feedback, and have disciplinary literacy as these were introduced in previous courses. However, the students' responses of being unfamiliar with research methodologies, lacking theoretical understanding, and lacking disciplinary literacy indicate that supervisors might be overestimating their students' literacies. These findings show that Lillis' (2003) call for a pedagogic practice that allows dialogic knowledge production and raises students' awareness of meaningmaking between them and teachers/institutions largely remains unanswered in the investigated contexts.
In accordance with Wingate's (2018) view, students' struggles to effectively communicate their arguments and choices through both speaking during supervision and writing in the form of draft revision show that the challenges in EAL students' thesis writing are more than language shortcomings. As Gee (2014) points out, students need literacy to effectively communicate in specific contexts (here, writing their theses, communicating ideas to supervisors, and defending their thesis in front of thesis opponents). This study illustrates that students require academic literacy development by integrating language, disciplinary, and sociocultural approaches. Castanheira et al. 's (2015) question on how far thesis supervision facilitates students to recognize, reflect on, and acquire different genres to develop their academic literacy is complex and cannot be answered based on the present findings. However, the present study supports Donovan and Erskine-Shaw's (2020) suggested priority of raising students' awareness of the supervision goals as a form of iterative academic socialization by creating shared academic literacy and developing a sense of belonging to the community. In addition, raising students' awareness of feedback's actual function as a meaning-making tool is continuously required. Students' emotional aspects need to be seriously taken into account when communicating the feedback since a simple question such as "Are you serious?" can drive students to emotional breakdowns (as reported by A3 and A4). This study also supports Choi's (2021) call for thesis supervision to reconsider students' cultural challenges so they can voice themselves and transform knowledge through the meaning-making of the feedback provision.

Conclusion
The present study has taken a novel approach to studying EAL thesis supervision experiences by using concepts from Habermas and Biesta to discern priorities related to feedback provision and students' reactions to feedback in two different countries. Our hope is that our approach can inspire future research into pedagogical aspects of supervision in different academic contexts. Although students in both investigated contexts react to feedback quite similarly due to their lack of feedback, academic, or disciplinary literacies, it is worth noting that each context has its own uniqueness that calls for viewing supervision as a situated pedagogical practice rather than a general one. A limitation of the study is that it only focuses on students' and supervisors' self-reported experiences. Studying on-going supervision may contribute further insights into specific supervision practices, for example, how the feedback is actually phrased in supervision sessions and responded to in students' revisions of their texts. Furthermore, with a large number of participants, the present study does not relate the findings to the individual biographies of the students and supervisors. A more limited selection would have enabled more fine-grained analyses or portraits of experiences relating to supervision but made it more difficult to draw broader conclusions about feedback priorities and reactions voiced in the two contexts. Also, this research employs exploratory multi-cases study which involves cross cultural contexts. The theory-exemplification model can be used with a more limited set of participants to demonstrate how a certain theory manifests at the individual level and how it contributes to the description, interpretation, or explanation of that person's engagement in a specific educational practice.
Based on the present findings, supervisors need to clearly communicate their intention to support the students' development of ideas and their critical evaluation of their own academic work. These socialization functions were emphasized by the supervisors in both contexts but were not clear to the students, which indicates a discrepancy in discourse models for supervision. Furthermore, it seems necessary to support the students' academic and disciplinary literacy throughout the education program to support their familiarity with necessary research paradigms and methodologies. This cannot be the supervisors' responsibility alone. With academic literacy as a clear strand throughout the programs, exposing students to various ways of writing and conducting research, the possibilities will increase for more students to develop subjectification as part of the learning process. This seemed reserved for just a few students in the present study. Finally, the role of thesis writing as part of the students' qualification needs to be made clearer. Students need to know how and why writing theses relates to their future professional and academic life and to their participation in society at large. To overcome students' strategic orientations to thesis writing and supervisor feedback, we need to address these questions in the process of supporting the students in their thesis writing.

Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement
Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. The patients/participants provided their written informed consent to participate in this study.

Author contributions
MN designed the study, constructed an interview protocol, collected participants' informed consent, interviewed participants in both contexts, transcribed the audio-recorded interview results, performed thematic data coding, analyzed data, and described Habermas' conceptual framework, and wrote the general manuscript. RW contributed to the study's conceptual design with respect to academic literacies and Biesta's concepts, discussed the thematic data reconstruction and data analysis, and reviewed the manuscript. All authors collaborated in discussing data analysis, writing, and revising the manuscript, and approving the submitted version.