INDONESIAN EFL STUDENTS’ THESIS CONCLUSIONS: ORGANIZATIONAL PREFERENCES AND PEDAGOGICAL IMPLICATIONS

This paper examines thesis conclusions written by the students of an English education program at a private university in East Java Indonesia. The central question guiding this study is “How do the authors of the theses organize their conclusions?” This study is a genre analysis. The data sources were 24 undergraduate thesis conclusions written from 2017 to 2019, consisting of 12 quantitative theses and 12 qualitative ones. The analysis was conducted in the framework of genre analysis by identifying the major moves and steps. Bunton’s (2005) framework for the generic structure of conclusions was used to understand the data. The results indicate that the majority of the conclusions follow the generic structure with its moves and steps as suggested by Bunton. However, different preferences in organizing the thesis conclusions were apparent between the two groups of thesis conclusions. These findings reflect the mastery of the particular academic genre by the students as well as their preferences in organizing moves and steps. Pedagogical implications of the findings are discussed in teaching English as a foreign language (TEFL).


Introduction
The thesis conclusions in these are the most important part of the academic works that the readers usually want to read first after knowing the topics. As foreign language learners, Indonesian students who are required to write their thesis in English often have difficulties mastering the genre of thesis conclusions as they are not familiar with the generic structure and the linguistic features. This current study is aimed at understanding the generic structure of thesis conclusions written by Indonesian students. By understanding the generic structures we expect that the findings can shed light on the pedagogical implications of the genre analysis of the thesis conclusions.
In the case of those who are required to write their theses in English, the challenges faced by the students are two-fold as they need to master the target language's written forms as well to master the discourse genre of the texts. Hopkins and Dudley-Evans (1988) suggested that students be able to produce texts that can perform "the function for which it was intended and that it is acceptable by the community for whom/within which it was produced" (p. 113). Therefore, for English language learners the first challenge is mastering the target language forms that include vocabulary and grammar. The second challenge is mastering the target discourse genre which may be different from the cultural tradition of the learners.
In this line of thought, this paper is aimed at exploring the product of academic writing, namely the thesis conclusions written by the students at an English education program of a private university in East Java Indonesia. We are interested in thesis conclusions because as the final section of academic writing, it is important in providing final remarks about the whole work discussed (Paltridge & Starfield, 2007). The conclusion, then, is the most likely part that will get the readers' attention and be read before the readers go to the more detailed sections in the body of the thesis.
Previous studies related to the academic genre are often framed in the terms of move analysis, which is part of discourse analysis to understand the generic structure of a particular genre (Flowerdew, 2015;Flowerdew & Forest, 2009;Hopkins & Dudley-Evans, 1988;Maher & Milligan, 2019;Paré, 2019;Starfield & Paltridge, 2019). From this perspective, academic texts can be seen as the product of communicative events which are intended to engage in the academic community discourse on the corresponding topic. Therefore, academic texts can be analyzed descriptively by looking at (1) the writers' messages, (2) the logical organization, and (3) the presentation to the intended audience (Hopkins & Dudley-Evans, 1988, p. 114). The moves to describe the structure of an academic genre, then, should be relevant to the three elements of the written discourse.
As this paper focuses on exploring the thesis conclusions, we are interested in finding out how they are organized in terms of the organizational moves of the academic genre. In particular, we want to know whether the thesis conclusions could reflect the students' mastery of the academic genre. Our interest, then, is beyond identifying and describing the generic structure of the thesis conclusion but also offering the pedagogical implications of the generic structure in the context of English language teaching in Indonesia.

Literature Review The genre analysis of thesis conclusions
This current study follows the tradition of genre analysis which has gained momentum since the early 1980s (Askehave & Swales, 2001;Swales & Najjar, 1987). The genre analysis movement has widened our understanding of genre membership identification and its taxonomy which brings us to understand various types of texts, from shopping lists to recipes, and various academic texts including research reports in academic journals and full-length theses. This genre analysis has been conducted in the framework of discourse analysis. It means that texts are understood as products of a discourse community and texts are created for certain communicative purposes.
From the literacy perspective, the genre of academic writing can be seen as language pedagogical practices with an emphasis on mastering the academic discourse and claiming membership in the academic community. In gaining the academic writing competency "Peraturan Pemerintah Republik Indonesia Nomor 19 Tahun 2017 tentang Perubahan atas Peraturan Pemerintan Nomor 74 Tahun 2008 tentang Guru" 2017) outline three conceptual approaches that can shed light on the nature of academic writing. The first approach is the study-skill model in which academic writing is viewed as a personal and cognitive skill. In this model, the focus is on the mastery of language features and forms. The cultural aspects and the generic structure of the academic genre are considered transferrable without explicit teaching. The second approach is academic socialization which is "concerned with students' acculturation into disciplinary and subject-based discourses and genres" ("Peraturan Pemerintah Republik Indonesia Nomor 19 Tahun 2017 tentang Perubahan atas Peraturan Pemerintan Nomor 74 Tahun 2008 tentang Guru," p. 369). Using this approach, the focus is the engagement with the discourse community to become competent in producing texts which can be understood and accepted by the academic community. Finally, the last approach is the academic literacy model, which focuses on gaining full membership in the academic discourse community. Related to academic texts, this third model "is concerned with meaning-making, identity, power, and authority, and foregrounds the institutional nature of what counts as knowledge in any particular academic context" ("Peraturan Pemerintah Republik Indonesia Nomor 19 Tahun 2017 tentang Perubahan atas Peraturan Pemerintan Nomor 74 Tahun 2008 tentang Guru," p. 369).
The three approaches are not exclusive models with clear boundaries but they are overlapping. They can be used for various pedagogical practices, reflecting the continuum of literacy levels in an academic context. Consequently, this model is more pedagogical and it provides a beneficial framework to understand non-native speakers' academic writing products. This conception echoes Bhatia's (2002) continuum that sees language, from text to genres and social practices. The first model is highly related to language mastery-the mastery of the grammatical features and relevant vocabulary of the academic genre. The second model is related to the understanding and mastery of the generic structure of the particular genre. Finally, the last model drives the learners to claim their full membership in the academic community, act out their identity as competent members, and negotiate meanings in the process of knowledge production and dissemination.
One of the important products of the academic community is a full lengththesis or dissertation, which is written to meet the requirements of undergraduate or graduate programs. At the graduate level, thesis and dissertation writing is considered important "as the rite of passage to an academic career, required by universities around the world" (Bunton, 2001, p. 57). Less strictly at the undergraduate level, thesis writing may not be the major requirement to graduate. In the context of undergraduate students in Indonesia, thesis writing in English is also meant to show the mastery of written English-indicating the expanding mastery of literacy among students majoring in English studies and English education.
Genre studies related to academic writing have been conducted by looking at various parts of the theses and journal articles as different sub-genres. Lamb and Wyatt (2019), for example, conducted a study on the writing of research articles (RA's) introductions by identifying various moves used by authors to organize their arguments and to build the case for their research. A move in genre analysis can be defined broadly "as a segment of text that is shaped and constrained by a specific communicative function" (Holmes, 1997, p. 325). The structure of a genre, then, can be distinguished by looking at the moves and how they occurred in the particular genre. Following Swales' works, Hopkins and Dudley-Evans (1988) investigated the discussion sections of RAs and dissertations and they found that the structure was more cyclical-not linear as suggested by Lamb and Wyatt (2019). They identified that there was only one obligatory move, the Statement of Results, which they called the Head Move.
Similar studies on exploring parts of the academic genre by looking at moves as the framework of analysis have been conducted by (Bunton (2001(Bunton ( ), 2005) and Grgurović, Chapelle, and Shelley (2013). The first study by Bunton (2001) focuses on PhD thesis introductions and he identified three major moves of thesis introduction: (1) establishing a territory, (2) establishing a niche, and (3) announcing the present research (occupying a niche). The latter study conducted by Bunton (2005) focuses on thesis conclusions. For Bunton, thesis conclusions are special as they usually can be found in a separate part from the discussion section. It is different from the discussion section in RAs, which usually include the conclusion as part of the discussion (Grgurović et al., 2013). In RAs' discussions, the generic structure can include up to seven moves: (1) background information, (2) reporting results, (3) summarizing results, (4) commenting on results, (5) summarizing the study, (6) evaluating the study, and (7) deductions from the research. In this model of a generic structure, the last three moves are the conclusion part. This echoes Yang and Allison's (2003) study of conclusion sections in RAs. They identified three major moves in their generic hierarchical structure: (1) summarizing the study, (2) evaluating the study, (3) deduction from the research four moves in the separate pedagogical implications.
Using the available framework identified by Yang and Allison, Bunton (2005) investigated 45 PhD thesis conclusions from various disciplines. From the conclusion parts with varied titles, Bunton identified six major roles of the conclusions: (1) summarizing the findings, (2) discussing the findings in broad terms, (3) coming to conclusions, (4) giving implications, (5) making recommendations, and (6) providing suggestions for future research. Upon closer analysis of moves, he also found that thesis conclusions related to science and technologies seem to be more straightforward with four identifiable moves: (1) introductory statement, (2) consolidation of research, (3) practical applications and recommendations, and (4) future research. On the other hand, thesis conclusions related to humanities and social sciences have five identifiable moves: (1) introductory restatement, (2) consolidation of research space, (3) practical implications, (4) future research, and (5) concluding restatement.
While Bunton (2005) distinguished the science and technology theses from humanities and social sciences, he did not distinguish between qualitative and quantitative theses, which in our consideration was the major drawback in his analysis of thesis conclusions. While subject areas and disciplines are important in a genre as they can mark various registers, the nature of research approaches is also important in determining the genre and its generic structure. Flowerdew and Forest (2009) inform us of the importance of research approaches in the academic genre of theses. When they searched for the collocations of the keyword research, they found that it collocates with words related to disciplines and with words related to approaches or methods. For the discipline of applied linguistics, for example, they found in their corpus data that research often collocates with the names of the specific sub-disciplines such as writing research, reading research, educational research, linguistic research, contrastive rhetoric research, SLA research, and L2 research. On approaches, the word research collocates with specific approaches or methods such as qualitative research, questionnaire-based research, action research, empirical research, experimental research, naturalistic research, test-based research, observationbased research, and language research.
What is suggested by Flowerdew and Forest (2009) regarding the importance of approaches and methods in determining the sub-genres in academic discourse is moving forward from Bhatia's (2002) conception of academic genre which focuses on the disciplinary culture. He realizes that the academic genre is dynamic and flexible in addressing changes and trends but he hardly discussed the importance of approaches and methods that can also influence the generic integrity of an academic genre. In addition, there is variability of structures within a discipline in academic discourse and there is little research on such variability (Holmes, 1997). This lack of attention to the variability within a discipline in the context of academic discourse can be understood as researchers often focus more on differences among different disciplines, the registers.
In this current study, under one discipline of language education, we distinguished undergraduate theses by looking at the approaches to the study, the quantitative and qualitative approaches. As suggested by (Elshadelin and Yumarnamto (2020); Gelo, Braakmann, and Benetka (2008); Yumarnamto, Widyaningrum, and Prijambodo (2020)), the distinction is not only on how data are collected and analyzed but also the philosophical foundations underlying both approaches. The quantitative approach is based on the positivists' perspective and the process of knowledging based on the objectivity of the observers. On the contrary, the qualitative approach is based on the acknowledgement of subjectivity and intersubjectivity in the process of knowledging. As we focus on one disciplinary subject, English language education, the distinction between the two approaches in looking at the integrity of the generic structure will be beneficial in understanding the dynamic and flexibility of the generic structure in thesis conclusions.

Method
Following the tradition of genre analysis (Askehave & Swales, 2001;Bazerman, Bonini, & Figueiredo, 2009;Dong & Lu, 2020;Hyland, 2015Hyland, , 2016Maher & Milligan, 2019;Starfield & Paltridge, 2019), this study explores thesis conclusions written by undergraduate students as a part of the requirements for their study completion. The students belonged to an English education program at a private university in East Java, Indonesia. They have trained in English, English literature, pedagogy, and applied linguistics to prepare them to become English teachers. By the time they wrote their theses, typically in the final year of their education, they were fluent in English and were familiar with various academic genres, both spoken and written. They had taken the academic writing course and the research method course, which introduced both quantitative and qualitative research methodology. They have also introduced the format and the academic convention of how to write a good thesis. In the courses they previously took, they were also exposed to other students' theses and academic articles that they needed to read to support their research. Therefore, the study they conducted and their thesis writing was their first effort to engage more meaningfully with academic discourses of research writing. Their work, then, can be characterized as the works of novice apprentices in the academic community.
For this study, the sources of data were 24 thesis conclusions written by the students of the English education program from 2017 to 2019. The reason why the writer only selected theses from 2017, 2018, and 2019 is that they were the most recent ones uploaded to the university's repository. In the repository, we found 46 theses completed in 2017, 42 in 2018 and 30 in 2019. From the theses available, we made a purposive selection by considering whether the theses were quantitative or qualitative in their approach to their study. The quantitative theses were ones that involve statistical analysis and testing hypotheses. On the other hand, qualitative theses did not involve any statistical testing of hypotheses and were more descriptive. However, qualitative theses might include descriptive statistics such as counting the number of occurrences and presenting them in terms of percentage. While data selection was purposive, we applied two main criteria in the selection: (1) the theses were completed in 2017, 2018, and 2019; (2) four qualitative theses and four quantitative ones were selected randomly from a list of theses completed in each year.
From the data sources, we identified thesis conclusions by looking at the conclusion section. The conclusion sections had various titles such as Conclusions, Conclusions and Recommendations, and Concluding Remarks. We read carefully each conclusion section and analyze it in the framework of Bunton's (2005) thesis conclusions that include three major moves: (1) introductory statement, (2) consolidation of the research space, and (3) recommendations and implications. Two minor moves in Bunton, future research and concluding restatement, become our reference but they were not central in our analysis as they can be included in the three major moves. We made this decision because, in Bunton's original research, the two minor moves occurred rarely (less than 25%).
In identifying the first major move, the introductory statement, there were four sub-moves or steps that could lead us to mark the boundary of the move. They were (1) restatement of the issue being researched, (2) work carried out, (3) purpose of the study, and (4) research questions or hypotheses. For the second major move, the consolidation of the research space, we used two steps, namely (1) summary and evaluation of methods and (2) summary of results/findings and claims. Finally, for the third major move, we applied three steps to indicate the move area, (1) future research, (2) practical applications, and (3) limitations of the study.
Using the three major moves and their steps, we analyzed each thesis conclusion, checked the occurrences of each step and present them in percentage. Further analysis was conducted by interpreting the findings by considering both the qualitative and quantitative thesis conclusions.

Findings
Our investigation of the generic structure of thesis conclusions written by EFL students in their undergraduate thesis was motivated by the assumptions that EFL learners may prefer different moves and steps in organizing their thesis conclusions. To benchmark the EFL the generic structure of EFL writers, we use Bunton's (2005) framework of the thesis conclusion generic structure preferred in social studies/humanities. We distinguished the students' theses based on the approach used in the research. Therefore, there are two groups of theses we observed, the qualitative thesis group and the quantitative one.
The result of the study shows that out of 24 thesis conclusions analyzed (12 qualitative studies and 12 quantitative studies), there are varied preferences in organizing the thesis conclusions. More interestingly, the organizing patterns between the two groups indicate differences in the use of moves and steps in organizing their conclusions as shown in Table 2 below. In the introductory statement move, it appears most of the quantitative theses prefer all the steps in organizing their statements. Over half of the quantitative theses use the first three steps in this first move. In contrast, less than half of the qualitative theses use each step in organizing their introductory statements. Only in the last step do both groups of these have the same portion. Two theses (16.7%) restated their research questions or hypotheses in each group. Therefore, it is apparent that the first three steps in this move are central to organizing introductory statements for quantitative conclusions. In contrast, the introductory statements in the qualitative thesis conclusion seem to have no fixed pattern.
For the second move, the consolidation of the research space both groups of theses show a similar pattern. Only 1 (8.3%) out of 12 quantitative theses state their summary and evaluation of the method and none of the qualitative ones state this step. Similarly, 11 (91.7%) out of 12 quantitative theses put their summary of the results or findings and all the qualitative ones (100%) put their summary in this move. Therefore, it is clear that in consolidating of research space, both quantitative and qualitative thesis conclusions are very similar in focusing on the summary of results or findings and claims.
Finally, for the third movement, the two groups of thesis conclusions make statements to recommend future research as the consequence of the findings. All 24 (100%) conclusions make this kind of statement. In the second step of this move, all quantitative thesis conclusions (100%) make statements to suggest practical applications of their findings. In contrast, 7 (58.3%) of qualitative thesis conclusions make similar claims. In this way, the recommendations and implications in quantitative thesis conclusions focus on the first two steps, future research and practical applications while qualitative thesis conclusion highlight more on the first and the third steps, future research and limitations of the study.
Overall, the quantitative thesis conclusions seem to be more consistent in using all the steps in each move compared to the qualitative one. This pattern indicates that quantitative thesis conclusions may have a more structured format, which is consistently followed by the writers. On the other hand, the qualitative thesis conclusions seem to have a less uniform format compared to the quantitative ones so only the main steps are consistently used to organize the conclusions. The different preferences in using the move and steps in thesis conclusions can be attributed to the different preferences of the authors and, possibly, the lack of understanding of the generic structure.

Move 1: Introductory statement
An introductory statement in a thesis conclusion is a restatement of the issue being researched (Bunton, 2005). To exemplify the first move in thesis conclusions, Table 2 provides instances for introductory statements found in both quantitative and qualitative thesis conclusions. The way the authors write the introductory statement varied, especially in the first and the second step, the restatement of the issue and the work carried out. The fact that authors composed these two steps in different ways may be influenced by the fact that the thesis conclusion began from different topics and issues. In the third step, the purpose of the study, a similar construction of expressing purpose is apparent. In the examples shown in Table 2, the expression "The study was conducted to find out ..." is a straightforward way to communicate the purpose of the study. In the last step, the restatement of research questions or hypotheses both quantitative and qualitative conclusions use a similar structure. As shown in the last part of Table 3, restating questions or hypotheses can consist of two parts, the introductory part and the restatement part.

Move 2: Consolidation of research space
The consolidation of research space moves includes the restatement of methods, findings or results, claims, and references to previous research to support the claims. In this move, there are two major steps: the summary and evaluation of research methods, and the summary of the results or findings and the claims. We found that in the first step, the entire qualitative thesis conclusions did not include the summary and evaluation of the methods. A must-have step in this move is the summary of the results or findings and the claims as the consequences of the findings. Examples of Move 2 can be seen in Table 3 below. (1) Summary and evaluation of methods Picture series facilitate students to get and develop ideas according to the sequences seen in the pictures. Picture series stimulate the students to add a better detail to their stories by looking at other object drawn in the picture to make their story better. Based on the analysis conducted some conclusion can be drawn. First, the results of t-test quasiexperimental method with two group pre-test and post-test design showed that the hypothesis (H1) was accepted … their friends in group. Third, collaborative writing is a bridge for the students to collaborate and contribute among themselves.
Fernandes (2017, p.79) From this study which has been conducted, the researcher found that the four maxims were violated for 57 times in total. Statistically, the frequency of each maxim's violation was illustrated as follows … the speaker's utterances. However, both speakers and hearers who were involved in the conversation were found to be oblivious towards the violation occurred during their conversation. (Qualitative) Monika (2017, p.39) The summary and evaluation of the research methods seem not to be part of the qualitative thesis conclusion. In this move, the qualitative thesis conclusions focus more on the summary of findings and the claims they made based on their findings. For a quantitative thesis conclusion, however, looking at the methods and providing an overview of them and relating them with the findings seem to be important.

Move 3: Recommendations and implications
Move 3, the recommendations and implications include the recommendation for future research, the practical application of the findings and the limitation of the study. The steps directly reflect the different parts of the move. As found in the previous section, the limitation part of this move is a minority in quantitative thesis conclusions while other parts are more commonly found in this move. In order to get the students immersed in the learning activity, it is better to choose the type of Jigsaw following the ability of the students in classroom … far from perfect. Therefore, the writer hopes that the other researchers could conduct a better study and give more meaningful contributions. Step 1 in this move is the recommendation for future research. This step is very important in thesis conclusions because it suggests the direction of future research as well as provides clues about the current research gap. In short, the recommendation for future research will expand the research community's understanding of the subject areas being investigated. This is the process of continuing curiosity to get a better understanding of the areas.

Discussions
The findings of this study confirm Bunton (2005) generic structure of thesis conclusions. The moves and steps in thesis conclusions in the areas of social sciences and humanity are reflected in the thesis conclusions written by Indonesian students. As a caveat, however, the preferences varied between the two approaches used in the discipline of applied linguistics, confirming the existence of sub-genres within the discipline. From this overall finding, three major implications can be drawn.

The importance of research approaches within a discipline
Research approaches within a discipline are important in distinguishing subgenres. Our decision to distinguish two different approaches in research within a discipline has shed light on different preferences in organizing thesis conclusions. Quantitative research is based on the assumptions of objectivity of the subject being studied and the researchers' roles are the neutral observers of the phenomenon (Elshadelin & Yumarnamto, 2020;Gelo et al., 2008;. On the other hand, qualitative studies are based on the acknowledgement that subjectivity may influence the subject of the study and from this lens researchers are not neutral observers of the phenomenon. The acknowledgement of subjectivity and intersubjectivity is important in generating knowledge in qualitative studies. The positivists' perspective and post-positivists' lenses can influence the generic structure and the rhetorical organization of the academic genre. Consequently, teaching academic writing should consider not only the topic areas but also the approaches the researchers investigate on different topic areas. The sub-genres within a discipline show different rhetorical organizations and different generic structures as reflected in the preference of the writer in quantitative and qualitative thesis conclusions in the discipline of applied linguistics in Indonesia. In Bhatia's (2002) terms, a genre can display "tensions of various kinds" (p. 31) and our findings show such tensions as suggested by different preferences in the use of moves and steps in organizing the conclusions.
This finding implies the importance of sub-genres within disciplines. Unfortunately, sub-genres within a discipline are often ignored as researchers focus more on investigating differences in genres among various disciplines. In Hyland's (2015) terms, the community practices in the discipline enable the reproduction of certain genres. In this way, Hyland suggests the importance of community practices within a discipline that determines the generic structure and grammatical conventions acceptable to the community. The community of a discipline, therefore, is expected to master various sub-genres within the discipline, which is often difficult considering the different ideologies and philosophical foundations in quantitative and qualitative research. The finding suggests more attention should be put forward to investigating genres by looking at the two opposite approaches to research.

The micro-social practices of thesis writing in the academic community
For EFL learners writing these may reflect micro-social practices in the academic community. The context of our study was in an EFL environment where English was a foreign language and the thesis writers were in the process of completing their studies to become English teachers in Indonesia. This context was unique in that the authors of the theses were still in apprenticeship in the disciplinary community. The different preferences in organizing thesis conclusions and the array of variations in the conclusions may suggest the existence of a sub-academic community of learning in the contexts of the study program. Therefore, membership in the sub-academic community could be confined to professors, lecturers, and students within the study program, and to some extent, the wider academic community found in journal articles and books. This is an intermediary community with more structured roles. In this community, professors and lecturers provide resources and models. Meanwhile, students can engage in learning as apprentices so that they can claim their full membership in a wider community of academic practices. This conception of such community as resources of learning echoes Hyland (2015, p. 2) when he wrote "... genres are community resources which allow users to create and read texts with some assurance that they know what they are dealing with." This sub-academic community on campus indeed can be a resource for the learners. While writing their theses, they consulted and discussed with their advisors as well as looked at previous theses they found in the library and online in the repository.
So far, this sub-academic community within a discipline with its unique practices seems to have been ignored as researchers focus more on the academic community within a discipline in a wider sense. Further research on sub-academic communities where novice members practices their apprenticeship and full members as resources is worthy of future investigation. How the novice members interact with the full members to create a specific academic culture? How the academic culture in the micro academic community is maintained in academic interactions? These questions could contribute to a better understanding of the micro-academic community.

Writing thesis as literacy practices and apprenticeship in the academic community
Teaching academic writing in EFL contexts should be focused on literacy practices and should be framed in terms of apprenticeship in the academic community. Our study was situated in an educational program that prepared and trained future English teachers in Indonesia. The findings suggest different preferences at the level of the two thesis groups as well as individual differences. The nature of differences can be interpreted at least from two perspectives: (1) real individual preference and (2) lack of understanding of the generic structures of thesis conclusions. Therefore, the pedagogical implication from this interpretation of the finding, especially the second one can be addressed by looking at how academic writing was taught. The traditional way of teaching academic writing that emphasizes language mastery may not be enough. It should be focused on literacy and framed in terms of apprenticeship in the academic community (Bhatia, 2002;Hyland & Wong, 2013;Richards & Pilcher, 2017;Xu, 2019). Therefore, the pedagogy of academic writing in the EFL context should move from language-oriented pedagogy to literacy-oriented pedagogy. While the former focuses on language mastery, the latter focuses on literacy practices in which learners engage in the academic community as apprentices.
In these academic practices, learners, facilitated by mentors engage meaningfully in the academic community. Practices and engagement in the academic community are central for the learners in terms of the mastery of the craft as suggested by Valtonen et al. (2017). We also agree with Hyland (2015) when he wrote: To work in a discipline, then, we need to be able to engage in these practices and, in particular, in its discourses. We need to proximate to the rhetorical conventions it routinely employs to claim membership and learn how to use these conventions to take positions on matters the community values. (p. 3) The role of the teachers of academic writing, then, should be more of mentors that help the apprentices to claim their full membership in the academic community. The way to do it is by engaging them with real academic literacy practices as suggested by Tardy (2011) when she suggested that writing pedagogy should be "grounded in the belief that helping students to demystify socially situated writing can facilitate the learning of privileged forms of discourse" (p. 2). In this way, the implications of the findings of our inquiry are in line with the shift of writing pedagogy, from language to genre, and literacy practices.

Conclusion
Our inquiry on thesis conclusions written by Indonesian EFL learners began with our curiosity about the generic organization marked by moves and steps in thesis conclusions. We distinguished the disciplinary area of applied linguistics into two groups by looking at the approaches: quantitative and qualitative research. The distinction between the two groups suggests the existence of subgenres within the thesis, one form of the academic genre in higher education. These sub-genres deserve more attention as they are often overlooked because researchers and writing teachers focus more on disciplinary differences-not the sub-genres within a discipline.
We also found that different preferences can be attributed to each group. The different preferences can be interpreted from two perspectives: individual preferences and the lack of understanding of the generic structure. This finding brings us to the pedagogical implication that academic writing should be treated as an apprenticeship to the academic community-shifting the focus from the mastery of language forms and features to a meaningful engagement with the academic community.