Perceptions of Indonesian Multilingual Scholars about Preparing and Publishing Research Manuscripts in International Journals

There has been a great deal of interest in issues multilingual scholars (henceforth: MLS) have in trying to gain publication in international mainstream English language journals (IEJs). However, little research has been published on the experience of MLS using their perspective, particularly how they perceive their competence (knowledge and skills) to publish their research internationally. The purpose of this study was to fill this gap by investigating what they perceived as the factors that inhibited them from publishing in IEJs. This study mainly used a quantitative method, but the results were supplemented with interviews and focus group discussion with some of the participants in the survey. Principal Component Analysis (PCA) tool of the SPSS statistical programme was used to conduct two levels of analysis: the PCA and Confirmation Factorial Analysis (CFA). The key finding was that the most critical factor for the participants was not a lack of funds as widely reported in many previous studies, but lack of competence to face the challenges of preparing and publishing research article (RA) in IEJs, which had dampened self-confidence. The low self-confidence and two solutions, * Corresponding author, email: zadnan@une.edu.au Citation in APA style: Adnan, Z., Arsyad, S., Purwo, B. K., & Sukamto, K. E. (2021). Perceptions of Indonesian multilingual scholars about preparing and publishing research manuscripts in international journals. Studies in English Language and Education, 8(1), 65-83. Received September 30, 2020; Revised December 16, 2020; Accepted December 22, 2020; Published Online January 3, 2021 https://doi.org/10.24815/siele.v8i1.18969 Z. Adnan, S. Arsyad, B. K. Purwo & K.E. Sukamto, Perceptions of Indonesian multilingual scholars about preparing and publishing research manuscripts in international journals | 66 suggested by the participants, implicitly confirmed the key finding. The participants were aware of the various benefits of research article publication in IEJs, but they were not strong enough to overcome the critical factor. The implication is that policymakers should consider providing regular training for staff with adequate practice and feedback and introducing the subject to undergraduate or, at least, MA students before they embark on starting their research projects.


INTRODUCTION
A great deal of attention has been paid to issues experienced by multilingual scholars (MLS) coming from the 'periphery of knowledge production' [henceforth 'the Periphery') in trying to gain publication of their research in international main-stream English language journals (henceforth: IEJs), known as 'the centre of knowledge construction' (henceforth the Centre), which suggests that the issues are of high significance. So important is the topic, a group of concerned scholars have established a conference called PRISEAL (Publishing and Publishing Research Internationally: Issues for Speakers of English as an Additional Language) held every two years.
However, little research has been published on the experience of Indonesian MLS using their own perspective, particularly how they perceive their competence (knowledge and skills) to publish their research internationally. The purpose of this study was to fill this gap by investigating the factors that inhibited them from publishing in IEJs. While previous studies have been mostly qualitative, on the contrary, this study mainly used a quantitative method. However, the results were supplemented with interviews and focus group discussions involving some of the participants in the survey and some other invited academics.
Further, no study employs a quantitative method to analyse the survey data. This present study fills this gap by involving scholars from four universities outside Java and one in Java, involving many participants relative to other studies. This research is also unique in that it used factorial analyses, which are hardly found amongst other studies reviewed earlier, and it illuminated the results using supplementary information gathered from interviews and focus group discussions (FGD).
The purpose of this study was to investigate the perceived reasons or factors influencing the low number of publications from the perspective of the players, the academics themselves, and the possible solutions so that the publication number can be increased. Hence, the central question was, 'What are the issues perceived by Indonesian MLSs which explain why they have not published in IEJs?' This study provided the following specific questions to supplement the central question: • What are the key factors or issues which inhibit Indonesian MLSs from publishing in IEJs? • To what extent are the critical issues reported in the literature confirmed or otherwise? Answers to these questions are significant for our understanding of why academics in Indonesia do not publish. Practically, they are also important for 67 | Studies in English Language and Education, 8(1), 65-83, 2021 policymakers as they can provide a reliable source of information for the government and universities as a guide to developing more appropriate policies and regulations. They may also be useful for other parties including governments of other nations who are keen to increase their research publication in IEJs.

LITERATURE REVIEW
Following the study conducted by St. John (1987), Uzuner's (2008) review alone cites 39 research articles (henceforth: RAs) which met her criteria, which include that it must be an empirical study, meaning some other related studies are not included. After the publication of this review (2008), there have been many more studies reported focusing on different countries, for example, Mainland China (Mu, 2019), Taiwan (Chien, 2019), India (Lahiri, 2011), Iran (Maniati & Jalilifar, 2018), Poland (Lopaciuk-Gonczaryk, 2016) and Sudan (Elmalik & Nesi, 2008). Due to limited space, it is impossible to review them all. Therefore, only the most relevant studies are surveyed. Uzuner (2008) found that of those 39 studies, the problems faced by MLS scholars in attempting to gain publication in IEJs have included English language problems, parochialism (failure to set a study in the context of international literature), failure to follow the 'accepted norms of research reporting', 'consuming and tedious nature of writing for publication in English', 'lack of connection with members of the core academic communities', 'potential bias against multilingual scholars' submission', and insufficient funds to conduct research (Man et al., 2004). Salager-Meyer (2008), also cites the issue of lack of funding for research. However, she, too, cites other issues such as discrimination by some international editors, restrictions by the academics' respective governments, low language skills, poor paper quality, unreliable mailing services, and under-resourced, absent, or unreliable communication means. Cargill et al. (2017), who trained a group of Indonesian university senior academics from the Agricultural Institute of Bogor, West Java, also found issues similar to those discovered by previous researchers. Other studies on Indonesian cases have also concentrated on funding, and on other major areas, including obstacles related to policy and regulations, broad attitude to research, policy, and practice at universities. The studies on policy and regulations examine how government policy and regulation hinders research activities, and discourage academics from doing research.
A study conducted by Nielsen (2010) looked at policies and regulations in economically emerging countries such as Brazil, Malaysia, and Singapore, and compare them with those of Indonesia. Nielsen (2010) found that funding for research in Indonesia is too little and much lower than the ideal amount, only 0.08% of the National Domestic Bruto (NDB), while the ideal standard is around 1%. This amount makes it difficult for researchers to obtain a decent amount of funding to do serious research. They found that the Government of Indonesia had no national framework for funding and development of high-quality research, and no incentives to develop one's own research. They also found funding issues (such as the low salary of academics) drive them to do a lot more non-research activities to supplement their income, which includes doing consultancy and teaching jobs outside their campus. Brodjonegoro and Greene (2012) also examined the funding related issue and found a similar issue, which is difficulty in getting funding. Other issues include no long-term funding for research centres, and that there is no job security for research jobs at the individual level. Therefore, the authors made a case for the establishment of 'Dana Ilmu Pengetahuan Indonesia' (DIPI) or the Indonesian Research Fund, which has now been supported by the government. It is yet to be seen how effective this funding scheme is. Although the issue of poor funding has been acknowledged by the Minister of Research and Higher Education and there is a promise to increase it, funding for the 2017 financial year has in fact been reduced. Apparently, other priorities are still higher than research. When examining the broad overview of the knowledge sector in Indonesia, Karetji (2010), found another issue regarding funding, that is government bias towards technology and engineering sectors at the expense of other fields. He also found issues with the absence of a clear career path for researchers, and a gap between research and government policy.
There is also an issue of lack of pro-research leadership from the top leaders. Similar findings were also made by McCarthy and Ibrahim (2010), who examined factors affecting the development of the social sciences. They found issues at three levels, macro, meso (middle), and micro (individual). He identified yet another important issue: the rigidity of the reporting system which had inhibited serious research. The study also found that the regulation is also over-restrictive, making it very difficult to obtain the already small amount of funding, and to produce reports after the research is conducted. The great difficulty in meeting the requirements of the regulation, at the expense of producing quality articles and publishing them, was also discovered by Brodjonegoro and Green (2012).
A new strand of studies has also been conducted, focusing on issues found in Indonesian RAs. The first study was conducted by Adnan (2009), who looks at the issue from a linguistic perspective, particularly Indonesian rhetorical patterns of RAs published by Indonesian scholars in Indonesia. He identifies and discusses some potential problems faced by Indonesian authors when sending their articles to international English journals in 'the Centre' of knowledge production. In 2014 he examined the prospects of Indonesian RAs when submitted to an international journal (Adnan, 2014). This study is followed by a similar study by Arsyad (2016).
More recently, two studies are being conducted by the University of Indonesia and the Centre for Innovation Policy and Governance sponsored by the Global Development Network. One is a study on reform of research in Indonesia. This study explores factors inhibiting research at macro, meso (middle), and micro (individual) levels, focusing on social sciences. It examines seven case studies in different regions in Indonesia, including Aceh and Papua. At a macro level, this study also examines government policy and regulation; at the middle level, how universities interpret this policy and regulation and develop their respective policy and regulation, and at a micro-level, how these middle-level policy and regulations affect individual researchers at the universities. The other study is conducted by the Indonesian Science Academy (Akademi Ilmu Pengetahuan Indonesia or AIPI) and sponsored by an Australian-Indonesian supported scheme called Knowledge Sector Initiative abbreviated as KSI (Nugroho et al., 2016). This study outcome is called 'Buku Putih Pendidikan Tinggi' (White Book of Higher Education). It concentrates on the issue of mono-disciplinary' or 'linearity' adopted in Indonesia versus the 'multi-disciplinary' approach to research. No publications of these studies have been found, but according to Nugroho et al. (2016), the consultative group of these two studies suggested that the studies should consider not only conceptual and philosophical issues, but also recommend an appropriate intervention to change the inhibiting factors. Amongst the issues to be investigated are the issues already identified in the previous studies, for example, the disharmony between research activities, researchers and contexts, the very small amount of funding, the lack of attention to the 'National Research Agenda (Agenda Riset Nasional or ARN)', the lack of career path for researchers, and the lower amount of salary for researchers compared to lecturers.
These projects also investigate factors inhibiting research at universities partnering with the KSI, i.e., Universitas Indonesia (UI), Gadjah Mada University (UGM), Universitas Islam Negeri (UIN) Jakarta, and the Atma Jaya University of Jakarta. As described above, all of these studies investigate structural and environmental problems inhibiting research at those universities. No doubt that these studies have provided major contributions to our understanding of problems of low research outputs. However, there is still much to be investigated, at least, on two fronts: the other 70 public and 452 private universities in Indonesia, and the actual problems individual researchers face when trying to publish their research. The urgent question is when the amount of research has been increased, then what? In the past, there have been many good quality research projects, but the low number of articles published in international journals suggests that they were not published. It is not easy to ascertain how many government-funded research projects have not been published in international journals. However, the statements made by the Minister for Research and Higher Education seem to imply that the number is not negligible. Consequently, the only outcome of the project is a report, which is often read by a few or no-one other than the researchers themselves.
In short, the studies reviewed earlier have raised many issues faced by MLSs when trying to published internationally. However, few (if any) studies focus on how MLSs perceive their knowledge and skills to face the challenge, use quantitative analysis, and involve a large number of participants from different universities and different provinces. The studies focusing on Indonesia reviewed earlier, in particular, have concentrated much on the issue of funding, government policy, regulations and bureaucratic procedures, and inappropriate rhetorical patterns as inhibiting factors contributing to low research publication output. While these studies have shed a significant amount of light on our understanding of issues faced by MLSs Indonesia, studies on Indonesia that have been published internationally are still limited. To the best of our knowledge, there have been no survey-based studies, specifically addressing the question 'Why has the contribution by Indonesian scholars to international research publications, been so low compared to smaller nations such as Bangladesh, Kenya, and Malaysia?'

THEORETICAL FRAMEWORK
The following theoretical framework guides this study to find answers to the research questions. As shown in the literature review, the issue considered as the dominant factor inhibiting publication of research outputs by Indonesian academics is lack of funding as it is found in most studies. We question this finding since we believe that knowledge and skills are the most fundamental factors for someone to be able to produce something (Zahra et al., 2020), including research articles. Without knowledge and skills, someone cannot produce a RA as it requires many and complicated requirements such the understanding of the subject matter, the research process, knowledge about the rhetorical patterns, and the skills to write the RA. Writing a RA for a reputable international English journal expects more complex knowledge and skills as they have more standards not required in an Indonesian language journal. This study intends to examine whether this belief is supported by the participants of the study.
Thus, this is a perception study of the absence of an action, which Indonesian academics are supposed to do. It is a perception study in that it examines the ways Indonesian MLSs perceive the reasons why they do not publish. Perception is defined as a process that allows an individual to make sense of their sensory information (Jenkinson, 2014). So, it is their brain that runs the process, which produces meaningful expressions of what their senses catch. In this process, the brain identifies, organises, and interprets the information. This process is largely influenced by expectation, experience, mood, and cultural norms. This happens when the stimulus is external (comes from outside the self). When the perception is about his/her own self, called self-perception, the individual processes their own knowledge and experiences in response to a question. The result of the process is the same as the externally triggered process; "... people come to 'know' themselves and their internal states by observing their own behaviour in a manner much like that of an external observer" (Bem, 1972, cited in Haemmerlie & Montgomery 1984. It is assumed, therefore, that the answers given by the participants of the study are the meaningful representation of what they perceive of their knowledge and experiences concerning the issue in question. This means that the information that they offer is not necessarily the exact replication of what happens externally nor everything they have in their knowledge and experience repertoire. Instead, it is the meaningful representation of what they know and experience in regard to the questions being asked. This theory guides the design of this study in that the participants would provide the best possible representation of what they know and experience concerning the reasons or variables which inhibit them from preparing RAs and publishing them in IEJs in their answers to the survey and interview questions. Further, perception is closely related to attitudes. Attitudes influence behaviours, whether someone decides to do something or otherwise. Someone with a good attitude about a task would tend to complete it positively, to the best of their ability (Flowerdew, 2001), while someone with a poor attitude may carry out the task, but the quality tends to be poor, except another factor(s) encourages him/her to do it otherwise, e.g., the necessity to achieve a more important goal. So, it is related to motivation, which is defined as a psychological process that creates 'an internal drive to satisfy an unsatisfied need' (Higgins, 1994); and 'the will to achieve' (Bedeian, 1993). Thus, negative attitudes can weaken the internal drive and will to achieve something.
A fundamental premise should be made that a researcher can infer and interpret a more appropriate answer to a research question from the different answers provided by the participants. Statistical applications can assist in discovering an association between the answers. More specifically, the applications can identify associations between the many items mentioned by the participants in their answers. This process is called 'data reduction' (Chan, 2014;Quantitative Specialist, 2014). The researchers can further investigate to verify, confirm, or otherwise, and explain the findings from further statistical analysis. The interviews and FGD provided supplementary information.

METHODS
This study employed a mixed-method, combining quantitative and qualitative methods, following Creswell (2009), to address the research questions. However, in this study, the primary method is quantitative, which employed a national survey, aiming to collect a significant number of variables which may contribute to the issue under question. The qualitative method, which employed interviews and a focus group discussion (FGD), was supplementary, in order to verify and clarify the findings of the national survey. For this, a randomly chosen five percent of those included in the national survey were interviewed.

Data Collection and Population
The study employed a multi-purposive sampling method, in which the population was selected based on the purposes already specified. Basically, they are active researchers who have created unpublished research project reports; they are from the disciplines already specified, namely linguistics, applied linguistics, and language education; selected from universities based on some geographical representation; and feasibility considerations.
Based on these criteria, the population was researchers whose names were collected from research centres of five universities in Indonesia. These universities were chosen based on financial feasibility and geographic spread representing the west, middle and eastern parts of Indonesia. We planned to select an equal percentage of participants from each university. However, the responses we received did not meet the plan. We received responses as follows: 12 participants from the University of Bengkulu (19.9%), 5 participants from Atma Jaya University of Jakarta (AJJ) (7.5%), 19 participants from UNM Makasar (The Makassar State University) (28.4%), 14 participants from the University of Padang (UNP) (20.9%), and 17 participants from Universitas Mataram, Lombok (25.4%). The total number of participants (with a valid response to the survey) was 67 MLSs.
The in developed countries such as Australia, the UK, and the US, and why? • How do they perceive their research results?
• How can their strategy and methods be improved?
• How do they rate their motivation to publish their research results in international journals and why? • What are the most effective ways to make researchers publish in reputable international journals?

Data Analysis
The primary purpose of the research is to get to the 'truth' about the issue being investigated, in this case, why Indonesian MLSs do not publish in IEJs. To ensure a high level of reliability of the survey analysis results, the authors conducted two levels of data analysis. They were Principal Component Analysis (PCA), confirmatory factorial analysis (CFA), which both are parts of SPSS Statistical software. The results are supplemented with some interviews and focus-group discussion.
The PCA has KMO and Barlett's Tests, which can determine whether the data is adequate for the analysis and minimises errors, and conducts the component analysis. The KMO and Barlett's tests can provide the maximum and minimum required scores that a set of data must achieve to allow a valid analysis. The PCA analyses the data based on the commonality of the high number of possible variables (coded 'items' in the PCA results) as supplied by the participants. It statistically analyses these variables and indicates the degree of their contribution to explaining the issue using Eigenvalue (henceforth: EV). Based on their respective EVs, it groups them into a meaningful set of factors (named 'component' in the PCA results), each of which can be given a covering name based on the commonality of the variables. For example, a factor could be named 'significant lack of knowledge and skills' if the variables are mainly about knowledge and skills. Hence, each factor has a set of associated variables, and each variable has an 'eigenvalue', which indicates the degree of its association with the factor or component (both factor and component will also be used interchangeably). In other words, an Eigenvalue is a value that indicates the degree of the seriousness of the factor in influencing the 'variance' (a technical term used in the PCA software) or the dependent variable, the higher the value, the stronger the influence. However, this procedure alone is, in our view, not reliable enough. Therefore, we employed a second statistical analysis, the CFA, which is capable of confirming whether or not the set of factors and their respective variables found in PCA are the accurate representation of the factors and the associated variables that explain the lack of publication in IEJs.

Results of the PCA Analysis
As mentioned earlier, PCA contains KMO, Barlett's Tests, and the PCA analysis. All the items being analysed are listed in the Appendix. The researchers collected these items from the questionnaire, completed by the participants of the survey, and, subsequently, keyed them into the PCA section of the SPSS statistical software for the tests and analyses. The next section presents the results of these two tests.

The results of the KMO and Barlett's Tests
The KMO and Bartlett's Tests of Sphericity or 'Measures of Sampling Adequacy (MSA)' tests are necessary to determine the adequacy of the data for both the principal component analysis (PCA) and the CFA. The minimum requirements are that the Kaiser-Meyer-Olkin (KMO) Measure of Sampling has to be larger than point five (>0.5), and the significance of Bartlett's Test of Sphericity must be 0.05 (5%) or lower. After conducting the tests, all the results exceeded the minimum requirements, the Kaiser-Meyer-Olkin Measure of Sampling Adequacy is 0.73, and the value of Bartlett's Test of Sphericity is 276.000 with the significance of 0.00, which is perfect (see Table 1). These results mean that the data can be further analysed using PCA and CFA.

The results of the PCA data analysis
As the results were analysed again using CFA, this section reports only the key findings of the PCA. The analysis found six components associated with the central issue of this study, each with a number of variables. Based on the nature of commonality found in the variables in each component, we analysed and classified the components as follows: 1) The inhibiting factor 1: Lack of knowledge and skills to produce a RA for IEJs; 2) The inhibiting factor 2: Negative self-attitudes; 3) Motivating factor: Awareness of the advantages of publishing RAs in IEJs; 4) Solution 1: Taking the initiative to acquire knowledge and skills; 5) Solution 2: Taking concrete actions to write and submit RAs to IEJs, and 6) Solution 3: Improving English and other RA presentation skills.
The next section presents the results of the CFA test of these factors with their respective associated variables.

Results of the CFA Test Procedures
The first set of results of the CFA test procedure are presented in Table 2. As shown in this table, the total number of factors (components) are confirmed, namely six (numbered from 1 to 6), each with a set of variables (coded 'item' followed by the item number) with their respective eigenvalues.  These CFA results, the components, and their respective set of items (variables) were further analysed and yielded the second set of results which shows several changes as presented in Table 3. As shown in this table, we found two main changes from the PCA analysis results (Column 2) to the CFA test results (Column 3), which involved two factors. The factors cover solutions and inhibiting factors. As shown in Column 3, only two solutions remained, while the Inhibiting factors increased from two (in Column 2) to three in Column 3 due to the emergence of Inhibiting factor 3: Negative attitude against IEJs (Component 5) replacing Solution 2 (see Column 2). This is because the CFA re-analysed the variables, producing a result (a set of variables) which makes it difficult to call this component Solution 2. After all, the (new) set of variables are all about negative attitudes. The variables associated with Solution 3 were moved to Solutions 1 and 2. Consequently, only two solution components (factors) remained as presented in Column 3, namely Solution 1 as factor number 4 and Solution 2 in factor number 6 as 'Solution 3' was renamed Solution 2.  The final CFA results, as shown in Column 3 of Table 3, show the ranking order of factors influencing the lack of publications by Indonesian academics in the language-related disciplines. The first and the most influential factor is Inhibiting factor 1: Lack of Knowledge and skills, followed subsequently by Inhibiting factor 2: Negative attitudes toward self, Motivating factor: Awareness of the advantages of publishing RAs in IEJs, Solution 1, Inhibiting factor 3: Negative attitudes toward IEJs and, finally, by 'Solution 2' (Number 6). Each of these components has a set of variables, as presented in the following tables. However, for a coherent and meaningful presentation, the factors are re-grouped into three categories: the 'inhibiting factors', the 'solutions', and the 'awareness of the advantages'.

The inhibiting factors
a.
Inhibiting factor 1: Lack of knowledge and skills As shown in Table 4 below, this factor is associated with seven variables (coded with 'Item + number') (see the Appendix). The most dominant variable contributing to this factor is lack of adequate knowledge regarding the publication of research articles (RAs) in reputable international journals, contributing most significantly (Item 8.1), with the EV of 0.893. This factor is associated with six variables, namely the perception that the quality of their research is inadequate; followed, subsequently, by the perception that writing an article that meets the international standards is too difficult, lack of funding support, lack of knowledge about the format of an acceptable RA, lack of references in their library and, finally, the absence of a financial incentive. There are two interesting points worth noting here, the variables that concern lack of financial support appears lower in the ranking order, only in Line number four (Item 8.7) and Line number seven (Item 8.8) (to be discussed later).  This factor is closely associated with and provides more evidence for Factor one, the perception of the lack of knowledge and skills. As presented in Table 5, the variables associated with this factor clearly support this point. The most vital variable 'I am not confident with my analysis skill' is indeed a reflection of the lack of knowledge and skills to produce a RA with acceptable quality. This point is supported by the next variable, fear of rejection by the journal editor, which could imply disappointment and wasted efforts if conducted. The fear is backed up by another expression of lack of confidence, now, in the results of their research, which is strengthened with a perception of the inadequate quality of their research article. Apart from the two dominant factors described earlier, Negative attitudes toward international journals (IEJs) also seemed to have contributed to the explanation of why the Indonesian academics did not publish RAs in IEJs. This factor has two variables associating with it as presented in Table 6. The first is a complete rejection of IEJs, saying that they are 'not needed'. This perception might have been expressed by senior academics who no longer wish to reach higher academic career as publishing RAs in IEJs was necessary to get a promotion to a professorship. We found some academics who could not gain promotion to professorial level due to lack of publication in IEJs. It could also be an expression of resentment against the threat of the then Minister for Research, Technology, and Higher Education to suspend professorial financial allowance if a professor does not publish in IEJs. The second variable expresses dislike against IEJs. As this factor appeared lower in the rank, it is not as influential as the other two factors. Further evidence for their recognition of the inhibiting factors, they nominated two solutions. Language and Education, 8(1), 65-83, 2021

a.
Solution 1 (Component/factor 5) As reflected in its variables, this factor constitutes efforts an academic should do to improve their knowledge and skills so that they can publish internationally. As presented in Table 7, the strongest variable is reading RAs published in international journals as much as possible, which clearly addresses the issue of lack of knowledge about the nature of RAs that have been published. With a great deal of exposure to such articles, they could have a good understanding of elements such as the structure, the language, and the content. The participants also suggested studying the format of RAs acceptable to IEJs through reading or writing guide books (Item 11.2), and finally attending workshops on how to write and publish in reputable international journals (Item 11.7) so that they can learn directly from experts. Indeed, these variables recognise the existence of a lack of knowledge issue. Reading RAs published in international journals as much as possible.

0.831
Item 11.2 Studying the format of articles acceptable to international journals, e.g., from a guide to RA writing books, and the like.

0.784
Item 11.7 Frequently attending workshops on publication in reputable international journals.
0.691 b. Solution 2 As reflected in the variables listed in Table 8, this factor addresses mainly the issue of lack of skills, and to less extent to overcome the lack of confidence issue. The most influential variables suggest a great deal of practice in producing and sending RAs to IEJs despite being accepted or not (Item 11.5 and Item 11.6). These efforts are to be strengthened by another suggestion, that is to seek mentoring from a researcher who has been successful in gaining publication in IEJs. In short, a lack of skills was indeed an inhibiting issue for the participants.

Item number
Variables Eigenvalues Item 11.5 I need to send as many RAs as possible to international journals (although there is no certainty that they will be accepted.)

0.768
Item 11.6 Write as many RAs as possible (even though in Indonesian) to be published in national accredited journals.

0.674
Item 11.8 Asking to be mentored by a successful researcher in publishing in reputable international journals. 0.593

Motivating factor: Awareness of the advantages of publishing in international journals
An interesting question was 'Weren't the academics aware of the benefits of publishing RAs in IEJs?' The findings suggested that they were generally aware of them. As shown in Table 9, there are three strong variables which should motivate RA publication in IEJs as indicated by their Eigenvalues, all of which are very high above 0.8 out of the maximum 0.99, plus a considerable one (0.583). The first and most influential one was that they were aware that the publication could gain selfsatisfaction or pride, with the highest Eigenvalue (0.873). This is followed by the second influential variable; namely, they were also aware that the international publication of their RAs could also improve the reputation of their institution. Thirdly, they even knew that international RA publication would incur credit points necessary for career promotion, and finally, it could lead to a strong personal reputation (Item 14.4). As the issue of lack of RA publication in IEJs, the awareness of these advantages did not seem to be strong enough to overcome the more dominant inhibiting factors. These factors and their ranking were discussed in a focus-group discussion with approximately 40 academics at the University of Bengkulu in November 2018; some of them participated in the research (completed the survey). When asked the participants were asked for their opinion regarding the findings, most of them supported them.

DISCUSSION
The first finding that the most important factor is the lack of competence with its associated variables is both expected and surprising. It is expected because they are related to many issues reported in previous studies which affect publication in international journals. These include English language problems (e.g., Cargill et al. 2017;Uzuner, 2008), and divergence from the accepted norms of research presentation (Flowerdew, 2001) as they did not know about them. However, this finding is also surprising because it contradicts many reports, especially those focusing on Indonesia, e.g., Nielson (2010) and Brodjonegoro and Greene (2012), which consider lack of funding as a dominant issue. In the present study, the issue of funding only appeared twice in Table 4, as Item 8.7, 'No funding to support to write', which ranks number four, and as Item 8.8 'No incentive to for those who publish' (ranked number 7). Hence, the issue of lack of knowledge and skills to produce a quality RA is much more critical for the participants. It may be because most of them are not trained nor oriented to publishing in international journals. There are two possibilities for a different result. First, the previous studies did not focus on many ordinary academics of languagerelated studies and their perceptions of their knowledge and skills. Instead, they concentrated on selected high achieving academics, e.g., head of departments, deans, and professors who have held important positions. The second possibility is that they concentrated on universities in Java, including Universitas Indonesia (UI) and Universitas Gajah Mada (UGM), while this study concentrated on those outside Java, who had fewer resources than their counterparts on Java.
The finding regarding the issue of the absence of funding support is interesting because universities and the Ministry of Research, Technology and Higher Education (Ristekdikti) had been providing financial incentive to publish RAs. Interviews with some of the participants seem to suggest several things. First, the incentive seemed to be beyond reach for many participants for a range of reasons such as the lack of confidence that they would get it because they have low confidence in the quality of their research. Second, they felt that rather than doing the complex research and publishing RA work, which would offer no sure success, it would be better to use the time and energy to take a teaching job, which promises much quicker and more certain financial benefits. "Accepting teaching work means cash in hand" as one participant put it. Third, Regarding the reward offered by the Ristekdikti, in particular, there was also a feeling of uncertainty because the participant had a RA published in a reputable international journal but was not given the reward for unclear reasons. Hence, for many of the academics, the reward was not convincing enough to spend a great deal of time and energy for an uncertain outcome.
The second factor, 'negative attitude toward self and own work', which includes lack of confidence, supports the hypothesis of the government as mentioned in the National Research Grand Plan or RIRN (Ristekdikti, 2016). It also seems to be a logical consequence of the first factor. The presence of factor number three, the motivating factor, awareness of the advantages (such as benefiting self-pride, career promotion, and personal reputation) does not seem to be strong enough to outweigh the first two factors. The issue may be worsened by a negative attitude toward international journals. Moreover, as many interviewees pointed out, the lure of many hours of teaching, which provide quick cash, may sway those MLSs, who do not have strong knowledge and writing skills, to choose to teach multiple hours instead of conducting research and producing RAs, which present them with complex and extremely difficult challenges but without promising certain financial benefits as the teaching does. Compare this point to the attitude of UK Management and Accounting academics who avoid sending RAs to USA journals because of uncertain outcomes (Brinn et al., 2001). Furthermore, in the Indonesian university performance appraisal, teaching has a much larger portion than research (multiple sources). Hence, research and its publication tend to be neglected.
The two solutions, the fourth and sixth factors (with their respective variables), implicitly confirm the issues associated with lack of knowledge and skills found in Factors one and two because the variables indicate the logical response to those two inhibiting factors, for example, the suggestions for adding more knowledge, more practice, and more training. An interesting point here, though, is the variable 'Asking to be mentored by a successful researcher in publishing in reputable international journals.' appears to be the most effective way of improving their knowledge and skills as reflected in the interviews, but this variable appeared under the lowest component/factor, and its ranking is the lowest amongst the variables. Many participants might have thought that it would be challenging to implement.

CONCLUSIONS
This study was motivated by a significant problem faced by Indonesia, that is the lack of RA publication by its academics. Since the finding clearly shows that the most challenging issue, faced by the Indonesian MLSs is lack of knowledge and skills (which harm their confidence to write RAs), this study has revealed the real problems perceived by Indonesian MLSs in language-related disciplines, which had not been considered a great deal previously, especially in the studies that focused on Indonesia. These problems are crucial because the inhibiting factors are so influential that rhetorical campaigns and promises of rewards and other benefits are not strong enough to defeat the inhibiting factors and the lure of the clear and direct financial benefit of taking teaching work. Moreover, if the academics, early in their academic career, were rewarded with more credit points for publishing RAs (than teaching and other administrative duties), then they would be more motivated to continue to publish RAs in the later part of their academic career. Internationally too, few studies have considered how MLSs (quite a large number of participants) perceive their knowledge and skills necessary for constructing RAs acceptable to IEJs, even though some of the issues the participants raised in the survey of this study have been discussed by some authors. Hence, this study has contributed to the literature in two ways. Firstly, to the studies focusing on Indonesia, it reveals an important hidden issue, perceived by academics in developing countries such as Indonesia, that seems to have been overlooked in the past. Secondly, the literature focusing on multiple countries supports some of the findings reported in previous studies.
The uncovering of the key issue has a crucial policy implication for the Indonesian government, particularly on how EAP (English for Academic Purpose) is treated, despite its strong rhetorical push to publish internationally. To address the key issue, the government should focus on and prioritise providing regular training for staff with the right practice. Ad hoc training so far seems ineffective since writing and publishing quality RAs is complicated and challenging work. Funding for research is undoubtedly critical, but the survey suggests that it is not the most important motivating factor. If the academics are knowledgeable and skilful, they would likely write, but if they do not have the knowledge and skills, they will not write, even if they have the funding. This point was confirmed by some of the interviewees at UNIB. Many of the interviewees who have published RAs admitted that they had already gained some talent in writing when they were in high school, which they kept developing during their tertiary education. Hence, knowledge and skills are indeed more important than funding. Strong rhetorical campaigns and punitive threats (against those who do not publish) seem to have raised considerable awareness of the benefits of publishing in IEJs. However, it is not sufficient for those who have weak or no knowledge and skills. It could even backfire in the form of having negative attitudes against IEJs as one of the variables shows. Therefore, it would be more effective if a course on publishing in international journals is included in the last year of completing a Bachelor's degree (S1), or at least at MA level (S2), where graduates are expected to publish before they can graduate. This point is vital as this degree (MA) is the minimum qualification for a lecturer position (in which publication of RAs is expected).
This article is a preliminary report of a three-year project. It involved a smaller data size relative to the number of MLSs in Indonesia. Therefore, the findings need further investigation using a bigger data size and involving more universities to verify the findings. The survey has now been improved, and the plan is to use a significantly bigger data size of over 200 participants from a more significant number of universities. The results will be reported in another article. Language and Education, 8(1), 65-83, 2021