PERSPECTIVES OF EFL DOCTORAL STUDENTS ON CHALLENGES OF CITATIONS IN ACADEMIC WRITING

PurposeCitation is vital in academic writing but particularly challenging for novice writers who use English as a second or foreign language. While much is known about citations types and functions, scarce knowledge is available about what makes citing a complicated procedure. Hence, this study explores the difficulties in citing and integrating information from academic sources into the literature review chapter of PhD proposals. MethodologyThe study, which falls under New Rhetoric studies, involved conducting individual discourse-based interviews with six male Arab doctoral students who were selected purposefully from the Information Technology department of a Malaysian public university. The participants’ literature review chapters were employed in a stimulated recall to retrieve their inner perceptions when writing their PhD proposals. Hand analysis, mind mapping, and visual mapping were used in the analysis of the data. Multiple codes were initially obtained and further reduced to form four major themes and several

Integrating the citation into one's writing requires advanced research and academic writing skills (Rabab'ah & Al-Marshadi, 2013).The process of citing information involves several significant issues, such as what to cite, who to cite, how to cite, and when not to cite (Polio & Shi, 2012).Hence, citing information from articles, theses or dissertations is challenging, particularly for researchers who write in English as a second or foreign language (Law et al., 2013;Nguyen & Pramoolsook, 2016;Rabab'ah & Al-Marshadi, 2013;Rowley-Jolivet & Carter-Thomas, 2014).Citation practices may vary across disciplines (Hyland, 1999;Jomaa & Bidin, 2016;Swales, 2014) and cultures (Rowley-Jolivet & Carter-Thomas, 2014).Al-Zubaidi (2012) argued that Arab postgraduates in Malaysia lacked sufficient basic academic literacy habits to build upon.In addition, EFL Arab postgraduates are known to experience serious difficulties in dealing with citations (Al-Zubaidi, 2012) or references (Al-Khasawneh, 2010;Al-Zubaidi, 2012).In this regard, several researchers have suggested exploring the citing behaviour of students from different disciplines (Jalilifar & Dabbi, 2012;Mansourizadeh & Ahmad, 2011;Petric` & Harwood, 2013) and nationalities (Jalilifar & Dabbi, 2012).Therefore, this study focused on the citation process during writing the literature review chapter of PhD proposals.It explored what made citations challenging for Arab non-native writers of English who were studying in an Information Technology department at a Malaysian public university.

LITERATURE REVIEW
Citations, according to Swales (1990), indicate a research space.In other words, when demonstrating that studies have been conducted in a certain field, citations could pave the way for the unknown, thus creating a space for more exploration and examination.Based on Swales ' (1990) taxonomy, citations are of two types: integral and non-integral citations.In an integral citation, the name of the cited author, represented by a number or a pronoun, is integrated with the structure of the citation to function as a subject, a complement, an agent, a part of the noun phrase or an adjunct (Jomaa & Bidin, 2016;Swales, 2014), as in the Example 1 below.
Example 1 : Hyland (2004) also employs this distinction. 1n this type of citation, the focus is put on the researchers' actions, and the name of the cited author is given more prominence.On the other hand, in a non-integral citation, where the findings of the research are highlighted, the name of the cited author is not integrated with the structure of the citation.This suggests that the role of the cited author is less emphasized, as in Example 2.
The integral citation was classified by Thompson (2001Thompson ( , 2002) ) into three sub-categories: verb controlling, naming, and non-citation.As for the non-integral citation, Thompson (2001Thompson ( , 2002) ) identified four sub-categories: source or attribution, identification, reference and origin.
Among the studies that have researched non-native Arab writers is Rabab'ah and Al-Marshadi (2013), in which five EFL MA theses by Saudi writers were compared with five EFL MA theses by native writers of English at universities in the United States of America.Swales ' (1990) classification and Thompson and Tribble's (2001) taxonomy were adopted to analyse the citation types and the frequency of citation types in the different sections of ten theses.Similar to a study on non-native Iranian MA student writers (Oskueia &Kuhi, 2014), Rabab'ah andAl-Marshadi (2013) revealed that Saudi writers used more citations than their native counterparts.Saudi writers were also similar to Vietnamese postgraduates in Nguyen and Pramoolsook's (2016) study, in that they lacked linguistic competence.Therefore, they were unable to rephrase information or add their own ideas to their writing.
Another study by Petric` and Harwood (2013) involved an in-depth discourse-based interview with only one successful female L2 postgraduate management student in two assignment writing tasks (a directed task and an open task).They found that several functions of using citations, including positioning, defining, supporting, application, topic relevance, disagreement, agreement, and acknowledgement.Nguyen & Pramoolsook (2016) who also used interviews, revealed that Vietnamese writers of TESOL MA theses were unaware of the rhetorical functions implied in citations and had difficulty in implicitly acquiring the conventions of citations.Consequently, these Vietnamese postgraduates overused secondary sources, used a question mark (?) instead of the year of publication, and invented unconventional ways of citing due to receiving insufficient guidelines on citation conventions.
International students studying in Malaysia seem to have difficulties related to L2 academic writing, English language proficiency, and the influence of prior academic culture, such as the effect of L1 on L2 academic writing (Al-Zubaidi, 2012;Manjet, 2016).Ethnolinguistic influences on citations have been demonstrated in a study by Hu and Wang (2014) since language in general and writing specifically, are associated with culture (Bian & Wang, 2016;Hyland, 2013).In Malaysia, difficulties relating to using citations are not limited to international students, but are also experienced by Malaysian students.For the latter, research has shown that the citation process seemed to involve mainly reporting other researchers' findings and their procedures.They lacked critical thinking when citing; they did not try to synthesise and criticise; and the attitude tended to be neutral, rather than opposing or supporting other scholars' claims (Manan & Noor, 2014).
A key feature when reviewing literature in academic writing is critiquing previous studies; students need to demonstrate their position in their academic community (Paltridge & Starfield, 2007).However, adopting a stance also forms a major problem for nonnative speakers of English (Hyland & Milton, 1997;Paltridge & Starfield, 2007).
Paraphrasing, which is vital in academic writing (Bailey, 2003), can be also challenging for non-native speakers of English (Al-Zubaidi, 2012).When citing information from previous studies, students have to paraphrase it by using different vocabulary without changing the original meaning (Bailey, 2003).Paraphrasing is directly associated with citations because it is used in order to avoid plagiarism.Shi (2012) emphasised that L2 student writers have problems in paraphrasing since it is a complicated procedure that depends on having knowledge of content and being familiar with the rhetorical purpose of citations uses in each discipline.Martin (1992) argues that people are usually socialised into language that expresses social class or gender identity.Student writers are socialised into the language of their academic disciplines.These disciplines can be classified into hard and soft domains.For example, natural sciences and technology are in the hard domain, whereas the humanities is in the soft domain.The social sciences mediate between these two domains (Coffin et al., 2003;Hyland, 2009).In hard disciplines such as Information Technology, it is assumed that one single objective reality exists.This reality is independent of human subjectivity, and is governed by consistent rules of cause/effect relations (Cohen et al., 2007).Hence, the purpose of a scientific investigation is to comprehend, validate, and display the external world empirically.This inquiry involves basically adopting a rigorous approach inductively and deductively to test systematically particular hypotheses in controlled settings by employing certain procedures and measurements (Becher & Trowler, 2001;Hyland, 1999).Hence, members of the science community focus on the originality of the idea, believing ''the core of science is facts and theories, not words'' (Bouville, 2008, p. 314).According to Pecorari (2008), writing in these domains needs certain skills, and writing from sources is an essential sub-skill.Hence, academic writing in each of these domains is different.For instance, writing in scientific disciplines is based on hypotheses that are verified through real experiments.On the other hand, writing in the soft domain is based on the strength of the arguments.Consequently, knowing how physicians, chemists, linguists, and other specialists read and interpret their disciplines requires exploring their perspectives through conducting interviews, which could lead to rich information about their performance.Miller (1984) argues that the focus must not be on the substance or the form of the discourse but on the action.Other scholars in the New Rhetoric fields have adopted ethnographic methods for analysing texts, such as interviews, document collection, and participant observation, rather than linguistic methods, thus presenting detailed descriptions of both the professional and the academic contexts of the genres (Schryer, 1993).
When students write, they attempt to reveal to their readers that they are writing grammatically and correctly.However, the focus of studies has shifted from students' syntax and sentence-level issues to students' work as language-in-use, particularly students' discourse and what students do with language (Al-Zubaidi & Richards, 2010;Canagarajah, 1999;Donahue, 2004;Knight, 2009).In relation to this, the use of citations among non-native writers of English has not been fully explored from the perspectives of Arab students in Information Technology as a hard domain.Therefore, the present study explored EFL Arab postgraduates' perspectives toward the citation process and identified the challenges faced by EFL Arab postgraduate students in an Information Technology department.The study involved an exploration of the challenges faced in citing information from original sources and integrating them into the literature review chapter of PhD proposals.

Methodology
The present study involved face to face interviews of six EFL Arab doctoral students at a Malaysian public university.Discourse-Based Interviews (DBIs) were used as a source of information.As Patton (1990) argues, the aim of interviewing participants is to explore their perspectives.Lancaster (2016) states that DBIs by Odell, Goswami, and Herrington (1983) are an effective research methodology that is used to explore writing choices in the performances of writers.In this aspect, Hyland advocates such interviews since they are able to capture the variations and regularities of the use of disciplinary language (Swales, 2013).According to Odell and Goswami (1981), DBIs can elicit beneficial types of information related to the perceptions of the writers towards their own tasks.

Sampling
It is typical in qualitative studies to study only a few cases or a few individuals, as the aim of the qualitative study is not to generalise the findings to a population, but to obtain a detailed exploration of a specific phenomenon (Creswell, 2008).Dörnyei (2007) argued that interviewing 6-10 participants might be suitable.With regards to discourse-based interviews, the number of participants have been very small.Petric` and Harwood (2013) conducted a discourse based interview with one female L2 successful student to explore her citing behaviour in two tasks.In another study, Harwood and Petric` (2012) conducted a discourse-based interview with two female L2 students to explore their citation uses.The present study chose to interview six participants.Creswell (2012) suggested that participants and sites in qualitative studies should be identified purposefully in order to assist in exploring the phenomenon comprehensively.Hence, the six participants were selected purposefully.Snowball sampling was another sampling strategy that was used in selecting some of the participants.Since the researchers were not familiar with all students in the IT department at the university, interviews were conducted with some participants based on the recommendation of their friends who were interviewed earlier.All the six male participants had defended their PhD proposals successfully.At the time of the study, some of them were involved in improving their literature review chapter based on the comments they had received, whereas others were involved in data collection and data analyses.Six criteria were followed in selecting the participants.First, all of them were EFL doctoral students.Second, they were from Arab countries that are classified as EFL countries (Kachru, 1990).Third, the first language of all the participants was Arabic.Fourth, they had successfully defended their PhD proposals.Fifth, the participants were selected based on the recommendations of their friends and one of the supervisors in IT department.Sixth, they were from Information Technology department.Although the six participants were from five Arabic countries, getting diverse perspectives, according to Creswell (2008), is essential in order to reveal the depth of the phenomenon and is a distinguishing feature of qualitative studies.
Table 1 presents information about the six male participants.This includes their country of origin, department of the study, degree, age, their L1, and the length of the interview for each participant.The participants were given names that were different from their real ones.

Data Collection
To enhance the validity and reliability of the findings, the participants were informed of the aim of the study and the technical terms used in the study, such as the integral and non-integral citation.The participants were also encouraged to retrospect freely and truthfully about their experience in writing the literature review chapter of their PhD proposals and the difficulties they had in citing information from original sources.Open-ended questions and semi-structured questions were formulated with reference to EFL Arab students' PhD theses at the university.These DBI questions were modified after a pilot study was conducted with two EFL Arab doctoral students from Iraq and Syria, which lasted 52 minutes and 58 minutes respectively.The data obtained from the pilot interviews were not included in the main findings of the present study.The types of the questions in the DBIs enabled the participants to express their experiences without being limited by the perspectives of the researchers.As mentioned earlier, when the interviews were conducted, the participants were still involved in writing, editing, and revising their literature review chapter.Hence, it was easy for them to recall and comment on their past experiences.Thus the DBIs were considered appropriate for these participants.The literature review chapters of the participants' PhD proposals were used as a stimulated recall in order to help them in retrospecting on their experience of using citations.The questions were directed towards their own citations, such as the types of the citations used, the types of reporting verbs, and the expressions employed to show their stance.
One of the recommendations made by Dörnyei (2007) was to conduct retrospective DBI "in the respondent's L1" (p.50), as in a study by Nguyen and Pramoolsook (2016).Hence, the Arabic language was used in order to give the participants more opportunity to express themselves, since speaking in English might impede them from expressing their inner thoughts.Occasionally, the participants mixed English with Arabic spontaneously when referring to certain items, such as 'citations', 'sources', 'author', 'researcher', 'previous studies', 'grammar', and 'punctuation'.Each interview was taperecorded after scheduling an appointment with each participant to select the time and the place that was most suitable for him.One of the researchers' native language was Arabic, and he had experience in teaching English and Arabic/English translation.This enabled him to translate the participants' responses to the interview questions.

Data Analysis
The data were transcribed and translated into English.Every single word was translated taking into consideration the differences in grammatical structures between Arabic and English.According to Bailey (2008), the process of transcription and the representation of the audible data into written form constitute an interpretive process.Hence, it was the first step in analysing data.Then, the data were marked by hand because the analysis involved a small database.A hand analysis strategy together with mind mapping and visual mapping were used in the analysis of the data.A preliminary exploratory analysis, as suggested by Creswell (2008, p.250), was employed in order to inductively explore the data to get 'a general sense'.The analysis of the interviews was achieved through using themes, which are similar codes that form the main idea.The initial analysis resulted in several codes which were reduced to form a few themes.The multiple codes obtained at the initial stage were minimised to four major themes with multiple sub-themes.The reason for having a small number of themes is to provide deep details about a few themes rather than presenting general information about several themes (Creswell, 2008).
Besides the pilot study with two participants, member checking was adopted in the present study to increase the trustworthiness of the findings, i.e., by asking two participants to check the accuracy of the findings.The member checking was carried out after categorising the themes.The credibility of the findings was also strengthened by reporting quotes from the participants' interviews, as suggested by Creswell (2008).

FINDINGS
Challenges in using citations by the participants in the present study are demonstrated in Table 2.

1) Credibility
The participants faced challenges in the citation process which involved reading sources.Specifically, citing information from these sources required validating the content of the sources.This necessitates validating the methodology adopted in these sources, the obtained findings, and the cited information used by the authors in the literature review section of these sources.This challenge arises due to lack of credibility in the cited information and students' inability to locate credible sources.

A) Credibility of Cited Information
In their preparation for writing their PhD proposals, the participants read the different sections of the published sources, including the abstract, the introduction, the related studies, the methodology, the findings, the discussion, and the conclusion.Hence, they cited information from the literature review section of these sources.However, lack of trust in the cited information made participant Ahmad obtain the original source in order to check the original meaning of the cited information.For Ahmad, there was a problem in identifying whether the cited information in the related studies' section of the article really reflected the idea of the original source from which it was cited or not.
"This is originally one of the problems that we notice and face.There is no credibility in citing the information from the original source.Among the problems was in not citing the information exactly on the part of the author.When you cite from a certain article, you do not really know if the author cites exactly the information from the original source or not, and this was one of the challenges, whereby we are forced to go back two times and three times to the original source in order to make sure of the credibility of the cited information."(Ahmad) Survey papers formed a significant source of information for students in IT.These types of articles are mainly critical and address a specific topic from several aspects.Thus students cited information from such survey papers and integrated it into the literature review chapter of their PhD proposals.However, the cited information in the literature review section of these types of articles may not be accurate, as Ahmad reported.Ahmad added that some authors may change the meaning of the cited information in their articles.In other words, authors' use of vocabulary to avoid plagiarism resulted in a change of the meaning of the original idea.The participants' limited familiarity with the vocabulary used in their area of specializations might be a reason for declaring that authors of sources used strange vocabulary.
"Some writers, when he tries to criticise in his article, he changes the whole sentence; he uses strange vocabulary in order to avoid plagiarism, thus he changes the meaning of the original source."(Ahmad) Although the literature review section of the published articles and survey papers constituted a rich source of information for students, lack of credibility in the citations posed a challenge for the participants, who were required to check the original sources of these citations.

B) Inability to Locate Credible Sources
The findings' section of research articles is also another main source which the participants depended on to cite the information, and employ it in the literature review chapter of their PhD proposals.However, as two of the participants claimed, some research articles which are published by some conference proceedings include findings that are incorrect.In other words, if the same methodological procedure was adopted by another researcher, different outputs would be obtained.This posed the challenge of whether or not to trust the information in such types of articles.For Khalid, there are no criteria to follow in order to check the validity of these results.
"Sometimes, some papers especially in the new areas, they have findings, and these findings are somehow nice, but to what extent are these findings credible?And how can we know that these findings are Ok or not?Frankly, this is one of the problems that faced me.The issue that is questionable for me here is how do we trust a paper?Is there a specific way to follow in order to trust the paper?Or shall we consider that there is a scientific credibility and accept it as everything is valid?This is one of the things that raised several questions during this stage."(Khalid) In writing the PhD proposal, students have to include information from recently published articles.However, the participants had challenges when citing the information from some of these articles since the findings might be incorrect.Mahmoud had to update his references and in-text citations with the most recent articles since a PhD study may take five years of preparation.On the other hand, even though the findings in some of these published articles are not trusted, he had to include them in his PhD thesis or proposal to be academically accepted.
"Some papers have vague findings, and some other papers, when you examine them carefully, you will notice that there is something wrong.But only because they are recent, I have to include them in the literature review or problem statement."(Mahmoud) The participants' low level of knowledge in their research topics at the beginning of their PhD study and their lack of awareness affected their writing and citing negatively.They were unable to determine the validity and the credibility of the information that was published in some research articles.In other words, they could not identify the strength, the weaknesses, and the correctness of the information.Therefore, Khalid is not sure of what to cite in order to support his claim and justify his methods.
"I have great challenges in such issue, the issue of writing, the issue of whether this information is strong enough or not?Is this information trusted or not?Because we had insufficient experience in such matters, thus challenges do exist."(Khalid) In conclusion, it could be argued that citation is directly associated with reading.More readings will guide students to identify the type of credible sources, the correct findings, and the experts of the field.On the other hand, less reading probably results in a failure to realise what types of articles to read and which information to cite.Consequently, writing the literature review chapter of the PhD proposal could become an obstacle for students, particularly novice writers who have insufficient reading and lack the depth of reading.

2) Adopting a Stance towards Citation
The participants also faced challenges in adopting a position when reporting information from sources.Using either type of citations implies evaluation.Hence, the citation is strongly associated with evaluation.Several factors affected the participants in adopting a stance, including facts, low level of knowledge, lack of evaluative vocabulary, as well as politeness and tension.

A) Facts
The field to which the topic belonged was a significant issue that affected the participants in adopting an attitude.IT is classified as a hard domain that is based on facts rather than arguments.Consequently, as some of the participants claimed, students in this domain are limited by their field, and they felt that they were not allowed to express their points of views towards the information they cited.For Faisal, IT was based on facts; therefore, he cited the information from sources and integrated it into his literature review chapter without revealing his point of view.Since IT deals with facts, the EFL Arab postgraduates focused mainly on citing the information from sources, with no clear or direct stance towards the citations, specifically during the PhD proposal stage.This could make the style of writing the literature review chapter involve only a summary of past studies with the absence of the writer's stance.

B) Low Level of Knowledge
When the participants started preparing their PhD proposals, they were not able to position themselves due to their low level of knowledge and unfamiliarity with their research topics.Consequently, the participant's position was not strong enough to enable him to criticise or argue about a certain issue.As the level of knowledge might be insufficient, there was also lack of awareness on whether the cited information was an established fact or an argumentative issue that could be opposed or supported.
For Mahmoud, during the PhD proposal preparation stage, he was not able to adopt a stance when citing since he was not really familiar with all aspects of his research topic.However, this stance changes at the final stages of the PhD.
" The weak position of the participant during the PhD proposal preparation was based on the fact that he was still involved in reading, and the research topic had not been covered entirely.Thus, this negatively affected the way of citing and evaluating the cited information.

C) Lack of Evaluative Vocabulary
Evaluating the citation requires students to utilize expressions that are used in criticising, emphasising or showing doubt.However, lack of knowledge in the evaluative expressions formed another obstacle that limited participants' attempt to evaluate the reported information.
"I do not know, I comment in general.I do not know if there are specific words that we use" (Suliman) Suliman lacked the vocabulary used to show degrees of certainty, usuality, and probability.Therefore, for him, evaluating the cited information was achieved generally by using multiple citations that support or oppose a certain idea.

D) Politeness and Tension
Politeness also formed a factor that limited the participants in evaluating the citation.One of the participants was hesitant to criticise other authors' findings.For Nabeel, it was impolite to underestimate other researchers' works since these studies consumed much of the authors' time and effort.Consequently, for him, evaluation of citations was based mainly on revealing the limitations of other studies, such as the population, the sampling, and the number of the respondents.In other words, the evaluation of the cited information was based on methodological limitations.
"I do not doubt the credibility of the study and the researcher.In order to be polite in criticism of others, I say Ahmad stated within his study.He found that this variable was significant, but the limitation with Ahmad's work was the number of the respondents was small.The scope of his study was only in one hospital; therefore, I cannot generalise the findings of Ahmad, and these findings reflect only the environment of only one hospital.I do not decrease the importance of his work, but I depend on the size of the population.For example, for this study, the scope of this study, if it was limited, I say that these findings do not form the majority because the study was limited within its scope.' (Nabeel) As Nabeel clarified, adopting a stance towards the citation was achieved through discussing the methodological limitations for two main reasons, which are showing respect to other researchers and revealing politeness.This could reveal that the participants had their own perceptions towards the issue of stance.In other words, their cultural background probably affected their way of critiquing.
Adopting an attitude towards the citation was also affected by the student's tension.That is to say, one of the participants was afraid that the cited work that he criticised might belong to one of the examiners.As shown in the extract below, Mahmoud assumed that criticising their works possibly evoked a sort of tension between the student and the examiner.Hence, it can be said that politeness and tension formed a reason for abstaining from criticising other researchers' works.Some participants thought that criticising a certain study implied criticising the authors.Their Arab cultural background might explain the reluctance to criticise other researchers' findings.In summary, it can be concluded that showing attitude towards the citations can be probably affected by field, the low level of knowledge of one's topic, lack of evaluative vocabulary, and showing politeness and tension.

3) Insufficient Knowledge in Using Citations
Students in IT depended on either the American Psychological Association (APA) author-date referencing style or the number system in using citations.Due to insufficient knowledge on using citations, the participants employed the two types of citations (integral and non-integral citations) randomly.In other words, they had insufficient knowledge about the different implications of using integral and non-integral citations.Consequently, their choice of either type of citations was according to their own perceptions, in order to change the monotonous presentation of the citation.

A) Limited Guidelines and Instructions
The majority of the participants attended courses in Research Methodology and Academic Writing at the beginning of their postgraduate study.In the academic writing course, although they received information about citation uses, it was insufficient and focused on author-date referencing system rather than the (Vancouver) number system.Some participants used the IEEE number system in citations, which was different from the APA system.For example, Faisal did not have enough information about using citations, particularly in using the number system since the information he received focused only on APA style.
' 'No, no, in this way, no.I only learned how to make citation using APA style.I remember this, in the master's degree only'' (Faisal) Ahmad also revealed that no information about using citations was received in his academic writing and research methodology courses.
"I repeat again, unfortunately, we did not study such issues at all.Even in academic writing, we learned how to write an essay.For example, citations, how to criticise, how to report, to be a critique, the verbs that should be used, how to use it, favouring verbs, all these, we have not learned."(Ahmad) For Nabeel, although he attended Research Methodology and Academic Writing courses, he did not obtain enough information on using citations.
"It was improved through reading.The benefit I got from research methodology and academic writing was limited.
Frankly, my writing was enhanced through reading continuously."(Nabeel) As reported by the participants, instruction on citations was received at the beginning of their PhD study, but only limited information focused on the use of citations.Hence, the supervisor's feedback on their writing and following the style of other PhD theses were the only two sources of knowledge for using citations.Consequently, the participants selected types of citations, i.e., integral or nonintegral citations, randomly.This action reflected that they were unaware of the evaluation implied in each type of citations, based on the participants' perspectives.

B) Students' Struggle to Self-regulate
Due to the insufficient knowledge on using citations, the participants used citations based on their own beliefs.For example, they used the integral citation to emphasise an idea.They depended mainly on the name of the cited author that can give more credibility to the cited information.For them, citing the information and providing details about the name/s of the cited author/s, date of the publication, and the number of the reference could probably make the reader and the examiner more confident of the significance of the reported information.Mahmoud believed that the integral citation gave more strength to the idea.
"Mentioning the name of the author with the number of the reference is a sort of more emphasis, and I feel that it is stronger information, I feel so.As I told you, I am in the academic writing and such matters, my information in such matters may not be completely accurate."(Mahmoud) Faisal also shared a similar belief with Mahmoud in terms of using citations to highlight ideas.He believed that the integral citation was used to emphasise the ideas.
"Frankly, in my thesis or in my study, I used example 3 (nonintegral citation) and example 4 (integral citation), but when I want to emphasise and highlight an idea, when I want to highlight and do not want to say that this is my idea, I use the author's name together with the reference."(Faisal) It can be argued, based on the participants' use and perceptions, that in writing instruction, citation had not received a lot of interest; rather, the focus was directed towards other aspects, such as essay writing, the organisation of thesis, and what to write in each chapter.
Consequently, academic writing courses should focus more on citations as an important issue in academic writing.

4) L2 Difficulties
The participants' challenges in using citations are also associated with L2 difficulties.These included paraphrasing and L2 academic writing.

A) Paraphrasing
The participants' problems in citation use are represented by their difficulty in paraphrasing citations.They needed to rephrase the cited information in a way that could keep the original meaning and avoid plagiarism.As Ahmad stated, "Frankly, there were many difficulties.We do not know the ideal way to represent the information.We do not know how to make a representation of the cited information."(Ahmad) Paraphrasing the cited information using the student's own vocabulary and keeping the original meaning of the reported information resulted in further associated challenges.These included students' inability to deliver the information accurately.
"You have the idea in your mind, but when you reread what you have written, you find that the idea is different from the original one".(Mahmoud) It can be said that paraphrasing formed another challenge for the participants due to difficulties in rephrasing the citation in an ideal way, and keeping the original meaning of the citation.

B) L2 Academic Writing
Since citation is an essential aspect of academic writing, the process of citing the information from its original source and integrating it into one's literature review chapter demands a variety of academic writing skills.However, the participants reported that they faced a challenge in using citations mainly because they were not proficient in L2 academic writing.As Suliman pointed out: "There is a difficulty in writing in English, using grammar in general, and also academic writing, how to write academically, in an academic manner, and citing in general."(Suliman) Writing in English formed the most basic challenge since it entailed several requirements, such as using correct grammar, using appropriate punctuations, and employing academic vocabulary and expressions.For example, the semi-colon punctuation mark (;) was not used or it was rarely used in the writings of these participants.
Ahmad did not know how to use the semi-colon.
"Frankly, for the semi-colon, I never used it.We have not learned such issues; our learning was based on following others.I could not find other researchers who used it in order to use it.The reason is I do not know how to use it and could not find any other researchers who used it.I represent others in their style of writing."(Ahmad) Using punctuation marks, such as the semi-colon, is essential when using citations with clause complexes.The absence of the semicolon may result in the low use of elaboration as a subtype of logico-semantic relationships in a citation with clause complexes.Suliman added another problem related to writing in English, which was using verbs in the citations, as well as using the tenses of these verbs.
"I lack a lot of information in academic writing and English language in general and also using verbs and tenses in writing."(Suliman) English is not their native language; therefore, the participants did not know which expressions and terminologies were more appropriate, stronger, and suitable for their academic context.This forced Mahmoud to memorise expressions through reading the different sources in his field.
"There might be other expressions, but you do not know which one is the most appropriate, which is the strongest, and which one is most appropriate in the academic field.I have become trying indirectly to memorise these expressions through reading the papers."(Mahmoud) Citing information from articles, survey papers, books, and theses and integrating it into one's writing was also affected by the L1.
All the participants' L1 in the present study was Arabic, but they had to write in English as a Foreign Language.In citing sources, the L1 had effects on L2 in two aspects.First, the idea in the citation was translated into Arabic to understand it, then it was written in English based on the Arabic interpretation.Second, there was the difficulty in understanding students' writing by the supervisor(s) and examiner(s).For Faisal, his L1 affected his writing of citations positively and negatively.The positive effect is represented by selecting English vocabulary that is equivalent to strong Arabic vocabulary.On the other hand, the negative effect of Arabic language is embodied in thinking in Arabic and writing in English.
"Yes, it affects both of them, to be honest.In Arabic, we use a lot of expressions.The Arabic language is no doubt a strong language, and when we compare it with English or translate literally, word by word, this is catastrophic and this is the way we use.We think in Arabic and speak in English.This is a disadvantage.The positive effect of the Arabic language sometimes gives you shortcuts.Instead of writing a long sentence in English to express a small idea, you can look for an equivalent word in Arabic.For example, instead of saying today is very crowded, in Arabic we say 'Zahma', and that is all."(Faisal) Processing the idea mentally in Arabic language and writing it in English resulted in a difficulty in understanding the students' writings by their supervisors as well as examiners.This problem had been commented on by the students' supervisors and examiners during the proposal defence session.This was reported by Suliman.
"Sure, you always try to write in your mother tongue, that is the Arabic way.You have a certain way of expressing, you find a difficulty in English, and Arabic language, mostly.For example, one of the comments was by the examiner during the proposal defence.She mentioned that the majority of my sentences are used in my own mother tongue, and she said that some sentences are not clear.''(Suliman) Writing long sentences was also one of the problems faced, particularly the overuse of the relative pronouns 'that' and 'which' and the effect of L1.For example, Mahmoud used long sentences inappropriately due to the frequent use of the relative pronouns.This resulted in an increase in the number of grammatical mistakes committed.
"I had a sort of inappropriate use of long sentences, then I tried to make the sentences shorter, I omit 'which' part.Sometimes, when I write, the sentence is three lines long, and when the sentence is long, the average of making mistakes may increase.You are using a compound sentence that has connectors, so I tried to use 'that' and 'which' only when necessary, but when I started writing, I use 'that ' and 'which' repeatedly." (Mahmoud) Khalid added that the long sentences had more details, but the idea was not communicated appropriately."We up till now, our thinking in structuring the sentence is Arabic thinking rather than English thinking in writing.Therefore, our sentences are always long, have a lot of explanations, and do not communicate the idea correctly and exactly."(Khalid) In summary, writing in English and using citations academically constituted a great challenge for the participants, since English academic writing required mastering several issues, such as academic vocabulary and expressions, correct grammar, and correct use of punctuation.Their low proficiency in English and the influence of their L1 resulted in inappropriate use of citations.

DISCUSSION
The present study explored the challenges faced by EFL Arab doctoral students in an Information Technology department, in citing and integrating information from different sources into the literature review chapter of their PhD proposals.The findings demonstrated that they lacked awareness and had insufficient knowledge about the use of citations in academic writing.Only limited instruction was received on using citations during their attendance in academic writing courses.Feedback received from supervisors and following other writers' style were the two main sources of knowledge on using citations.However, these are considered insufficient because, as Nguyen and Pramoolsook (2016) argue, learning how to use citations implicitly is insufficient.The participants' preference for the integral citation over the non-integral citation was due to their perception that the integral citation was stronger in presenting reliable and credible information.They also believed that the integral citation made the examiner and the supervisor more confident about the cited information, as well as attracted the attention of the reader to something important.However, this finding contrasts with the citations practices in the hard domain.For example, in the hard domains such as Information Technology, the non-integral citation is over emphasised, and the role of the cited author is downplayed (Hyland, 1999;Swales, 2014;Thompson, 2001) The participants used long sentences in citations because these long statements enabled them to express ideas in more detail.In spite of this advantage, using citations with clause complexes may increase the number of mistakes due to insufficient knowledge in using conjunctions, punctuations, and other discourse markers.Using such long statements also seems to be the effect of their Arabic mother tongue language.This finding is similar to the effect of the French language on using citations, as demonstrated in a study by Rowley-Jolivet and Carter-Thomas (2014) on 40 pre-publication uncorrected drafts of articles written in English by expert French writers in Engineering, Science, and Computational Linguistics.Using reporting verbs, attribution and stance represented a problem for expert French writers of English due to their lack of awareness of English structures used for reporting, insufficient mastery of English and/or the influence of the French language.The French language may have other forms of citation, and writers of this language carried the forms of their native language into their English academic writing.In the current study, citing and evaluating information were also affected by participants' low level of knowledge on their research topics, their weak position to argue and claim, insufficient knowledge on evaluative expressions, and the effect of their field.This latter element, i.e., field, supports one of Halliday and Martin's (1993) register variable in Systemic Functional Linguistics that affects the use of language.In other words, these students faced difficulty in adopting a stance toward the citation because Information Technology is a factual field that limits a student in using evaluative expressions or playing on words.Consequently, the reluctance of the students in the IT discipline to adopt a stance is consistent with the findings of Hyland (2005).In his study, Hyland stated that the highly formalised reporting system adopted in the hard disciplines minimised the presence, the voice, and positioning of the writers in their texts.
Overall, the participants showed insufficient knowledge about citation practices.Their lack of competence and performance in using citations limited their ability to write cohesively and argue strongly.Little knowledge was gained from taking an academic writing course at the beginning of their PhD study due to its focus of on essay writing, the structure of the proposal and thesis, and the use of the author-date referencing style.Researchers have started to realise that university students have numerous problems working with sources, projecting a voice, and developing ideas, and such issues have not received much focus in terms of the disciplinary conventions (Donahue, 2004).In this regard, Liu and Huang (2005) argue that the increase in accessibilty to information forms exceptional challenges, because selecting credible sources has become a daunting task for students.Thus effective evaluation of the information is an educational necessity for all students in order to raise awareness of search strategies and criteria to find credible information (Currie, Devlin, Emde, & Graves, 2010).Students' weaknesses in citations and other related issues, such as reading reliable sources, paraphrasing, evaluating, writing academically (Hei & David, 2015) and citing sources efficiently and appropriately (Rabab'ah & Al-Marshadi, 2013) need to be further enhanced through developing appropriate academic writing teaching materials.
The contents of such materials should focus more on using citations, using evaluative expressions, hedges, boosters, using tenses, and reporting verbs.

CONCLUSION
This study contributes to the growing interest in research on academic writing, focusing mainly on the aspects of citations.By applying DBIs under the New Rhetoric studies to explore EFL Arab postgraduates' perspectives on using citations, beneficial outcomes could be obtained.Analysing the citations types in the different genres led to the diversity in the findings of the previous studies (Hyland, 1999;Nguyen & Pramoolsook, 2016;Polio & Shi, 2012;Rabab'ah & Al-Marshadi, 2013;Swales, 2014;Thompson, 2001Thompson, , 2002;;Thompson & Tribble, 2001).Some findings of these previous studies were in line with the norms of the hard domain or the soft domain (Hyland, 1999), whereas other findings were in contrast.
Other studies focused on identifying the functions of integral and non-integral citations (Petric` & Harwood, 2013;Samraj, 2013).Some other researchers focused on the role of the cited author in the integral citations (Jomaa & Bidin, 2016;Swales, 2014).The interpretations of the findings in these studies proposed that nonnative writers of English have challenges in using citations and lack skills of academic writing.However, in the present study, the focus was mainly on what makes citations challenging for them.Hence, the present study involved identifying the problems faced by EFL Arab postgraduates in the different stages of citing the information, such as reading scientific texts (Fatonah, 2014;Hei & David, 2015), paraphrasing, evaluating, integrating and writing (Hei & David, 2015).
Although the present study is limited to only six male students from the Information Technology discipline, the deep exploration of citations with the participants presented significant outputs.Thus, including more students from other disciplines may enrich the results and present us with more knowledge about the challenging issues in citations.Second, by including other students from the outer circle, such as ESL postgraduates and the inner circle, such as native speaker writers, more insights could be obtained about these challenges.Third, the study is limited to EFL Arab postgraduates who shared relatively similar socio-cultural and educational backgrounds, which had an effect on citation processes, mainly in the influence of the Arabic language on paraphrasing and writing.
Hence, having other EFL postgraduates who belong to other cultural and educational backgrounds may present more evidence on the effect of language backgrounds on using citations in academic writing.Fourth, challenges in using citations were experienced in writing the literature review chapter of PhD proposals.Thus, involving other genres such as research articles and dissertations might result in more perspectives about the challenges in the citation process.As a whole, conducting retrospective interviews that are discourse-based under the New Rhetoric studies presented significant findings of the citing behaviors of EFL Arab doctoral students in writing the literature review chapter of their PhD proposals.In future studies, combining the perspectives of New Rhetoric studies that focus on the context of the text and Systemic Functional Linguistics that focuses on the text in context could lead to more important information about students' use of citations and perceptions.

Table 1
Background Details of Six Male Participants

Table 2
Challenges in Using Citations by EFL Arab Postgraduates tension for the examiner, specifically, imagine that by accident, one of the works belongs to the examiner.In such a case, the tension will be higher, you know."(Mahmoud) When you see a student whose writing is full of criticism and uses strong words, (such what?) bad performance, poor performance, lack of efficiency, for example.I am not against using these words because there exist such mechanisms.But when you only focus on such expressions and forget what they have contributed even if we cannot find it now, this may create a sort of