Chinese undergraduate EFL learners’ perceptions of Plagiarism and use of citations in course papers

Abstract Source-based writing research has received much attention in recent years, which generally shows that both novice and expert EFL (English as a foreign language) writers have difficulties in writing from sources. As many Chinese institutes of higher education attach increasingly more importance to publications in international journals, citation and plagiarism become critical issues for both student and teacher researchers as well as the institutes. Nevertheless, not much research can be found on the issues with Chinese students, especially undergraduate students. The present study thus investigated Chinese undergraduate EFL learners’ perceptions of plagiarism and use of citations in their course papers. A total of 141 students from a highly prestigious university answered an open-ended questionnaire and 97 of them submitted course papers. Major findings were: (1) the students had a (fairly) good knowledge of plagiarism and identified various reasons for plagiarism in academic writing, (2) they used summary the most often when citing from sources, followed by paraphrase and quotation, and (3) they mainly used single-source citations to primarily support their own ideas, position an author’s opinions and/or findings, and acknowledge the author’s ideas. These findings reveal a general overview of students’ perceptions of plagiarism and use of citations in their course papers, thus providing implications for formal classroom instruction of writing from sources.

ABOUT THE AUTHORS Dr. Meihua Liu is professor of Applied Linguistics at the Department of Foreign Languages and Literatures, Tsinghua University, China. Her research interests mainly include EFL teaching and learning, reticence and anxiety, EFL writing, and International education. Presently she is working on anxiety and sources of anxiety among Chinese university teachers and young learners of English and second language writing. The present paper reports on the results of a study on citation practices in academic writing by Chinese college students and their perceptions of plagiarism.
Dr. Yong Wu is a faculty of School of Humanities, Beijing University of Posts and Telecommunications. Her research interests mainly include EFL writing and Educational technology. Presently she is working on second language writing with her colleagues.

PUBLIC INTEREST STATEMENT
Writing from sources is a typical feature of academic writing. Yet writing from sources, especially in the target language, is challenging due to various reasons such as writing conventions, language proficiency, and discipline requirement. The present study examined Chinese undergraduate English learners' perceptions of plagiarism and use of citations in their English course papers.
Course papers and open-ended questionnaires collected from 141 students from a researchoriented university revealed that: (a) The students had a (fairly) good knowledge of plagiarism and identified various reasons for plagiarism in academic writing, (b) they summarized the most often when citing from sources, followed by paraphrase and quotation, and (c) they mainly used single-source citations to primarily support their own ideas, position an author's opinions and/or findings, and acknowledge the author's ideas. These findings further pinpoint the difficulty of writing from sources to non-native speakers and justify the need for continuous research on the issue.

Introduction
A recent trend in writing research is source-based writing, in which writers read one or more source texts and write with the information from the sources (Weigle & Parker, 2012). In source-based writing, reading and writing blend and influence each other (Bracewell et al., 1982) so that it is a recursive, complicated, and challenging procedure (Spivey, 1990). Closely linked to source-based writing is the issue of how to borrow or cite source content, especially in second/foreign language (SL/FL) writing.
As writing from sources has caught increasing attention of researchers, more and more studies have been done on it (Hirvela & Du, 2013;Howard, 1999;Hyland, 1999;Pecorari, 2003;Wette, 2017Wette, , 2018. These studies have shown that both novice and expert L1 and L2 writers have difficulties and tend to use paraphrasing, patchwriting, summarizing, and synthesizing when citing from sources though more expert writers use summarizing and synthesizing more often than less experts ones. Less proficient writers often cite to display knowledge and attribute information more than for other purposes (Hyland, 1999;Mansourizadeh & Ahmad, 2011;Petrić, 2007;Wette, 2017;Zhang & Deng, 2019). If citing inappropriately due to such reasons as cultural differences, low proficiency, and lack of practice and training, writers may be likely to plagiarize in their sourcebased writing (Borg, 2000;Campbell, 1990;Hu & Lei, 2012;Kroll, 1988). Even so, most studies target postgraduate students in English-speaking environments and/or expert writers and few have been done on ESL/EFL 1 (English as a second/foreign language) undergraduates, as reviewed in Cumming et al. (2018). Studies are even fewer on Chinese EFL undergraduates, which deserves more research considering their large population and huge diversity (Cumming et al., 2018). It deserves research also because it is normal for Chinese students to model on good essays and use others' ideas and sources without clear citations in their Chinese writing. This habit is often transferred to their English writing, as observed by the authors of this paper in their teaching and discussed in Qin (2013). Nevertheless, this habit of writing is different to the convention of English writing, which usually requires clear citations (Gu & Brooks, 2008;Kroll, 1988). Consequently, the present research, targeting Chinese undergraduate novice EFL writers, aimed to explore their perceptions of plagiarism and use of citations in their course papers, hoping that the findings would be able to enhance instruction and improve students' ability to cite source materials in academic English writing (Cumming et al., 2016).

Use of citations: Frequency, type, and function
A typical feature of academic writing is to write from sources (Hyland, 1999). As found in Wette (2018) and Cumming et al. (2018), students often turn to internet, books, dictionaries, research papers, textbooks, classroom lectures and notes, documents, and so on when writing from sources for academic purposes. This source-based writing is apparently a complicated and challenging process in which writers integrate what he/she reads into what he/she wants to express (Campbell, 1987). To accomplish this task, patchwriting, paraphrasing, summarizing, and synthesizing are commonly used strategies by both expert and novice writers (Hirvela & Du, 2013;Howard, 1999;Hyland, 1999;Pecorari, 2003;Wette, 2017). During this process, various factors may affect students' use of sources such as language proficiency (Kim, 2009), difficulty of source materials (Wette, 2017), ability to paraphrase effectively (Campbell, 1990;Wette, 2018), cultural differences (Shi, 2006), knowledge of appropriate use of source materials (Pecorari, 2003;Shi, 2006Shi, , 2008Wette, 2017Wette, , 2018, training and practice (Grabe & Zhang, 2013;Kim, 2009;Petrić, 2012), and so on. For example, less proficient students engage in much more copying than more proficient students and depend heavily on the language of source texts (Grabe & Zhang, 2013;Kim, 2009); students can do better in summary writing when the source materials are easier (Kim, 2009). Lack of knowledge of academic norms may also lead to students' failure of identifying the sources (Chandrasegaran, 2000;Hu & Lei, 2012).
Research on citation practices in source-based writing has primarily concerned with different types and functions of citations and writer stance (Harwood, 2009;Hyland, 1999;Petrić, 2007;Swales, 1990). For example, Swales (1990) categorized citations into integral (i.e., citing the author within the grammar of a sentence) and non-integral (i.e., referring to sources in brackets) citations. Hyland (1999) grouped citations into four types-direct quotation, block quotation, summary, and generalization. Studies have shown that experienced writers are able to cite from multiple sources and know how to position source materials to best support their arguments (Harwood, 2009;Hyland, 1999;Mansourizadeh & Ahmad, 2011), while inexperienced undergraduates and postgraduates tend to use single sources and paraphrase citations, primarily to attribute a finding or idea to a particular source (Abasi & Akbari, 2008;Hirvela & Du, 2013;Lee et al., 2018;Li & Casanave, 2012;Petrić, 2007;Petrić & Harwood, 2013;Wette, 2017). For example, Hyland's (1999) analysis of 80 published research papers written by experienced writers from leading journals in eight disciplines found that all the writers cited a lot in their papers and used summary the most often while citing, followed by generalization, with little quotation or paraphrasing. Lee et al.'s (2018) study of source-based research papers produced by L2 undergraduates in their first-year writing also showed that summary was the most commonly used citation form, but followed by quotation, as found in Wette (2017). Based on analyses of 66 questionnaires, 13 texts and 13 interviews, Wette (2018) investigated source text use of first-year undergraduates in health science courses in terms of: Placement, source type, citation type, reporting forms, formatting, density, rhetorical purpose, and quality. She found that: (a) Most citations (92%) were located in the body part, with few in the beginning and conclusion parts; (b) disciplinary and assignment topics influenced source type choices; (c) the density of citations per 1000 words was 11.75, similar to the findings of expert writers in Hyland (1999), Mansourizadeh and Ahmad (2011), and inexperienced writers in Petrić (2007) and Wette (2017); (d) nearly two-thirds of citations were made to attribute information to a particular source or sources without any evaluative comment, typically characteristic of novice writers' works; (e) no extensive copying was identified and less than 20% of citations were patchwriting; (f) 65% of citations were accurate and complete; and (g) students cited largely to transmit views rather than evaluate others' views or advance their own ideas.
To evaluate Chinese students' writing from sources, Cumming et al. (2018) surveyed the practices and abilities of 103 students at 4 universities in China to write from sources in English and analyzed citations in a subsample of 41 matching sets of course papers in their first and second years of Bachelors' and Masters' programs. The citations were analyzed in terms of type (e.g., quotation, paraphrasing, and summarizing), relative accuracy, presence, and discourse functions (e.g., position, definition, supporting, acknowledging, agreeing, disagreeing, application, topic relevance, Petrić & Harwood, 2013). The results were: (a) The students wrote with some proficiency in English and emerging competencies in writing from academic sources, (b) little developmental difference in frequency, accuracy, and functions of citations was observed for undergraduate students but more for graduate students from their first to second years, and (c) most students reported greater success in acknowledging source materials in their second years. Thus, though half the students had taken courses that helped them to write from sources, the researchers still recommended formal instruction to help students write better from sources. Other researchers generally found that Chinese writers, especially novice writers, had difficulty citing sources in their academic English writing (e.g., Liu, 2005;Qin, 2013;Zhang & Deng, 2019). This is normally attributed to the traditional Chinese education which emphasizes the importance of modeling on others' works, especially master works, with acknowledgments but not necessarily with clear citations. Therefore, most Chinese teachers and students are generally not used to giving clear citations in writing, even when writing from sources. Lee et al. (2018) analyzed rhetorical functions of citations and writers' stances of 100 sourcebased research papers produced by ESL undergraduates. They found that the major rhetorical function of citations (approximately 87%) was attribution, followed by evaluation, exemplification, and establishing links between sources, respectively, consistent with the finding in Petrić (2007). Lee et al. (2018) also found that acknowledgement (67%) was the dominant stance that L2 undergraduate writers took in source-based writing, followed by distance, contest, and endorse, respectively. This finding further showed that inexperienced students mainly cited to attribute, as found in earlier studies (Bereiter & Scardamalia, 1987;Dong, 1996;Hirvela & Du, 2013).

Perceptions of plagiarism
Closely related to source-based writing is the issue of plagiarism. Youmans and Evans (2000, p. 114) defined plagiarism as "involv[ing] the use of someone else's writing as one's own without giving credit to the author, whether the source be a published work or a classmate's essay . . . Asking or hiring another person to write a paper or buying a paper written by another to submit under one's own name is included within this definition". Howard (2000) considered clearly intentional deception as plagiarism.
Plagiarism has been observed in the writing produced by both native speakers and ESL/EFL students, who have different opinions about plagiarism due to reasons like cultural difference, lack of training, discipline culture, lack of personal authority, and low proficiency (Borg, 2000;Campbell, 1990;Hu & Lei, 2012;Kroll, 1988;Mohan & Lo, 1985;Szadziewska & Spigarska, 2019). Kroll (1988) found that American university ESL students viewed plagiarism as unfair to authors and believed that students should be responsible for their work independently. However, their Chinese counterparts thought differently. Chu et al. (2019) explored 433 Hong Kong secondary school students' perceptions and actual understanding of plagiarism. Analyses of the questionnaires revealed that the students had different understanding towards "obvious" and "obscure" plagiarism, with more difficulty on obscure plagiarism, and that students' perceptions and understanding of plagiarism differed across grade levels. Campbell (1990) compared the source text use of English native speakers and ESL undergraduates. The major difference between the two groups was that ESL students showed more reliance on source materials in the beginning paragraphs. Moore (1997) and Shi (2004) found that L2 undergraduates relied more on source texts than their L1 counterparts in summary writing. Deckert (1993) found that Chinese college ESL learners perceived plagiarism to be wrong and thought that plagiarism negatively influenced their learning and personal integrity. Gu and Brooks (2008) investigated 10 Chinese students' perceptions of plagiarism during their MA study in a UK university. The data included two formal rounds of semistructured interviews focusing on their difficulties on academic writing and their attitudes towards plagiarism over time. They found that students were confused about British academic conventions, which "places a much higher value on the demonstration of profound collective knowledge and the presentation of the essay than on citation and referencing" (p. 344). For example, students lacked knowledge of why referencing was needed and the rules of referencing. The students also found it hard to distinguish their own original ideas and others' opinions. These findings echo with the finding discussed earlier that many Chinese writers have difficulty citing sources in academic English writing due to traditional Chinese culture and education (Qin, 2013;Zhang & Deng, 2019). Qin (2013) found that Chinese EFL learners had difficulty identifying implicit plagiarism in writing and believed that it demonstrated writers' good use of what they had learned. Thus, the researcher claimed that there existed differences between Chinese EFL learners' perceptions of plagiarism and Western academic writing norms and suggested that academic institutes should clearly define plagiarism and specify citing norms in academic writing. Zhang and Deng (2019) also found that Chinese postgraduate EFL learners had difficulty identifying implicit plagiarism and inadequate citations. Moreover, they found that writing practice significantly enhanced students' ability to cite from sources but helped little with their perceptions of plagiarism. They hence claimed that explicit instruction and writing practice were necessary to improve students' ability to cite from sources and avoid plagiarism.
As reviewed, quite many studies have been conducted on plagiarism and citation practices in writing from sources. Nevertheless, most of such studies target postgraduate students in English-speaking environments and/or expert writers and few have been done on ESL/EFL undergraduates, as reviewed and discussed in Cumming et al. (2018) and Wette (2017). Studies are even fewer on Chinese undergraduate EFL learners, which deserves more research considering their large population and huge diversity (Cumming et al., 2018). Moreover, as discussed in Introduction, it is often considered normal to model on good essays and use others' ideas and sources without clear citations in Chinese writing, while English writing usually requires clear citations (Gu & Brooks, 2008;Kroll, 1988). Thus, Chinese EFL learners, especially untrained learners, often transfer their habit of writing without clear citations in Chinese to their English writing. This, however, often puts them in an embarrassing situation of plagiarism which they themselves are not aware of. Clearly, perceptions of plagiarism and use of citations are closely related to each other, which, nevertheless, has been seldom examined on the same learner population. For these reasons, the present research, targeting Chinese undergraduate novice EFL writers who had limited instruction of plagiarism and limited practice of writing from sources in English, aimed to explore their perceptions of plagiarism and use of citations in their course papers. The following two questions were of particular interest: (1) How do Chinese undergraduate EFL students perceive plagiarism?
(2) How do Chinese undergraduate EFL students use citations in their course papers in terms of frequency, type, and function?
In the present study, plagiarism refers to the use of someone else's writing as one's own without giving credit to the author, whether the source be a published work or a classmate's essay, intentionally or unintentionally. Inappropriate or wrong citation is not considered plagiarism. Nevertheless, plagiarism occurs if direct quotation accounts for more than one fourth of an essay even if with appropriate citation.

The present study
In the recent decade, as many Chinese institutes of higher education strive for international reputation, more and more colleges and universities attach importance to research ethics, including citations in research papers, although the focus is often on postgraduate students in nationfamous institutes. The present study was conducted in a highly prestigious research-oriented university in Beijing, which has been attaching more importance to publications and patents and been more aware of the importance of research ethics than most other universities in China.
Because of the increasing focus on research, the university has been offering seminars and/or courses on research ethics to students on campus in recent years. Plagiarism has also become a topic of discussion in various courses for both undergraduate and postgraduate students, although there was no written description of how to avoid plagiarism in research paper writing when the study was conducted. For example, researchers would be punished if they are reported to have plagiarized in their publications, and all degree theses are subjected to repetition checks via a software developed by the university prior to submission for committee evaluation. Although undergraduate students are normally not required to write for publication, they do need to write essays from sources. In this context, it is important to explore how they perceive plagiarism and cite source materials in their academic English writing.

Participants
Altogether 141 (71 male and 70 female) undergraduate students of hard disciplines (e.g., civil engineering, media and communication, pharmacy, biomedicine, energy and power, etc.) of the university enrolled in the same course which required them to write three 400-word papers from sources participated in the present study. They were all intermediate to high proficient English learners, most of whom were first-year students (89), with 26 second-, 9 third-and 17 fourth-year students. Thus, their proficiency in English was considered as a control variable in the present study.

Data
The data gathered in this study included student texts and an open-ended questionnaire, as detailed below.

Student texts
A total of 97 course papers on global warming written by the participants were collected. The essays, ranging from 258 words to 964 words long with an average length of 453.63 words (SD = 117.18), focused on different aspects of global warming (e.g., "the effects of global warming on agriculture" and "causes for global warming"). This was the second completed task when the present study was conducted. Considering that students might not be so familiar with rules of citing when writing up their first task yet gained such experience more or less when working on their second task, essays of this task were used for analyses in the present study.

The open-ended questionnaire
The open-ended questionnaire had two general, open-ended questions: what is plagiarism and what causes plagiarism in academic writing.

Procedure
A research assistant emailed the questionnaire as well as a consent form in Chinese to students who registered in English writing courses in the last 3 weeks of a 16-week semester, which resulted in 141 responses. After that, the respondents were asked to submit the essays that they wrote for the courses, with a final total number of 97 essays collected.

Data analyses
The collected student papers were analyzed in terms of total number of words, frequency, types, presence, and functions of citations. The framework for presence, types, and functions of citations was adopted from that used in Cumming et al. (2018) which categorizes citations into three types (i.e., quotations, paraphrase, and summary). The framework for functions of citations was adopted from that in Petrić and Harwood (2013), which classifies functions of citations into 8 types: position (to present an author's viewpoints and/or findings), defining (to help define or explain a concept or approach), supporting (to help justify the writer's claim or idea), application (to apply a concept, approach, or theory to the writer's own analysis), topic relevance (to help justify the writer's choice of topic), disagreement (to express the writer's disagreement with an idea), acknowledgement (to acknowledge the author of an idea or term), and agreement (to express the writer's agreement with an idea). In real writing, one citation might serve more than one function.
The responses to the open-ended questions were analyzed according to themes (Richards, 2009) by two research assistants, with an inter-rater reliability .936 and .952 for responses to questions 1 and 2, respectively. As discussed in Richards (2009), any unit like word, phrase, sentence, and idea can be a theme, according to which, qualitative data can be coded, analyzed, and grouped. In the present study, example themes of responses to question 1 were "copying", "citation", "acknowledgement" and "permission". Example themes of responses to question 2 were "being lazy" and "lack of innovative ideas". To protect privacy, a number was assigned to each respondent, which was then used when reporting their remarks in this paper.

Perceptions of plagiarism
When asked what plagiarism was, all the 141 respondents of the open-ended questions offered definitions. Most of these definitions shared the following components: (a) Using published ideas of others or their own without any citation, (b) copying others' ideas partially or completely without acknowledgment, (c) copying blocks of others' expressions, (d) using others' copyrighted diagrams, data, patents, technologies, and so on without permission, (e) copying others' homework and assignments, (f) copying classroom work, and (g) citations accounting for a large portion of a paper. For example, as defined by S17, ". . . Plagiarism means that a person unduly or even maliciously cites others' ideas or opinions without acknowledgement, or copies and pastes what is others' to be his/her own, or directly copies others' papers and gets them published in his/her name". According to S99, plagiarism "happens when a person purposefully or accidently copies sentences, paragraphs, sections, and even chapters word by word or with a few changes in wording, without any statement or acknowledgement". "Plagiarism is a kind of academic dishonesty, is often conceived as fraudulent behavior that diminishes the intellectual property of the original author and rewards plagiarists for their work" (S60). Though generally consistent with the definition used in the present research, some respondents had a broad definition of plagiarism, which defined such behavior as plagiarism like not citing the sources correctly, citing a massive part directly from another paper without proper acknowledgement, completely citing the results of authoritative publications, using others' views with wrong citations, citing others' views in an incorrect style.
In addition, they voiced various reasons for plagiarism, as summarized in Table 1. According to Table 1, being lazy was identified by 34.75% of the respondents to be No. 1 reason for plagiarism in academic writing. To gain benefits (27/19.15%) was another major reason, followed by seeking quick success and instant benefits (25/17.73%), lack of innovative ideas (16/11.35%), not knowing citing rules (14/9.93%), and the big research environment in China (13/9.22%), respectively, along with many other reasons such as poor use of English (e.g., poor paraphrasing ability), and little confidence in one's own research ability. As they reported, "many students and researchers plagiarized in their papers purposefully or even maliciously for various benefits like course credits, smooth graduation, scholarship, promotion, fame, and money" (S5). "Some people are eager to have papers of their own so that they can get better jobs, better positions or higher pay. But they do not want to spend time and effort on researching or are not good at it. Then they might plagiarize in their works." (S13). ". . . In many universities and institutes, there is a rule to encourage more academic publications that a researcher has to quit if he/she doesn't have enough publications or can't be promoted" (S47). A good illustration of the big research environment in China came from S21's statement, ". . . In most cases, students (researchers) are required to write up research papers and get them published in top-ranking journals within a short period of time.
Publications seem to be so important, exceeding other things. And because fake things exist almost everywhere, the academic society seems to be tolerant to plagiarism in that the punishment for plagiarism is rather light. All these lead to plagiarism". The lack of a sound regulatory mechanism (7/4.96%) and light punishment (7/4.96%) were also identified to be important reasons for plagiarism. ". . . There is no standard for judging what constitutes plagiarism. Some students don't know what plagiarism is. And though there is punishment for copying others' works, the cost is (much) less than the gains. All these lead many students and researchers to plagiarize" (S89). Apart from these reasons, rapidly developed internet was considered to contribute to plagiarism, because "it makes it so easy to search for relevant literature and thus makes it convenient to copy others' works" (S69).

Frequency, presence, types, and functions of citations in students' papers
As shown in Table 2, the total number of words of student papers ranged from 258 words to 964 words, with an average of 453.63 (SD = 117.18) words. On average, there were 4.22 (SD = 1.91) citations in a student paper, with an average of .28 quotations, 1.74 paragraphs, and 2.20 summaries, respectively. When using citations, the participants preferred summary, followed by paraphrase and quotation, respectively. Of the citations, all were single-source citation except for a few multiple-source citations, with an average of 3.34 being present with a signal and .86 being not. This shows that most participants knew how to cite and identified the sources of citations in their writing, although many citations might not be complete in providing all necessary information (namely, both in-text and aftertext citations were complete). This is illustrated by the following two examples. Example 1 indicated the person who made the remark but provided no further information like time and place of the remark either in or after the text. It was the same with example 2, whose source information 'S. Res, 98ʹ was confusing.
For both examples, it was difficult for readers to locate the sources.

Example 1:
According to Liangyan Xu, a senior engineer at the National Climate Centre of the China Meteorology Administration, warm winters caused by global warming make for the occurrences of extreme weather. (W21, paraphrase)

Example 2:
But soon after that, the U.S. government announced that the United States had not ratified the Protocol, said that the Protocol "would result in serious harm to the economy of the United States" (S. Res, 98). (W33, quotation) In terms of functions of citations, the means for position, defining, supporting, application, topic relevance, disagreeing, acknowledgement and agreement were 1.16 (SD = 1.29), .103 (SD = .34), 1.49 (SD = 1.22), .27 (SD = .62), .61 (SD = .69), .103 (SD = .39), .76 (SD = .93), .11 (SD = .38) respectively. This means that the participants used citations for diverse purposes, though mainly for the purposes of positioning, supporting, acknowledgement and topic relevance. The following are a few examples.

Example 3:
Global warming refers to the phenomenon that the rise in atmosphere and ocean temperature in a period of time. (W53, definition)

Example 4:
Some researches claimed that global warming may increase the temperature in South Africa and lead to the bigger possibilities of the malaria. (W61, position)

Example 5:
As the temperature rises, the sea becomes warmer and warmer, and the melting of the polar and alpine ice, leading to rising sea levels. The rate of 0.3 degrees per year as increasing temperature to calculate the next century, sea level will rise every 10 years 6 cm, which will have a great impact on the accounting for eight percent of the earth's surface and coastal areas (Guo YinYan, 1995). (W85, acknowledgement) In addition, as shown in Table 2, most skewness and kurtosis scores were less than 10, with a few exceeding 10 but not more than 16. This indicates that the scores for each measured scale of citations were generally in a normal range, with a few abnormalities being observed in functions of citations such as defining, application, disagreeing, and agreeing.

Perceptions of plagiarism
Consistent with the definitions reviewed in the literature (Howard, 2000;Youmans & Evans, 2000) and that used in the present study, most respondents in the present study viewed copying others' work unduely or intentionally, whether it was a published work or a classroom work, as plagiarism. This indicates that they generally had an appropriate knowledge of plagiarism, which might be attributed to the increasing awareness of and training on research ethics and conventions in the university. Different from the finding in Kroll (1988) but similar to that in Deckert (1993), this finding not only shows that context may be an influential factor affecting students' perceptions of plagiarism but further confirms that culture varies and should not be stereotyped, as implied in Liu (2005). Meanwhile, it is worth noting that some respondents had a broad definition of plagiarism, who considered not citing the sources correctly, using others' views with wrong citations and citing others' views in an incorrect style to be plagiarism as well. This understanding might be due to discipline difference in what constitutes acceptable source borrowing (Borg, 2000;Hyland, 1999), or to the issue of morality (Chandrasoma et al., 2004), or students' over-awareness. In any case, this suggests that the term plagiarism should be cautiously used.
When asked about why students/researchers plagiarized in academic writing, the respondents voiced a number of reasons like not knowing citing rules, poor English, little confidence in their own research ability, similar to those identified in the current literature (Campbell, 1990;Deckert, 1993;Gu & Brooks, 2008;Moore, 1997;Shi, 2004). What was special was that Chinese students/researchers were reported plagiarizing primarily for being lazy and gaining benefits in academic writing, especially in research papers. This was rarely reported though probably existent in other contexts. Lack of innovative ideas, lack of a sound regulatory mechanism, and light punishment worsened the situation. Moreover, due to technology, internet made plagiarism easy(ier) nowadays. Moreover, as Pennycook (1996) reminded us to be cautious about the understanding of author, ownership and plagiarism, and the relationships between different approaches to texts based on his teaching experiences in China and Hong Kong, not knowing what plagiarism was, lack of training, unfamiliarity with Western conventions of academic writing and the convention of modeling on good/authoritative texts (Gregg, 1986;Pennycook, 1996) might make the issue of plagiarism even more complicated. Consequently, this issue deserves more attention and appropriate training is needed.

Frequency, presence, types, and functions of citations in students' papers
Analyses of citations showed that the participants made a mean of 4.22 citations in the present research, most of which were signaled clearly, higher than the number for undergraduate students in Cumming et al. (2018). Nevertheless, the density of citations (i.e., the average ratio of citations per words written) in student papers in the present study was essentially the same (.01) as that for undergraduate students in Cumming et al. (2018) and inexperienced writers in Petrić (2007), higher than that for Malaysian and Chinese students in their third year at a New Zealand university in Wette (2017), but lower than that reported in Wette (2018) with first-year students in health science. This might be because a few papers in the present research were particularly long but not many citations were employed in them. Another possible explanation was that students were required to provide a reference for every citation in Wette's (2018) study.
Regarding types of citation, the participants adopted summary the most frequently, followed by paraphrasing and quotation, similar to Lee et al. (2018), Wette (2017Wette ( , 2018 and Cumming et al. (2018) who found that students summarized more frequently than paraphrased. This finding suggested that the students in the present research did not rely on direct borrowing from sources, which might be thanks to their high English proficiency and the training they had received. Nevertheless, as novice writers, they still had difficulty in citing from multiple sources and using multiple-source citations.
Meanwhile, the present research revealed that the students cited for a variety of purposes in their course papers, primarily to support their own ideas, position an author's opinions and/ or findings, and acknowledge the author's ideas, demonstrating that they voiced their personal authority rather than just present their reading. This finding was different from that in earlier studies (Lee et al., 2018;Petrić, 2007;Wette, 2018), which found that novice students cited largely to transmit views or tell old knowledge rather than advance their own ideas in that displaying knowledge did not require high-level rhetorical skills (Mansourizadeh & Ahmad, 2011). This might be because the participants had received training and practice in source-based writing in this prestigious research-oriented university. This finding provides evidence against the argument that Chinese students lack personal authority (Li, 1985) and depend on established authorities (Gregg, 1986;Sowden, 2005), but lends further support for training and practice of the conventions of academic English writing (Chandrasegaran, 2000;Hu & Lei, 2012).
Text analyses indicated a mean of 3.34 out of 4.22 citations per essay was present, with most citations signaled by words or phrases like "according to", "as said by", and "a study showed that". In other words, appropriately 52% of students' citations were complete, slightly lower than that in Wette (2018). The incompleteness might be because the participants were still not (much) familiar with the conventions of academic English writing which require full information of citation used in writing and were still under the influence of Chinese writing conventions which do not require clear citation in writing. Even so, it further indicates the importance of training and practice.

Conclusions
The present study investigated Chinese undergraduate EFL learners' perceptions of plagiarism and their citation practices in course papers in a prestigious research university in China. Analyses of the data revealed that the participants generally had a good knowledge of plagiarism and identified various reasons for plagiarism, though some students had a (rather) strict definition of the term. The results also showed that the students were able to cite sources in their course papers, mainly in the form of summary, followed by paraphrase and quotation, though around one fourth of the citations were not complete. They cited primarily to support their own ideas, position an author's opinions and/or findings, and acknowledge the author's ideas. These findings indicated that the students had a good knowledge of plagiarism and use of sources and were able to cite from sources by adopting various strategies. These findings were different from those in earlier studies (Shi, 2004(Shi, , 2006Sowden, 2005) which found that Chinese students tended to borrow directly from sources without acknowledgement, partially due to cultural differences. This means that more research on the issue is needed to avoid stereotyping Chinese students' citation practices considering the large and diverse EFL/ESL population in China. The ability to cite appropriately is developmental (Campbell, 1987), especially for novice writers from different cultures. Appropriate instruction and writing practice will (gradually) help enhance students' awareness and understanding of plagiarism and improve their ability to cite source materials in academic English writing.
Meanwhile, the findings show that the students were not so good at synthesizing and generalizing ideas from different sources and following citation conventions, which suggests that they needed further training and practice to become more efficient academic English writers. In order to help students, it is better for teachers to explicitly and consistently instruct students what appropriate and inappropriate source uses are and spend more time teaching students source use skills, as discussed in the current literature (Hirvela & Du, 2013;Li & Casanave, 2012;Pecorari & Shaw, 2012;Petrić, 2012;Storch, 2012;Zhang, 2013). Zhang's (2013) experimental study showed that explicit instruction helped improve synthesis writing in terms of overall quality, measures of organization, and using source reading materials. Teachers can also provide students with ample opportunities to practice source-based writing so that they become more and more familiar with and confident in citing source materials (Storch, 2012). Concurrently, it is important for institutions to regulate what constitutes plagiarism and penalties against plagiarism in academic writing (Cumming et al., 2018), so that students are more aware of the importance of citing when writing from sources (Hu & Sun, 2017).
The present study recruited a large sample and revealed a general overview of students' perceptions of plagiarism and use of citations in their course papers, thus providing implications for formal classroom instruction of writing from sources. However, the participants in the present study were predominantly from hard disciplines (science and engineering), while earlier research has found that students in different disciplines have different knowledge of plagiarism (Borg, 2000;Hu & Lei, 2012;Hyland, 1999;Wette, 2018). For example, Hu and Lei (2012) found that discipline (humanities and social sciences vs. engineering) was a significant predictor of detecting plagiarism. Meanwhile, as discussed in the current literature (Abasi & Akbari, 2008;Hyland, 1999), students' writing is shaped by disciplinary culture, epistemological assumptions, socially constructed academic literacy, and training. For example, the convention of academic writing in Chinese is different from that in English (Deckert, 1993;Gu & Brooks, 2008;Kroll, 1988), and the convention in arts is different from that in science and engineering (Wette, 2018). Thus, a more diverse sample might reveal more diverse perceptions of plagiarism and different citation practices.