A search for paraphrasing and plagiarism avoidance strategies in the context of writing from sources in a foreign language

Writing from sources is a keystone in academic education. Studies show that it can be problematic for students and in extreme cases may result in plagiarism. This article is devoted to one of the many skills necessary to write from sources, namely paraphrasing. The study described here aims to identify and categorise the paraphrasing and plagiarism avoidance strategies applied by students when writing their Master of Arts (MA) dissertations in English as a Foreign Language (EFL). The data were collected via questionnaires and were based on students’ reports concerning their paraphrasing behaviours. The study enabled an array of before-, whileand after-paraphrasing strategies to be collected. The results may contribute to the literature on writing from sources by drawing greater attention to strategic behaviours of students connected with paraphrasing and plagiarism avoidance.


Introduction
Connections between reading and writing in an academic writing setting remain largely unexplored (Hirvela, 2004;Plakans, 2009).Their investigation is vital, as tasks combining reading and writing are an important part of academic education.Paraphrasing, together with direct quotation, summarising, and translation, are core skills that students need to develop to be able to write from sources.Although studies on writing from sources in Foreign Language (FL) and L2 contexts have intensified in the last decade, little is known about strategies and behaviours that allow students to paraphrase sources in ways that avoid plagiarism.The aim of the current study is therefore to investigate what strategies students apply while writing their MA dissertations to avoid the potential for plagiarism when paraphrasing texts from FL sources.

Defining paraphrasing
Paraphrasing may be defined as "restating a passage from a source in fresh language" (Howard, Serviss, & Rodrigue, 2010, p. 181); however, studies (Hirvela & Du, 2013;Shi, 2012;Yamada, 2003) show that in the context of academic writing commonly known definitions are not enough to signal to students what paraphrasing really embodies.For example, the phrase 'restatement' may be problematic as it suggests that students should report what is in the source text in a different way which, apparently, is not enough (Yamada, 2003).An analysis of paraphrasing in academic writing shows that it goes beyond a mere restating of ideas in other words and frequently requires substantial inferencing and interpreting skills combined with elements of discipline knowledge (Yamada, 2003).
There have been some attempts to identify different types of paraphrasing.Shi (2004) distinguished between 'slightly modified' and 'syntactically reformulated' paraphrases.However, both these types of paraphrases involved modifications of source text of a local character and could be seen as bearing traces of plagiarism.The literature on writing from sources frequently mentions a division of paraphrasing into 'superficial' and 'substantial' (Keck, 2006;Roig, 1999;Shi, 2012).Superficial paraphrasing encompasses minor text modifications, mainly word substitution, deletion, addition of single words, and rearrangement of a sentence structure (e.g.Keck, 2010;Roig, 1999), whereas substantial paraphrasing involves major modifications of the source text (Keck, 2006).The problem with this division is that it is based on a high degree of subjectivity and fluid boundaries.The most detailed and least subjective division of paraphrase types was proposed by Keck (2006Keck ( , 2014)).She distinguished four paraphrasing types which differ according to the amount of words copied from the original and number of syntactic and lexical transformations performed on the source text: (1) "Near copies […] contain copied strings of five or more words" and "simplification through synonym substitution and deletion" (Keck, 2014, p. 9); (2) 'Minimal revisions' comprise copied strings of three to four words and numerous substitutions of synonyms; (3) 'Moderate revisions' may copy one to two word phrases and involve substitution of synonyms and change of clause structures; and (4) 'Substantial revisions' involve the borrowing of individual words and revision of clause structures.

Paraphrasing in a context of writing from sources
Writing from sources necessitates several decisions when a "student locates, and reconstructs, or appropriates material" (Hirvela, 2004, p. 94).It is a complex process because, as Campbell (1990) remarks, it involves "reading, understanding, learning, relating, planning, writing, revising, editing and orchestrating" (p. 211).Apart from understanding the sources, students need to select relevant excerpts in an original text that would serve some particular rhetorical function in their writing.They need to decide on the form of citation, for example whether they want to paraphrase or quote.They need to integrate a cited excerpt in such a way that readers understand the purpose of the citations (e.g.Petrić, 2012).What is more, writers need to make sure that each time readers know "whose voice is speaking" (Groom, 2000, p. 15) and that the boundaries between their own words and the words adapted or copied from sources are clearly marked (e.g.Pecorari, 2003).Students need to acknowledge the sources properly by using a selected citing system, and not only relate the content of the source accurately but also, if needed, relate the author's stance to the presented ideas.In some cases students may also need to indicate their own stance, so-called 'writer stance', to the cited materials (for author and writer stance see Thompson & Ye, 1991), which is connected with building their own authority as writers (Abasi & Akbari, 2008).Relating a writer or author's stance also necessitates appropriate use of reporting verbs (Hyland, 2002;Thompson & Ye, 1991).Furthermore, students need to be familiar with the conventions used within a given discipline and be aware of what constitutes plagiarism in that discipline, as practices of writing from sources and understanding of plagiarism vary across different academic disciplines (Bloch, 2012;Shi, 2012).
Paraphrasing as part of writing from sources has been found to be challenging for students (e.g.Pecorari, 2003Pecorari, , 2008;;Shi, 2012) and having to write in their L2 or FL may pose an additional challenge for them.Students, especially novice ones, were observed to extensively rely on copying from sources (Keck, 2006).Students' attempts to paraphrase were found to be based on superficial text modifications which stayed too close to the original text (Howard et al., 2010;Pecorari, 2008;Pecorari & Shaw, 2012;Shi, 2012).This is sometimes referred to as patchwriting, which in turn may qualify as plagiarism (Howard, 1995).The studies show that students' superficial paraphrasing may result from problems with source text comprehension (e.g.Howard et al., 2010).Superficial paraphrasing was also found to be applied by students as a strategy of academic survival (Abasi & Akbari, 2008) -the only resort for students who have to write in their L2 as part of their academic assignments but are new to academic discourse and academic writing.Patchwriting may also be applied as a strategy of learning of how to write academic texts, as by copying the language of sources and rhetorical devices of authors students learn to construct academic texts (e.g.Abasi & Akbari, 2008;Howard, 1995;Pecorari, 2003Pecorari, , 2008)).Shi (2012) observed that L2 students had difficulties in understanding how a paraphrase should look and how to paraphrase without plagiarising.Similarly, Roig's (1997) study found that in some instances the students' main criterion for qualifying an excerpt as plagiaristic was whether it contained an author's name and not the extent of text transformation (which resulted in their qualifying paraphrases based on minor modifications as non-plagiaristic).Wette (2010) observed that L2 students had difficulties in selecting citation-worthy text extracts, indicating boundaries between citations and their ideas, incorporating citations in their writing, and developing a "questioning, evaluative stance towards the authority of published texts" (p.168).Hirvela and Du (2013) found that EFL students had no major problems with paraphrasing when it was performed as an isolated activity but it became problematic for them when writing longer texts.Students' understanding of paraphrasing was found to be rather superficial, and their paraphrases did not reach a rhetorical or conceptual level.They did not perceive themselves as powerful speakers and treated paraphrasing as a "linguisticallyoriented rearrangement tool" (Hirvela & Du, 2013, p. 96) rather than a rhetorical device which serves some purpose for their writing.Due to uncertainty about how to paraphrase while writing academic texts they avoided paraphrasing altogether and resorted to direct quotation.

Paraphrasing as part of tasks combining reading and writing 2
Paraphrasing combines reading and writing.These processes constantly overlap and interact, as in integrated reading/writing tasks "writing provides a way into reading, extends reading and consolidates understanding of a text just as reading sustains writing and furnishes, for the writer, the counterpart of another voice" (Carson & Leki, 1993, p. 2).In order to paraphrase, students not only need to understand a source text and incorporate it in their writing (a reading-to-write direction), but they also need to approach reading from a writing perspective (a writing-to-read direction).If students have awareness of what they want to achieve through their writing and approach reading with this in mind, they will have a clearer sense of direction and study sources in a way that is selective and relevant to the function and topic of their writing (Hirvela, 2004).
As both reading and writing skills are needed in order to paraphrase, it may be assumed that the strategies applied by students during FL or L2 reading-only tasks 2. Initially, I had planned to use the term 'reading-to-write' tasks as it is very commonly used in the literature to refer to tasks which combine reading and writing.However, the book by Hirvela (2004) made it very clear that during such tasks students both read-to-write and write-to-read as it is not only reading that influences their writing but also writing that significantly impacts how they analyse and read sources.Hence the term 'reading-to-write' could be slightly misleading in this context as during tasks involving writing from sources there is an intensive bidirectional interplay of reading and writing.
(e.g.Phakiti, 2003) and FL or L2 writing-only tasks (e.g.De Silva, 2010) may to some extent also be applied in integrated reading/writing tasks.Students need to apply a set of FL reading strategies in order to read FL texts effectively, and a set of FL writing strategies to plan, write, and revise their own texts.However, the fact that students need to read-to-write and write-to-read simultaneously makes writing from sources a very specific endeavour which may involve a unique set of strategies typical only for FL reading-writing constructs.Few studies have focussed on this issue: Cohen (1994), Esmaeili (2002), and Plakans ( 2009) investigated reading strategies applied by students during integrated reading/ writing tasks.Plakans's (2009) study resulted in the proposal of a taxonomy of such strategies comprising five major categories: (1) goal-setting, for example checking the task to integrate sources; (2) cognitive processing, for example slowing the reading rate (pausing), breaking lexical items into parts/using phonological cues, rereading passages; (3) global strategies, encompassing for example asking questions, recognising text structure/rhetorical cues; (4) metacognitive, for example recognising lack of comprehension; and (5) mining strategies, for example scanning texts for ideas to use in writing (pp. 257-258).
The study also showed that students who achieved higher scores for their essays used more global-and mining-type strategies while the lowest scoring students employed more word-level reading strategies.
As far as writing from sources is concerned, there have also been studies that investigated strategies and behaviours typical for summarising (e.g. Brown & Day, 1983;Johns & Mayes, 1990;Taylor, 1984;Yang & Shi, 2003).For example, very careful reading of the text, spending a considerable amount of time reflecting on the subject of the text and on what to write, thorough analysis of a text's structure, close monitoring of accurate reporting of a source text (Taylor, 1984), constant referring back to the source text, spending extra time on planning and monitoring (Yang & Shi, 2003), and verbalising what is being written (Yang & Shi, 2003) have been identified as strategies or behaviours typical for high quality summaries.
Little is known however about what strategies students use when they paraphrase while writing from sources and what strategies they apply when they try to avoid plagiarism while paraphrasing.Investigating paraphrasing strategies seems worth pursuing as paraphrasing has been found to be challenging for students.
Identifying paraphrasing and plagiarism avoidance techniques could help effective instructions in paraphrasing FL source texts to be developed.Providing students with proper instructions on paraphrasing would seem to be a very important aspect of training in academic writing as information available to students on how to write from sources without plagiarising, even those available on highly informative and student-friendly webpages, is definitely inadequate.It was found by Bloch (2012) that students who were thoroughly informed how to write from sources and avoid plagiarism still had problems with putting these rules into practice when writing from sources.Hirvela and Du (2013) observed that "while the procedures involved in paraphrasing source text material may appear simple, the enactment of these procedures is often a complex and elusive experience for L2 writers" (p.87).Hence this study looks at the strategies students adopt when they paraphrase and try to avoid plagiarism while writing from sources in a foreign language.Its aim is to identify strategies used by students, recognise some features these strategies share, and organise them into categories.This article reports on the first stage of this study of the paraphrasing and plagiarism avoidance strategies used during academic writing tasks in FL, the ultimate aim of the study being to create a scale measuring student's strategic behaviour while paraphrasing.

Participants
A hundred and ten MA students from three public universities in the south of Poland took part in the study.All were second year students in English philology and were in the process of writing their MA dissertations in English.

Instrument
This study reports on the data acquired via a questionnaire in which students were required to reflect on the paraphrasing and plagiarism avoidance strategies used by them while writing their MA dissertations.The first three questions were formulated in order to elicit from students information on strategies applied during the first three stages of paraphrasing3 , namely before writing a paraphrase, while writing, and once it has been written.This division of paraphrasing strategies was partially modelled on the division of writing strategies by Petrić and Czárl (2003) into 'Before I start writing an essay in English', 'When writing in English', and 'When revising' and on a taxonomy of writing strategies proposed by De Silva (2010) based on beforewriting, while-writing, and after-writing strategies.The fourth question aimed to elicit from students information on whether they check their paraphrasing for potential plagiarism during all three stages of paraphrasing.
The students filled in the questionnaire during their classes.Participation in the study was voluntary and the questionnaire anonymous.

Data analysis
Qualitative analysis was applied in the study in order to investigate strategies reported by students.Students' responses from questionnaires were closely and recursively analysed by the author of the article in order to identify interconnecting categories and dimensions and consequently form a typology (Woods, 2006).Following Yang (2014, p. 80), it may be said that "analytic induction" and "constant comparison" were applied to categorise the strategies reported in the questionnaires.

Results
The study identified an array of strategies employed by students when they paraphrase and try to avoid plagiarising.The strategies they apply before writing down a paraphrase are mainly connected with how they read and analyse source materials.Students set goals for their reading, read in a selective way, work on source material using graphic devices, create separate files with excerpts worthy of citing, add their comments to the marked excerpts, and practise paraphrasing.Some analogies may be observed between Plakans's (2009) taxonomy of reading strategies applied during reading-to-write tasks and the strategies reported in this study.As a result, two categories from Plakans's (2009) strategies have been adopted in order to categorise before-writing strategies of paraphrasing, namely 'goal-setting' strategies and 'cognitive strategies' (details are provided in Table 1).
As far as while-writing paraphrasing strategies are concerned, the study showed that these vary from strategies comprising text modifications not going beyond paraphrased excerpts (restating), text modifications comprising mediation of a source text, text transformations going beyond paraphrasing by the addition of comments and conclusions, writing from memory, the use of external resources, to a strategy of giving-up.For details, see • I look for the gist of the whole paragraph.
• I underline key words in an excerpt to be paraphrased.Practising paraphrasing (silently, in one's mind) • I close my eyes and say it in my own words.
• I paraphrase the text in my mind when I read.Selecting excerpts for paraphrasing by the use of graphic devices • I put brackets around an excerpt I could use in my thesis.
• I highlight excerpts for quotation in one colour and for paraphrasing in the other.Selecting excerpts for paraphrasing by copying excerpts (into a separate file, into a thesis) from electronic sources for subsequent paraphrasing • I use books in PDF so I copy some extracts to my thesis and I later paraphrase them.
• When I read I create a separate file with the fragments I like.

Notes and comments
• I take notes (on the margin) 4 .
• I add my comments to the material I want to put in my dissertation.

Table 2. While-writing paraphrasing strategies
Writing from memory • I write from memory.
• I do not look at the text.Text transformations (lexical and syntactic) • I do not resort only to synonym substitution.
• I look for synonyms of the key terms • I mix a few sentences, I combine them and keep their sense 4. The purpose of such behaviour was not given.If these strategies were used for understanding purposes they may be qualified as cognitive, but if notes were made in order to put them into the dissertation then they can be qualified as belonging to mining strategies.The study also aimed to identify strategies once a paraphrase has been written (the third question in the questionnaire) and strategies connected with checking paraphrases for potential plagiarism (the fourth question).The questionnaires showed that for students, these two categories were synonymous, and they therefore provided the same answers to Questions 3 and 4. Most students (70) 5 prioritised the fourth question over the third one and chose to report how they tackled the problem of plagiarism while paraphrasing.Under the third question they simply wrote 'see below', 'see Point 4' etc. Hence this category includes the strategies applied while checking paraphrases for potential plagiarism.As far as this group of strategies is concerned, the study showed that students applied various strategies including 'comparing a paraphrase with the source', strategies connected with checking 'whose voice is speaking' (reporting for intertextual transparency at source level), delaying self-evaluation to get some perspective on the text, using external resources like the Google search engine or free software for text matching, and asking others for advice; for details see Table 3.
There were also some students ( 11) who said that they did not use any strategies to check whether a paraphrase was plagiaristic as, for them, restating something with their own words automatically meant elimination of any plagiarism potential.As one student wrote: "It is obvious for me -if I used my own words to restate a piece of a text it means it is not plagiaristic.There is no need for checking".
5. 70 out of 99 students, as 11 did not report any strategies in the third and fourth question.Although it was not a primary aim of the study, students were also asked about possible problems with paraphrasing.An attempt to identify some of the problems that students at this level of FL academic writing development might have was rather to set the context for the use of paraphrasing strategies and plagiarism avoidance.The students reported difficulties with paraphrasing arising out of comprehension problems during the pre-writing stage: "I sometimes have access only to a sample of a text (the rest is paid), which makes paraphrasing difficult".
"It's difficult sometimes for me to understand original and that is why my paraphrase may be wrong".
Some pointed to problems with text transformations during the while-writing stage: 6. Students reported that they make sure they do not copy more than three or four words in a row into their writing.
"It is difficult to change the sentence structure".
"Some sentences are impossible to paraphrase".
"There are no synonyms to some concepts -which is difficult to deal with".
As far as the post-writing stage is concerned, students indicated that they are sometimes uncertain whether their paraphrases are properly formulated ("Sometimes I don't know whether it's plagiarism or paraphrase") or have doubts about the originality of their own paraphrasing: "I think that my paraphrase may not be original (not that it copies exactly from a source, but somebody else might have paraphrased that piece in a similar way".

Discussion
The study allowed to identify a diversity of strategies that are applied by students while paraphrasing source materials in an FL.They can be divided into three groups that mirror the stages of writing, namely the before-writing, whilewriting, and after-writing stages.The strategies were reported by second year MA students who were in the second year of writing their MA dissertations, therefore it may be assumed that the reports came from quite experienced writers.The reports also show what strategic behaviours students had developed during their five-year experience of academic writing.
The before-writing paraphrasing strategies used while reading and analysing sources focussed on macro-level understandings with an emphasis on an understanding of the main ideas and memorising key information (for comprehension at a macro or higher level see e.g.Koda, 2004, or Grabe, 2009).These strategies seem to present a global approach to reading which is crucial in efficient paraphrasing and writing from sources (Howard et al., 2010;Shi, 2012) and which has been found to be characteristic for high-scoring students in reading-to-write tasks (Plakans, 2009).It may be contrasted with local text analysis that focusses only on 'good sentences' worth putting in one's writing which is not a recommended approach for academic writing as it often results in patchwriting and plagiarism (Howard et al., 2010).
The study also showed that, while reading, students decide which excerpts to paraphrase and which to quote verbatim.This may partially explain the use of the strategy of 'silent paraphrasing' at a before-writing stage.It may be applied as a testing device via which students discover whether they are able to paraphrase an excerpt successfully or not and consequently impacts their decision whether to quote or to paraphrase (this assumption, however, requires verification by interviews with students).
Adding notes and comments while reading source materials (although at this stage of the study it is not known what exactly the comments concerned) shows that students reflect on and draw inferences from what they read.This strategy may be highly beneficial as inferencing has been identified as crucial in good paraphrasing (Shi, 2012;Yamada, 2003) and adding one's voice to paraphrased excerpts is also crucial in writing as it shows a writer's authorial voice (Abasi & Akbari, 2008) and maturity (Hirvela & Du, 2013).
During writing, apart from strategies strictly connected with text transformations, students also reported numerous strategies that went beyond merely restating what was in the source text.For example, they pointed to a very important aspect of paraphrasing: mediating academic discourse.Students used this strategy to show that they understand a source text and to make a source text understandable for readers.What is more, mediating academic discourse shows students' authority as writers, as they are no longer reporting but transforming a source in order to fit their own writing style and fulfil some rhetorical function.
Students also reported a strategy that is crucial for maintaining intertextual transparency at a propositional level (for details about types of intertextual transparency see Pecorari & Shaw, 2012).It was a strategy of keeping the propositional content of a source which necessitates what students called "balancing rewording with keeping the sense of the original".Finally a strategy of giving up appeared in students' reports.Similarly to Hirvela and Du's (2013) findings, students gave up paraphrasing and resorted to quotation.In Hirvela and Du's (2013) study, this resulted from students' insecurity as writers, their lack of authorial voice, and confusion as to the rhetorical functions of paraphrasing, which may also be true for some participants of this study.On the other hand, as the investigated students of the current study were in the second year of their MA programme and have had some experience in academic writing, such a strategy may be also seen as a sign of students' authority and empowerment as writers.
They assess what is possible for them to paraphrase and consciously withdraw from their initial plan of paraphrasing when there might be a chance of violating various aspects of intertextual transparency.
As far as strategies applied while checking paraphrases for potential plagiarism are concerned, students reported resorting to comparison at various textual levels.For example students used a strategy of comparing from a perspective of intertextual transparency at content level as they were aware that rewording bears the risk of changing the propositional content of the source ("nuances are lost").They also reported having used a strategy of checking the boundaries between their own input into the text and the input of sources.This is vital as students have been found (e.g.Pecorari, 2003Pecorari, , 2008;;Pecorari & Shaw, 2012) to blur the borders between sources and their own words.Delaying self-evaluation was also identified among paraphrasing strategies; this seems a highly valuable strategy as it allows to distance oneself from one's own writing and have a fresh look at whether paraphrases are properly formulated.
Although students used a variety of strategies to monitor their paraphrases for plagiarism, they also resorted to help from external resources; Google or text matching programmes.On the one hand it may show students' resourcefulness in applying all possible devices to make sure that they are not plagiarising, but on the other hand it may reflect their insecurity about whether their paraphrasing is non-plagiaristic (which was also signalled by students in the 'problems' section of the questionnaire "I am not sure whether it's enough", "There must be someone who paraphrased it in a very similar way").Students also admitted that they ask others or their supervisor for advice, which may signal their awareness of the complex nature of paraphrasing and the need to take great care in order to paraphrase efficiently.
It should be added that although the strategies reported by students were not verified for their effectiveness in writing from sources, most of them seem worth recommending to students as strategies to enhance their writing from sources, to help to keep the intertextual transparency of their writing, and help them to avoid plagiarism.Just to name a few, conscious reading of sources in order to select excerpts for quoting or paraphrasing, writing down comments and making notes while reading, making sure that propositional content has not been changed as a result of semantic and syntactic modifications, checking whether it is clear for readers whose voice is speaking at that moment, delaying revision, or consulting others in case of doubt -all of these behaviours have the potential of being highly valuable while writing from sources.The strategies identified in this study could be presented to students during academic writing classes for at least the following two reasons: to make students realise that paraphrasing is a complex undertaking that cannot be approached lightly, and to guide them through the process of paraphrasing, right from reading sources up to checking a paraphrase for potential plagiarism.

Continuation of the study devoted to the questionnaire construction
A think aloud procedure is needed to follow up this part of the study as it would give some insight into the actual application of strategies when paraphrasing source text in an FL.This seems vital, as students' perceptions of the strategies they use while paraphrasing may differ from the actual strategy they use while writing from sources in practice.To extend the study, an analysis of literature on writing strategies is needed in order to identify some further strategic behaviours that might fit into a paraphrasing setting.Finally, a pilot version of a questionnaire investigating paraphrasing strategies while writing from sources needs to be created that might take a form of a Likert scale, similar to the Oxford Strategy Inventory for Language Learning (Oxford, 1990), with the issues to be assessed on a scale ranging from 'never true of me' to 'always true of me'.Consequently, reliability and validity of the instrument needs to be established.In order to verify the reliability of the instrument, a test-rest method (Hatch & Lazaraton, 1991) may be used; the Cronbach alpha coefficient may not be efficient in this case.Following Petrić and Czárl's (2003) rationale for excluding this method from a writing strategies' context, it may be said that internal consistency is appropriate for instruments which are to measure a single underlying construct.As a consequence, internal consistency may be applicable to a specific group of strategies but not to a questionnaire which is a combination of clusters of strategies.As far as the validity of the questionnaire is concerned, its content may be validated by consultations with faculty members, and members of the target population (students).Construct validity could be evaluated by a comparison with theory, as factor analysis may be inadvisable in this case due to problems with interpreting results (Petrić & Czárl, 2003).Response validity may be established via a think aloud procedure in which participants are asked to verbalise their thoughts while filling in a tested questionnaire (Converse & Presser, 1986;Petrić & Czárl, 2003).

Further studies of paraphrasing strategies
There are many ways this study can be followed up.One path would be to verify paraphrasing strategies in terms of their effectiveness by conducting a survey of the paraphrasing strategies used by highly-rated and low-rated EFL writers or by verifying the use of strategies in terms of the quality of the final product, which in this case would be the quality of students' writing based on the integration of source texts.A closer look at the paraphrasing strategies reported by students shows that they may not be limited exclusively to FL writing.All of them seem potentially applicable to an L1 context.As there are some studies which indicate both differences and similarities between L1 and L2 writing (e.g.Çandarlı, Bayyurt, & Martı, 2015;Silva, 1993;Kobayashi & Rinnert, 2008), it might be interesting to investigate the relationship between the use of paraphrasing strategies in L1 and FL writing and to observe possible interactions between paraphrasing performed in L1 and FL.

Conclusion
The aim of this study was to search for strategies that students apply when paraphrasing source texts in an FL and try to avoid plagiarising.The study allowed an array of strategies to be identified that can be categorised into three groups: before-writing paraphrasing strategies, while-writing paraphrasing strategies, and after-writing paraphrasing strategies, which mainly included checking paraphrases for potential plagiarism.The results of the study could have high educational value as the strategies could be presented to students during academic writing classes in order to help them with effective paraphrasing while writing from sources in an FL and to make them realise that paraphrasing in academic writing entails far more than merely restating a source text in their own words.
Mediating a source text • I try to present information in a more simple and a more clear way.Staying close to the propositional content • I try to balance my own wording with keeping the sense of the original.Reaching beyond paraphrasing • I add my observations and conclusions.Use of external resources • I use dictionary with synonyms.Referencing • I immediately give source.Giving up • If very difficult I turn to quotation • If it is not possible I give up.
The questionnaire consisted of open questions which were formulated in the following way: 3. What do you do once you have written your paraphrase?4.Do you check whether your paraphrase is plagiaristic?If so what do you?There was also one additional question in the questionnaire: 'Do you have any problems with paraphrasing?If so, what do you find problematic?'.The aim of this question was to gather additional data that would help to contextualise strategies reported by students.