A Preliminary Study to Identify the Extent of Self-Plagiarism in Australian Academic Research

E-mail: [email protected] ; [email protected]

Abstract

This paper reports on the preliminary findings from a pilot project which sought to identify self plagiarism in Australian academic publications. Ten Australian authors were selected at random from top published authors on the Web of Science (Social Science and Humanities) database. Evidence of textual re-use was collected from 269 electronically available published journal articles using the text-matching software program, Turnitin. Self-plagiarism was defined for this study as “10% or more textual re-use of any one previous publication by the author without attribution”. The preliminary findings suggest that textual re-use is widespread in academic research, with 60% of the authors in the sample having committed self-plagiarism in at least one of their published papers in the period 2003-2006.

Introduction

While plagiarism by students has become a widely discussed and researched topic in recent years (Angelil-Carter, 2000; Bretag, 2005, 2007; Carroll, 2003; Devlin, 2003; Howard 1995, 2000; McCabe 2005; Wilson, 1997; Zobel & Hamilton, 2002), self-plagiarism by academics is a taboo issue, with academics and administrators divided about what constitutes original research. Opinion exists along a continuum. At one end of the continuum is the view that not citing your own work when you have used large chunks of text from one or more papers in a new paper “verges on the fraudulent”. At the other end of the continuum is the view that it is “it is perfectly acceptable...to write up research for two or three conferences...and to get at least a couple of journal articles out of it as well” (see Monash University 2006, ‘Citing your own work’ for video vignettes describing these two extreme views ).

The issue of self-plagiarism by academics has only recently begun to be explored. Green (2005) refers to it as a “relatively new evil”. Schein and Paladugu (2001) examined 660 articles published in three surgical journals in 1998, using the online search engine PUBMED, and found that 3% were dual publications and 7.6% were virtually identical. A number of other researchers have also looked at the issue of duplicate publication, particularly in medicine and the health sciences (Bailey 2002; Barnard & Overbeke 1993; Blancett, Flanagin & Young 1995; Bloemenkamp, Walvoort, Hart & Overbeke 1999), and reported similar findings. Collberg and Kobourov (2005) conducted a small study using the software tool SPLaT (Self Plagiarism Tool) to examine the publications found on computer science web sites from 50 university departments in the United States. They found substantial evidence of textual re-use (self-plagiarism), particularly between conference papers and published journal articles.

In the current Australian context, where the introduction of the Research Quality Framework (RQF) has placed pressure on academics to increase both the quality and quantity of their research outputs, there are no guidelines regarding what constitutes original research. Different disciplines have loose conventions which are shared via an informal academic apprenticeship model. In addition, peer reviewed journals usually provide guidelines which specify that the submission must be original, previously unpublished work and that the work is not under consideration for publication by another journal (see Cheung & Driver 2004). In recognition of the tendency of some authors to submit virtually identical work to two or more different journals at the same time, the New England Journal of Medicine, insists that “Authors should submit to the Editor copies of any published papers or other manuscripts in preparation or submitted elsewhere that are related to the manuscript to be considered by the journal” (Kassier & Angell 1995).

Sheikh (2000) argued that the Research Assessment Exercise (RAE) introduced in the United Kingdom in 2001, and on which the RQF is modeled, had the potential to foster unethical publication practices, including but not limited to plagiarism and duplicate and redundant publications. Brice and Bligh (2004) agreed that the increasing pressure to publish has resulted in ‘author misconduct’, stating that the most frequently encountered example of misconduct is redundant and duplicate publication. The authors included in this category ‘salami-slicing’ (publishing ‘thin slices’ of research in multiple papers), cutting and pasting whole sections from one manuscript to another, publishing in a minor journal and then submitting it to a larger journal without revealing its previous publication, and attempting to have a paper published in two journals simultaneously (Brice & Bligh 2004).

However, despite the experience of the RAE in the UK, the Australian higher education sector has not addressed the need to provide systematic, cross-disciplinary guidelines to academics who publish their research in a number of forums or media. This is important when research funding by bodies in Australia such as the Department of Education, Science and Training (DEST) is implicitly tied to new, original research.

Literature Review

Self-plagiarism is a complex and delicate issue. Samuelson (1994) suggests, “It is, in truth, a far more complex matter to determine what is self-plagiarism versus what is fair use” (p. 25). Scanlon (2007) agrees, and recommends both clear definitions self-plagiarism and appropriate editorial guidelines. Some academics argue that self-plagiarism is impossible because plagiarism is theft and you cannot steal from yourself. However, Hexham (1999) argues that just as it is considered unacceptable for students to submit the same essay for credit in different courses, self-plagiarism by academics does exist, particularly when “the author attempts to deceive the reader”. This reference to the author’s intention has parallels with the research on student plagiarism, where a number of writers argue that an essential element in determining penalties for plagiarism is the student’s intention to deceive (Carroll, 2003; Devlin, 2003; Hiller & Peters, 2005). According to Scanlon (2007) duplicate publication and ‘salami-slicing’ are not only deceptive, but in some cases could even be called fraudulent.

Collberg and Kobourov (2005) focus on the phrase ‘textual re-use’ rather than self-plagiarism to refer to texts or sections of text which are published a number of times. They define self-plagiarism as “reusing one’s own previously published ideas in a new publication without adequate attribution”. Boisvert and Irwin (2006) state that while it is quite acceptable for authors to re-use any portion of their copyrighted works in other works of their own, “what is not ethical is the practice of reusing one’s own work in a way that portrays it as new when, in fact, it is not” (p. 24). The authors maintain that reusing significant portions of your own work is only acceptable when the previous work is appropriately cited. Anderson (2006) agrees, stating that that “self-plagiarism occurs when you reproduce portions of your work without citing the original work” (p. 623), while the Editor of Obstetric, Gynecologic and Neonatal Nursing, is adamant that authors who use exact passages from previous manuscripts need to either use quotation marks with the full citation, or else paraphrase and self-cite as appropriate (Lowe 2003).

The major types of self-plagiarism include redundant and duplicate publications, partitioning a larger study into smaller published studies, copyright infringement and text recycling (Roig, 2006). Whether these practices constitute a breach of academic integrity is subject to debate (see Scanlon 2007 for details of a case where charges of self-plagiarism were dropped after a committee determined that verbatim text in the context section of two papers was acceptable practice). According to Samuelson (1994) self-plagiarism is transgressive from both legal and ethical standpoints. From a legal perspective, self-plagiarism breaches copyright law because the owner of the text (the journal or publishing company) is deprived of financial benefit. In ethical terms, self-plagiarism is transgressive because the author misrepresents the work as new and original. Lowe (2003) states, “It is just as unethical for an author to quote herself or himself without referencing as it is for the author to quote another individual’s material without appropriate citation” (p. 145). With opinion and practices divided, some academics contend that there is a need for publishing bodies to articulate and enforce a broad policy against self-plagiarism (e.g.Gotterbarn, Miller, & Impagliazzo, 2006; Schein & Paladugu 2001).

Previous work in this field, emanating from North America and Europe, has focused largely on duplicate publications in medicine/health sciences, with little attempt to define ‘self-plagiarism’ in more narrow terms. The current study aimed to first define self-plagiarism, including developing parameters for identifying it, and then to determine whether self-plagiarism is prevalent in research publications in the Humanities and Social Sciences disciplines in Australia. It is our intention to use the results of this preliminary study to assess the need for more extensive research in the area.

Methodology

Ten Australian authors were selected as a stratified random sample from the Web of Science database. The selection of Web of Science was made as the limiters yielded a large number of author records (15,804). The database was sophisticated in its output records and allowed for the results to be analysed according to the needs of the study. The articles on the Web of Science were also available to a large extent in electronic format, which was a feature needed for Turnitin [1] analysis. The sample chosen was based on the following limiters:

The Web of Science database was searched for full text articles for each Australian author. In total, 269 published articles were analysed. Turnitin was used to match articles against other publications on the Turnitin Database, the Internet, other articles used in the study, and available full-text electronic articles written by the same author.

This study was limited to electronically published, peer reviewed articles by Australian authors. It seems likely that a much higher rate of self-plagiarism would be detected if conference papers and other non-peer reviewed papers had also been included in the sample (as occurred in the study by Collberg & Koborouv 2005). This reasoning is based on the academic convention across disciplines which encourages a publishing cycle that begins with conference presentation. Articles were considered to contain self-plagiarism if they contained 10% or more of any one of the author’s previous publications without appropriate attribution.

Although Samuelson (1994) suggests that “a rule of thumb that if one reuses no more than 30% of one’s prose in another article, that’s OK” (p. 24), none of the literature provides any guidelines on how this 30% is determined. For example, does this mean 30% from one source, or is it 30% as a total of all re-used text from multiple sources? Furthermore, the Copyright Act permits copying of a reasonable portion of works for specific purposes, and defines a ‘reasonable portion’ as “10% of one chapter of a book, or one article from any one issue of a magazine”(University of South Australia 2003). This, therefore, goes some way to providing explicit guidelines to authors wishing to re-use their work. So, while 10% may at first appear overly stringent, in light of the ‘chain of textual re-use self-plagiarism’ and the ‘cut and paste textual re-use’ identified in this study (see Findings below), 10% was deemed to be an appropriate starting point for identifying self-plagiarism. For example, even where an author ‘cut and paste’ small parts of text from a range of sources, resulting in an overall text-match ranging from 23-52%, this was not classified as ‘self-plagiarism’. These issues are described in more detail in the Findings section.

Findings

The text-matching software Turnitin only permitted analysis of electronic sources. For this reason, it was not possible to analyse every article published. As Table 1 below indicates, in 9 out of 10 Author Records [2] 60% or more of the articles were able to be analysed. Of these nine Author Records, an average of 73% of articles were analysed using Turnitin. Self-plagiarism was analysed as potentially occurring when Turnitin reports showed a text match of 10% or more in any one paper with another or others by the same author. However, if authors provided in-text citations of their previous work, they were not considered to be self-plagiarising, regardless of the percentage of textual re-use.

Table 1. Extent of Self-plagiarism Detected Using Turnitin

The Turnitin text-matching reports provided the first step in the analysis. The researchers then manually examined each Turnitin report and corresponding article to make qualitative judgments on the appropriateness of textual re-use in an article. For example, it was surmised that researchers may use the same methodology in a number of papers, but it was considered to be unacceptable if the text was exactly the same in each without self-citation.

As Table 1 indicates, evidence of substantial self-plagiarism was found in Author Records 2, 5, 7 and 8, ranging from 25% to 53% of the articles analysed. Self-plagiarism was also identified to a lesser extent in Authors Record 3 and 4 (less than 5% of the articles analysed). Table 1 also provides information on the maximum percentage of self-plagiarism found in any one article in the Author Records. Where found, this ranged from 16% in Author Record 5 to 55% in Author Record 7.

Using both the Turnitin text-matching records and a manual qualitative check, a pattern of self-plagiarism, which we refer to as a ‘chain of textual re-use’, was able to be identified. Many authors also used a ‘cut and paste’ method of textual re-use, which we have not defined as self-plagiarism, where small sections (less than 10%) of text from a number of previously published publications were used in the so-called ‘new’ publication. This is shown in Table 2, using a single example from all the Author Records that used ‘cut and paste’ textual re-use. Table 2 demonstrates that each article was constructed to a large extent (23% - 52%) from previous publications, resulting in a ‘new’ publication which contained little original material. Please note that we are not suggesting that any of the examples in Table 2 are instances of self-plagiarism, or that small text matches of 1% or 2% constitute a breach of academic integrity. Rather, the point of Table 2 is to illustrate the ‘cut and paste’ method of textual re-use identified in the research.

Table 2. ‘Cut and Paste’ Method of Textual Re-use

Analysis

Hexham (1999) makes the case that self-plagiarism “must be distinguished from the recycling of one’s work that to a greater or lesser extent everyone does legitimately”. It is this ‘legitimate recycling’, where researchers re-use parts of a literature review or methodology from a paper or papers on a related topic, that causes the confusion about what constitutes self-plagiarism. ‘Legitimacy’ is attained through strict adherence to academic conventions such as in-text citation and clear indication of quoted sources. As Green (2005) maintains, the simplest way to avoid charges of self-plagiarism is to include appropriate self-citation in all published work.

The overall finding from the present study is that self-plagiarism (i.e. textual re-use without self citation) occurred in the majority of articles analysed. A further finding was that even in Author Records with no self-plagiarism, most authors used ‘cut and paste’ textual re-use. The following section looks at particular examples of both ‘chain of textual re-use’ self-plagiarism, and ‘cut and paste’ textual re-use.

‘Chain of textual re-use’ self-plagiarism

Author Record 7 was the most blatant case of both self-plagiarism and inappropriate academic practice in the sample. This Author Record represented a collaborative writing effort with the four authors’ names sometimes being published in different order, and on occasion with other authors’ names added. Articles by a collaborative team such as this one were located using just one of the authors’ names. Out of a total of the 28 articles naming this author submitted to Turnitin, seven articles were found to contain substantial self-plagiarism ranging from 10% to 55% from one previously published source. Every other paper in Author Record 7’s sample also contained ‘cut and paste’ textual re-use, although these fell outside of our definition of self-plagiarism. The percentage of ‘cut and paste’ ranged from 16% to 30% of the article. The following analysis provides close examination of the most substantial examples of self-plagiarism.

Article 29 was found to have a 55% text match with Article 30. These two articles were published within one day of each other, by two separate international peer-reviewed journals in the same field. Close analysis showed the following:

• Both papers contained the same abstract, with the exception of a few words.
• The Introduction of Article 29 was a piecing together of text from a number of different sources, including large sections from Article 30. In one sentence towards the end of the Introduction, the authors included one self-citation to Article 30, stating that it was ‘in press’.
• The Method section of Article 29 was a complete match with Article 30, with the exception of two words. No self-citation was given in this section.
• The Procedure section of Article 29 matched Article 30, with the exception of the last paragraph (this paragraph matched another paper [Article 24] by the authors which was published six months later). No self-citation was provided in this section.
• The entire section entitled ‘Statistical Analysis’ was later used in Article 24 without self-citation.
• The Results section of Article 29 included large portions from Article 30, as well as sections that were later used verbatim in Article 24. The only self-citation provided in this section stated ‘These results are remarkably similar to our previous findings’.
• In the extensive Discussion section, only 18 lines represented new, previous unpublished material.

In analysing the data from all Author Records, we needed to make qualitative judgments about whether the citation was appropriate. We considered that adding a reference at the end of long section which had been re-used from another source, or generally referring to a previous study without providing the full reference, was not appropriate.

Figure 1 below demonstrates the way that Author 7 re-used text in Articles 29 and 30 and how text from both of these articles was then re-used in future publications. Note, each of these articles was manually checked and the text in question was not self-cited appropriately in any of them.

Figure 1. Chain of Textual Re-use Self-plagiarism, Author 7

In Figure 1, the date inside each box refers to the publication date of the Article (which takes its number from the list on the Web of Science); the direction of the arrow shows which is the earlier published article; and the percentage next to the arrow is the percentage of textual match between the two articles, as detected by Turnitin.

While the textual re-use apparent in Articles 29 and 30 is substantial, it could be argued that it occurred because two similar articles were submitted within a short time frame. Analysis of other articles by the same author/s, however, demonstrated that they consistently use the same ‘chain of textual re-use’ approach, even when there is a time gap in the publishing cycle. For example, Article 7 published in January 2005 has a 55% text match to an article by the same author/s published in September 2004. Both articles were published by the same international peer-reviewed journal.

Substantial self-plagiarism was also identified in Author Record 2. This Author Record represented a collaborative authorship with a number of different authors in the writing team, depending on the article. Of the 40 articles submitted by this author to Turnitin, 14 were found to contain self-plagiarism according to our definition. This author used a similar ‘chain of textual re-use’ approach to publishing as Author 7. Figure 2 below demonstrates the interconnections between this author’s various publications and how parts of one article was used in a number of other publications during the time period under analysis.

Figure 2. Chain of Textual Re-use Self-plagiarism, Author 2

In another chain of textual re-use self-plagiarism for Author Record 2, Article 41 published in June 2003 had a 31% text match, and Article 34 published in April 2004 had a 9% text match with Article 28, published in January 2005. In addition to the 14 cases of substantial self-plagiarism in Author Record 2, there were also 12 cases of ‘cut and paste’ textual re-use.

Author Record 8, like Author Records 2 and 7, represented a collaborative writing effort with seven authors’ names sometimes being published in different order, and on occasion with other authors’ names added. Author 8 used the same ‘chain of textual re-use’ approach to publishing. Of the 24 articles submitted to Turnitin, 10 articles contained substantial self-plagiarism. For example, Article 23 had a 44% match with a report published on the Internet, with no self-citation. In addition, this article contained 12% of ‘cut and paste’ textual re-use from other sources by the same author. In another example, Article 4 contained 50% textual re-use from 12 different publications by the same author. In addition to the 10 cases of self-plagiarism, four articles used ‘cut and paste’ textual re-use.

Other than the Author Records discussed in detail above, two other Author Records in the random sample also contained self-plagiarism in at least one published paper in the period 2003-2006. Author Record 3 represented a collaborative writing effort, with five or six authors listed, depending on the publication. Article 23 had a 23% match with a previously published Internet article for which the authors did not provide an appropriate citation. In addition to the identified self-plagiarism, Article 23 comprised ‘cut and paste’ textual re-use from two other sources comprising 11%. Therefore Article 23 had a 34% match with previously published articles.

Of the 32 articles for Author Record 4, 23 articles used ‘cut and paste’ textual re-use, and one article was identified as containing self-plagiarism. Twenty-four articles for Author 5 were analysed by Turnitin, with 8 articles containing self-plagiarism and 3 articles using a ‘cut and paste’ method of textual re-use.

This preliminary study found that 60% of the authors in the stratified random sample have self-plagiarised in at least one published paper in the time period under analysis, and 70% of authors used ‘cut and paste’ textual re-use. In addition to these two practices, other forms of attribution, such as referring generally to a previous study, or adding a citation at the end of a long section which was taken directly from a previously published source, was considered to be inappropriate. Other forms of questionable research practices included: submitting virtually identical papers to two journals at the same time; using text from previously published collaboratively written papers without citation (not only self-plagiarism, but also plagiarism of the other person’s work); and not citing other papers submitted for review as ‘in press’ or ‘under review’.

The above findings need to be considered in the context of the limitations of the study. First, not all published papers were available to be analysed by Turnitin. Second, while a large number of papers were analysed (269), the actual sample size of 10 authors was very small. Third, the study was limited to Social Sciences, Arts and Humanities and therefore the results cannot be generalised to other disciplines. Fourth, the study was limited to peer-reviewed journal articles by Australian authors. A more comprehensive assessment would need to involve other publications such as conference papers, book chapters and reports. Fifth, this study was limited to articles published during a three year period. To explore further the concept of a ‘chain of textual re-use’ self-plagiarism, a study which takes in a longer time frame would be needed. Finally, the manual examination of articles by the reseachers to make a qualitative assessment of self-plagiarism requires a higher degree of resource commitment which must be taken into account for larger studies.

Conclusion

Previous research in this area has mostly focused on duplicate publications, and there has been no research specifically aimed at identifying self-plagiarism in Australian publications. A clear gap therefore exists in the literature on self-plagiarism in Australian publications which has been addressed by this current study. This exploratory investigation has found that self-plagiarism is a common practice in academic research, with over 60% of authors in the sample having re-used text from a previous publication without appropriate citation. An analysis of the articles with substantial self-plagiarism showed that the authors use a strategy which we have called ‘a chain of textual re-use’ which results in the so-called ‘new’ publication containing only minimal original research. In addition, the study identified a writing method which we refer to ‘cut and paste’ textual re-use, used by 70% of the authors, which entails piecing together small sections of previously published material. While evidently both are common academic practice for many authors, self-plagiarism or textual re-use to this extent seems an unethical use of limited research funding intended for the generation of original work.

This study has raised the question of what is acceptable practice for textual re-use in the publishing of academic research. It has also raised a concern regarding what appears to be a relatively common practice of making dual submissions of identical or nearly identical papers to different journals. Furthermore, the study identified self-plagiarism in papers published only six months apart in the same journal, which supports the question raised by Collberg and Koroubov (2005) who asked ‘who is responsible to ensure the original of research – is it authors, publishers or reviewers?’ Clearly there is a need for guidelines regarding textual re-use, not just in relation to self-plagiarism, but also in terms of evaluating the originality of the research.

This preliminary study has important implications for the way that funding bodies allocate resources for research. In the Australian context, the Department of Education, Science and Training (DEST) provides universities with funding directly corresponding to the number of peer-reviewed articles published in a given year. While the Research Quality Framework, modelled on the Research Assessment Exercise in the UK and due to be introduced in Australia in mid-2007, will attempt to look at both the quantity and quality of research publications, the issue of self-plagiarism and textual re-use has yet to be addressed.

REFERENCES

Anderson, M. P. (2006). Plagiarism, copyright violation, and dual publication: Are you guilty? Ground Water, 44(5), 623.

Angelil-Carter, S. (2000). Stolen language? Plagiarism in writing. Essex: Pearson Education Limited.

Bailey, B.J. (2002). Duplicate publication in the field of otolaryngology-head and neck surgery. Otolaryngol Head Neck Surgery, 126(3), 211-216.

Barnard, H. & Overbeke, A.J. (1993). Duplicate publication of original manuscripts in and from the Nederlands, Ned Tijdschr Geneeskd, 137(12), 593-597.

Blancett, S.S., Flanagan, A. & Young, R.K. (1995). Duplicate publication in the nursing literature, Journal of Nursing Scholarship, 27(1), 51-56.

Bloemenkamp, D.G., Walvoort, H.C., Hart, W. & Overbeke, A.J. (1999) Duplicate publication of articles in the Dutch Journal of Medicine in 1996, Ned Tijdschr Geneeskd, 143(43), 2150-2153.

Boisvert, R. F., & Irwin, M. J. (2006). Plagiarism on the rise. Communications of the ACM, 49(6), 23-24.

Bretag, T. (2005, 2-3 December). Implementing plagiarism policy in the internationalised university. Paper presented at the 2nd Asia-Pacific Educational Integrity Conference, University of Newcastle.

Bretag, T. (2007). The emperor’s new clothes: Yes, there is a link between English language competence and academic standards. People and Place, 15(1), 13-21.

Brice, J. & Bligh, J. (2004). Author misconduct: Not just the editors' responsibility. Medical Education, 39, 83-89.

Carroll, J. (2003, 21-22 November). Six things I did not know four years ago about dealing with plagiarism. Paper presented at the Educational Integrity: Plagiarism and Other Perplexities Conference, Adelaide.

Cheung, M. & Driver, D.C. (2004). Self-plagiarism as a social work concern. The Hong Kong Journal of Social Work, 38 (1-2), 3-13.

Collberg, C., & Kobourov, S. (2005). Self-plagiarism in computer science. Communications of the ACM, 48(4), 88-94.

Devlin, M. (2003, 21-22 November). Policy, preparation, prevention and punishment - one faculty's holistic approach to minimising plagiarism. Paper presented at the Educational Integrity: Plagiarism and Other Perplexities Conference, Adelaide.

Gotterbarn, D., Miller, K., & Impagliazzo, J. (2006). Plagiarism and scholarly publications: An ethical analysis. Paper presented at the 36th ASEE/IEEE Frontiers in Education Conference, San Diego, CA.

Green, L. (2005). Reviewing the scourge of self-plagiarism, M/C Journal (Vol. 8).

Hexham, I. (1999). The plague of plagiarism. Retrieved March 6, 2007, from http://c.faculty.umkc.edu/cowande/plague.html

Hiller, M. D., & Peters, T. D. (2005). The ethics of opinion in academe: Questions for an ethical and administrative dilemma. Journal of Academic Ethics, 3, 183-203.

Howard, R.M. (1995). Plagiarisms, authorships, and the academic death penalty, College English, 57(7), 788-806.

Howard, R.M. (2000). Sexuality, textuality: The cultural work of plagiarism, College English, 62(4), 473-491.

Kassirer, J.P. & Angell, M. (1995). Redundant publication: A reminder, New England Journal of Medicine, 333(7), 449-450.

Lowe, N.K. (2003). Publication ethics: Copyright and self-plagiarism, Journal of Obstetric, Gynecologic and Neonatal Nursing, 32(2), 145-146.

McCabe, D. (2005). Cheating among college and university students: A North American perspective, International Journal for Educational Integrity, 1(1).

Roig, M. (2006). Avoiding plagiarism, self-plagiarism, and other questionable writing practices: A guide to ethical writing. Retrieved 5 January 2007, from http://facpub.stjohns.edu/~roigm/plagiarism/Index.html

Monash University. (2006). Citing your own work. Retrieved 20 March 2007, from http://calt.monash.edu.au/staff-teaching/plagiarism/acknowledgement/module2/sc/intro.html

Samuelson, P. (1994). Self-plagiarism or fair use? Communications of the ACM, 37(8), 21-25.

Scanlon, P.M. (2007). Song from myself: An anatomy of self-plagiarism. Plagiary: Cross-Disciplinary Studies in Plagiarism, Fabrication and Falsification, 2 (1), 1-10.

Schein, M. & Paladugu, R. (2001). Redundant surgical publications: Tip of the iceberg? Surgery, 129 (6), 655-661.

Sheik, A. (2000). Publication ethics and the research assessment exercise: Reflections on the troubled question of authorship. Journal of Medical Ethics, 26, 422-426.

University of South Australia. (2003). Using information appropriately. Retrieved 30 May 2007, from http://www.library.unisa.edu.au/infogate/08/08b2.htm

Wilson, K. (1997, 8-11 July). Wording it up: Plagiarism and the interdiscourse of international students. Paper presented at the Annual Conference of the Higher Education Research and Development Society of Australia, Adelaide.

Zobel, J., & Hamilton, M. (2002). Managing student plagiarism in large academic departments. Australian Universities' Review, 45(2), 23-30.

NOTES

1. Turnitin (http://www.turnitin.com) is a text-matching software program which produces ‘originality reports’ from written text submitted to the Turnitin program and matched against existing material on the Turnitin database or on the Internet. Importantly, Turnitin cannot identify plagiarism (of others or own work); it simply matches text to facilitate manual analysis. Papers were entered into Turnitin twice to allow them to be matched against each other, in addition to being matched against other sources.

2. ‘Author Record’ refers to the list of available full-text electronic articles published by a specified author in the sample. For reasons of confidentiality, no author names are used in this research.

Acknowledgements: The authors wish to acknowledge the generous support of the Hawke Research Institute for Sustainable Societies at the University of South Australia which enabled the data collection for this study; the opportunity to discuss the ideas in this paper and the encouragement to pursue the research provided by Dr Howard Harris, Professor Tony Winefield, Associate Professor Chris Provis and Professor Richard Blandy; and the assistance of Carole Gibbs at the University of South Australia Mawson Lakes Library.

Tracey Bretag, BA (Hons), MA, Ed.D, teaches a range of communication courses in the School of Management at the University of South Australia. She is the Editor of the International Journal for Educational Integrity, and the Co-Chair of the 3rd Asia-Pacific Conference on Educational Integrity: Creating a culture of integrity (University of South Australia, 6-7 December 2007). Her research interests include internationalisation of higher education, academic integrity, teaching English to Speakers of Other Languages (TESOL), developing learning communities, computer mediated communication, interculturality, and gender equity.

Saadia Carapiet, MBA (University of Karachi), MPhil (University of Cambridge), PhD (University of South Australia) is a lecturer at the School of Management at the University of South Australia. She complements her passion for management research with industry involvement as a financial planner holding a Graduate Diploma in Financial Planning (Finsia). Her research interests include self-organisation and adaptive capability, academic integrity and management of technology.