Metaphors for University Teaching
Supporting First Year Undergraduates through Blended Learning
Margaret Kendall and Alicia Prowse
Three Years of eLearning - the guinea pigs bite back!
Liz Marr and Guy Lancaster
The JISC Plagiarism Detection Services Revisited
Sociology Students Submit! to the JISC PAS
A 3D Response to reducing cut-and-paste plagiarism using the JISC PAS
Ian Martin and Mark Stubbs
The PlaySMART Research Project: promoting thinking through physical education
| View this article as a .pdf file
Ian Martin and Mark Stubbs
Manchester Metropolitan University Business School
A 3D Response to Reducing Cut-and-paste Plagiarism Using the JISC Plagiarism Detection Service
This article describes an action research project undertaken in the Business Information Technology (BIT) subject group at the Business School to combat the growing menace of cut-and-paste plagiarism. The authors regard plagiarism—the passing off as one’s own, the words and ideas of another—as an academic malpractice that should be deterred , detected and dealt with appropriately (Park, 2004, JISC 2005). We use these three themes, or dimensions, to structure our account of an expanding portfolio of tools and techniques we have deployed over a period of two and a half years. Recently the Joint Information Systems Committee’s (JISC) Plagiarism Advisory Service (PAS) has become central to our efforts, and whilst it is certainly useful in structuring student perceptions, detecting and highlighting sections of cut and paste, and providing professional disciplinary evidence, we advise it should be used carefully as part of a more holistic approach to student plagiarism rather than as a quick and easy panacea.
Pilot studies carried out across six undergraduate and postgraduate units have revealed a growing awareness, amongst both academic and student enthusiasts, of the strengths and limitations of this service. Potentially, these limitations, combined with the restricted sanctions available according to university regulations, could constitute a small risk that some students may calculate and be willing to take (Woesnner, 2004). We feel it is important therefore, whilst working within this framework, to adopt other complimentary strategies in order to make the wholesale or part copying of another’s work an irrational choice, even for the desperate student. This article draws upon current plagiarism literature and field observations from our ongoing action research to frame a ‘3D’ response that attempts to share best practice in deterring, detecting, and dealing appropriately with cut-andpaste plagiarism.
The authors’ use of information systems to supplement and enhance the judgement of lecturers making decisions regarding the authenticity of student work has its beginnings in the Emerging Technologies and Issues 1 (ETI1) unit. Here a bespoke computer system was constructed to analyse completed programming assignments and use pattern matching algorithms to detect similarities between submissions in order to prompt tutors to examine work further and possibly call for a face-to-face explanation (viva). The focus in that unit was to reduce the significant overhead of running vivas for every student whilst ensuring that students were aware that plagiarism could be detected and would be acted upon in a unit culture that promoted code sharing, but ultimately stressed individual understanding (Stubbs & Martin, 2003).
Indeed we began to observe that not only was the student culture informed by code sharing at a local level, but also by wider issues that could potentially encourage plagiarism in other units and had been reported elsewhere. These were: the ease with which material could be copied and pasted from Internet resources; the growing filesharing culture and blurring of ownership and copyright issues; the increase in incidences of graduates selling dissertations on eBay and access to ‘paper mills’; a possible lack of confidence with the English language; and previous educational environments failing to emphasise plagiarism as deviant behaviour. See Bennett (2005) for a more comprehensive review.
Whilst we acknowledge that incidences of plagiarism are sometimes more sophisticated than simply cutting and pasting work from unreferenced sources, and that others have quite rightly debated the definition of plagiarism (see for instance Johnston , 2003), the fuzzy nature of the concept is not of central concern in this paper. Within BIT, we made an early decision that the direct copying and pasting of another’s text or code as a shortcut to writing and understanding the work oneself was the most clear cut of crimes and for pragmatic reasons the most sensible to tackle. The electronic detection tools at our disposal both pre- and post- the JISC PAS could readily highlight instances of this type of plagiarism and as such structured our approach. Although this was recognised as a less than 100% solution for all instances of plagiarism, it did however neatly circumvent the numerous referencing style issues that students were finding particularly difficult and which were being tackled by other initiatives elsewhere in the Business School and across MMU. The increasing potential of software other than the Turnitin UK software used by the JISC PAS to provide a more expert comparison between submitted work (Clough, 2000; Lancaster & Culwin, 2004) was also acknowledged and we have been watching these developments with interest. Nonetheless, the packaged JISC service and the impressive growing scope of its document base meant that, although we were aware of the long-term risks of over-reliance on proprietary software, for the short-term this service provided the most effective information system to augment our 3D response to the cut-and-paste plagiarism problem.
Whilst a student may unwittingly commit academic malpractice through ignorance of rules and norms, the increasingly commodified nature of Higher Education in the UK means that a student might well knowingly plagiarize when the apparent rewards for breaching rules and norms outweigh the sanctions perceived for being caught (Saltmarsh, 2004). These perceived rewards include time saving, mark improvement, failure mitigation, or language improvement. A review of the literature suggests this type of plagiarism is becoming more common and that those students studying at business faculties could be most inclined toward this type of offence (Park, 2003). It is important therefore to discourage students from plagiarizing not only by reinforcing institutional policy sanctions consistently, but also by designing assessments at the unit level that make plagiarism the difficult option and so act as an effective deterrent.
With this in mind, our first year unit ETI1 made use of a holistic assessment redesign that emphasised individual understanding and a brief that, by the very nature of its currency (emerging technologies and issues ) , was changed year upon year. The brief also included a marking grid that spelt out plagiarism sanctions, in line with institutional tariffs, as percentage penalties for a failure to demonstrate understanding when requested. This was reinforced collectively in lectures and individually in tutorials to stress that personal understanding of the work presented was paramount and this understanding would be tested at a viva if necessary. This, coupled with the requirement to submit the work electronically, and repeated written and verbal reinforcements that sophisticated tools would be used to compare submitted work acted as a useful initial deterrent. It was found on this first unit that simply requesting electronic submission raised the perception of detection. Further demonstrating to the students, by means of a selective viva process, that their submissions were being scrutinised systematically ensured that student-to-student communication, both horizontally across the unit and vertically back to subsequent years, reinforced the intended message that plagiarism was taken seriously by tutors on this unit.
Following in the footsteps of good plagiarism practice reported in other technical units outside of the Business School we then broke down a second assessment in a more advanced level unit, ETI2, to include a summative assessment that tested knowledge using a multiple choice questionnaire (Eskins, 2004). This test was scheduled early on after the distribution of the assignment brief with the aim of encouraging students to start work early and hopefully militate against the last minute submission culture that could encourage students to plagiarise out of panic. Students were also provided with comprehensive written and electronic resources in tutorials with which they could check their progress with tutors. This carrot and stick combination proved effective for those students who were willing and able to engage fully.
The use of electronic submission to structure students’ perceptions of plagiarism was rolled out to the doubleweighted final year Dissertation unit where the perceived rewards for plagiarism in terms of time and effort appeared greater than any others, and also the postgraduate Information Systems in Organisation (ISO) unit where a greater number of overseas students, for whom English was not their first language, were enrolled. Varied methods of submission were trialled including email to tutor, saving to network drive, submission to WebCT, and submission direct to the JISC PAS. Whilst all were not without teething problems, submission direct to the JISC service proved to be most effective in raising the profile of serious plagiarism detection and therefore acted as the most effective deterrent.
This year, the Electronic Commerce (EC) final year option was redesigned to include submission of a 3,000-word business report direct to this service and, in addition, JISC advisory material was distributed to students at lectures and via the Business School Intranet. The best practice recommendations of Carroll & Appleton (2001) were incorporated into an assignment brief that focused on application of concepts to a specific business problem, which encouraged individual ownership and creativity and resulted in 85 unique solutions within the same assignment framework. Deterrence was not only reinforced through assignment design, electronic submission, and penalties on the marking grid, but an in-class exercise required that students agree upon a shared definition of plagiarism. The students literally drew the line under what was deemed unacceptable (Swales & Feak, 1994). As long as this line was well below our limits of cut-and-paste plagiarism then it was stressed—in order to ease other students’ heightened anxieties—that for this unit referencing was important, but small syntactical mistakes did not carry the same level of penalties as verbatim copying either with or without acknowledgement.
Having outlined key initiatives undertaken to deter plagiarism, we now describe efforts directed towards its detection as a credible deterrent requires a real risk of being caught.
Detection within paper submissions very much depends on an assessor’s knowledge of and access to relevant texts or their ability to recognise plagiarism signatures, such as style changes, within a text. Whilst the increased use of the Internet has afforded easier opportunities for student plagiarism it has also allowed assessors ready access to relevant texts via search engines for comparison and plagiarism detection purposes. Tutors have in the past often used this mechanism informally when their suspicions have been aroused whilst marking an assignment. Electronic submission makes this comparison more efficient and so routinely extends an assessor’s powers of comparison and pattern matching to include a wider base of source material. The automation of this comparison process in order to detect suspected instances of plagiarism was first attempted in the ETI1 unit. A bespoke set of utilities compared all submissions with each other and then against all previous years’ submissions for this unit. Copied material was highlighted and presented to assessors within an easily navigable web-based marking system that allowed assessors to select suspected students for viva. These powers of detection were further enhanced in ETI2 to combine the results from the early multiple-choice test with both virtual and physical attendance statistics in order to build up a profile of a likely plagiarist. Research has shown that it is weaker students that do not engage and identify with the learning outcomes that are most likely to plagiarize (Bennett, 2005). It must be stressed however that this profile was used only to give more attention to those students in order to examine their work more carefully for evidence of plagiarism; under no circumstances did it lead to the tutors pre-judging a student as guilty.
Whilst a comparison of student submissions within the unit both horizontally and vertically was achievable with this bespoke system, it had its limitations. These were: the absence of an ability to do external comparisons; it was only designed to compare programming assignments; and it relied heavily on the technical expertise of the authors. The advent of the institution’s subscription to the JISC PAS provided a convenient solution for automating the detection process to encompass both technical and non-technical units and widened the scope of comparison to: 3.5 billion websites; a subset of research paper databases; and submissions from all other subscribing institutions. However for non-technical assignments the culture of electronic submission and potential viva was not well established and we found that careful management of the submission process was required to ensure that an electronic copy was received for detection purposes. To be effective this meant that students were informed via the brief that penalties would be applied unless both paper and electronic submissions were received. We had trialled electronic-only submission on the ISO unit, but had found that tutors generally found marking online more restrictive than traditional paper-based marking and often resorted to printing out student submissions which led to an unacceptable administrative overhead and frequently a loss of formatting for the majority of submissions that were in Microsoft’s Word rather than Adobe’s Portable Document Format (PDF).
It was also observed that some students became aware not only of the sophistication of the detection process, but also of the limitations given our requirement for both electronic and paper submissions. General weaknesses inherent in this method and some specific to the functionality of the JISC PAS meant that tutors could be lulled into a false sense of security if they relied solely on the JISC plagiarism reports for detection purposes whilst marking. Currently, in particular, the JISC service can stall when it reaches embedded spreadsheets and graphics within a document and only allows a student to submit one file per assignment. Students who lacked sophisticated document production skills often produced a paper assignment from more than one electronic document and so, without the support of any prior advanced document design and production training, were advised just to submit their largest file. Even those students who presented one document found on occasion that the Turnitin UK engine had successfully processed only a subset of their work. Conscientious students would worry about these limitations and we spent considerable time managing this process because we had repeatedly stressed the importance of electronic submission to the students as part of the deterrence phase. More worryingly though, were reports of at least one student who had removed offending plagiarized material from his electronic submission safe in the knowledge that we could not realistically compare all paper versions with electronic copies with a staff/student ratio of 1/25. Until enhancements are made to the service to include digital watermarks that can reassure tutors of the authenticity between paper and electronic submissions, or a move is made towards robust PDF generation and electronic-only submission with industrial-strength secure printing then we advise caution in using the JISC service as a primary method of detection. Rather, in BIT, we have gently introduced the JISC PAS as a tool to enhance rather than replace a tutor’s expert judgement.
This softly, softly approach is not without its own issues. General findings indicate that some academic staff feel uncomfortable confronting plagiarism or simply do not have time to deal with the perceived increased administrative overhead (Park, 2003). Leaving the exploitation of the JISC service to a few enthusiasts does mean that some students, depending upon which units they take, are scrutinised more closely than others. This is a disparity that can only be addressed by institutionalising the use of the JISC PAS together with good practice plagiarism recommendations for assessment design so that the detection of plagiarism is as consistent as the penalties imposed for plagiarism offences set out to be.
We will return to some of the issues of detection and submission in our conclusions, but it is important to say something first of actions taken in response to plagiarism.
Use of the JISC service provided tutors with a valuable extra tool to supplement their professional judgement, and whilst disciplinary hearings have been relatively few and far between, producing evidence has now become a matter of colour printing the JISC reports. The extent of plagiarism is now readily apparent, but hearings also consider intent. It is important at this stage that the evidence is acted upon in a consistent and well-publicised manner. An assessor’s willingness to manage the plagiarism deterrence and detection process and subsequently escalate instances of cut-and-paste plagiarism depends upon the gathered evidence being dealt with in an appropriate way that not only deters a student from re-offending, but also communicates to the student body at large that the penalties for academic malpractice are not worth the risk. Some would argue that the limited sanctions available according to university regulations would not appear to send this message (Woesnner, 2004). However, working within this sanction framework we recommend tutors reinforce institutional policy within the assignment brief and ensure all students engage with a shared definition of plagiarism so there is no room for doubt about intent if an incident does occur. We would encourage the university to keep its penalty regime under review as efforts to design out plagiarism become more sophisticated.
While it has been reported elsewhere that some staff do not act on suspected plagiarism because of the extra work involved (Park, 2004), we feel that the time saved by using the JISC PAS to generate reports for a plagiarism hearing, far outweighs the small amount of extra time required to review the initial reports. Of course further work is required to redesign units holistically to deter plagiarism and to manage the electronic submission process, but we hope in the first instance the pedagogical rationale of an application of concepts rather than explanation wins out, and in the second that the submission process can be improved to become part of an integrated submission service to provide further benefits for both staff and students.
The Way Forward
The ad-hoc enthusiast model we have described thus far as our ‘3D’ response to cut-and-paste plagiarism needs to be replaced with something more convenient for mainstream use. We recommend that in the longer term an integrated online submission service be developed with e-submission going to e-portfolio, e-print and e-detection services. This local development will of course require appropriate investment in both students and staff to make it work. A cultural shift may also be needed to re-frame assignment submission from loaning a piece of work to an assessor to receive a mark, to making an individual contribution to a community of practice, in which the contribution is held in perpetuity to uphold academic norms of integrity and originality.
In the medium term, if the JISC PAS is to become part of the university’s infrastructure for assignment submission, then just as with any other key information system, consideration must be given towards its accuracy, reliability and transparency (Lancaster & Culwin, 2004) and the levels of support offered to support 24x7 assignment submission. Improvements are required to the software to allow manifold multi-type file submissions per assignment and more flexibility is required to allow multiple assessors on a unit team to view the same plagiarism reports. While the JISC PAS is well suited to identifying cut-and-paste plagiarism it is hoped that its sophistication will continue to grow to match that of its rivals. To mitigate these improvements not taking place, the university needs to give consideration to creating an e-detection framework that will prevent the institution becoming locked into the JISC service and allow a graceful move to alternatives should the need arise.
In the short term although we do recommend the JISC service as a useful tool as part of a more holistic approach to combating cut-and-paste plagiarism, it should by no means be viewed as a universal remedy. The consequences of adopting the service in terms of affecting not only student but also staff perceptions of plagiarism require careful consideration. Whilst we wish to see the use of the service institutionalised to ensure equity in the treatment of students, we do not want to see assessors removed from the process of deterring, detecting and dealing appropriately with instances of plagiarism. It is our belief that not only is it important to preserve professional judgements like this within the realms of those best placed to make them, but it is also important that the message ‘plagiarism is unacceptable’ is delivered by those closest to the students whilst they are here rather than being institutionalised and potentially lost amongst many other impersonal communications.
Bennett, R. (2005) ‘Factors associated with student plagiarism in a post-1992 university.’ Assessment and Evaluation in Higher Education , 30(2), pp. 137-162.
Carroll, J. & Appleton, J. (2001) Plagiarism: a good practice guide. Retrieved June 16, 2005.
Clough, P. (2000) Plagiarism in natural and programming languages: an overview of current tools and technologies. Retrieved June 16, 2005.
Eskins, R. (2004) ‘Designing out plagiarism and supporting widening participation: a strategy implemented to support a technical unit, with online submission.’ Learning and Teaching in Action, 3(3).
JISC (2005) Deterring, Detecting and Dealing with Student Plagiarism . Retrieved July 18, 2005.
Johnston , W. (2003) ‘The concept of plagiarism.’ Learning and Teaching in Action, 2(1).
Lancaster, T. & Culwin, F. (2004) Using freely available tools to produce a partially automated plagiarism detection process. In R. Atkinson, C. McBeath, D. Jonas-Dwyer & R. Phillips (Eds), Beyond the comfort zone: Proceedings of the 21st ASCILITE Conference (pp. 520-529). Perth , 5-8 December.
Park, C. (2003) ‘On other (people’s) words: plagiarism by university students—literature and lessons.’ Assessment and Evaluation in Higher Education , 28(5), pp. 471-488.
Park, C. (2004) ‘Rebels without a clause: towards an institutional framework for dealing with plagiarism by students.’ Journal of Further and Higher Education , 28(3), pp. 291-304.
Saltmarsh S. (2004) ‘Graduating tactics: theorizing plagiarism as consumptive practice.’ Journal of Further and Higher Education , 28(4), pp. 445-454.
Stubbs, M. & Martin, I. (2003) ‘Blended learning: one small step.’ Learning and Teaching in Action , 2(3).
Swales, J. M. & Feak, C. B. (1994) Academic writing for graduate students: essential tasks and skills: a course for non-native speakers of English. University of Michigan Press, Ann Arbor .
Woessner, M. C. (2004) ‘Beating the house: how inadequate penalties for cheating make plagiarism an excellent gamble.’ PS: Political Science and Politics , 37(2), pp. 313-320. Ian Martin
Tel: 0161 247 3894
Tel: 0161 247 3739
top of page