Writing as criminologists, Hodgkinson et al (2014) propose that: “There are three broad categories of academic dishonesty: plagiarism, test cheating, and collusion”. Unfortunately, this is not a sharp distinction, because one category can be embedded in another. However distinct and familiar these labels are, the phenomena they refer to rarely sit apart in a convenient manner. Not only are boundaries blurred, but there exist gradations of offence for the practices commonly attached to these labels.
We have built the present review around a further category, namely, ‘social plagiarism’. This label has been chosen to help concentrate attention on a family of practices that (i) capture the ‘failure of attribution’ that is so resonant with the term ‘plagiarism’, while (ii) also capturing the ‘construction in partnership’ that is resonant with ‘collusion’. Viewing offences through the lens of ‘social plagiarism’ helps counter some of the rigid reasoning that can arise when too much faith has been put into allocating a practice to one of Hodgkinson et al’s three primary categories alone.
Collusion: definitional problems
In university statements for regulating integrity offences, ‘collusion’ is sometimes regarded as a sub-set of plagiarism. Both ‘collusion’ and ‘plagiarism’ can thereby refer to authorship tainted by undeclared borrowing of material from others. With classic ‘text plagiarism’, the “other” is a document. Accordingly, there is a visible record of that borrowed material – a record that is independent of (but preceding) the assignment submitted. So, comparisons can be documented, and accusations can be evidenced. However, this need not be the situation for all plagiarism offences termed ‘collusion’. The controversial material that is ‘borrowed’ might exist only at the level of conversational exchange and not evidenced by independent document ‘drafts’. The advantage of working with the phrase ‘social plagiarism’ is that it starts from the basic offence of unattributed borrowing – but highlights a large family of cases that involve borrowing mediated by some kind of collaborative partnership.
If collusion is defined at all in regulations, it is usually as a variant of: “The work submitted has resulted from collaboration with others whose contribution has not been acknowledged” (Nottingham). It is positioned as a subset of plagiarism and, for some cases, this will feel right – because, as with text plagiarism, there may be residual documents arising from that “collaboration with others”. But often there will be no such records and so the plagiarised/borrowed ‘material’ is rather more ephemeral. And, most troublesome, this means that the standard methods for identifying such collusion as being a case of plagiarism (e.g., Turnitin analyses) cannot be applied.
We take up this tricky matter of detection later. In the present section, we are more concerned with how the general practice of “collaboration with others” is regulated. The often-ephemeral nature of illicit collaboration makes defining the offence of collusion a daunting task. Sadly, labelling it as a class of plagiarism may distract institutions from noticing this. In our own research we have studied regulations as published by more than 30 UK universities; we considered how they address this definitional challenge (Crook and Nixon, 2019).
It is common for institutions to have a shallow depth of field when viewing collusion. For example, there may be an assumption that the colluding agent must be a course peer, as in this definition: “a student or students collaborate with another student or students, as an individual or group to gain a mark or grade to which they are not entitled” (Manchester). This suggests a lack of attention towards the need to rule on all the other agents who might be illicit partners in claimed authorship. Or, more likely, it is assumed that those other partners are covered by the single category ‘essay mill’ – which is usually discussed in its own section – and where the rules are crystal clear (“don’t purchase assignments!”).
Major definitional problems surround specifying exactly when innocent collaboration mutates into something that is an offence – namely, ‘collusion’. Some regulations deal with this by invoking “collaborate only if permitted” rules. For example, “Unless it is made explicitly clear that you have been given a joint assessment, you should never work directly with other students on your module or anyone else when creating your work” (Sussex). Other approaches urge that collaboration must be declared: “If in doubt, state that the work has been peer reviewed, when and by whom, and be prepared to furnish proof of the extent of the contribution of the reviewer” (Leeds). Yet, the identification of other voices in a piece of work and the documentation of their reach could be a disheartening responsibility. Integrity guidelines risk undermining their own credibility with such demanding requirements. The worry would be that challenging obligations may erode the respect given to guidelines by students.
Such obligations have to be reconciled with collegial advice of an apparently opposing kind: “Sharing ideas with others is often an excellent way to learn, as you can bounce ideas off each other and find alternative points of view that you wouldn’t have thought of yourself” (Wolverhampton). Yet such messages are easily contradicted. For example: “Do not discuss an assignment in detail with other students if you are working on the same question. Even if they do not work directly alongside you, they may come to exactly the same conclusion following a conversation they had with you, and this could be classed as collusion” (Sussex). It may be supposed that such proscriptions are intended only for the study category ‘assignment’. But such tasks my be set at the very start of a course: It is optimistic to suppose that they can be readily cordoned off from normal discussion habits.
Research on students ‘native’ understanding of academic integrity stress how this is viewed in terms of mutual respect and trust (Ashworth et al., 1997). Yet these are values that may sometimes seem to be threatened by institutional guidance:
Do not lend your work to course-mates or leave it where they can access it. It is an offence to knowingly allow others to use your work. Even if you trust them, you could get a nasty surprise if it turns out they copied some of your ideas. (Sussex)
If your friends ask you how to approach an assignment and you give them a copy of your work, the chances are they will copy it, even if they claim they won’t. YOU and the other student(s) can be penalised if this happens. (Wolverhampton)
A member of staff in our research interviews expressed the tension experienced by students in such contexts of regulation
It’s difficult maybe then for students to understand how much they can talk to each other and how much they can share their ideas. I think sometimes students do end up getting quite cagey with their ideas, so they don’t want to share their thoughts with somebody else in case it shows up in their – which doesn’t really quite allow for the discussion of ideas that you might want.
On studying this variety of institutional advice, our impression was that inappropriate collaboration was not clearly enough discussed as a form of plagiarism. Moreover, advice that was given on good practice was variable, sometimes contradictory, and sometimes potentially divisive. McGrail and McGrail (2015) have raised similar concerns from their study of plagiarism regulations at research intensive US universities; while Bermingham et al (2010) do so for the particular academic family of UK Law schools. In short, many studies have reported student uncertainty around the collaboration/collusion distinction (Sutherland-Smith, 2013; Velliaris, 2015). More worrying, other research has revealed students can limit their collaboration with peers on the grounds that they fear accusations of cheating (Stacey, 2016; Sutton and Taylor, 2011).
Student and staff responses to regulations
It may be judged inevitable that there is variation across institutions in the definitions given by the sector to key concepts of academic integrity. It is surely less acceptable if staff and student understanding of regulations varies significantly within an institution. Of course, this is likely to happen if the regulations proscribed by an institution are simply never looked at. Various studies suggest that this may often be so (e.g., Sutton and Taylor, 2011). For instance, Gulifer and Tyson (2014) surveyed 3405 students at an Australian university that believed it had a strong emphasis on academic integrity. Only half of this student sample had read the plagiarism policy, thereby displaying significant uncertainty as to how plagiarism should be defined. These researchers attribute this neglect to “information overload”. In other words, putting stress on such regulations at the start of a degree programme may risk having it overwhelmed by too much other induction information. If this is credible, it implies that integrity principles should be more firmly integrated with those teaching contexts or those teaching moments where their recognition is relevant. However, Gulifer and Tyson’s research is just one study of one institution. The story may differ elsewhere: access and consultation is likely to vary across institutions, depending on how they approach this challenge of gaining attention.
Even if regulations and guidelines are consulted, this does not promise that they will be understood in the way intended. Variation of interpretation has been researched for both staff and students. Risquez et al (2013) surveyed 787 first and second year students at an Irish university. Self-reported understanding of plagiarism did not correlate well with performance in tests of the ability to identify it in case study scenarios. Barrett & Cox, (2005) considered commonalities of understanding among staff and students in one UK university, finding that while text plagiarism was understood by both in institutional terms, collusion was not. Although both staff and students were inclined to see collusion as less serious (because learning was thought more likely to take place within it).
In deeper and more conversational explorations with students concerning integrity understanding, both Ashworth et al., (1997) and Adam et al., (2017) found students expressing confusion around interpreting institutional proscriptions, along with confusion as to their rationale, and a strong preference for discourses of fairness in guiding moral judgement. Such personal readings of integrity principles are also evident in studies of staff interpretations. Flint et al (2006) report this from interviews with 26 academics. The researchers make the important point that variability of understanding at staff level can lead to divisive differences in how those staff exercise integrity rules. Borg (2009) documents how such differences are meaningfully related to the particular disciplinary alignment of staff. Sheard et al (2018) develop this point around the case of Computer Science.
These signals of variability in accessing regulations – and variability of their interpretation – evidently raise questions as to how best to construct guidelines: how they can be designed with consensus, and how they can be then effectively made visible. The consensus question has received scant attention. More effort has been invested in considering effective methods for reaching students – considering, for example, how this material is best integrated with the individual student’s curriculum. It may help that there is now an academic integrity MOOC aimed at students and another that covers integrity matters for new international students. Individual institutions are also starting to design and evaluate such integrity programmes/modules for students (e.g., Benson et al, 2019). Resourcing staff engagement in a similar way does not seem to have been considered. However, a study by Ransome and Newton (2018) considers the texts commonly used in PGCHE courses for new academic staff. They conclude that “Overall, academic integrity was poorly addressed in these texts, and the language used (‘cheating’, ‘plagiarism’) was not reflective of an academic integrity-based approach. Newer issues, such as the purchasing of custom written assignments (contract cheating), were barely addressed”.
Policing regulations
Institutions adopt a range of solutions when it comes to acting on reported instances of integrity offence. They also have a range of approaches to the ways in which offences are evaluated and, then, what the consequences for the student should be. Glendinning (2014) summarises a European Union project which articulates some of the considerable variation that is apparent internationally.
Institutions differ as to whether polices are expressed in terms of ‘academic misconduct‘ or ‘academic integrity‘. Arguably, this choice matters, because the language that frames student cases communicates rather different kinds of messages to both the students and staff involved. Misconduct approaches deploy a vocabulary of criminal law and thus will refer to such concepts as ‘offence’, ‘penalties’, ‘theft’, and ‘guilt’. Whereas an integrity framing is more likely to refer to the language of ethical and professional ‘responsibilities’, ‘fairness’, and ‘respect’. Sutherland-Smith (2010) reports a study of plagiarism regulations as published by a set of Russel Group UK universities: they all expressed practice in misconduct terms.
Some commentators regret the ‘retributive’ approach to justice that dominates current disciplinary practices in higher education. Kara and MacAlister (2010) have outlined the case for a more ‘restorative’ justice approach. They suggest that the deterrence argument underpinning retribution is clearly not working in this case. Moreover, exclusionary punishments may dislocate offenders from the very remedial support that they require. In short, current solutions fail to reflect the goals and values of the HE sector. The restorative approach appears to hinge upon the concept of ‘integrity’ (not ‘discipline’) boards. They would have significant peer membership and would dwell on the student’s awareness of responsibility for action and the victims’ experience of that action. There are scarce examples of such methods in practice – although some practitioners have reflected positively on its value.
The choice for characterising a putative offence is not exclusively a matter of academic misconduct discourse versus academic integrity discourse – a regulatory matter versus a moral matter. Occasions of the issue may sometimes be approached in a third way – as a matter of academic literacy. For example the patchwriting or the shallow paraphrasing of a source may merely signal that the student has not yet developed the expressive repertoire required for the production of an academic text (Pecorari, 2015); this has sometimes been referred to as “unintentional plagiarism”. Any process of adjudication or efforts towards re-mediation will therefore have to consider this reach of this interpretation.
Summary
In sum, the research literature reveals problems of individuals failing to consult institutional guidelines, but also variability – both within and between staff and student communities – as to how they are then interpreted. Macdonald and Carroll (2006) derive a checklist of seven questions that institutions need to ask to consider if they are taking a suitably holistic approach to integrity management. Some progress can surely be made in key areas: for instance, tighter specifications of what is expected, more vivid translation of such principles into practice examples,, more consideration as to the best time to discuss these imperatives to students and faculty development programs to support staff (see Morris and Carroll (2015) and Serviss (2015) for examples). However, particularly in the definition of social plagiarism, there is a lingering tension: one that exists between the integrity expectations of an academic community and the more ‘native’ understandings of its student membership. The latter may be defined more around respect, trust and fairness – but towards peers, family and others, as much as to institutions. This tension is likely to remain a significant challenge.
