Staff views on commercial contract cheating in higher education: a survey study in Australia and the UK

Contract cheating is, potentially, a serious threat to the quality of higher education around the world. Prior research has focused on student perspectives and the companies themselves, but the staff view is poorly understood, despite staff being a major stakeholder with considerable influence over strategies designed to address contract cheating. We surveyed staff in Australia and the UK about their views on contract cheating. We asked staff to estimate the extent of contract cheating, how much it cost students to buy assignments, and whether proposed strategies to tackle contract cheating would work. We also asked about the factors which may motivate students to engage with contract cheating. Staff in both countries estimated high costs for an assignment from an online essay mill. Staff believed that low numbers (5 – 10%) of students are using these services, although this could represent approximately a quarter of a million students across the two countries. A large proportion of staff had had some experience with student cases of contract cheating at their university and reported that outcomes were lenient. The most prominent reasons which staff believed contributed to contract cheating were ‘ Studying in a non-native language ’ and ‘ Fear of Failure ’ . Over half of the respondents were aware of companies selling work to students on campus. There was strong support for the view that contract cheating services should be illegal, and that creative assessment strategies could reduce contract cheating. There was also modest, qualified support for the criminalising of student use of these services, and an increased use of examination-based assessments. Suggestions are made for how these data can inform the ongoing debate around contract cheating, including increasing staff awareness of contract cheating and the development of more appropriate assessment strategies.


Introduction
Contract cheating, as defined here, is the act of students submitting work for academic credit or formative assessment, which they have purchased from an essay mill or other service selling work to students. This is developed from the original definition by Clarke and Lancaster (2006) which focused on students who 'paid contractors to write for them'. The work can be custom-written to very detailed specifications (Lancaster and Clarke 2016;Newton and Lang 2016). The services are cheap and quick (Wallace and Newton 2014) and pose an obvious threat to academic quality and standards. In response to this, the organisations which define educational standards in the UK and Australia have been developing strategies to specifically tackle contract cheating (e.g. QAA 2017; TEQSA 2017). The provision of such services is currently legal in most countries, although New Zealand and some states in the USA have laws against it (Newton and Lang 2016) while others, such as the UK, are actively seeking the creation of new laws or exploring the use of current legislation to tackle contract cheating .
David Matthews reporting for the Times Higher Education Supplement (2013) suggested that contract cheating is a widespread problem in the UK and quoted Jennifer Wiss, Business Manager of UKessays.com, as saying that the company produced 11,470 essays during 2012 (Matthews 2013). By 2016, the company stated that it had sold 16,000 assignments and had a team of 3500 researchers (see Usborne 2017). McNeilage and Visentin (2014), reporting on the MyMaster essay writing site in Australia, found that over AUD 160,000 was paid to the site during 2014, equating to over 700 assignment purchases. In 2014, the high court in New Zealand investigated a couple under suspicion of having received over $1.1 m (NZD) from selling work to students from the website Assignment4U over a period of 7 years (see Savage 2014).
Research in Australia has assessed the quality and costs of papers produced by contract cheating services. Lines (2016) revealed how easy it was for purchased assignments to be marked and awarded a pass grade with no suspicion, by markers, of contract cheating. She purchased one undergraduate and one postgraduate assignment from each of 13 essay mills and asked academics, in the discipline of history, to mark the papers. Only 2 of 13 undergraduate and 6 of 12 postgraduate papers failed. Some plagiarism was picked up in three of the assignments, but none of the markers failed the work through suspicion of them having been bought (the academics thought they were marking real student work). Dawson and Sutherland-Smith (2017) undertook a similar study; however, they informed staff that some of the papers had been purchased and they were asked to detect which ones they were. Sixty-two percent of the time the markers were correctly able to identify those papers which were purchased (Dawson and Sutherland-Smith 2017). This is a worryingly low number when one considers that markers knew there were a proportion of papers which had been purchased. The apparently high quality of the essays may reflect the fact that writers may be former HE students (Sivasubramaniam et al. 2016).
The reasons why students engage in contract cheating via essay mills and other commercial services have not been studied in detail, although a recent study in Australia suggests that a general outsourcing (e.g. including to friends and family) is driven by three main factors: dissatisfaction with the learning and teaching environment, speaking a language other than English at home and the perception that there are lots of opportunities to cheat . Hundreds of self-report studies, conducted over the last century, have measured student self-report of misconduct in general and have attempted to identify factors associated with it. A full review of this literature is beyond the scope of this paper, but there are some common themes, as summarised in recent reviews (e.g. Liebler 2016;Whitley 1998;Megehee and Spake 2008), for example that misconduct is more common where students perceive that cheating is normal, lack motivation, or are fearful of failure. Wilkinson (2009) undertook a comparative study which considered staff and student opinions on the motivations for plagiarism. She noted that staff tended to favour harsher penalties for plagiarism than students did, and that there was a disparity in beliefs of the seriousness of cheating behaviours between the students and staff. As work by Carroll (2007) has stated, it is important to have a transparent and clearly understood policy, and associated processes, in an institution so that everyone has the same understanding of the definitions of cheating and plagiarism and an awareness of what any possible actions should be for this, as well as expected academic conventions. Without shared agreement on the definitions, seriousness and elements of academic integrity and misconduct policies, educational practice to enhance integrity and minimise misconduct are unlikely to succeed (Brimble and Stevenson-Clarke (2005), Wilkinson (2009)).
Many studies undertaken on contract cheating have looked at student engagement with an understanding of this form of cheating. Jones (2011) found that 100% of students surveyed believed that handing in someone else's work as their own was plagiarism, but only 75% of students thought that downloading an assignment from the Internet could be considered plagiarism. Specifically, in regard to essay mills, Christensen Hughes and McCabe (2006) surveyed students in Canada and asked them to self-report cheating behaviours: when asked whether or not obtaining work from an essay mill (free or purchased) was cheating, students stated that it constituted moderate to serious cheating. However, around 10% of students did not think it was cheating at all (Christensen Hughes and McCabe 2006). Similar results have been found more recently in the UK, where student suggested outcomes from a contract cheating scenario was significantly more lenient than that which would be imposed by their institution (Newton 2015).
A recent survey study from Australia ) examined the views of staff towards contract cheating. The study used a broad definition of contract cheating that included outsourcing to any third party, including friends and family, rather than the restricted focus on commercial contract cheating studied here. Staff were concerned about contract cheating, more so than students, but viewed contract cheating as symptomatic of wider structural problems in higher education. This meant that suspected contract cheating often went unaddressed or resulted in lenient penalties.
A number of strategies have been proposed to address contract cheating (reviewed in Lancaster and Clarke 2016;Newton and Lang 2016). These include altered assessment strategies, such as making assignments more reflective and personalised, increasing the use of exams (including oral vivas), and legislative approaches. There is already legislation against contract cheating in some jurisdictions; in some states in America, legal cases can be taken against companies found to have sold work to students for assessment. In New Zealand, companies can be prosecuted and fined up to NZ$10,000 if proved that they provided third party assistance to cheat (Newton and Lang 2016;. However, it is not currently known whether there is support amongst staff for the use of a legislative approach.
The existing literature on contract cheating has largely focused on students, the companies themselves, and ways to address it. When this survey was designed, the staff perspective into contract cheating was lacking from the literature. It is critically important to gather staff perceptions about how common contract cheating is, and why they think it is happening. They are the persons tasked with dealing with this on a regular basis; they are tasked with attempting to detect any instances of contract cheating, determining relevant outcomes for those caught cheating and designing assignments to encourage engagement and dissuade cheating behaviours. Staff see this behaviour happening 'on the ground' and so have, as a fundamental stakeholder group, a unique insight into why it might be happening. Gathering staff views to help understand student academic misconduct; defining what it is and what to do about it, has formed an important thread of the research literature on academic integrity previously (for examples, see Bennett et al. 2011, Bruton and Childers 2016, Flint et al. 2006, Li 2015, Serviss 2015, and here we extend that approach to commercial contract cheating. Staff are also one stakeholder group that will be charged with the design and enacting of any new strategies to address contract cheating (QAA 2017;TEQSA 2017). If staff are not informed on student-reported motivations for contract cheating, prevalence or effective detection methods, then their understanding of contract cheating may differ from that of students, putting possible prevention strategies at odds with the problem itself. For example, one possible set of findings here could be that university staff think that contract cheating is uncommon, expensive, undertaken largely by students who do not understand academic requirements and should be addressed by a return to examinations. An alternate set of findings could be that staff think contract cheating is common, cheap, driven by fear of failure and should be addressed by structural changes to the business model of higher education. Given the central role that staff play in higher education, the implications for the HE sector are very different in either case. The gathering of these views is important regardless of whether they are accurate; in fact, it is important to understand, where possible, how accurate staff views are as it will identify areas for future work to address contract cheating. For example, if staff think that contract cheating is incredibly rare and very expensive (when the literature suggests the opposite), then this would make it less likely that they recognise it as a problem and so are less likely to take steps to address it. Staff will be significantly affected by the outcome of the ongoing debate about the legality of contract cheating services ) as this will, potentially, put them in the position of being the first to report a 'crime'. If staff are also influencing these debates into criminalisation, it is important to understand whether or not their perceptions of the extent and reasons behind contact cheating are just that, perceptions; rather than being based upon reported figures given by students in published studies. As an important stakeholder, staff are also able to give us some important insights into why, in their view, contract cheating happens.
We used an online survey-based methodology to determine staff views on all these important aspects of commercial contract cheating in higher education. We sought to understand their perspectives on why it happens, their perceptions about how much it costs students, how many students are doing it, and what the response should be to reduce the problem, including specifically addressing the issue of whether or not it should be criminalised.

Survey design
Staff view on the use of commercial contract cheating services is a topic that, at the time of this research, has not been formally studied previously. Thus, we were not able to use a previously validated survey instrument and had to design one ourselves. To do this, the authors drew upon their published background in contract cheating research along with their respective workplace experiences as an academic and professional staff member both tasked with dealing with contract cheating by students. We discussed this in the context of the existing literature on contract cheating and academic misconduct generally, to clarify the aims of the study: 1. To gather the views of staff on what could/should be done about the use of ghostwriters and online essay services by students in higher education 2. To raise awareness of the use of ghostwriters by students in higher education We then used a review of the literature, as summarised in the introduction, to identify specific survey items that could be used to address these questions. We refined these through frequent discussion until we arrived at an instrument to take forward to a pilot stage.

Survey structure
The survey gave respondents information, upon commencement, about the purpose of the survey, that anonymity was guaranteed and that they were free to exit the survey at any point if they did not want their responses captured in the final data collection and any research outputs. There then followed a series of open-ended and closed questions, with each question providing information to respondents about what was required of them. The questions asked participants their views on various aspects of contract cheating, the reasons why it happens and strategies to address it. The majority of the closed questions were mandatory, with the open-ended text fields being voluntary. The tool was created with two sets of questions for Australian and UK participants to provide prices in the appropriate currencies well as utilising different language around the terms 'grading', 'GPA', '2.1, 2.2 etc'. The list of job titles and disciplines was tailored to be meaningful to both countries' respondents.

Survey piloting and release
Eight university staff (5 in Australia and 3 in the UK) piloted the survey prior to release and were asked to comment upon the functionality and clarity of the survey. Comments were returned regarding some potential rephrasing of the questions, as well as re-ordering a section to make the flow of the questionnaire more meaningful and clear for respondents. The survey was released in English through an online survey tool, Limesurvey™, and distributed indirectly through networks and individuals known to the two authors; contacts were asked to send the survey on to their university networks and associated email distribution lists. Participants were thus not recruited directly by the authors and were a convenience sample, recruited via snowballing following the initial targeted distribution. The survey was open from September 1, 2016, to December 23, 2016, to cover the final teaching period of the calendar year, in each country. The survey received ethical approval from the University of Canberra Human Research Ethics Committee (project number 16:71).

Data analysis
Quantitative data were largely derived from Likert scales and so were analysed using nonparametric methods. Many questions asked participants to rate their 'agreement' with specific statements with a standard 5-point Likert scale ('strongly disagree', 'disagree', 'neither agree nor disagree', 'agree', 'strongly agree'). In order to test whether the sample, as a whole, agreed or disagreed with a particular statement, the distribution of response was converted to a 1-5 ordinal scale, where 3 represents the midpoint, neither agreement or disagreement. One-sample Wilcoxon tests were undertaken to determine whether the distribution of response was significantly different to '3', as in previous work (Newton 2015;Newton and Miah 2017). A similar approach was undertaken to determine whether participants considered a factor to be 'important' or 'not important', again as in previous work. Specific tests are described in the relevant results section.
Qualitative data from the free-text questions were analysed using a bottom-up thematic analysis (Braun and Clarke 2006) using NVivo10™. Both authors independently read and reread through the data and identified broad common themes. These were then discussed, and the themes were refined, consolidated and organised into parent and child. Four free-text questions were asked, and these all related to prior quantitative questions, and thus, key relevant themes are presented alongside the appropriate quantitative data. In keeping with the 'bottom-up' principle, unforeseen themes emerged and these are also presented.

Participant demographics
Due to the distribution method used, we cannot calculate the response rate. Two hundred thirty participants started the survey, and 196 fully completed, giving an 85% completion rate. Seventy-eight percent were from Australia and 22% from the UK. Table 1 displays the areas which respondents worked in.
Participants were also asked BHow many years have you been working in Higher Education?^Mean response was 14.5 years (SD = 9.2, range 0-46). These data were not normally distributed; the median response was 13 years and 44 respondents had been teaching for less than 5 years.  (11) 17.3 Governance, secretariat (12) 4.6 Business, finance, HR (13) 6.6 Other 12.2

Results
How common is the use of ghostwriters by students?
At the start of the survey, we wanted to understand whether respondents were already aware of ghostwriting and how students use them in their university work. The majority of respondents agreed or strongly agreed that they understood how ghostwriting services are used by students, only 15% (combined) disagreed and strongly disagreed. Staff were then asked to estimate 'Across your institution in general, what percentage of students do you think are using these services to gain academic credit?' The mean rating was 10.3%, although responses ranged from 0 to 50%, and the standard deviation was 10.4. The data were not normally distributed and were skewed towards the lower end of the range (see Fig. 1). The median response was 5% and 55/192 participants (29%) answered 5% while 42/ 192 participants (22%) answered 10%. Three responses were excluded as they were outside the 0-100% range, one participant did not answer this question and one participant answered zero.
We also asked about individual participant experience of students using contract cheating services; 41.8% answered 'yes' that they had 'personally ever been aware that a student has been using one of these companies' while 59.7% answered yes that their 'institution ever caught a student using one of these companies'.
Participants were then asked, 'If possible, please provide some detail of the detection of this and what the outcome was'. Specific outcomes were classified into groups as shown in Table 2. The most common outcome was that students were failed for that unit/module. While most outcomes involved serious sanction of the students, a few responses indicated more lenient outcomes such as a warning or failing the assignment. One respondent specifically stated that students were marked within the normal parameters and then failed as they did not answer the question. One respondent stated that only a warning was given.
There were seven themes relating to detection in the qualitative data. A total of 44 comments were coded into these themes, some being relevant to more than one theme. The most common theme (16 comments) we called 'Mismatch', staff detected cases of contract cheating as a result of some kind of salient mismatch between, for example, a student's written Fig. 1 Participant responses to the question 'Across your institution in general, what percentage of students do you think are using these services to gain academic credit?' The percentage of respondents providing answers within each 1% range (bin) is shown work vs their performance in examinations, references not relating to the unit/module information or the work being a mismatch to the student's previous standard or language abilities. Example comments included: 'The student obtained 52% for a poorly written paper in comparison to a perfectly worded > 85% graded essay'. 'Individual academics have noticed the tone, language or referencing looks strange'. 'A student's work is in high quality but just not followed the instructions given by the unit convener/lecturer'. 'The explanation of the basic concepts or theories is very odd or not the same as what the students were taught'.
The other themes were related to specific means and characteristics of detection, including the use of a viva or other investigation (9 comments), external bodies (including contract cheating services) alerting institutions to an incident of contract cheating (5 comments), detection (or not) through text-matching software (5 comments), the difficulties with detection (3 comments) and active monitoring of cheat sites by staff (3 comments).

Factors causing students to use ghostwriters
Participants were asked 'Below is a list of factors which have been shown to contribute to students committing 'normal' plagiarism (e.g. copy and paste). Tell us how important you think they are in motivating students to use ghost writers'. Factors were derived from literature as described in the introduction, in particular Brimble (2016) and Park (2003). Each factor was ranked on a 5-point Likert-type scale from 'Not at all important' to 'Very important'. The results are shown in Fig. 2.
A one-sample Wilcoxon signed rank test showed that academics rated all the factors as important. Immediately following this question, we asked participants 'If you wish, please describe why you think these are the strongest motivators. Are there any other motivators that you feel contribute to students buying from essay mills that were not listed here?' Many responses to this question were simply to support, and add detail and experience to, the motivations identified in the quantitative section. However, some additional motivations emerged, both in this section and elsewhere. One theme, found across each set of qualitative data, articulated the idea of higher education having become transactional and 'like a business', whereby the focus for students is on paying fees for the end goal of getting a certificate and for universities, that students fees are required for survival (particularly international fees), and that reputational damage is more important than processing cases which may be considered as a negative impact on the business. For example, 'Many (not all) international students do not come to Australia to get an education, they come to get a degree'. 'Universities themselves have become transactional organisations rather than places of learning and reflection, so why wouldn't students take the same approach to assessment'. 'It is a long and tedious process to prove that someone bought their essay and there is a reluctance to alienate our 'customer base' and a desire to avoid reputational damage for the organisation which results in people getting away with stuff like this'.
Other motivators which staff talked about were students feeling they are perhaps not good enough, or not being confident, in part due to a lack of contact and engagement with overworked and underappreciated teaching staff: 'Another motivator might be feeling that your own work isn't good enough and not receiving formative feedback that allows you to improve. A succession of low marks with little support could cause a crisis of confidence'. 'I believe that lack of contact between students and those teaching and lack of personalising the subject and tasks required is important in students accessing any means to get through their units including using essay mill sites. Universities are big and not personal unless the student makes it that way -if a student isn't good at this or comfortable in this environment a means to a end (using essay mill sites) can be useful' 'I recently ran a unit… with 69 students and I was spending more than an hour providing feedback on each of their Reports, when I was being paid for about 5-10 minutes of effort. The good news is that I saw significant improvement -but this approach was not supported by the payment regime and I am unlikely to do this again…'

Estimated price of an assignment
Participants were asked to estimate how much it would cost to have an essay written in their discipline. The essay would be 3000 words long, delivered in 14 days, and be written for a final year undergraduate at the credit (AUS) or 2:1 (UK) level. Data from the UK Fig. 2 Staff views on factors contributing to the use of ghost writers by students. Each item was rated on a 5point Likert-type scale. There was significant agreement with every factor (i.e. they are all considered important) were converted to AUD to facilitate a comparison (at the time of writing survey, September 2016, $1 AUD = £0.60 GBP, source xe.com). Participants in Australia (n = 152) estimated a mean price of 273 AUD (range 10-1250, SD = 236). Participants in the UK (n = 44) estimated a mean of 1066 AUD (666.50 GBP). A Mann-Whitney test showed this difference to be significant (P = 0.001, U = 2259), with Australian respondents believing it to be far cheaper to purchase an assignment than their UK counterparts. One UK response was a considerable outlier (15,000 GBP), but the difference between countries was still significant with this response excluded (P = 0.002). Although precise figures on these costs are subject to change and dependent on the source, figures available in the literature are substantially cheaper than the estimates supplied by participants here (e.g. Clarke and Lancaster 2013;Newton and Lang 2016).
Should essay-writing companies, or their use by students, be illegal?
Participants were asked to rate their agreement, on a standard 5-point Likert scale, with the following two statements 'A student who purchases an essay that has been written by someone else and then submits it as if it was their own is committing an act that should be illegal' and 'A company that sells an essay, knowing that a student intends to submit it as if it is their own, is committing an act that should be illegal'. Results are shown in Fig. 3. A one-factor Wilcoxon signed-rank test showed that participants agreed with both statements; group ratings were significantly different to the 'indifference' midpoint (3) on the scale. However, a Wilcoxon matched pairs signed rank test showed a significant difference between the two questions, with higher agreement for the question regarding the behaviour of the company (P = 0.0375, W = 634). As the data in Fig. 3 show, the magnitude of this difference is small. Eighteen free-text comments directly related to the issue of legality of essay mills. Despite the quantitative data showing agreement with criminalising the behaviour of students who engage in contract cheating, one theme of the free-text data was to argue against this. Six participants commented on this, observing, for example, that students were young and still learning and that they should not be penalised for this in a criminal way, others stated that the focus should be placed on preventing or resolving the motivations towards this, and promoting ethics with students over criminalising the act. One person noted that Fig. 3 Should essay writing companies, or their use by students, be illegal? Participants were asked to rate their agreement with descriptions of the behaviour of a company supplying students with essays, and a student who submits such an essay for academic credit. There was overall agreement that both behaviours should be illegal, but this was stronger for the behaviour of the company (*P < 0.05 Wilcoxon matched pairs test) 'Universities are educational institutions not correctional facilities. The key constituency is aged 17-25, a time of risk taking and experiencing consequences in a safe environment, without it completely ruining their career prospects' Another theme, specifically articulated by three participants, was that criminalisation in any form would just not work, with one participant stating: 'the law is well behind modern technological affordances, and that students who wish to buy these essays will do so regardless of the law. It is not hard to imagine these kind of transactions making their way to the darknet and online networks that go beyond that which is easily enforceable'.
Of interest was the theme, articulated by four participants, that the behaviours were already illegal, which, at the time of writing, is not actually the case in either the UK or Australia .

Assessment strategies to tackle contract cheating
Participants were asked to rate their agreement with the following questions: 'Creative assessment strategies could make it much harder for students to use these services' and 'Reducing the number of written assignments and increasing the proportion of examinations (including open book) may help reduce this type of cheating'. There was overall agreement with both, as shown in Table 3. Of participants, 83.1% agreed that creative assessment strategies could make contract cheating harder, while fewer (64.7%) agreed that reducing written assignments and increasing examinations would reduce contract cheating. This pattern was also reflected in the qualitative data, with respondents volunteering numerous comments discussing their opinions of what the most suitable assessment strategies would be to help reduce contract cheating, alongside a fairly negative view of examinations (Table 4).
The 47 comments in this area were coded according to seven themes (under a broad parent category of 'Assessment and Learning') which included recommendations for better learning and assessment strategies and those who thought more exams would or would not make a difference, and whether assessments could help reduce contract cheating, see Table 4. Ten comments gave opinions on assessment strategies more generally, both positively and negatively. Suggestions of better assessment practices included work-in-progress, presentations, take-home exams, mini essays due weekly, personal reflections, in-class writing tasks, vivas to support large written assignments, and increased use of practical tasks. Many of the comments which talked about the importance of not focusing on exams (16 stated that exams were bad forms of assessment; 7 that exams would not make a difference in reducing contract cheating). One quote captured many of these themes: 'I don't feel that exams are the answer. We may remove the risk of students outsourcing the assignment, but we reduce the quality of assessment practices. We still face academic misconduct in exam environments'.
Four respondents added that education institutions need to focus more on teaching students the importance of integrity, ethics and learning, rather than placing the focus on redesigning assessment methods.

Value of the survey
At the end of the survey, respondents were asked to state whether they agreed or disagreed with the statement 'Answering their questionnaire has helped me understand or has expanded my understanding of the knowledge of the use of ghostwriting by students in Higher Education', 35% agreed with the statement and 6.6% strongly agreed. The free-text comments at the end of the survey mirrored this, with several persons volunteering comments that they felt the research was very useful, and/or had helped them become aware of the problem.

Discussion
We surveyed staff opinions on commercial contract cheating. On the whole, we found that staff rated themselves as having a good understanding of the issue, although this was not fully supported by their responses to subsequent questions around the legality of contract cheating and approximate prices. We found broad support for pursuing a legal approach to contract cheating, although this included, perhaps surprisingly, modest support for the criminalisation of students who engage in contract cheating. Stronger support was found for the use of creative assessment measures to minimise contract cheating, and many descriptions were made of flags which may facilitate detection. This is clearly a subject about which participants felt strongly; we received a large number of free-text responses, 322 over the four free-text fields.  Table 4 Numbers of themes extracted from the comments provided on whether more exams and better assessment strategies could help to reduce contract cheating Theme Number of comments Assessment design is positive to promote reduction 3 Assessment design will not work 4 Exams are bad form of assessment 16 Exams will not make a difference (in reduction of contract cheating) 7 Exams (more) would work 3 Promote integrity and student learning 4 Suggestions for assessment design 10 Participants estimate that 5-10% of students at their institution are 'using these services for academic credit'. Other studies have returned similar figures; for example, Wilkinson (2009) asked staff what proportion of students they thought were paying someone to write assignments for them; 12.8% thought that this behaviour was common. These figures are higher than the single figure percentage estimates from recent student self-report studies (see for example, Curtis and Clare 2017). How serious would this 5-10% figure be if it were accurate? This is clearly a small minority of students. However, there are currently 2.28 million students in UK higher education (Higher Education Statistics Agency 2017) and 1.25 million students in Australian higher education (Australian Government Department for Education and Training 2017). Thus, 5-10% of these 3.53 million students would mean there are currently approximately a quarter of a million students using these services in these two countries alone.
Much of the aforementioned media coverage, along with recent literature Newton and Lang 2016;QAA 2017) has included a consideration of a legal approach to tackling contract cheating. There are many actors in contract cheating, all of whom contribute, in different ways, to the submission of purchased work that is not the student's own. Two of the most obvious, and whose actions may be subject to legal challenge, are the companies who sell the work and the students who submit it. Whilst any legal debate is being seriously considered, the staff opinion on whether this would work (as those who will ultimately be a gatekeeper to legal proceedings), need to be weighed up with student opinion on whether this is or should be illegal (providing implication for whether this may be a deterrent or not) and particularly in relation to whether legal sanctions should be pursued against students, companies or both. This will be the subject for future study.
Our data showed overall support for developing laws against the actions of both groups. However, while the quantitative data supported action against students, the qualitative comments contained many caveats and qualifiers to that view. For example, some participants noted that the criminalisation of students was not the preference due to students often being at a vulnerable time of their life when they may be learning new academic conventions, as well as learning and developing their knowledge in their discipline. There were also qualifying comments regarding the difficulties of developing legal action against the companies. Any criminalisation of contract cheating must be very careful to distinguish between parties who could be considered to be found guilty and how cases would be managed. For example, any pursuance of cases by universities may require staff to become witnesses to a crime, and for students to be complicit in the act. Discussions of who the 'victim' may be would also arise. Legal jurisdictions and the actions against the companies would require specific disclaimers for work being sold across international borders.
There was stronger support for the use of creative assessment strategies to tackle contract cheating, and the qualitative data contained many suggestions for how this might be achieved. Although pedagogically, the suggestions made for assessment design should promote learning and engagement in assessment tasks, few were not 'contractable'. The sites selling work to students provide a plethora of assignments and can provide drafts and hand-written notes. Students do have to be in attendance for exams and vivas, and this may provide some strength to these assessment strategies, but many participants commented on the poor pedagogic value of examinations and the fact that there are many ways to cheat in examinations. It is essential for us all to focus on the most pedagogically strong way to test learning and allow students to demonstrate and apply their knowledge.
Many participants described the ways in which suspicions of contract cheating had been aroused and these supported those presented by Dawson and Sutherland-Smith (2017) and Lines (2016), such as the question not being directly answered, inconsistencies in language or the use of phrasing (potentially due to synonyms being inserted), lack of reference to previously used materials or lacking of relevance and specific discussion of the disciplinary focus. However, it is unlikely that these alone would be sufficient to pursue an allegation of misconduct against a student, with the attendant workload, and damage to the staff-student relationship.
Perhaps the most important to addressing contract cheating is to understand why it happens. The views of staff are important here as they will be one stakeholder group tasked with designing and implementing initiatives to address contract cheating. Their views on why contract cheating happens will influence the nature of those initiatives, as well as decisions about who they should be targeted at. Staff may have a different perspective to students and other stakeholders, and ultimately, it will require triangulation of views from multiple stakeholders to get the most accurate picture. Participants showed strong agreement with all the proposed motivations, which were derived from earlier work regarding the motivation for 'regular' plagiarism and other forms of misconduct. Of staff, 85.2% responded that studying in a non-native language was either very important, or important, in being a motivator, showing strong agreement with results from a recent self-report study on academic outsourcing by students in Australia where 'speaking a language other than English at home' was a factor associated with self-report of outsourcing generally . The qualitative data supported this, and the discussion of international students and non-native language was a running theme throughout the qualitative data.
The comments regarding second-language students were dispersed amongst conversations around having detected purchased work, the motivations leading students to do it, and final comments on the topic. Some respondents raised it in relation to students not understanding the academic requirements or referencing conventions; not being able to effectively present their own ideas in another language; the pressures which may be placed upon them by family members and future job offers; and fear of failure in doing well in something which they struggle to understand in a different language. It seems reasonable to tie these observations into the prior paragraph regarding detection; the issues described regarding second-language students seem more likely to result in the sort of 'mismatch' situation which triggers suspicion of contract cheating. It also seems reasonable that these may affect the results of attempts at contract cheating, if the purchased work is in a non-native language. In short, the aforementioned difficulties associated with studying in a second language may make studying harder, but they may also make contract cheating harder, and thus easier to detect, leading to suspicion that the incidence is higher in second-language students.
A final motivator identified in the qualitative data was the commercialisation of higher education in the two countries, which may result in students themselves taking a more transactional approach to their learning.

Limitations
There are some limitations to the methodology used here which should be considered when interpreting the findings. We had a modest sample size, although it was comparable to many other studies on academic integrity (Löfström et al. 2015;Ramzan et al. 2012) and provided rich qualitative data. We used a convenience sample, derived via snowballing from an initial distribution list, with a guarantee of anonymity for the participants. Thus, we cannot be sure that the survey has captured a representative sample of staff working in HE. This also means we have no way of knowing the number of institutions represented in the dataset. We also only surveyed in Australia and the UK. We did not anticipate that there would be differences between the two, but the price estimations were significantly different. We therefore are cautious about extrapolating our findings to other countries. The use of commercial contract cheating services is a borderless phenomenon; the student, their university, the writer, the company and the servers hosting all the activity could all be in different countries. There is currently a limited literature on the writers themselves but what we do know about them suggests that they write for clients in multiple countries (Sivasubramaniam et al. 2016), and so it seems reasonable to hypothesise that country-specific comparisons are perhaps less relevant for this issue than for others.
Finally, the novel contribution made by this study is that it is one of few to investigate staff perceptions regarding commercial contract cheating. However, this is also a limitation; it provides only one side of the story. What staff believe to be the motivations are based on their experience with students involved in cases of misconduct, and/or through their own assumptions. Full interpretation of the findings requires consideration of the views of other stakeholders, in particular students. This is the subject of ongoing research.

Conclusion
This survey was carried out to explore staff opinion into commercial contract cheating, to help fill in a gap in existing literature on the topic, and to allow a comparison of staff perceptions with student self-reporting. Staff have significant responsibility for designing assessments, marking and consequently are asked to detect outsourced work, as well as those who can influence institutional and sector policy responses to the problem; we wanted to determine whether there seemed to be disparities in the staff opinion, to existing literature on student opinion.
There was a running theme in the qualitative data relating to international students and students studying in a second language, and this reflects findings from similar literature where the finding was specifically that it is studying in a second language, rather than international student status, that was associated with a higher likelihood of self-reported misconduct . Clearly it is harder to study in a second language, and HE providers need to ensure that appropriate support is offered to second-language students to enable them to reach their full academic potential. Staff responses were varied in relation to pedagogical directions; therefore, it seems necessary to continue to place an emphasis on assessment design and on the detection of contract cheating through various methods, including moving away from purely textual-based assignments, increasing promotion of practical, in-person assessment, including the presentation of written work, enhancing and promoting learning and teaching activities that explicate the importance of integrity and ethics in students' lives, alongside staff being trained more effectively in how to identify contract cheating.
Consideration of a legal approach will need to carefully consider the knock-on effects of the criminalisation of contract cheating. It is recommended that the initial focus should be placed on criminalising the acts only of the companies selling the work to students.
Finally, staff raised the possibility that higher education itself has become transactional and therefore a natural progression of this was that they are seen as businesses within which students may feel they can buy their way. If institutions really have become more transactional in nature, either the methods of delivering education to students and expectations placed on the student body need to be redressed, or the transactional nature minimised so that students do not feel that they are simply a number in a system and consequently respond by disengaging and completing their studies through transactions of their own. However, we are unaware of literature which asks students for their opinions on a possible commercialisation of higher education, specifically in relation to subsequent commercial contract cheating behaviours. A future study of student perceptions of contract cheating which also explores this possible transactional nature and business model of universities, and their purpose for learning, as well as a larger staff survey based on this study, would be a useful comparison, particularly if such studies can achieve high response rates from representative samples. It will continue to be imperative that the issue of commercial contract cheating is explored from a variety of angles to allow as full a picture as possible to be painted by the staff, students, institutions, companies and writers. Without this, any sector or local responses are likely to be ineffective.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.