View in EDS HTML Full Text PDF Full Text

Moving beyond Self-Reports to Estimate the Prevalence of Commercial Contract Cheating: An Australian Study

Saved in:
Bibliographic Details
Title: Moving beyond Self-Reports to Estimate the Prevalence of Commercial Contract Cheating: An Australian Study
Language: English
Authors: Curtis, Guy J. (ORCID 0000-0002-4174-6955), McNeill, Margot, Slade, Christine (ORCID 0000-0002-2197-2824), Tremayne, Kell, Harper, Rowena (ORCID 0000-0002-5330-525X), Rundle, Kiata (ORCID 0000-0002-4207-9760), Greenaway, Ruth (ORCID 0000-0002-8707-1797)
Source: Studies in Higher Education. 2022 47(9):1844-1856.
Availability: Routledge. Available from: Taylor & Francis, Ltd. 530 Walnut Street Suite 850, Philadelphia, PA 19106. Tel: 800-354-1420; Tel: 215-625-8900; Fax: 215-207-0050; Web site: http://www.tandf.co.uk/journals
Peer Reviewed: Y
Page Count: 13
Publication Date: 2022
Document Type: Journal Articles
Reports - Research
Education Level: Higher Education
Postsecondary Education
Descriptors: Measurement Techniques, Incidence, Cheating, Contracts, Foreign Countries, College Students, Outsourcing, Writing Assignments, Predictor Variables, Incentives, Ethics
Geographic Terms: Australia
DOI: 10.1080/03075079.2021.1972093
ISSN: 0307-5079
1470-174X
Abstract: The highest estimates of the prevalence of commercial contract cheating in Australia come from self-report surveys, which suggest that around 2% of students engage in commercial contract cheating during their higher education studies. However, self-report surveys are limited in that participants under-report socially-undesirable behaviours. In this study, we used an incentivised truth-telling method and surveyed 4098 students from six universities and six independent higher education providers in Australia. We found that 2.46 times more students admitted to commercial contract cheating, via submitting ghost-written assessments, when truth-telling was incentivised (via a Bayesian Truth Serum methodology) rather than when normal self-report survey instructions were used. Using prevalence estimation formulae that are combined with the incentivised truth-telling method, we estimate that 7.9% of students buy and submit assignments from commercial contract cheating services. Additionally, 11.4% outsource assessments via obtaining pre-written work from commercial file-sharing sites. These are substantially higher percentages of commercial contract cheating than self-reports suggest. Furthermore, having a first language other than English was the strongest demographic predictor of Australian students' engagement in commercial contract cheating. We conclude that commercial contract cheating is a more common problem than suggested by self-report surveys. We argue that academic integrity researchers should consider methods beyond standard self-reports to estimate the prevalence of academic misconduct and that efforts to curb commercial contract cheating must be increased.
Abstractor: As Provided
Entry Date: 2023
Accession Number: EJ1367888
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwFTb1Daa2xAF-T1h0J9-mIgAAAA4jCB3wYJKoZIhvcNAQcGoIHRMIHOAgEAMIHIBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDJzrs3dFsGj6L5SFoQIBEICBmupL6ImyM36T6Ty4UWy-teHWOjhy35N7ZskQaIx89HvsIDNR9oy8W0Lq9MQXH3y7umB6K0oIw579aga-q9daVn0lNrC9lWbuImwmf9iYZwj7rgj1oMWTsc02zWyeVfcf0GwDykcwQSbZyZZhCAxTq56KnZWVuGpJAe6XSDKfvyH8HDQATz9mfENxDhQGOXYtmF9XgPioT7WY-1s=
Text:
  Availability: 1
  Value: <anid>AN0158963237;she01sep.22;2022Sep09.05:00;v2.2.500</anid> <title id="AN0158963237-1">Moving beyond self-reports to estimate the prevalence of commercial contract cheating: an Australian study </title> <p>The highest estimates of the prevalence of commercial contract cheating in Australia come from self-report surveys, which suggest that around 2% of students engage in commercial contract cheating during their higher education studies. However, self-report surveys are limited in that participants under-report socially-undesirable behaviours. In this study, we used an incentivised truth-telling method and surveyed 4098 students from six universities and six independent higher education providers in Australia. We found that 2.46 times more students admitted to commercial contract cheating, via submitting ghost-written assessments, when truth-telling was incentivised (via a Bayesian Truth Serum methodology) rather than when normal self-report survey instructions were used. Using prevalence estimation formulae that are combined with the incentivised truth-telling method, we estimate that 7.9% of students buy and submit assignments from commercial contract cheating services. Additionally, 11.4% outsource assessments via obtaining pre-written work from commercial file-sharing sites. These are substantially higher percentages of commercial contract cheating than self-reports suggest. Furthermore, having a first language other than English was the strongest demographic predictor of Australian students' engagement in commercial contract cheating. We conclude that commercial contract cheating is a more common problem than suggested by self-report surveys. We argue that academic integrity researchers should consider methods beyond standard self-reports to estimate the prevalence of academic misconduct and that efforts to curb commercial contract cheating must be increased.</p> <p>Keywords: Contract cheating; academic integrity; file-sharing; prevalence; Bayesian-truth-serum; methodology</p> <p>Commercial contract cheating occurs when students buy outsourced assessment work and submit it as their own (Clarke and Lancaster [<reflink idref="bib6" id="ref1">6</reflink>]; Newton [<reflink idref="bib22" id="ref2">22</reflink>]). This can include hiring exam impersonators as well as outsourcing written work. Contract cheating poses a serious threat to the integrity of educational assessment and scholarly research (Bretag [<reflink idref="bib4" id="ref3">4</reflink>]). Contract cheating is notoriously difficult to detect in non-proctored written work, because bespoke ghost-written assignments can evade detection by common techniques for monitoring academic integrity, such as text-matching software. Academics, administrators, and higher education leaders need to know the prevalence of contract cheating to determine if their detection rates are consistent with the extent of this practice among students. In this paper we note that, to date, the highest current estimates of commercial contract cheating prevalence in Australia come from self-report surveys, which typically underestimate the prevalence of socially-undesirable behaviours (Tourangeau and Yan [<reflink idref="bib31" id="ref4">31</reflink>]). We then present the results of a large-scale survey of commercial contract cheating, comparing a typical self-report methodology with an incentivised truth-telling methodology.</p> <hd id="AN0158963237-2">Literature review</hd> <p>Various estimates have been made, or can be made, of the prevalence of commercial contract cheating. Importantly, the estimated prevalence appears to be dependent on the measurement used. Objective measures, such as cases detected by university staff, typically produce vanishingly small percentages, often under 1% of students. On the other hand, self-reports routinely produce estimates many times higher than the detected incidence, at around 2%−3.5% of students who have ever engaged in contract cheating (e.g. Curtis and Clare [<reflink idref="bib7" id="ref5">7</reflink>]; Bretag et al. [<reflink idref="bib5" id="ref6">5</reflink>]). This difference between objective detection and self-report methods suggests that only a fraction of commercial contract cheating is detected. In Australia, various public scandals and subsequent survey research (e.g. Bretag et al. [<reflink idref="bib5" id="ref7">5</reflink>]) have focused the minds of university leaders and government on the problem of commercial contract cheating.</p> <p>In 2014, the Australian media reported on the activities of the <emph>MyMaster</emph> website, which offered custom-written assignments for sale and online test imposters to Chinese-speaking international students (McNeilage and Visentin [<reflink idref="bib21" id="ref8">21</reflink>]). Journalists discovered 700 receipts for payment to <emph>MyMaster</emph> from students in a range of university courses for commissioned assessment work. This scandal prompted the question, 'How widespread is such cheating?'</p> <p>Drilling down into the figures from the <emph>MyMaster</emph> scandal, the five most effected universities had 450 identified requests for assessments in 10 months of 2014 (McNeilage and Visentin [<reflink idref="bib21" id="ref9">21</reflink>]). These universities had a total enrolment of over 265,000 students, and thus if each <emph>MyMaster</emph> request was from a unique student, the prevalence of contract cheating would be just 0.17% of students. However, Curtis and Clare ([<reflink idref="bib7" id="ref10">7</reflink>]) have shown that students who engage in contract cheating tend to do so more than once, and, indeed, one <emph>MyMaster</emph> customer accounted for five of the 450 requests (McNeilage and Visentin [<reflink idref="bib21" id="ref11">21</reflink>]). This being the case, it is possible that the MyMaster orders were from less than one-tenth of one percent of students even at the most effected universities.</p> <p>More recently, enhanced detection methods enabled academic integrity investigators at the University of New South Wales (UNSW) to substantiate an increase in the number of cases of contract cheating by nearly 500% in one year (Fellner [<reflink idref="bib11" id="ref12">11</reflink>]). Still, their figure of 168 substantiated cases of contract cheating in 2019 represents just 0.27% of UNSW students, or about one-tenth of the rate at which students admit to contract cheating in self-report surveys.</p> <p>As noted, self-report methods suggest that the real prevalence of contract cheating may be substantially more than has been detected in cases such as <emph>MyMaster</emph> and UNSW. Curtis and Clare ([<reflink idref="bib7" id="ref13">7</reflink>]), for example, found that 2.1% of students reported ever engaging in commercial contract cheating in the four Australian anonymous self-report surveys included in their small-scale meta-analysis. In a subsequent landmark study by Bretag et al. ([<reflink idref="bib5" id="ref14">5</reflink>]), 2.2% of over 14,500 students in Australia reported that they had obtained an assignment written by someone else (not necessarily a paid ghostwriter) to submit as their own. Another sizeable, and predominantly Australian, survey also found a 2% admission rate of engagement in commercial contract cheating (Rundle, Curtis, and Clare [<reflink idref="bib26" id="ref15">26</reflink>]). These estimates of the prevalence of contract cheating from self-report surveys are notably lower than those from more diverse international samples (e.g. Newton [<reflink idref="bib22" id="ref16">22</reflink>]). Nonetheless, a figure or around 2% is in the range of 10–20 times higher than the percentage of cases suggested by the <emph>MyMaster</emph> scandal. Nonetheless, it is almost certain that the prevalence estimates of commercial contract cheating that come from self-report surveys are also lower than the true prevalence among students (Tourangeau and Yan [<reflink idref="bib31" id="ref17">31</reflink>]).</p> <p>Self-report surveys are widely used in educational, marketing, and psychological research. There are, however, several recognised limitations of self-report surveys (Siev and Kliger [<reflink idref="bib27" id="ref18">27</reflink>]). The most important limitation, in the context of estimating the prevalence of commercial contract cheating, is the tendency for survey respondents to under-report socially-undesirable behaviour (Sudman and Bradburn [<reflink idref="bib28" id="ref19">28</reflink>]). Commercial contract cheating is widely understood as both 'serious' and 'wrong' among higher education students (Bretag et al. [<reflink idref="bib5" id="ref20">5</reflink>]; Curtis and Tremayne [<reflink idref="bib9" id="ref21">9</reflink>]), and therefore it is clearly understood as being socially-undesirable by students. Moreover, researchers have argued for some time that self-reports underestimate the prevalence of various forms of academic integrity breaches (Whitley, Nelson, and Jones [<reflink idref="bib33" id="ref22">33</reflink>]; McCabe [<reflink idref="bib20" id="ref23">20</reflink>]).</p> <p>There are several reasons why survey participants may not disclose socially-undesirable behaviours, even in anonymous surveys (Krumpal [<reflink idref="bib17" id="ref24">17</reflink>]). These reasons include: questions being perceived as intrusive by survey takers; concerns that their answers will not remain anonymous; the desire to maintain a favourable self-concept by not admitting to bad behaviour; and the lack of an incentive to be truthful (Prelec [<reflink idref="bib23" id="ref25">23</reflink>]; Tourangeau and Yan [<reflink idref="bib31" id="ref26">31</reflink>]; Ariely [<reflink idref="bib1" id="ref27">1</reflink>]). Various methods have been proposed to overcome people's reticence to answer sensitive questions truthfully in anonymous surveys. In the context of academic integrity, Jann, Jerke, and Krumpal ([<reflink idref="bib13" id="ref28">13</reflink>]), for example, estimated higher rates of plagiarism among students when using a 'crosswise' survey methodology than when using a standard anonymous self-report survey. Unfortunately, the crosswise methodology is a form of 'random response task', which tends to provide survey respondents with difficult instructions that are often poorly understood or poorly followed (Boeije and Lensvelt-Mulders [<reflink idref="bib3" id="ref29">3</reflink>]). Furthermore, such methods fail to provide an incentive for survey respondents to tell the truth.</p> <p>An alternative method for more accurately estimating the prevalence of socially-undesirable behaviours provides participants with an incentive for responding truthfully while maintaining anonymity (John, Loewenstein, and Prelec [<reflink idref="bib14" id="ref30">14</reflink>]). The incentivised truth-telling methodology applies Prelec's ([<reflink idref="bib23" id="ref31">23</reflink>]) Bayesian-Truth-Serum (BTS) formulae for estimating the truthfulness of answers in surveys where the true prevalence rate of the behaviour in question is unknown. For example, participants in John, Loewenstein, and Prelec's ([<reflink idref="bib14" id="ref32">14</reflink>]) study were randomly assigned either to a control (survey-as-usual) or BTS (incentivised truth-telling) condition. All participants were told that a charitable donation would be made for each participant who completed the survey, and they chose their preferred charity from a list provided by the researchers. In the incentivised-truth-telling condition, before answering items about questionable research practices, participants were told that the BTS would be used to estimate the truthfulness of their answers and that more truthful answers would increase the size of the donation to their preferred charity. BTS participants were provided with a link to Prelec's ([<reflink idref="bib23" id="ref33">23</reflink>]) paper and could not continue the survey until they indicated that they understood that more truthful answers would increase donations to their preferred charity. Apart from participants being unaware that they had been assigned at random to the BTS or control conditions, no other deception occurred, thus, the BTS formulae was applied to survey responses and used to determine charitable donations. Using this method, John, Loewenstein, and Prelec ([<reflink idref="bib14" id="ref34">14</reflink>]) found that nearly three times more research psychologists in the BTS condition admitted to data falsification than did research psychologists in a control condition.</p> <p>In addition to the comparison of admission rates between the BTS and control conditions, John, Loewenstein, and Prelec ([<reflink idref="bib14" id="ref35">14</reflink>]) used additional data derived from participants' estimates of peers' behaviour to estimate the prevalence of questionable research practices. In addition to self-admitted behaviours such as data falsification, they asked participants to estimate the percentage of peers who engaged in the behaviour (peer prevalence estimate) and the percentage who had engaged in the behaviour who would admit to doing so (estimated admission). This allowed for the calculation of four prevalence estimates; (<reflink idref="bib1" id="ref36">1</reflink>) self-admission, (<reflink idref="bib2" id="ref37">2</reflink>) peer prevalence estimate, (<reflink idref="bib3" id="ref38">3</reflink>) the ratio of self-admission to estimated admission, and (<reflink idref="bib4" id="ref39">4</reflink>) the mean of the first three estimates. Notably, the impersonal nature of peer prevalence estimates and estimated admission rates means that these should be unaffected by the reasons why students would normally avoid self-admission of cheating, e.g. concerns about anonymity.</p> <p>John, Loewenstein, and Prelec ([<reflink idref="bib14" id="ref40">14</reflink>]), as well as numerous papers by Prelec and his colleagues (e.g. Prelec [<reflink idref="bib23" id="ref41">23</reflink>]; Weaver and Prelec [<reflink idref="bib32" id="ref42">32</reflink>]), discuss the rationale for the BTS methodology in estimating the prevalence of otherwise-unknown rates of hidden behaviours, and in using the four prevalence estimating methods noted above. An in-depth discussion of these methods is beyond the scope of the current paper, and we direct interested readers to John, Loewenstein, and Prelec ([<reflink idref="bib14" id="ref43">14</reflink>]) for a fuller discussion of these methods. More importantly, for the purposes of this paper, given the success of the BTS methodology in eliciting more admission and thorough-going estimates of socially-undesirable behaviours in the context of questionable research practices, we decided to apply this method to estimate the prevalence of commercial contract cheating among higher education students in Australia.</p> <hd id="AN0158963237-3">The current study</hd> <p>It is important to provide some historical context for our study. The current study was conducted between September and December 2020. It was originally planned for April-June 2020, but was delayed by the COVID-19 pandemic. As a result of the pandemic, a higher percentage of students than is typical in Australian higher education institutions were predominantly studying online; 44% among our survey respondents. Data has since emerged suggesting that the COVID-19 pandemic has corresponded with increased contract cheating (e.g. Lancaster and Cotarlan [<reflink idref="bib18" id="ref44">18</reflink>]), which is possibly attributable to causes such as rapid changes to online teaching, assessment practices (Reedy et al. [<reflink idref="bib24" id="ref45">24</reflink>]), and student distress (Eaton and Turner [<reflink idref="bib10" id="ref46">10</reflink>]).</p> <p>Contract cheating had become an issue of public concern in Australia because of the <emph>MyMaster</emph> scandal, and the large-scale research of Bretag et al. ([<reflink idref="bib5" id="ref47">5</reflink>]) and Harper et al. ([<reflink idref="bib12" id="ref48">12</reflink>]). In 2019-2020, the Tertiary Education Quality Standards Agency (TEQSA), commissioned nation-wide academic integrity training for higher education leaders, the creation of an academic integrity toolkit, and enacted legislation allowing commercial contract cheating providers to be geo-blocked and/or prosecuted (Curtis et al. [<reflink idref="bib8" id="ref49">8</reflink>]). These actions were in addition to earlier work to inform the higher education sector about contract cheating, such as the development of the <emph>Good Practice Note: Addressing contract cheating to safeguard academic integrity</emph> (TEQSA [<reflink idref="bib29" id="ref50">29</reflink>]). Thus, over the preceding years, and at the time of the study, there had been increasing efforts, both at institutional and government levels in Australia to increase awareness of, and counteract, commercial contract cheating.</p> <p>As stated, the aim of our study was to apply the BTS methodology employed by John, Loewenstein, and Prelec ([<reflink idref="bib14" id="ref51">14</reflink>]) to study the prevalence of commercial contract cheating in Australia. The rationale, methods, and measures for the study were pre-registered with the Open Science Framework (see: https://osf.io/jx5rm/). In the survey, students were asked about two forms of commercial contract cheating: submitting custom ghost-written work and submitting pre-written work obtained from file-sharing sites. Students were randomly assigned to either receive, or not receive, BTS-based truth-telling incentive instructions.</p> <p>Our pre-registered hypotheses were as follows: H1. Students in the BTS condition will report significantly more engagement in contract cheating than students in the control condition. H2. Estimates of the prevalence of engagement in contract cheating using BTS methodology will exceed those reported in existing large-scale surveys, reviews, and meta-analyses.</p> <hd id="AN0158963237-4">Materials and methods</hd> <p></p> <hd id="AN0158963237-5">Participants and sampling procedure</hd> <p>In the pre-registered protocols of the study, we aimed to recruit a minimum of 2000 and a maximum of 12,000 students. Students were recruited from six universities and six independent higher education providers in Australia. Human Research Ethics approval was obtained from The University of Western Australia (RA/4/20/5939) and reciprocal ethics approval was obtained, where required, from the participating institutions. High-level management support was obtained at each institution to distribute the survey to students. Distribution of the online survey varied depending on the policies and permissions at each institution regarding communication with students, surveying of students, and research recruitment. The survey was sent to students at some institutions by email, at others via learning management system announcements, e-newsletters, or other electronic forms of communication that allowed them to click a link to the online survey. In most cases, invitations to participate in the survey were sent at times where the institutions were not running other large-scale surveys and, where possible, reminder invitation notices or emails were sent two to three weeks after the initial invitation. Students were offered an entry to a $AU500 gift-voucher prize draw at each institution for completing the survey. A separate survey was created for each institution so that it could be branded for the institution, and a separate survey for collection of student email addresses for the prize draw was also created for each institution. The separate prize draw link meant that students' survey answers and identifying contact details could not be connected, and students were informed of this protection of their anonymity.</p> <p>There were initially 6780 logins to the survey in total across the 12 institutions. However, at one institution, 1361 logins were identified as 'spam' and were deleted, as were 16 spam logins from among the other institutions. Two-hundred-and-four logins were from people not identifying as a student. A further 1101 students who commenced the survey did not answer one or more of the questions concerning contract cheating, and were omitted. This left a total of 4098 complete responses.</p> <p>The survey randomly assigned students to the Bayesian Truth Serum incentivised truth-telling (BTS) condition or the survey as normal (control) condition. Of the completed surveys, 2035 were in the BTS condition and 2068 were in the control condition.</p> <p>For the overall sample of students, the mean age was 27.54 years (median = 25, mode = 20, <emph>SD</emph> = 10.16). More students were undergraduates (<emph>n</emph> = 3096) than postgraduate (<emph>n</emph> = 1002). There were 1473 male students, 2612 females, 9 reported another gender, and 4 did not respond to the gender item. Most students were studying at universities (<emph>n</emph> = 3910) with 186 students from independent higher education institutions. Students were predominantly domestic (<emph>n</emph> = 3284) rather than international (<emph>n</emph> = 812, 2 not reported). Similarly, most students spoke English as a first language (<emph>n</emph> = 3098) rather than as a second or subsequent language (<emph>n</emph> = 999, 1 not reported). Students came from a range of academic disciplines with the top five most common being: health sciences (21.4%); business/commerce (13.5%); psychology (10.1%); education (9.4%); and arts/humanities (8.1%). Independent-sample <emph>t</emph>-tests showed that the control and BTS groups did not differ significantly in their demographics.</p> <hd id="AN0158963237-6">Measures and procedure</hd> <p>Once they accessed the study online, students were initially presented with information about the study and their choice to proceed with the study implied their consent. Next, they were presented with a question asking them their status as a student (undergraduate full-time; undergraduate part-time; postgraduate full-time; postgraduate part-time, or not a student); non-students were directed to the end of the survey automatically. Further demographic data (e.g. gender, major, etc) were then collected. After the demographic items, students were reminded about the compensation outlined in the study information i.e. the $500 prize draw and charitable donation for each participant. Students then selected their preferred charity from a list of seven provided.[<reflink idref="bib1" id="ref52">1</reflink>]</p> <p>After the charity selection, students were randomly assigned to the BTS or control condition. In the BTS condition, students were presented with the following instructions adapted from John, Loewenstein, and Prelec ([<reflink idref="bib14" id="ref53">14</reflink>]).</p> <hd id="AN0158963237-7">The importance of truthful answers</hd> <p> <emph>Your answers to the questions in the remainder of the survey will let us apply a mathematical formula called the Bayesian Truth Serum, which will be used to determine the size of the donation made to the charity that you selected</emph>.</p> <p> <emph>The key property of the formula is that it rewards truthful answers. This means that truthful answers about your behaviour will increase the donation made to your preferred charity. For the purpose of this survey, it is not necessary for you to understand how the formula works, although the theoretical paper from</emph> Science, <emph>which includes a short abstract, is available here:</emph>https://science.sciencemag.org/content/306/5695/462.abstract.</p> <p>As a check of their understanding, students assigned to the BTS condition were next presented with a statement: 'Giving truthful responses on this survey ________________ the amount of money donated to charity on my behalf.' Students could not proceed in the survey until they correctly selected 'increases' to complete this statement. Students in the control condition were not presented with the BTS instructions and understanding check.</p> <p>In the next stage of the survey, we examined students' estimates of commercial contract cheating prevalence and admission to engagement in contract cheating (see Supplementary Materials for the full survey). All students were first presented with a definition of 'academic cheating services' (TEQSA [<reflink idref="bib30" id="ref54">30</reflink>]). They were then presented with a scenario describing a student contracting a ghost-writer and submitting the ghost-written assessment. Students were then asked for two estimates on a scale of 0% to 100%: (<reflink idref="bib1" id="ref55">1</reflink>) The percentage of students who have purchased an assignment and submitted it as their own (<emph>peer prevalence estimate</emph>), and (<reflink idref="bib2" id="ref56">2</reflink>) The percentage of students who did this who would admit to it (<emph>admission estimate</emph>). Next, students were asked if they had ever purchased and submitted an assessment (<emph>admission</emph>) and, if they answered 'yes', they were asked whether they had purchased and submitted an assessment in the last 12 months. Students were then presented with a second scenario describing a student who had traded course material on a file-sharing site, and submitted an assignment (lightly edited) that they had downloaded from the file-sharing site. Students were then asked for a peer prevalence estimate, admission estimate, and if they had ever engaged in this behaviour for the file-sharing scenario. Thus, we obtained peer prevalence estimates, admission estimates, admission (ever and last 12 months), for two forms of commercial contract cheating i.e. submission of custom ghost-writing and submission of an assessment obtained via file-sharing.</p> <p>At the end of the survey, students were automatically redirected to a second anonymous link that allowed them to enter the gift-voucher prize draw.</p> <hd id="AN0158963237-8">Results</hd> <p>There were few statistical assumptions for our analyses because mostly non-parametric statistics were used; all statistical assumptions were satisfied in our analyses. We followed the analysis procedures of John, Loewenstein, and Prelec ([<reflink idref="bib14" id="ref57">14</reflink>]) to examine the impact of the BTS instructions (binary logistic regression), and to estimate the prevalence of contract cheating from students' self-admissions, peer estimates, and admission estimates (from participants in the BTS condition only). In addition, we analysed the potential impact of demographic variables on self-admission of contract cheating using Chi-squared Hierarchical Automatic Interaction Detection (CHAID). CHAID was used because we made no prior predictions of the impact of these variables. CHAID analysis is appropriate for large samples where researchers are interested in post-hoc analysis to create a description of predictors of category membership (e.g. admitted cheating vs no admitted cheating) (Kass [<reflink idref="bib16" id="ref58">16</reflink>]).</p> <hd id="AN0158963237-9">Impact of incentivised truth-telling</hd> <p>To examine the impact of the BTS condition on self-admission of contract cheating we compared the rates of self-reported contract cheating in the BTS and control conditions. These results are reported in Table 1.</p> <p>Table 1. Mean self-admission rates comparing BTS and control groups.</p> <p> <ephtml> <table><thead valign="bottom"><tr><td /><td>BTS %</td><td>Control %</td><td>Odds ratio</td><td><italic>p</italic></td></tr></thead><tbody><tr><td>Custom ghost writing (CGW)</td><td char=".">1.8</td><td char=".">0.7</td><td char=".">2.46</td><td char=".">.004*</td></tr><tr><td>CGW (last 12 months)</td><td char=".">0.7</td><td char=".">0.2</td><td char=".">1.68</td><td char=".">.072</td></tr><tr><td>File-sharing</td><td char=".">3.1</td><td char=".">3.1</td><td char=".">1.01</td><td char=".">.937</td></tr><tr><td>File-sharing (last 12 months)</td><td char=".">1.7</td><td char=".">1.1</td><td char=".">2.16</td><td char=".">.034*</td></tr><tr><td>Either or both</td><td char=".">4.2</td><td char=".">3.5</td><td char=".">1.20</td><td char=".">.253</td></tr><tr><td>Either or both (last 12 months)</td><td char=".">2.0</td><td char=".">1.3</td><td char=".">1.61</td><td char=".">.008*</td></tr></tbody></table> </ephtml> </p> <p>The odds ratios in Table 1 indicate how much more likely a student in the BTS condition was to admit to the form of contract cheating listed than students in the control condition. For custom ghost-writing, file-sharing, and 'either or both', the odds ratios were all >1, and for three comparisons they were statistically significant (custom ghost-writing, file-sharing in the last 12 months, and either or both in the last 12 months). There results partially support the hypothesis (H1) that self-reported contract cheating rates would be higher in the BTS condition.</p> <hd id="AN0158963237-10">Prevalence estimation</hd> <p>Although truth-telling was incentivised, there are reasons to suspect that the rate of commercial contract cheating is higher than the admission rates among students, even those in the BTS condition. As noted earlier, perceived intrusiveness, distrust of anonymity, and self-concept maintenance may all reduce disclosure of sensitive information with or without an incentive to be truthful (Tourangeau and Yan [<reflink idref="bib31" id="ref59">31</reflink>]). Thus, we followed John, Loewenstein, and Prelec's ([<reflink idref="bib14" id="ref60">14</reflink>]) method of deriving prevalence from students' admission rates, peer prevalence estimates, admission divided by admission estimates, and the geometric mean of these, using only data from the BTS condition. As the data are not reported elsewhere in this paper, we should note that students in the BTS condition estimated on average that 23.46% of students engage custom ghost-writing and 14.93% would admit to doing so. Furthermore, they estimated on average that 30.12% of students engage in file-sharing and 14.93% would admit to doing so. The prevalence estimates for custom ghost-writing and file-sharing are shown in Figure 1. Note that we did not ask students for peer prevalence or admission estimates for engaging in or admitting to 'either or both' forms of contract cheating.</p> <p>Graph: Figure 1. Results for the BTS condition. The graph shows the admission rate, the ratio of self-admission rate to admission estimate, peer prevalence estimate, and geometric mean of these three percentages (numbers above the bars).</p> <p>The geometric mean (above the bars in Figure 1), as compared with an arithmetic mean, is less sensitive to extreme scores. Thus, the geometric mean provides a circumspect overall estimate while taking information from the other three estimates weighted equally. For this reason, we contend that the geometric mean provides the best estimate of the prevalence of custom ghost-writing and file-sharing among students in the survey. Strikingly, as hypothesized (H2), the estimated mean prevalence of custom ghost-writing and file-sharing among students in the survey are in the order of three to four times higher than estimates derived purely from the incentivised admission rates, and admission rates in other self-report surveys and review papers (e.g. Curtis and Clare [<reflink idref="bib7" id="ref61">7</reflink>]; Bretag et al. [<reflink idref="bib5" id="ref62">5</reflink>]).</p> <hd id="AN0158963237-11">CHAID and demographic analyses</hd> <p>For the overall sample, regardless of BTS or control condition status, 1.2% and 3.1% of students admitted to ever submitting work obtained from custom ghost-writing and file-sharing sources respectively. We entered all demographic variables, except discipline of study, into two CHAID analyses predicting admission to custom ghost-writing and file-sharing as criterion variables (see Supplementary Materials for CHAID models).</p> <p>For admission to submitting custom ghost-written assessments, there was only one significant predictor, having English as a first language (<emph>χ<sups>2</sups></emph> (<reflink idref="bib1" id="ref63">1</reflink>, _I_N_i_ = 4098) = 19.82, <emph>p</emph> <.001). Only 0.8% of students with English as a first language admitted to submitting custom ghost-written work as compared with 2.6% of students with English as a second or subsequent language.</p> <p>For admission to submitting work obtained via file-sharing, having English as a first language was a first-level predictor (<emph>χ<sups>2</sups></emph> (<reflink idref="bib1" id="ref64">1</reflink>, _I_N_i_ = 4098) = 47.05, <emph>p</emph> <.001). Again, the rates were lower for students with English as a first language (2.1%) than for students with English as a second or subsequent language (6.4%). For students with English as a first language, submitting work obtained via file-sharing was more frequent among younger students (aged 23 or less; 2.9%) than older students (aged 24 or more; 0.9%; <emph>χ<sups>2</sups></emph> [<reflink idref="bib1" id="ref65">1</reflink>, _I_N_i_ = 3099] = 15.62, <emph>p</emph> <.001). Finally, among younger students with English as a first language submitting work obtained via file-sharing was more frequent among international (8.2%) than domestic students (2.7%; <emph>χ<sups>2</sups></emph> [<reflink idref="bib1" id="ref66">1</reflink>, _I_N_i_ = 1768] = 7.43, <emph>p</emph> =.019).</p> <hd id="AN0158963237-12">Discussion</hd> <p>The results of this study demonstrate the impact of incentivising truth-telling in a survey of serious academic integrity breaches, in the form of commercial contract cheating. For self-admission questions, students in the BTS condition admitted to more contract cheating, significantly so for half of these. Remarkably, student admission to ever submitting custom ghost-written work was nearly two-and-a-half times more in the BTS than control condition. Additionally, the geometric mean estimate of the prevalence of submitting custom-written assessments was over 10 times higher than the self-reported admission rate in the control condition. In fact, the geometric mean estimate of the prevalence of submitting custom-written assessments and assessment obtained via file sharing were in the order of three to four times higher than most estimates of the prevalence of contract cheating in Australia that only come from self-report data (e.g. Curtis and Clare [<reflink idref="bib7" id="ref67">7</reflink>]; Bretag et al. [<reflink idref="bib5" id="ref68">5</reflink>]).</p> <p>According to other work using the BTS to estimate the prevalence of sensitive behaviour (e.g. John, Loewenstein, and Prelec [<reflink idref="bib14" id="ref69">14</reflink>]), a higher admission rate to a sensitive behaviour under BTS instructions is, <emph>prima facie</emph>, a more accurate figure than standard self-reports. Thus, the higher rates of self-admitted contract cheating in our study when truth-telling was incentivised could be taken as a more accurate estimate of the prevalence of this behaviour than the rate reported in the control group. In addition to this, the impersonal estimates of peer admission and how likely students who engage in contract cheating are to admit to doing so, add helpful evidence to allow for a more thorough estimate of prevalence.</p> <p>It is reasonable to conclude that people know themselves better than they know their peers, thus self-reports have a particular utility in educational and psychological research. Still, as we have pointed out, there are various reasons why people may under-report contract cheating in surveys – only one of which is overcome by the BTS instructions – the lack of an incentive to tell the truth. Because peer estimates are impersonal, they do not suffer from the problems of self-reports that reduce the disclosure of sensitive information. However, peer estimates may also be biased by the false consensus effect (people over-estimate how similar other people are to themselves, Ross, Greene, and House [<reflink idref="bib25" id="ref70">25</reflink>]).[<reflink idref="bib2" id="ref71">2</reflink>] In addition, the government and institutional emphasis on combating contract cheating in Australia may have led some students to over-estimate the prevalence of this behaviour. Still, peer estimates give students the opportunity to report on what they see among other students without implicating themselves in undesirable behaviour. Furthermore, students may also have a good level of awareness of the behaviour of their classmates, for example, some of the highest peer estimates of contract cheating in our study were from students studying in disciplines identified as having the highest rates of contract cheating in Bretag et al.'s ([<reflink idref="bib5" id="ref72">5</reflink>]) study (e.g. engineering, IT, and business; see Supplementary Materials). Harper et al. ([<reflink idref="bib12" id="ref73">12</reflink>]) reported peer estimates of contract cheating in ten-percent increments (e.g. 0-10%, 10-20%, etc.). Their study found that most students estimated contract cheating to be in the range of 0-30% of students. This accords with our sample's peer estimates for submitting custom-written assessments and assessments obtained via file-sharing.</p> <p>In Harper et al.'s ([<reflink idref="bib12" id="ref74">12</reflink>]) study, self-admission rates for contract cheating and peer estimates were examined. However, Harper et al. ([<reflink idref="bib12" id="ref75">12</reflink>]) did not ask students to estimate admission to contract cheating by those who do it. The admission estimate, which was included in our study, adds helpful new data in assessing the prevalence of this behaviour. Hypothetically, if 25% of students who engage in contract cheating <emph>will</emph> admit to it, and 2% of students in the survey <emph>do</emph> admit to it, that self-admission rate is likely 25% of the true prevalence, e.g. 8%. Thus, our study goes beyond Harper et al.'s ([<reflink idref="bib12" id="ref76">12</reflink>]) in providing two new prevalence estimates (admission/admission estimate and the mean of the estimates), while also doing so in the context where an incentive was provided for truthfulness. However, as noted earlier, the geometric mean of the estimates nevertheless provides a conservative average that is less influenced by extreme scores than the arithmetic mean. Because the geometric mean cautiously combines three other estimates, we contend that these means are the most accurate assessment of the prevalence of commercial contract cheating provided by our study.</p> <hd id="AN0158963237-13">Practical implications</hd> <p>The geometric mean estimates of students submitting custom ghost-writing and assessments obtained via file-sharing suggest contract cheating rates that are three to four times higher than reported in previous self-report Australian studies. These results are concerning because assessment integrity is critical to academic integrity, and academic integrity is critical to the whole enterprise of higher education. If students are awarded grades for academic assessments that are completed by another person, the assurance of their skills and the value of their qualifications are undermined.</p> <p>As we noted earlier in this paper, in the Australian context, there has been significant action to attenuate contract cheating in recent years. To some extent, this may partly be reflected in our findings of lower admission rates to engagement in contract cheating in the 'past 12 months' than 'ever'. Still, 'the past 12 months' versus 'ever' admissions also reflect a difference in the number of opportunities students have had to cheat as well as any difference in the recent higher education environment.</p> <p>Importantly, when our study was conducted, laws came into effect allowing TEQSA to geo-block contract cheating providers' websites. Geo-blocking is a sensible measure as it reduces opportunities to cheat, and opportunities to cheat are both noted in research as a predictor of contract cheating (Bretag et al. [<reflink idref="bib5" id="ref77">5</reflink>]), and theoretically seen as a main cause of unethical behaviour in the criminological literature (Rundle, Curtis, and Clare [<reflink idref="bib26" id="ref78">26</reflink>]). Our prevalence estimates suggest that cheating via submitting work obtained by file-sharing is more frequent than submitting custom ghost-written work. Thus, geo-blocking of file-sharing sites that operate for commercial gain seems like an important step for TEQSA to take if such sites' practices technically meet the legal standards for site-blocking set out in the new Australian legislation.</p> <p>Consistent with Bretag et al. ([<reflink idref="bib5" id="ref79">5</reflink>]), we found that students having a first language other than English was a key demographic predictor of self-admitted engagement in both forms of contract cheating in our study. Plainly, these results suggest that higher education institutions need to be mindful of the challenges facing students who are undertaking higher education in their non-native language. Institutions must ensure sufficient language competency for students who they admit to their courses and provide additional educational support for non-native speakers of the language of instruction.</p> <hd id="AN0158963237-14">Limitations and future research</hd> <p>Given that our study was conducted during the COVID-19 pandemic, and research indicates an increase in contract cheating at this time (e.g. Lancaster and Cotarlan [<reflink idref="bib18" id="ref80">18</reflink>]), it is possible that commercial contract cheating rates we observed are not directly comparable with other studies. Where we can directly compare results with past research is the rate of admission to ever having engaged in either or both forms of contract cheating in our standard (control) self-report survey. We found a prevalence rate of 3.5% (non-incentivised) self-admission to either form of commercial contract cheating, which is higher than the ∼2% in other Australian studies (e.g. Curtis and Clare [<reflink idref="bib7" id="ref81">7</reflink>]; Bretag et al. [<reflink idref="bib5" id="ref82">5</reflink>]; Rundle, Curtis, and Clare [<reflink idref="bib26" id="ref83">26</reflink>]). This difference between 3.5% and 2% may be attributable to the circumstance of the COVID-19 pandemic. However, there is still reason to think that our control condition's self-admission is not highly dissimilar to pre-pandemic self-report research on contract cheating. For example, Curtis and Tremayne ([<reflink idref="bib9" id="ref84">9</reflink>]) also observed a self-admitted contract cheating rate of 3.5% in an Australian survey of over 1000 student conducted the year before the pandemic</p> <p>In order to increase recruitment and completion, the survey was deliberately brief and focused on only two contract-cheating behaviours. Bretag et al. ([<reflink idref="bib5" id="ref85">5</reflink>]), in contrast, surveyed seven different forms of assessment outsourcing, including behaviours omitted from our survey, such as students hiring others to sit their exams. Thus, this survey is limited to estimating the prevalence of only two commercial contract-cheating behaviours, albeit the most widely-researched forms of commercial contract cheating (Newton [<reflink idref="bib22" id="ref86">22</reflink>]). Future research applying methods such as the BTS should be conducted to obtain better estimates of the prevalence of other forms of sharing, outsourcing, contract cheating, and plagiarism behaviours.</p> <p>Although the BTS instructions significantly increased self-admission of engagement in some commercial contract cheating, it must be noted that the incentive offered for truthfulness in the BTS condition did not provide a direct personal benefit to students. Like John, Loewenstein, and Prelec ([<reflink idref="bib14" id="ref87">14</reflink>]), we linked truthfulness to charitable donations in order to keep the survey completely anonymous. However, it is possible that the students who are most likely to cheat are also those who respond more to self-serving, rather than altruistic, benefits (Jonason, Li, and Teicher [<reflink idref="bib15" id="ref88">15</reflink>]). The 'Dark Triad' personality traits of psychopathy, Machiavellianism, and narcissism are related to academic cheating (Lee, Kuncel, and Gau [<reflink idref="bib19" id="ref89">19</reflink>]), dishonesty, and selfishness (Jonason, Li, and Teicher [<reflink idref="bib15" id="ref90">15</reflink>]). In short, students who may be disposed to cheat and disposed to not be truthful about it, may also not be incentivised to be truthful when the benefit is given to someone else. Taken together, these facts suggest that the significant increase in reported commercial contract cheating under BTS instructions in our study is still likely to underestimate the extent of the behaviour.</p> <p>Our survey focused on Australian higher education institutions. We aimed to look at a similar cohort of students to Bretag et al. ([<reflink idref="bib5" id="ref91">5</reflink>]), as they had conducted the largest survey of contract cheating at the time of our research. Similarly, estimates of the prevalence of commercial contract cheating in other studies are also mainly from standard self-report surveys (e.g. Curtis and Clare [<reflink idref="bib7" id="ref92">7</reflink>]; Newton [<reflink idref="bib22" id="ref93">22</reflink>]). Nonetheless, it is important to note that self-reported estimates of the prevalence of commercial contract cheating are often higher in surveys from European and developing countries than from Australia (Awdry [<reflink idref="bib2" id="ref94">2</reflink>]). For example, Newton ([<reflink idref="bib22" id="ref95">22</reflink>]) reported an average of 15.7% of students engaging in commercial contract cheating from 13 studies from 2014-2016, but only two of these studies were from English-speaking Western universities. We would argue that the limitations of self-report surveys are similar in all cultures. Thus, although the estimated prevalence of contract cheating that we have presented in this paper cannot be assumed to be the same in other countries, it is safe to assume that figures elsewhere that derive from self-report surveys may underestimate the prevalence of contract cheating in those nations. Therefore, we would encourage researchers elsewhere in the world to use methods such as the BTS approach to estimate the prevalence of commercial contract cheating.</p> <hd id="AN0158963237-15">Summary and conclusion</hd> <p>This study sought to overcome the problems of self-report methodologies that are commonly used in studies from which the prevalence of commercial contract cheating has been estimated. To do this, we applied the BTS methodology outlined by John, Loewenstein, and Prelec ([<reflink idref="bib14" id="ref96">14</reflink>]) in a survey of contract cheating among a large sample of higher education students from 12 Australian institutions. We found that the BTS incentivised truth-telling methodology significantly increased self-admission by students of having ever submitted ghost-written assessments and having submitted pre-written assessments sourced from file-sharing sites in the past year. Moreover, the averaging of self-admission and impersonal estimates of contract cheating suggests that these behaviours may be many times more common than self-report-only surveys suggest. Additionally, being a non-native speaker of the language of instruction was the principal predictor of self-admitted contract cheating.</p> <p>Our results are concerning in the context of an Australian higher education system that has already taken significant educational, regulatory, and legal steps to combat commercial contract cheating. Of particular concern is the rate of students submitting assessments obtained from file-sharing websites, as the extent to which such sites breach new Australian laws is less clear than for custom ghost-writing providers. Our survey suggests that higher education institutions and governments need to re-double their efforts to combat the threat to higher education integrity posed by commercial contract cheating.</p> <hd id="AN0158963237-16">Acknowledgements</hd> <p>Our dear friend and colleague Professor Tracey Bretag provided invaluable support in planning this study. Sadly, Professor Bretag passed away before the study was complete. With heavy hearts, we acknowledge her contributions to this study and the field of academic integrity more broadly.</p> <hd id="AN0158963237-17">Disclosure statement</hd> <p>No potential conflict of interest was reported by the author(s).</p> <hd id="AN0158963237-18">Declaration of interest statement</hd> <p>We do not have any conflict of interest regarding this paper.</p> <ref id="AN0158963237-19"> <title> Notes </title> <blist> <bibl id="bib1" idref="ref27" type="bt">1</bibl> <bibtext> Receipts for charitable donations paid from this study can be found here: https://tinyurl.com/3th2zj4e</bibtext> </blist> <blist> <bibl id="bib2" idref="ref37" type="bt">2</bibl> <bibtext> See Supplementary Online Materials.</bibtext> </blist> <blist> <bibl id="bib3" idref="ref29" type="bt">3</bibl> <bibtext> Supplemental data for this article can be accessed https://doi.org/10.1080/03075079.2021.1972093.</bibtext> </blist> </ref> <ref id="AN0158963237-20"> <title> References </title> <blist> <bibtext> Ariely, D. 2012. The (Honest) Truth About Dishonesty. New York, NY : Harper Collins.</bibtext> </blist> <blist> <bibtext> Awdry, R. 2020. "Assignment Outsourcing: Moving Beyond Contract Cheating." Assessment & Evaluation in Higher Education, 1-16. Advance online publication. doi: 10.1080/02602938.2020.1765311.</bibtext> </blist> <blist> <bibtext> Boeije, H., and G. Lensvelt-Mulders. 2002. " Honest by Chance: A Qualitative Interview Study to Clarify Respondents' (Non-) Compliance with Computer-Assisted Randomized Response." Bulletin of Sociological Methodology 75 (1): 24 – 39. doi: 10.1177/075910630207500104.</bibtext> </blist> <blist> <bibl id="bib4" idref="ref3" type="bt">4</bibl> <bibtext> Bretag, T. 2019. " Contract Cheating Will Erode Trust in Science." Nature 574 (7780): 599 – 600. doi: 10.1038/d41586-019-03265-1.</bibtext> </blist> <blist> <bibl id="bib5" idref="ref6" type="bt">5</bibl> <bibtext> Bretag, T., R. Harper, M. Burton, C. Ellis, P. Newton, P. Rozenberg, S. Saddiqui, and K. van Haeringen. 2019. " Contract Cheating: A Survey of Australian University Students." Studies in Higher Education 44 (1): 1837 – 56. doi: 10.1080/03075079.2018.1462788.</bibtext> </blist> <blist> <bibl id="bib6" idref="ref1" type="bt">6</bibl> <bibtext> Clarke, R., and T. Lancaster. 2006. " Eliminating the Successor to Plagiarism: Identifying the Usage of Contract Cheating Sites." Paper presented at the second international Plagiarism conference, Gateshead.</bibtext> </blist> <blist> <bibl id="bib7" idref="ref5" type="bt">7</bibl> <bibtext> Curtis, G. J., and J. Clare. 2017. " How Prevalent is Contract Cheating and to What Extent are Students Repeat Offenders? " Journal of Academic Ethics 15 (2): 115 – 24. doi: 10.1007/s10805-017-9278-x.</bibtext> </blist> <blist> <bibl id="bib8" idref="ref49" type="bt">8</bibl> <bibtext> Curtis, G. J., C. Slade, T. Bretag, and M. McNeill. 2021. " Developing and Evaluating Nationwide Expert-Delivered Academic Integrity Workshops for the Higher Education Sector in Australia." Higher Education Research & Development. Advance Online Publication, doi: 10.1080/07294360.2021.1872057.</bibtext> </blist> <blist> <bibl id="bib9" idref="ref21" type="bt">9</bibl> <bibtext> Curtis, G. J., and K. Tremayne. 2021. " Is Plagiarism Really on the Rise? Results from Four 5-Yearly Surveys." Studies in Higher Education 46 (9): 1816 – 1826, doi: 10.1080/03075079.2019.1707792.</bibtext> </blist> <blist> <bibtext> Eaton, S. E., and K. L. Turner. 2020. " Exploring Academic Integrity and Mental Health During COVID-19: Rapid Review." Journal of Contemporary Education Theory & Research 4 (2): 35 – 41. doi: 10.5281/zenodo.4256825.</bibtext> </blist> <blist> <bibtext> Fellner, C. 2020. "Ghost Writers Helping UNSW Students to Cheat on Assessments, Leaked Report Reveals." The Sydney Morning Herald. May 5 <ulink href="http://www.smh.com.au">http://www.smh.com.au</ulink>.</bibtext> </blist> <blist> <bibtext> Harper, R., T. Bretag, C. Ellis, P. Newton, P. Rozenberg, S. Saddiqui, and K. van Haeringen. 2019. " Contract Cheating: A Survey of Australian University Staff." Studies in Higher Education 44 (11): 1857 – 1873. doi: 10.1080/03075079.2018.1462789.</bibtext> </blist> <blist> <bibtext> Jann, B., J. Jerke, and I. Krumpal. 2012. " Asking Sensitive Questions Using the Crosswise Model: An Experimental Survey Measuring Plagiarism." Public Opinion Quarterly 76 (1): 32 – 49.</bibtext> </blist> <blist> <bibtext> John, L. K., G. Loewenstein, and D. Prelec. 2012. " Measuring the Prevalence of Questionable Research Practices with Incentives for Truth Telling." Psychological Science 23 (5): 524 – 532. doi: 10.1177/0956797611430953.</bibtext> </blist> <blist> <bibtext> Jonason, P. K., N. P. Li, and E. A. Teicher. 2010. " Who is James Bond? The Dark Triad as an Agentic Social Style." Individual Differences Research 8 (2): 111 – 120.</bibtext> </blist> <blist> <bibtext> Kass, G. V. 1980. " An Exploratory Technique for Investigating Large Quantities of Categorical Data." Journal of the Royal Statistical Society: Series C (Applied Statistics) 29 (2): 119 – 127. doi: 10.2307/2986296.</bibtext> </blist> <blist> <bibtext> Krumpal, I. 2013. " Determinants of Social Desirability Bias in Sensitive Surveys: A Literature Review." Quality & Quantity 47 (4): 2025 – 2047. doi: 10.1007/s11135-011-9640-9.</bibtext> </blist> <blist> <bibtext> Lancaster, T., and C. Cotarlan. 2021. " Contract Cheating by STEM Students Through a File Sharing Website: A Covid-19 Pandemic Perspective." International Journal for Educational Integrity 17 (1): 1 – 16. doi: 10.1007/s40979-021-00070-0.</bibtext> </blist> <blist> <bibtext> Lee, S. D., N. R. Kuncel, and J. Gau. 2020. " Personality, Attitude, and Demographic Correlates of Academic Dishonesty: A Meta-Analysis." Psychological Bulletin 146 (11): 1042 – 1058. doi: 10.1037/bul0000300.</bibtext> </blist> <blist> <bibtext> McCabe, D. 2016. "Cheating and Honor: Lessons from a Long-Term Research Project." (pp). In Handbook of Academic Integrity, Edited by T. Bretag, 183-198. Singapore : SpringerReference.</bibtext> </blist> <blist> <bibtext> McNeilage, A., and L. Visentin. 2014. "Students Enlist MyMaster Website to Write Essays, Assignments". The Sydney Morning Herald. November 12 <ulink href="http://www.smh.com.au">http://www.smh.com.au</ulink>.</bibtext> </blist> <blist> <bibtext> Newton, P. M. 2018. " How Common is Commercial Contract Cheating in Higher Education and is it Increasing? A Systematic Review." Frontiers in Education 3 (August): 1 – 18. doi: 10.3389/feduc.2018.00067.</bibtext> </blist> <blist> <bibtext> Prelec, D. 2004. " A Bayesian Truth Serum for Subjective Data." Science 306 (5695): 462 – 466. doi: 10.1126/science.1102081.</bibtext> </blist> <blist> <bibtext> Reedy, A., D. Pfitzner, L. Rook, and L. Ellis. 2021. " Responding to the COVID-19 Emergency: Student and Academic Staff Perceptions of Academic Integrity in the Transition to Online Exams at Three Australian Universities." International Journal for Educational Integrity 17 (1): 1 – 32. doi: 10.1007/s40979-021-00075-9.</bibtext> </blist> <blist> <bibtext> Ross, L., D. Greene, and P. House. 1977. " The 'False Consensus Effect': An Egocentric Bias in Social Perception and Attribution Processes." Journal of Experimental Social Psychology 13 (3): 279 – 301. doi: 10.1016/0022-1031(77)90049-X.</bibtext> </blist> <blist> <bibtext> Rundle, K., G. J. Curtis, and J. Clare. 2019. " Why Students Do Not Engage in Contract Cheating." Frontiers in Psychology 10 : 2229. doi: 10.3389/fpsyg.2019.02229.</bibtext> </blist> <blist> <bibtext> Siev, S., and D. Kliger. 2019. "Cheating in Academic Exams: A Field Study." In Dishonesty in Behavioral Economics, edited by A. Bucciol & N. Montinari, 111-140. London: Academic Press. doi: 10.1016/B978-0-12-815857-9.00008-X.</bibtext> </blist> <blist> <bibtext> Sudman, S., and N. M. Bradburn. 1982. Asking Questions. San Francisco : Jossey-Bass.</bibtext> </blist> <blist> <bibtext> Tertiary Education Quality and Standards Agency [TEQSA]. 2017. "Good Practice Note: Addressing Contract Cheating to Safeguard Academic Integrity". Available at: https://<ulink href="http://www.teqsa.gov.au/sites/default/files/good-practice-note-addressing-contract-cheating.pdf?v=1507082628">www.teqsa.gov.au/sites/default/files/good-practice-note-addressing-contract-cheating.pdf?v=1507082628</ulink> (accessed May 27, 2018).</bibtext> </blist> <blist> <bibtext> Tertiary Education Quality and Standards Agency [TEQSA]. 2019. Amendment (Prohibiting Academic Cheating Services) Bill. Available at: https://<ulink href="http://www.aph.gov.au/Parliamentary%5fBusiness/Bills%5fLegislation/Bills%5fSearch%5fResults/Result?bId=r6483">www.aph.gov.au/Parliamentary%5fBusiness/Bills%5fLegislation/Bills%5fSearch%5fResults/Result?bId=r6483</ulink> (accessed February 16, 2021).</bibtext> </blist> <blist> <bibtext> Tourangeau, R., and T. Yan. 2007. " Sensitive Questions in Surveys." Psychological Bulletin 133 (5): 859 – 883. doi: 10.1037/0033-2909.133.5.859.</bibtext> </blist> <blist> <bibtext> Weaver, R., and D. Prelec. 2013. " Creating Truth-Telling Incentives with the Bayesian Truth Serum." Journal of Marketing Research 50 (3): 289 – 302. doi: 10.1509/jmr.09.0039.</bibtext> </blist> <blist> <bibtext> Whitley, B. E., A. B. Nelson, and C. J. Jones. 1999. " Gender Differences in Cheating Attitudes and Classroom Cheating Behavior: A Meta-Analysis." Sex Roles 41 (9): 657 – 80.</bibtext> </blist> </ref> <aug> <p>By Guy J. Curtis; Margot McNeill; Christine Slade; Kell Tremayne; Rowena Harper; Kiata Rundle and Ruth Greenaway</p> <p>Reported by Author; Author; Author; Author; Author; Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib22" firstref="ref2"></nolink> <nolink nlid="nl2" bibid="bib31" firstref="ref4"></nolink> <nolink nlid="nl3" bibid="bib21" firstref="ref8"></nolink> <nolink nlid="nl4" bibid="bib11" firstref="ref12"></nolink> <nolink nlid="nl5" bibid="bib26" firstref="ref15"></nolink> <nolink nlid="nl6" bibid="bib27" firstref="ref18"></nolink> <nolink nlid="nl7" bibid="bib28" firstref="ref19"></nolink> <nolink nlid="nl8" bibid="bib33" firstref="ref22"></nolink> <nolink nlid="nl9" bibid="bib20" firstref="ref23"></nolink> <nolink nlid="nl10" bibid="bib17" firstref="ref24"></nolink> <nolink nlid="nl11" bibid="bib23" firstref="ref25"></nolink> <nolink nlid="nl12" bibid="bib13" firstref="ref28"></nolink> <nolink nlid="nl13" bibid="bib14" firstref="ref30"></nolink> <nolink nlid="nl14" bibid="bib32" firstref="ref42"></nolink> <nolink nlid="nl15" bibid="bib18" firstref="ref44"></nolink> <nolink nlid="nl16" bibid="bib24" firstref="ref45"></nolink> <nolink nlid="nl17" bibid="bib10" firstref="ref46"></nolink> <nolink nlid="nl18" bibid="bib12" firstref="ref48"></nolink> <nolink nlid="nl19" bibid="bib29" firstref="ref50"></nolink> <nolink nlid="nl20" bibid="bib30" firstref="ref54"></nolink> <nolink nlid="nl21" bibid="bib16" firstref="ref58"></nolink> <nolink nlid="nl22" bibid="bib25" firstref="ref70"></nolink> <nolink nlid="nl23" bibid="bib15" firstref="ref88"></nolink> <nolink nlid="nl24" bibid="bib19" firstref="ref89"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1367888
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Moving beyond Self-Reports to Estimate the Prevalence of Commercial Contract Cheating: An Australian Study
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Curtis%2C+Guy+J%2E%22">Curtis, Guy J.</searchLink> (ORCID <externalLink term="http://orcid.org/0000-0002-4174-6955">0000-0002-4174-6955</externalLink>)<br /><searchLink fieldCode="AR" term="%22McNeill%2C+Margot%22">McNeill, Margot</searchLink><br /><searchLink fieldCode="AR" term="%22Slade%2C+Christine%22">Slade, Christine</searchLink> (ORCID <externalLink term="http://orcid.org/0000-0002-2197-2824">0000-0002-2197-2824</externalLink>)<br /><searchLink fieldCode="AR" term="%22Tremayne%2C+Kell%22">Tremayne, Kell</searchLink><br /><searchLink fieldCode="AR" term="%22Harper%2C+Rowena%22">Harper, Rowena</searchLink> (ORCID <externalLink term="http://orcid.org/0000-0002-5330-525X">0000-0002-5330-525X</externalLink>)<br /><searchLink fieldCode="AR" term="%22Rundle%2C+Kiata%22">Rundle, Kiata</searchLink> (ORCID <externalLink term="http://orcid.org/0000-0002-4207-9760">0000-0002-4207-9760</externalLink>)<br /><searchLink fieldCode="AR" term="%22Greenaway%2C+Ruth%22">Greenaway, Ruth</searchLink> (ORCID <externalLink term="http://orcid.org/0000-0002-8707-1797">0000-0002-8707-1797</externalLink>)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Studies+in+Higher+Education%22"><i>Studies in Higher Education</i></searchLink>. 2022 47(9):1844-1856.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Routledge. Available from: Taylor & Francis, Ltd. 530 Walnut Street Suite 850, Philadelphia, PA 19106. Tel: 800-354-1420; Tel: 215-625-8900; Fax: 215-207-0050; Web site: http://www.tandf.co.uk/journals
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 13
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2022
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22Higher+Education%22">Higher Education</searchLink><br /><searchLink fieldCode="EL" term="%22Postsecondary+Education%22">Postsecondary Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Measurement+Techniques%22">Measurement Techniques</searchLink><br /><searchLink fieldCode="DE" term="%22Incidence%22">Incidence</searchLink><br /><searchLink fieldCode="DE" term="%22Cheating%22">Cheating</searchLink><br /><searchLink fieldCode="DE" term="%22Contracts%22">Contracts</searchLink><br /><searchLink fieldCode="DE" term="%22Foreign+Countries%22">Foreign Countries</searchLink><br /><searchLink fieldCode="DE" term="%22College+Students%22">College Students</searchLink><br /><searchLink fieldCode="DE" term="%22Outsourcing%22">Outsourcing</searchLink><br /><searchLink fieldCode="DE" term="%22Writing+Assignments%22">Writing Assignments</searchLink><br /><searchLink fieldCode="DE" term="%22Predictor+Variables%22">Predictor Variables</searchLink><br /><searchLink fieldCode="DE" term="%22Incentives%22">Incentives</searchLink><br /><searchLink fieldCode="DE" term="%22Ethics%22">Ethics</searchLink>
– Name: Subject
  Label: Geographic Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Australia%22">Australia</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1080/03075079.2021.1972093
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0307-5079<br />1470-174X
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: The highest estimates of the prevalence of commercial contract cheating in Australia come from self-report surveys, which suggest that around 2% of students engage in commercial contract cheating during their higher education studies. However, self-report surveys are limited in that participants under-report socially-undesirable behaviours. In this study, we used an incentivised truth-telling method and surveyed 4098 students from six universities and six independent higher education providers in Australia. We found that 2.46 times more students admitted to commercial contract cheating, via submitting ghost-written assessments, when truth-telling was incentivised (via a Bayesian Truth Serum methodology) rather than when normal self-report survey instructions were used. Using prevalence estimation formulae that are combined with the incentivised truth-telling method, we estimate that 7.9% of students buy and submit assignments from commercial contract cheating services. Additionally, 11.4% outsource assessments via obtaining pre-written work from commercial file-sharing sites. These are substantially higher percentages of commercial contract cheating than self-reports suggest. Furthermore, having a first language other than English was the strongest demographic predictor of Australian students' engagement in commercial contract cheating. We conclude that commercial contract cheating is a more common problem than suggested by self-report surveys. We argue that academic integrity researchers should consider methods beyond standard self-reports to estimate the prevalence of academic misconduct and that efforts to curb commercial contract cheating must be increased.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2023
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1367888
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1367888
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1080/03075079.2021.1972093
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 13
        StartPage: 1844
    Subjects:
      – SubjectFull: Measurement Techniques
        Type: general
      – SubjectFull: Incidence
        Type: general
      – SubjectFull: Cheating
        Type: general
      – SubjectFull: Contracts
        Type: general
      – SubjectFull: Foreign Countries
        Type: general
      – SubjectFull: College Students
        Type: general
      – SubjectFull: Outsourcing
        Type: general
      – SubjectFull: Writing Assignments
        Type: general
      – SubjectFull: Predictor Variables
        Type: general
      – SubjectFull: Incentives
        Type: general
      – SubjectFull: Ethics
        Type: general
      – SubjectFull: Australia
        Type: general
    Titles:
      – TitleFull: Moving beyond Self-Reports to Estimate the Prevalence of Commercial Contract Cheating: An Australian Study
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Curtis, Guy J.
      – PersonEntity:
          Name:
            NameFull: McNeill, Margot
      – PersonEntity:
          Name:
            NameFull: Slade, Christine
      – PersonEntity:
          Name:
            NameFull: Tremayne, Kell
      – PersonEntity:
          Name:
            NameFull: Harper, Rowena
      – PersonEntity:
          Name:
            NameFull: Rundle, Kiata
      – PersonEntity:
          Name:
            NameFull: Greenaway, Ruth
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 01
              Type: published
              Y: 2022
          Identifiers:
            – Type: issn-print
              Value: 0307-5079
            – Type: issn-electronic
              Value: 1470-174X
          Numbering:
            – Type: volume
              Value: 47
            – Type: issue
              Value: 9
          Titles:
            – TitleFull: Studies in Higher Education
              Type: main
ResultId 1