The document provides an overview of contract cheating prevention and detection strategies within the computing discipline. It discusses 12 strategies for preventing contract cheating, including designing original assessments, using team-based work, and educating students about academic integrity policies. It also outlines 10 strategies for detecting contract cheating, such as monitoring essay writing websites, using individualized assignments, and checking students' work against writing samples. Finally, the document discusses how academic policies can help or hinder prevention/detection and provides examples of policies to consider, such as those related to examinations, attendance, and training of academic staff.
3. Students are being done a disservice when staff do not make it
clear that they are against contract cheating.
Good assessment design is crucial.
Need to design out opportunities to cheat.
How much effort is put into assignment design?
Students are expected to put hours into their own work.
Is the same effort put into developing the assessment materials?
Are they written clearly and originally?
Why Prevent Contract Cheating?
4. Many universities devalue exams.
A well-invigilated exam is hard to cheat on, and can accurately
assess learning outcomes:
Exams can be designed to assess more than memory.
Many STEM subjects benefit from practical tests and exams.
Students arrive at university now who are less experienced at
coursework:
Much GCSE assessment is exam based, or is work produced under classroom conditions.
An examination component, even if small, means that students
cannot pass by contract cheating alone.
Prevent #1 – Type Of Assessment
5. A module, with multiple pieces of assessment, can be assessed as a
coursework and a linked test.
The test further assesses an element that the students completed in the assessment.
For instance, in a computer based subject, this could require making a small modification to
the coursework under test conditions.
In a report based subject, this could involve reflecting on one element of the report.
The preparation required for this test is relatively low if students
have completed the work for themselves.
The module marking should be constructed so that students who
do not demonstrate understanding at this point cannot pass.
Prevent #2 – Linked Test
6. In many ways, this can be considered a variant of the Linked Test
method.
Students are given a short individual viva and asked questions
about their work in order to assess if they have done it.
This is a very good way of checking deeper understanding, but can
be more difficult with larger classes.
It is also possible for a viva to cover more than one module for a deeper check.
Prevent #3 - Viva
7. “There are students being caught at Leeds purchasing essays through several of the online
sources mentioned above. Where this is reported by a department to the University
(Office of Academic Appeals & Regulation) and the allegation is upheld, such students
invariably face exclusion from the University, even for a first offence.
Where a student is suspected of submitting work that may have been commissioned or
bought, and is therefore beyond electronic detection, the University supports a
department in questioning the student about the content of the work as part of
the formal investigation at departmental level. Such questioning usually reveals in an
obvious and often painful way that the student is not the author.”
University Of Leeds Support
http://www.lts.leeds.ac.uk/bulletin/issue18/page3.php
8. Most of the time, contract cheating is an individual activity.
By getting students to work in teams, it is much harder for one
student to disappear and not contribute.
Such assignments still need to be carefully designed, so that students cannot end up just
splitting work into multiple tasks that are completed in isolation.
Opportunities need to be given for students to report suspicion
without the chance for reprisals.
It is best for teams to be assigned randomly to eliminate
complacency and to simulate the workplace.
Prevent #4 - Teamwork
9. Assignment briefs should never be reused, taken from other
university websites, or taken from textbooks.
This is unfair to students and this is tempting them to cheat.
Often, model answers end up easily available.
Assignment briefs can be personalised.
Include recent news stories or academic research.
Get the class to decide on an appropriate topic as a cohort.
Structured assignment briefs can be generated
Topics can be slotted in each year, to reduce the workload for redelivery.
Students can be allocated their own topic out of a large selection.
Prevent #5 – Original Assessments
10. Assignment briefs can be set which require students to have a
local understanding.
For instance, it can be required to use a certain lab, or to follow a computer programming
style taught in class.
An attendance component can be included.
The student must show their logbook at 5 out of 6 sessions.
Ask for changes based on what is shown.
Students can still pass on these changes as they go through, but the effort required by them
is now not far off from doing the work themselves.
Prevent #6 –Local Knowledge
11. Assess more than just a single core element.
For instance, in software development, assess more than just the source code produced.
Critical reflection can be difficult to fake.
What problems did I encounter?
What would I do differently next time?
What are the next skills I am going to develop?
Prevent #7 – Require Reflection
12. Honour Codes are much used in the United States to ensure
good academic behaviour.
Students are required to report people who they know are not behaving, otherwise they
are considered equally guilty.
It may be possible to implement these, or some form of system
where students are able to report their suspicions:
For instance, could a reward be offered is a student finds an assignment specification on an
auction site?
Many students are already keen to have the opportunity to report things that are amiss.
Prevent #8 – Honour Codes
13. An electronic copy of work can be easily checked for indicators
who completed it:
Look in the Properties.
Look at the file time stamp and time of submission.
Open in an ASCII editor.
This can also help to identify students when an assignment
specification is found on a contract cheating site.
Prevent #9 – Electronic Submission
14. Automated plagiarism detection tools should always be used.
This is generally Turnitin in the UK due to its large international database of previous
work.
Students need to be informed that their work is being checked for
plagiarism.
Although detection of contract cheating using these services is
difficult, indicators are possible.
Students do not always understand how these tools work, and this can be a deterrent.
Third parties can cut corners and plagiarise themselves.
Prevent #10 – Detect Plagiarism
15. Make sure that students know that you are aware of contract
cheating and plagiarism.
Many students assume that their tutors are not technically astute, and that they can use
these methods of cheating.
Ensure that plagiarism and contract cheating education is included
in your course.
Prevent #11 – Educate Students
17. Ultimately, this is so the honest students aren’t penalised.
The “don’t let them get away with it” attitude.
Detection can be difficult, but there are some steps that can be
taken to make this easier.
Why Detect Contract Cheating?
18. On 11 January 2012, this message was posted on the JISC
Plagiarism Mailing List.
When applying for a hardship loan, why would you include bank statements as evidence
that include reference to payment to “customessays.co.uk”?
Oops. Had to laugh.
Derek
Sometimes, Detection Is Easy…
19. Monitor the main auction sites.
This can be an individual job, or assigned as a role within a
department.
Look for jobs posted locally, or where the university is
mentioned.
A Google search can be powerful, but will not access internal
databases.
Detect #1 – Monitoring
20. Set up a Google Alert for terms used in the assignment.
This will provide a regular email when new pages appear on the Web which mention that
term.
An alert can also be set for a given academic.
Detect #2 – Google Alerts
22. This technique is related to Google Alerts and monitoring.
Where possible, add unusual words or terms to the assignment.
This makes monitoring much easier.
Made up town name.
Foreign words.
Detect #3 – Unusual Terms
23. Embed your own details into every assignment specification.
These include your name, subject, email address and university.
This means that, when an assignment is spotted online, it can be
traced and a method of notification is available.
For advanced results, also add this assignment specification to
plagiarism detection engines (such as Turnitin).
These are often used by people trying to detect contract cheating to identify unusual
fragments within the assignment.
Detect #4 – Embedded Details
24. This is primarily a method for traceability, when a student places
an assignment specification online.
It can also prevent plagiarism, where students copy from one another.
Every student is given a unique problem to work from.
This can be a unique set of questions or topics.
Unique numbers for mathematical problems.
An individual set of requirements to be met.
For best results, embed this within a personalised PDF.
Use the Word Mail Merge functionality or similar.
Upload the individualised assignments to Turnitin.
Detect #5 – Individualisation
25. A more advanced individualisation technique places visible and
invisible indicators within a file.
Visible indicators could include the name of the student, or the topics that they had been
allocated.
Invisible indicators could include sequences of blank spaces or hidden characters (those
appearing as if they were blank to a user).
A student stripping away the visible indicators is less likely to
notice the invisible ones, leaving that element of traceability.
Detect #6 – More Individualisation
26. Linguistic techniques exist that can check the author of a
particular work.
These require that a sample of work produced by that author is
already available.
These can be collected by early assessed pieces of writing completed in class.
Continuing to collect work electronically through the career of a student is also helpful.
Where the writing style differs greatly, this can be checked by
observation.
There is an on-going research problem to automate this process.
Detect #7 – Writing Samples
27. Use manual and automated checks for plagiarism.
Look for indicators that the work was not produced by the
student:
Unusual writing conventions, examples or place names.
Textual indicators shown by Turnitin (look at all results, not just the obvious ones).
Detect #8 – Check Everything
29. Policies can help and hinder the prevention and detection of
contract cheating.
These can be on a departmental or institutional level.
A number of policy decisions are presented, intended to open up
discussion, or provide opportunities for improvements.
Help Or Hinder?
30. One of the easiest ways to reduce contract cheating opportunities
is to use examinations instead of coursework.
Does this method accurately assess the student ability?
Is widespread use of exams even permissible?
Examination Policy
31. Students can be required to pass all items of assessment, which
may be linked in some way.
Could students pass a module purely by passing a single item of assessment which they
could contract cheat on?
Where two assessment items are linked - if students fail one of the two items, can that be
used as evidence for them to fail the whole module?
Pass Policy
32. Students who are attending classes regularly, and engaging with
taught material, are much likely to produce work for themselves.
They may also learn something in the process.
Is there a way to mandate attendance?
Can it be a requirement that some work is produced during taught sessions?
Attendance Policy
33. Help people actively looking for contract cheating by making
assignment specifications available.
Are your assignment specifications available outside of a university intranet or password
protected site?
Do they contain details of the university, module and tutor?
Upload the assignment specifications to Turnitin to provide content to match
contract cheating work against.
Make Data Available Policy
34. Some methods to detect contract cheating, such as vivas, are a
potentially onerous assessment technique – but they give much
higher levels of confidence of the originality of work.
Does the Workload Allocation Model allow different modules to carry varying amounts of
effort?
Are all modules required to adhere to a standard delivery plan (e.g. set number of hours)?
Is time given to pursue cheating cases?
Workload Allocation Policy
35. Knowing the writing style of a student is important to identify
work that they have not written.
Does an anonymous marking policy prevent this?
Anonymous Marking Policy
36. The university academic misconduct regulations need to explicitly
mention contract cheating.
Is a clear statement about contract cheating included?
Are agency sites, essay writing sites and third parties explicitly listed as unacceptable?
What level of evidence is required for a contract cheating case?
Are the penalties appropriate?
Are the academic misconduct regulations reviewed on a regular basis to take account of
developments in the cheating world?
University Regulations
37. Students should be explicitly educated about contract cheating.
Do students ever hear that the contract cheating regulations are taken seriously, and not
just ignored?
How often are students educated on both plagiarism and contract cheating?
“Name And Shame” Policy
38. Staff at all levels should be trained to ensure that they assess in a
contract cheating friendly manner.
Do all staff receive this training and are they receptive to it?
Are checks made that staff have implemented what they were trained in?
Do management understand the importance of avoiding contract cheating, and how this
influences the value of academic qualifications?
Training Policy
39. Staff should be required to indicate the steps they have taken to
minimise contract cheating.
Does assignment specification moderation explicitly check that this is original and the
methods are suitable?
Is there explicit reflection on the success of those techniques after the module has been
delivered?
Is good practice reported back to other staff by management?
Moderation Policy