Cheating in tests and examinations and plagiarism in assignments
have a long history in educational institutions around the world.
The Internet and wireless communications have made what some students
call ‘outsourcing’ into a form of e-commerce and subject matter
for a growing number of publications. Some authors report on their
strategies for catching ‘offenders’, while others suggest that
better teaching and assessment will reduce the ‘need’ for plagiarism.
International Survey: I was part of an Association for Computing
Machinery working group which produced a report on cheating based
on a survey of tertiary computing academics in Australia, New
Zealand, the UK and the US, conducted in May/June 2002. The survey
asked tertiary educators about incidents of cheating, methods
used to detect cheating and suggested improvements.
Most respondents mentioned copying in assignments: from texts
or the Internet (without referencing), and from other students
with their knowledge (straight copy or unauthorised collaboration)
or without (from disks, printouts or unattended computers). Less
common were cheating in exams (smuggling in notes, looking at
other students' answers or communicating with other students)
and handing in the same assignment for different courses. One
respondent identified "giving false reasons for extensions" (including
false medical certificates), another complained about "late delivery
of work" and a third mentioned "fake IELTS certificates" (IELTS
is an English language test used for admission to tertiary education).
One student accidentally found the answers to an assessment and
gave them to his friends.
Measures taken to prevent (reduce) cheating include:
• closed book exams
• disabling email in practical exams
• explanations and warnings
• group assignments
• individualised assignments
• open book exams
• oral presentations on written work
• requiring a minimum number of references
• requiring signed "own work" declarations
• separating students in exams
• showing how referencing should be done
• supplying special resources for exams.
Measures taken to detect cheating include:
• checking disk properties
• commercial plagiarism detectors
• looking for changes in style
• putting invisible text in program code
• strip marking
• web searches.
Many respondents believe that the incidence of cheating has increased.
Reasons offered include changing student population; declining
moral values; family pressure; larger classes; peer pressure;
Most respondents stated that their institution had published
policy on cheating. Cases of cheating were usually referred to
the appropriate authority (adjudicator, mediator, programme leader,
head of department or CEO). Possible penalties included warnings,
having to complete another assignment (an essay on plagiarism
in one case), reduced marks, failing the course, and expulsion
from the programme or institution.
Personal Experience: When I taught programming to beginners,
I stressed the importance of students writing their own programs
and fixing the bugs themselves. Nevertheless, I often detected
plagiarism and was usually able to get the culprits to confess
and then reinforced the importance of ‘doing it yourself’. Now
I teach a master's course involving written assignments, and emphasise
the importance of learning to analyse and critique what others
have written before synthesising original ideas with the clearly
acknowledged ideas of others. I use plagiarism detection software
(turnitin.com) to show (in glorious technicolour) evidence of
any failures to acknowledge other writers. Only one student has
argued with the visual evidence!
As programme leader for a bachelor's degree I took part in a
disciplinary hearing involving three students who had obviously
worked together over two semesters, handing in minor variations
on a programming assignment for one course, a written assignment
for a second course and another written assignment for a third
course. Initially they all denied cheating, then one admitted
they had completed the assignments together and introduced minor
variations to make them look different. The second owned up, but
the third insisted (despite the other two having confessed in
his presence) that they had only ever "talked about the assignments".
The evidence was clear to the disciplinary committee who suspended
all three for twelve months. Not long afterwards we began to receive
obscene emails that obviously related to the case. They stopped
after we told the emailer's Internet Service Provider what that
account was being used for!
Donald Joyce is Associate Professor, Computer
Studies, UNITEC. This article is adapted from a paper presented
at the July 2002 NACCQ Conference in Hamilton.