Campus Review Vol 29. Issue 7 July 2019 | Page 28

TECHNOLOGY campusreview.com.au Evolution of cheating Could bespoke interventions through AI stop students cheating? By James Thorley W hile technology has had a positive impact on the student learning experience, it has also posed a challenge for educational institutions in upholding academic integrity. New technologies make it easier for students to conduct academic misconduct, such as plagiarising or contract cheating (where a student recruits a third party to complete their academic work). This is largely due to the increase of opportunities through better access to information, improved communication and access to cheap ‘labour’. As Newton suggests in his 2018 study, there has been an increase over recent years in academic misconduct across the board. While he doesn’t make any conclusions in his work, it’s very likely that increased access to technology is a contributing factor. So technology can be said to have had both a positive and negative impact on academic integrity. However, it is my belief that the next wave of technological changes could greatly diminish academic misconduct. Before we get into that though, it’s important to understand the evolution of misconduct. While academic misconduct has been around since the start of education, the first instance where technology had a major influence was the introduction of the World Wide Web, and the ability to copy and paste. As content moved from offline to online sources, more students would pull information from academic resources, 26 insert it into their essays and submit it as their own work. While some students continue to copy and paste today, similarity checking technology now gives teachers the ability to detect and police all unreferenced sources. The next challenge to academic integrity saw the rise of the essay mill, where students would upload essays to earn credits that gain them access to other submitted essays. The idea of pre-written essays being available made academic misconduct much easier. But again, technology has stopped this, to a large extent, as the database of essays submitted began to build up, allowing software to compare essays submitted across various academic institutions. Fortunately, essay mills have almost completely died out. But once again, academic misconduct evolved. Educational institutions are now faced with a larger problem: bespoke essay writing. The internet facilitated the rise of e-commerce, which saw the introduction of online marketplaces and the gig economy, matching writers to students willing to pay someone else to write their assignments. As Ellis, Zucker and Randall showed in their 2018 study, technology around e-commerce has allowed for a highly developed contract cheating site structure to emerge. This is currently the most sophisticated method of misconduct. In any teaching environment, there needs to be a balance between education and quality assurance. Students first need to be made aware of the risks, and educated about the importance of academic integrity. While some misconduct is inevitable, technologies have already significantly reduced the amount of instances across educational institutions. Here are the three stages where technologies have been applied to maintain integrity and reduce academic misconduct: 1. Policing the issue – the challenge posed by copy and paste and essay mills are easily policed through the use of plagiarism software. This technology can analyse a student’s work to see if they’re using online content or using past assignments. However, while policing is an effective measure to detect misconduct, it doesn’t address the root of the issue, which is why education is important. 2. Education aims to stop the misconduct before it occurs. Technology comes into the picture as a learning resource throughout the drafting and development of a written piece of work, providing the student with an opportunity to learn what’s considered misconduct, while not being penalised. It also opens up a line of communication between the teacher and the student around the topic of academic integrity. 3. Intervening through traits of misconduct – contract cheating is the hardest form of academic misconduct to identify because it is original work, just not by the student. However, we’re close to the stage where technology can be used to understand and detect unoriginal work. By using machine learning and artificial intelligence (AI) to study multiple written submissions from a student, the technology can identify patterns and traits, and recognise an essay that does not fit those patterns. While technology can still be used for policing and education, AI has the ability to intervene when the student begins to show signs of struggling with the course or program and provide them with the correct resources to help them stop before it results in academic misconduct. If done right, this is where technology has the ability to make academic misconduct all but disappear through a bespoke intervention, on an individual basis, to students who need this level of support. Most students who go to university don’t start a course wanting to plagiarise, download work from an essay mill or engage a contract cheater. They often get to that stage because they don’t have the skills or understanding, or have external pressures that force them down that path. This is where technology has an opportunity to highlight opportunities for teacher intervention, which ultimately should result in better academic results, higher graduation rates and greatly reduced levels of academic misconduct. ■ James Thorley is regional director, APAC, at Turnitin.