By Serhii Tkachenko, CEO at Unicheck.
Teaching consistently shows to be one of the most rewarding career paths. Which is not saying it is full of challenges, stress and downright hassles. Wouldn’t it be great if we put EdTech to the service of educators, simplify their workloads and allow them to become allies and examples of integrity to students?
Being a teacher also means safeguarding academic integrity. It’s only human to look for easier ways to carry out tasks we are afraid of, or just reluctant to do. Students, in their process of consolidating a moral compass and subject to —sometimes unduly— stress, are especially susceptible to short-term thinking. Industries such as essay mills take advantage of this vulnerability, offering students contract cheating services. This type of cheating is very hard to detect and prove.
However, Artificial Intelligence (AI) is here to change that.
Why do students cheat?
86% of high school students admit to have cheated at some point, according to this year’s Academic Cheating Fact Sheet by Stanford University. Younger students tend to copy answers from their classmates, while middle and high-schoolers often opt to let someone do their tasks instead of them.
The younger the students, the simpler the motives. Some of them appear not to be ready for the school curriculum, which makes the educational process look too overwhelming. Other factors behind cheating attempts in school are caused by a lack of interest in a particular subject, or fear of failing in the eyes of adults.
Among college students, the reasons for cheating get more serious. Modern work requirements are demanding and competition is severe. At the same time, higher education often can’t keep up with the changing requirements of the contemporary job market.
The particularly hairy problem of contract cheating
High aspirations can devolve into academic dishonesty and cheating. But of all its forms, contract cheating is arguably the most tricky one. There are two main reasons why.
First, the companies that offer these kinds of services, or as they call it “academic research writing,” target students with their marketing campaigns, pushing them toward dishonesty and making money in the process. They provide sophisticated methods to ease whatever moral conflict a student might have preventing them from reaching out to them.
Second, essay mills do pay attention to classic plagiarism. Taking advantage of modern plagiarism software, they make sure the works are original so plagiarism checkers will not be able to detect them.
Some countries now call on students to report peers who use essay-writing services. But neither such measures nor the efforts by companies like Google to stop advertising essay mills have been successful. Millions of university students worldwide keep paying essay mills to write their academic papers. Here are some of their reasons:
- Wrong image of education often broadcasted by mass media. Education is not something fancy and exclusive for children of wealthy parents alone. It’s a repository of knowledge which is worth every dedicated minute.
- Non-native students tend to cheat more than those studying in their home cities or countries. Unfortunately, universities often fail to provide support for language studying.
- Personal qualities. Being lazy is natural, but excessive idleness in educational matters shows disregard for personal responsibility.
AI as an Anti-Cheating Solution
Detecting, fighting and preventing cheating is one of the ultimate tasks of teachers, and technology is here to assist them. Modern-day cheating detection is more than comparing answers with other students or noticing unfamiliar handwriting. Since students started resorting to essay mills for contract cheating services, it used to be almost impossible to prove that someone cheated.
Teachers and institutions spend too many hours and money trying to ascertain whether students’ works were ghost-written, plagiarized, purchased, repurposed, traded or not. Even if they are convinced that the student had in fact cheated there is usually no evidence to hold them accountable.
Luckily, what is too complicated for people can be within the power of machines. Implementing AI for contract cheating detection, authorship verification and overall academic integrity can help teachers across schools, colleges and universities.
How contract cheating detection works: Breaking down the breakthrough
Machine Llearning and AI are perfect for pinpointing the subtle differences people cannot discern. They show impressive results in handling discrepancies in texts, inconsistencies in writing and language ambiguity. By investigating the choice of words, the sentence structure, the length of words and sentences, AI tools can verify if the author is indeed the one who had submitted the taskm almost instantly.
Take Emma, for example. Developed by Unicheck, this AI-based assistant helps educators analyze students’ works against the metrics that often remain unnoticed to the human eye. Emma uses Natural Language Processing (NLP) and Stylometry —the study of an individual's writing patterns— to learn and measure specific linguistic characteristics and writing practices typical to a particular person. This allows the software to analyze the writing of each text, categorizing it into linguistic and stylistic silos.
All Emma needs to do the job is data: At least three pieces of original work written by the student in question, of about 300 to a thousand words in length each. That would be enough to "learn" the writing style of the student, categorize the text into linguistic and stylistic structures and share the outcome with the teacher.
Using a machine learning tool in the classroom might sound overly complicated, but Emma and similar tools are increasingly available and very user-friendly. Furthermore, they can be embedded into the LMS of your organization and be seamless part of the teacher's workflow.
Powerful AI-enabled authorship verification tools, supplied with extended capabilities like those provided by Emma, can help school, college and university teachers to fight the problem of academic dishonesty more comprehensively. Over time these tools can become more dependable, and allow educators to go beyond mere plagiarism checking which can show word-for-word copying only.
Before punishing: AI keeping cheating incentives in check
Technologically-enabled cheating detection is not about trying to harm or punish students. Instead, AI-powered education tools can point out the right way toward educational integrity by detecting and preventing. Learning Analytics could even help us identify those users more likely to contract cheat.
To be sure, AI will not absolve students nor educators from the moral and ethical elements involved in cheating. Teachers should still be explicit about the implications and consequences of contract cheating among the academic community.
Before punishing students for inappropriate conduct, educators should encourage them and keep them motivated, making their students happy. According to Teacher Impact Statistics, 83% of students shared that a teacher they had interacted with boosted their self-esteem and confidence. That is why modern technologies should let educators do what their noble title implies: Being mentors and role-models who give students a chance to find their mistakes and correct them themselves. At the end of the day, the role of these technologies should be to extend these powerful pedagogical moments and abilities.
And when students know that their teacher is using tools like Emma, they will definitely think twice before even trying to cheat.
If used not to punish students, but to prevent essay mills from profiting off the students’ weaknesses, AI-based authorship verification assistants can make a real difference for academic integrity and the entire educational process.