Artificial Intelligence (AI) holds incredible potential when it comes to online assessment. Since its initial implementation in the sector in the 1990s, AI has played an increasingly bigger role in helping organisations create better online tests, improve proctoring and scoring, and clamp down on cheating. In the process, the technology has saved organisations a lot of money. AI is here to stay!
While these benefits are great, many are concerned about the impact of AI on the human experience of education. Will AI eventually replace humans? It’s a legitimate concern, but we believe the short answer is “no”, or at least not in the foreseeable future. Let’s see why.
How does AI work?
In simple terms, artificial intelligence is where computers simulate human behaviour, predominantly in decision-making and perception. It uses machine learning to ‘learn’ from past data without the need for programming. Developers use complex algorithms and so-called models to give the computer the ability to analyse data and identify patterns. They then feed it large chunks of data, known as datasets, to set the process in motion. The better the algorithm and the bigger the initial dataset, the more effective and accurate the AI will be. This is very important to remember when looking at the role of artificial intelligence in assessment.
AI in online assessment
When it comes to online tests, AI is employed in a variety of ways to make the online assessment process better, and a big part of that is maintaining exam integrity and combatting cheating. This includes things like facial recognition to validate student identities, audio recording to listen for irregular or suspicious sounds, eye movement detection to spot erratic or unusual eye movements, and environmental scanning to ‘look’ for potentially illegal objects like books or notes in the examination location.
It can also assist with the scoring of exams, give feedback to online exam takers, and provide examiners with analytics that will aid them in setting better, more effective exams.
The future of AI in online assessment
Artificial intelligence technology is improving by the day and modern AI can do some incredible things. But where does it leave human teachers? The good news is that while AI technology is extremely advanced, there are still things that require human intervention.
As academic integrity expert Dr. Ashley Norris puts it, “A smoke detector spends all day and night looking for one thing, when it finds it, an alarm is sounded. But a human then has to decide whether someone left the meatloaf in the oven too long or whether it’s time to gather the kids and dogs and get out.”
So where does AI fall short (and humans excel)?
Open-ended questions
A big hurdle for artificial intelligence in assessment is open-ended questions. It can identify spelling, grammar, clarity and how a piece of writing flows to a very high degree, but it struggles with interpreting nuances in language, thought processes and reasoning. Humans take into account things like cultural differences, language peculiarities and external references to accurately assess open-ended questions.
Flaws in facial recognition
Facial recognition technology is something that AI can struggle to get perfect. Despite being trained on millions of images, there are examples where facial recognition software has mistaken innocent people for wanted criminals due to inaccuracy.
How does this impact exams and how do humans help? The flaws in facial recognition technology mean that exam-takers may be misidentified or have certain eye movements or actions deemed suspicious. Humans, on the other hand, are extremely good at recognising faces and also spotting irregular head movement, like someone looking at something off-screen.
Wrongly identifying behaviour or environments
Similar to the facial recognition flaws, there have been instances where AI misinterpreted certain movements, or misidentifying objects in an environment, leading to flagging exam-takers for potential cheating.
Human teachers are significantly better at identifying suspicious behavior and are better at accurately perceiving surrounding environments.
Artificial intelligence has had an enormously positive impact on education and testing for both educators and students. However, the point at which technology can completely replace human teachers is still quite some way off, and humans will play a critical role in creating the best possible experience for everyone for many years to come. So teachers, don’t put down your red markers just yet, you are still very much needed.