AI technology used in online education spares tutors and instructors the trouble of manually reviewing and grading assignments and monitoring students during online tests so instructors can have more time to develop quality eLearning content and interact with students.
AI-powered reviewing and grading software can independently “read” and grade assignments building on the grades given by human instructors. After training, such software can bulk-check student papers with the precision of a human.
Grading software has already
proved its effectiveness in science, technology, engineering, and math classes, unbiasedly grading two times more assignments than a human professor during the same time. However, grading systems are more challenging to apply in social sciences and humanities. At least 21 states in the US
use technology to evaluate essays, from middle school to college level, yet AI algorithms powering the adopted software remain biased. For example, using advanced words tricks algorithms into thinking an essay is well-written when in fact, it doesn’t have any substance. The algorithms have also turned out to be biased against students from different nationalities and cultures, deeming their essays as poorly written.
When it comes to preventing cheating at online exams, AI-powered remote proctoring software can help. Educators may rely on proctoring solutions to automatically detect suspicious behaviors by recording students during tests via their computer’s camera or preventing students from switching browser tabs.
An Educause poll
suggests that 54% of institutions are already using eLearning tech for proctoring, and another 23% are considering using it. However, some educational organizations failed to weigh out the ethics of remote proctoring, which caused severe resistance from students and parents. Still, there are institutions whose approach to remote proctoring can be named a best practice. Florida State University, for example, encrypts all of the data collected by proctoring software, and it is professors, not the AI algorithms, that make the final decision on whether a student cheated or not.