3 reasons a hybrid proctoring approach makes sense
AI needs to be combined with the efforts of human proctors and instructors to effectively detect and prevent cheating
As online learning increasingly becomes part of the higher-ed experience, it is imperative that the methods and technologies used to deliver and monitor online learning maintain high levels of integrity, fairness and usability. When it comes to online assessments and proctoring, there are multiple models—AI-only monitoring (fully automated), human-only monitoring or a combination of both: hybrid proctoring.
Each model offers a slightly different approach, and each has benefits. AI and machine learning tools utilize technology to improve monitoring efficiency, provide better exam insights, and strengthen approaches to online teaching and learning. Human oversight allows for decisions to be made using an individual’s knowledge, logic, and reasoning, making students feel more confident that they are not being assessed solely by a machine.
But the most effective online proctoring model blends these benefits by combining AI, machine learning, and human intervention to detect unexpected behaviors, ensure fairness, and uphold integrity. The convenience that AI and machine learning provide, combined with a human’s ability to make decisions based on complex and unforeseeable situations, improves the entire testing experience for both students and instructors.
Here are three ways that a hybrid proctoring model—where AI and human oversight are used in tandem—creates the ideal testing experience for both instructors and students.
1. A Less Invasive Experience for Students
With fewer distractions during exams, students are able to focus on showing what they have learned. However, proctoring AI tools that are overly sensitive can trigger flags and create unnecessary interruptions.
Most proctoring AI can detect sound, but some platforms have AI with smart voice detection. While these terms are often used interchangeably, how they impact the test experience is very different.
For example, sound detection can trigger a flag for practically any sound, such as a cough or a doorbell. Whereas smart voice detection triggers for specific keywords or phrases, such as “Hey Siri” or “OK Google.”
While having AI in place to monitor and flag actions, some situations can be tricky. For example, a student may be talking to themselves to work out a question and may say, “Okay, Google was founded in 1998.”
In situations similar to this, human proctoring combined with AI can prevent unnecessary interruptions. The human proctor is able to review the footage and see that the student was not trying to cheat, eliminating the need to intervene and disrupt the student.
2. AI Becomes Smarter With Human Feedback
Human proctor feedback, potential suspicious behavior, potential violations, and instructor input can be collected by the proctoring platform and reported to the AI. This collection of data helps the AI learn, adapt, and refine to provide students, instructors, and educational institutions with better testing experience insights.
The feedback loop varies from institution to institution. This loop helps customize the proctoring platform based on institution and instructor preferences and policies. Additionally, this feedback loop allows AI to make better correlations between behaviors that may indicate true academic dishonesty and those that are false flags.
This also helps the AI to learn so that it can adjust to reduce the number of false or unimportant flags due to oversensitivity allowing proctors and instructors to focus on important situations related to academic dishonesty and streamlines the review.
Combining the benefits of AI with the benefits of a human proctor into a hybrid proctoring approach creates a fair and equitable learning experience for students.
3. Supports a Variety of Non-traditional Exam Formats
AI is a great resource for most traditional testing situations, but what if instructors want to test through a demonstration? Or an open book exam? In these cases, combining AI with human input is crucial for protecting academic integrity, minimizing unnecessary flags, and improving the student online testing experience.
In a non-traditional test format, certain situations may be allowed that are not typically allowed during a traditional test. In these situations, the AI needs to be told to not flag activities that may typically be seen as cheating, such as two people in the room, unexpected voices or movements, restricted browsers, and leaked test questions.
Instructors can adjust test settings and decide which proctoring features to use (or not use) before issuing a non-traditional exam. They can also give the proctor instructions to further customize what is allowed during the exam and to provide accommodations for specific students.
For example, a non-traditional exam may ask students to complete a math problem using pen and paper. The instructor would ask the proctor to turn on ‘Scratch Paper Allowed’ so students won’t be flagged for looking down at papers. Instructors can require students to use a sheet of blank white paper, a black or blue pen, and to show the paper before and after they complete the problem. This ensures that students can complete the exam without interruption and cuts down on the need for human proctors to review flags.
AI is a powerful tool that can provide an abundance of benefits for the online learning experience. However, it isn’t always perfect and shouldn’t be used alone. AI needs to be combined with the efforts of human proctors and instructors to effectively detect and prevent cheating, ensure academic fairness, and uphold integrity.