Maintain detailed records of exam sessions and track candidates' progress and workflow thoroughly
Conduct regular reviews to ensure accuracy, relevance, and adherence to academic standards
Use insights from these reviews to improve strategies, protocols, and overall performance
The Record & Review proctoring model provides test-takers the option to choose their preferred exam times. This AI-powered proctoring solution automatically spots and marks suspicious actions for closer examination. Students receive alerts about behaviors that could lead to exam termination before and during the test.
All recorded data is securely stored and accessible for review anytime, enabling thorough analysis and detailed conduct reports for each candidate's session. Each exam session is individually reviewed by the proctor, who identifies any violations and creates a final behavior report.
The main benefit of the Record & Review system is that candidates have the flexibility to take exams whenever they want, instead of being restricted to specific pre-scheduled times set by the test organizer.
During the test-taking, the candidate is monitored according to administrator-set rules, with AI-powered detection of suspicious activity. The system captures random photos of the candidate's face, which are stored for review. Additionally, short videos of violations detected by the system are available for viewing. Each video is accompanied by a chart displaying violations that occurred during that time frame.
Depending on the setting of AI proctoring, the system can track different types of violations. In fact, 15 different types of violations can be detected:
b1 – Focus changed to a different windowc1 – Webcam is disabledc2 – Face invisible or not looking into the camerak1 – Atypical keyboard handwritingm1 – Microphone muted or its volume is lowm2 – Conversation or noise in the backgroundn1 – No network connections1 – Screen activities are not sharedc3 – Several faces in front of the camerac4 – Face does not match the profilec5 – Found a similar profilen2 – No connection to a mobile cameraEach metric is assigned a default weight of 1, which contributes to forming a confidence rating. This weight is multiplied by the fixed indicators in the session.
A proctoring credibility report is automatically generated. This document provides an objective evaluation of the proctoring process, including any violations observed during the exam or assessment.
The credibility in the exam results is calculated using the following formula:
where E ∈ [0, 100] is the credibility score (if E < 0, then E = 0), xk is the session-averaged value of the metric k, wk is the weighting coefficient of the metric k, M ∈ {b1,b2,c1,c2,...} are the metrics.
Thus, the credibility score shows the total average percentage of violations for the entire session, the duration of the session is taken as the actual one, i.e. the difference between when the student started the session and when he/she finished it.
After a proctored test, the recordings are uploaded to the YTM Proctoring platform for review by proctors. They use a standardized rubric to ensure consistency in evaluations, considering factors like the test-taker's behavior, audio and video quality, and test accuracy. Proctors then provide a final report and assign a status (positive, negative, or neutral) based on the test-taker's behavior.
This process ensures fair and consistent evaluation of each test-taker's performance. By using the same criteria, proctors maintain fairness and accuracy in their assessments.
Once the final report is entered, results become available to the test-taker and relevant authorities. This allows informed decisions and necessary actions to be taken based on the results.
