Managing scoring settings
Scoring plays a central role in online proctoring by providing a systematic way to evaluate the integrity of a monitored exam session. It quantifies observed violations and helps distinguish between sessions with compliant behavior and those requiring further review. By offering a transparent framework, scoring supports fair assessments and helps maintain trust in the examination process.
What is scoring?
Scoring is the process of assigning points to rule violations or suspicious behaviors detected during a proctored session. Regardless of the proctoring method used (live, AI-assisted, or automated) each session is assigned a score that reflects its level of compliance. The higher the score, the more serious or frequent the violations.
This score serves as a flagging mechanism, helping administrators quickly identify candidates who may have breached exam rules.
The scoring color scale
To simplify interpretation, each session is visually tagged using a three-color system: green, yellow, or red. This intuitive scale gives proctors and administrators a quick overview of a session's risk level:
- 🟢 Green — the session is low-risk. Violations, if any, are minor and fall well below the alert threshold. The candidate’s behavior is considered compliant.
- 🟡 Yellow — the session shows moderate risk. The cumulative score is nearing the threshold for a violation. These sessions may warrant a closer look.
- 🔴 Red — the session is high-risk. The score exceeds the defined threshold, indicating strong evidence of suspicious or non-compliant behavior.
This color-coded system enables efficient triaging, helping teams prioritize reviews and take appropriate action.
How is the score calculated?
Scoring is based on a specialized formula developed by our engineering team. This formula takes into account several key factors, including the duration of the video, the type of each violation, their assigned coefficients, and the total number of violations recorded during the session.
Each violation is given a specific coefficient, which reflects its severity. These coefficients are pre-set on our servers but can be customized in the Dashboard for each exam. For greater flexibility, some coefficients can also be passed via the API (see Constructor Proctor API in the Proctor Developer Guide).
The final score for a session is calculated using several parameters:
Violation weight — this is expressed as a percentage and is based on the duration of the violation relative to the session length. If the session’s end time is unknown, the current duration is used instead.
Scoring points for the violation — each violation contributes to the final score based on its weight, its coefficient, and a duration multiplier. These values are combined to determine the total score for the session.
Together, these calculations provide a fair and consistent way to assess the overall integrity of a proctored session.
See also: