Analytics & Trends

The Analytics & Trends surface provides institutional administrators and lead proctors with a high-level overview of examination integrity. It transforms raw proctoring signals—such as gaze deviations, tab switches, and audio anomalies—into actionable institutional metrics and historical trends.

Performance Summary

The top-level metrics offer an immediate snapshot of the current examination cycle's health. These cards track four primary KPIs:

Total Sessions: The cumulative number of exam attempts recorded within the selected timeframe.
Average Risk Score: A mean calculation of the risk levels across all students, providing a baseline for typical behavior.
Total Violations: The absolute count of flagged events (e.g., identity mismatches, unauthorized device detection).
Flagged Rate: The percentage of sessions that reached the threshold for manual review, indicating the overall integrity of the exam.

Each metric includes a Trend Indicator (e.g., +12% or -8%) comparing the current data to the previous period, allowing for rapid identification of emerging integrity issues.

Integrity Visualizations

The dashboard utilizes recharts to provide multi-dimensional views of proctoring data:

Risk Score Over Time

An Area Chart that tracks the fluctuation of risk scores across the day or month. This visualization helps administrators identify specific time windows where suspicious activity peaks, which is critical for identifying coordinated cheating attempts or technical vulnerabilities.

Violation Category Distribution

A Bar Chart that breaks down violations by their specific type. This helps in understanding how students are attempting to bypass security:

Gaze Deviation: Sustained movement of eyes away from the screen.
Tab Switching: Attempts to access external browser resources.
Audio Anomaly: Detection of voices or suspicious background noise.
Face Absence: Instances where the student leaves the camera frame.

Risk Distribution (Segmentation)

A Pie Chart that segments the student population into four distinct risk categories based on their calculated score:

Low (0-25): Standard behavior with minimal to no alerts.
Moderate (26-50): Minor alerts (e.g., occasional tab switches).
High (51-75): Significant behavioral anomalies requiring attention.
Critical (76-100): Multiple high-severity violations; immediate review recommended.

Session History & Audit Trail

The Recent Sessions table provides a granular drill-down into individual student attempts. This table serves as the primary jumping-off point for manual audits.

Filtering and Time-Series Data

Users can toggle between different time ranges to observe broader trends:

Today: Real-time monitoring of ongoing or just-completed exams.
Week: Short-term trends used for reviewing specific course modules.
Month: Long-term institutional reporting used for accreditation and policy adjustments.

Data Structure (Mock Implementation)

The analytics view consumes data structured around session history and violation counts. For developers extending the platform, the metrics are derived from the following schema:

// Sample Data Structure for Analytics
interface SessionRecord {
  id: string;
  studentName: string;
  riskScore: number;    // 0-100
  violations: number;   // Count of events
  status: "flagged" | "clean";
  timestamp: string;    // ISO format
}

interface ViolationTrend {
  type: string;         // e.g., "Tab Switch"
  count: number;
}

Usage in Decision Making

Administrators use these charts to:

Adjust Thresholds: If the "Flagged Rate" is too high due to false positives (like background noise), proctors can adjust the sensitivity of the Audio Intelligence system.
Audit High-Risk Peaks: Use the "Risk Over Time" chart to focus proctoring resources during peak violation periods.
Institutional Reporting: Export summary metrics to demonstrate the effectiveness of remote proctoring guardrails to stakeholders.