Task 2.2 — Instructor-Facing

Learning Analytics Dashboard

Aggregate student interaction data into pedagogically meaningful metrics that inform real-time teaching decisions and long-term curriculum refinement.

32
Active Students
+4 this week
847
Scenario Attempts
+126 this week
14.2min
Avg. Time-to-Exploit
+2.1 min vs last week
203
Tutor Hints Used
-18% vs last week

Scenario Difficulty & Performance

Time-to-exploit and trust-boundary identification accuracy across scenarios.

Average Time-to-Exploit by Scenario

Trust-Boundary Identification Accuracy

Red Team vs. Blue Team Analysis

Comparing student attack success rates against their defense performance.

Attack vs. Defense Success Rate

Module Progression Correlation

Common Failure Patterns

Top misconceptions and failure modes identified from student interaction logs.

Module Failure Pattern % Students Recommended Action
Module 1 Used visible text injection instead of hidden channels 72% Review trust boundary between rendered vs. parsed content
Module 1 Identified attack success but couldn't name the vulnerability class 58% Add pre-lab exercise on trust-boundary taxonomy
Module 2 Failed to trace API response → tool invocation chain 45% Introduce tool-call visualization before API scenario
Module 2 Confused email social engineering with architectural exploitation 38% Clarify agent vs. human social engineering in lecture
Module 3 Could not trace reasoning chain past the 2nd hop 67% Add intermediate scaffolding for multi-step attack tracing
Module 3 Successful attack but defense proposal was trivially bypassable 24% Include adversarial robustness criteria in defense rubric

Instructional Insights

Auto-generated recommendations based on current class performance data.

💡

Recommended Teaching Adjustments

Based on this week's data across 32 students and 847 scenario attempts:

  • 1. Revisit zero-click attack vectors before Module 2 — 72% of students used visible injection in the email scenario, suggesting they haven't internalized the hidden-channel concept from Module 1.
  • 2. Add scaffolding for multi-hop reasoning — the DB→Email chained attack in Module 3 has a 67% failure rate at the second hop. Consider adding an intermediate exercise that traces 2-hop chains before attempting 3-hop scenarios.
  • 3. Introduce defense-first exercises — students show a significant attack/defense asymmetry (78% attack success vs. 41% defense success). A "defend before you attack" exercise order may improve defense reasoning.
  • 4. Reduce tutor hint dependency for the web injection scenario — students requested hints 3.2x more often than for file poisoning, despite similar difficulty ratings. Consider providing better initial framing for web scenarios.