PERSONALIZED SUBNETTING TUTOR
Intelligent tutoring system preparing community college cybersecurity students for industry success through personalized learning
CLIENT
CMU Future of Work Initiative
+
Community College of Allegheny County
MY ROLE
- • Instructional Design
- • User Research
- • Tutor Development
- • Testing & Analysis
TOOLS
- • Figma
- • CTAT Platform
- • Think-aloud Protocols
TIMELINE
Oct 2024 - May 2025
(8 months)
THE CHALLENGE
3000+ Students Every Year
enroll in introductory IT courses in Community College of Allegheny County (CCAC), but many struggle with foundational skills like subnetting.
Technical Competencies
are increasingly prioritized by Pittsburgh employers over degrees. These students at the threshold to secure local IT jobs once they master essential skills.
THE SOLUTION
Four Key Features Powering Personalized Learning
The Personalized Subnetting Tutor leverages intelligent tutoring system (ITS) technology, deployed on the Cognitive Tutor Authoring Tools (CTAT) platform, to provide adaptive learning through:
1. Error-Specific Feedback
Targeted explanatory feedback addresses different misconceptions at each step, helping learners effectively correct specific conceptual gaps based on 14 identified error patterns.
2. Multi-Tiered Hint System
Hints layered from definitions to guided instructions to complete solutions, allowing learners to access just the right level of support as needed without overwhelming them.
3. Easy Strategy Planning
Drag-and-drop solution steps before actual calculations, cultivating learners' metacognition skills and building confidence for future independent problem-solving.
4. Flexible Entry Points
Freedom to answer questions in learner-preferred order, simulating the real-world flexibility needed in cybersecurity troubleshooting scenarios.
THE RESULTS
Validated Effectiveness Across Two Student Populations
Study 1: CMU Validation
20 novice learners, 10 in each condition
Experimental Group:
Tutor with layered hints and error-specific feedback
→ 75.7% success rate
Control Group:
Basic tutor without adaptive feedback
→ 53.8% success rate
Result: 41% higher effectiveness with adaptive feedback features (statistically significant, p<0.05)
THE PROCESS
Theoratical & Empirical Cognitive Task Analysis (CTA):
I conducted comprehensive cognitive task analysis with multiple stakeholders to understand how students learn subnetting and where they struggle:
- 1 instructor as subject matter expert for learning objectives and theoretical framework
- 3 learners across skill levels (novice, advanced, professional) for empirical think-aloud protocols
Critical Research Findings:
Research Insights
Differentiated learning needs across skill levels
Novice learners easily forgot terminology, while advanced learners got stuck at procedural steps.
Systematic error pattern identification
Comprehensive analysis revealed 14 distinct misconceptions and error steps in problem-solving.
Binary/decimal conversion as prerequisite skill
All participants relied on binary/decimal conversion throughout problem-solving; a prerequisite knowledge rather than core learning objective
Design Decisions
→ Multi-tiered hint system
Layered hints starting with terminology definitions, progressing to instructional scaffolding, and culminating in complete solutions
→ Error-specific feedback system
Tailored corrective responses addressing each identified misconception to target specific conceptual gaps
→ Integrated conversion tools
Built-in converter to support problem-solving without detracting from primary subnetting learning goals
User Testing & Iterative Design - From Concept to Refined Solution
Pilot Testing with 4 CMU Students revealed critical usability insights:
User Testing Insights & Iterative Improvements:
User Testing Insights
Layered hints with definition is helpful
Learners found layered hints, particularly terminology definitions, highly helpful for learning support
Conversion tool effectiveness with workflow friction
Binary/decimal converter functionality was identified as essential ("life saver") for problem-solving success, but manual copy-pasting is laborous.
Design Response
→ Seamless auto-conversion (TPA)
Implemented transparent background conversion that maintains support while eliminating manual transfer steps
The Result: A significantly more intuitive and user-friendly learning experience that maintained educational effectiveness while improving usability.
Technical Development: From Prototype to Production
Development Approach:
- Rapid Prototyping: Initial behavioral graphs for quick design iteration
- Production Scaling: Developed nools file for efficient hint and feedback delivery
- Knowledge Component Tagging: Structured system for future data-driven improvements
Stakeholder-Responsive Enhancement
CCAC instructors requested integration of the magic number table — a shortcut method for subnetting calculations.
Solution: Redesigned tutor with two new features:
- Interactive magic number table for guided practice
- Drag-and-drop sequencing module for procedural step internalization
This required comprehensive updates to hints and error feedback, all systematically tagged for future analysis.
Real-World Deployment
Successfully deployed at CCAC following instructor approval, with ongoing data collection to inform continuous improvement.