Research Results
This page presents the comprehensive results from our systematic testing of AI capabilities in academic writing across multiple university modules and disciplines.
Results Overview
The table below displays results from 4 completed tests across various academic disciplines and levels. Each entry represents a complete assessment cycle where AI systems were tasked with producing work that meets university standards.
Test Results Data
| Module Code | Title | Grade | Time | Intervention | Level | Resources |
|---|---|---|---|---|---|---|
| PHIL301 | Philosophy of Mind: Consciousness and AI | 78% | 4 hours 23 minutes | Minimal | Undergraduate Year 3 | |
| SOCI450 | Digital Society and Technology Ethics | 82% | 3 hours 47 minutes | Guided | Graduate | |
| ENGL275 | Contemporary Literature and AI Narratives | 75% | 5 hours 12 minutes | Collaborative | Undergraduate Year 2 | |
| COMP580 | AI Ethics and Algorithmic Fairness | 85% | 2 hours 56 minutes | Expert | Graduate |
Data Interpretation
Grade Distribution
The results demonstrate varying levels of AI performance across different academic contexts, with grades ranging from 75% to 85%.
Intervention Impact
The data reveals significant relationships between the level of human intervention and the quality of final submissions:
- Minimal Intervention: Basic prompting with limited human input
- Guided Intervention: Strategic direction and structured prompting
- Collaborative Intervention: Substantial human-AI partnership in content development
- Expert Intervention: Extensive human expertise applied to AI outputs
Time Efficiency
AI-assisted academic writing demonstrates significant time efficiency compared to traditional student work, with most submissions completed in under 6 hours regardless of complexity.
Methodology Notes
Each test followed a standardized protocol:
- AI system provided with authentic assignment briefs
- Access to relevant course materials and reading lists
- Human intervention limited to predefined categories
- Submissions evaluated using standard academic criteria
- Complete process documented via video recording
Video Documentation
Click the 🎥 icon in the Resources column to view process documentation for individual tests. These videos demonstrate the AI-human interaction patterns and decision-making processes involved in producing each submission.
Data Downloads
For researchers interested in analyzing our complete dataset:
This data is made available for academic research purposes. Please cite the Plagiar-Ezy project in any published research using this data.