Analytics Dashboard
Aggregate data visualisation across participant sessions
Note: This dashboard currently displays illustrative mock data for demonstration purposes. Real analytics will be available once the platform is deployed with participant data collection.
Framework Preferences
Which framework resonated most with participants
Response Length by Framework
Average words and tokens per response
Framework Effectiveness by Scenario Category
Participant satisfaction scores (0-100) by scenario type and framework
Key Observations (Illustrative)
Patterns emerging from framework comparison data
- HSI excels in boundary scenarios: The containment-first approach shows highest satisfaction when clear limits are needed.
- Constitutional AI preferred for extended conversations:Character-based engagement resonates more in longer interactions.
- Hybrid shows consistent middle-ground performance:Rarely the strongest but avoids the weaknesses of either extreme.
- Response length correlates with framework philosophy:Constitutional AI produces longer, more detailed responses on average.