Spaces:
Sleeping
Sleeping
| title: AI Trading Trust Experiment | |
| emoji: 📈 | |
| colorFrom: blue | |
| colorTo: green | |
| sdk: gradio | |
| sdk_version: 4.44.0 | |
| app_file: app.py | |
| pinned: false | |
| license: mit | |
| # AI Trading Trust Experiment | |
| A psychology research game studying trust in AI advice under varying conditions. | |
| ## About | |
| This is a behavioral experiment disguised as a trading simulation. Participants: | |
| 1. Configure an AI advisor's characteristics (confidence, explanation depth, risk tolerance) | |
| 2. Receive AI-generated trading advice for fictional companies | |
| 3. Make trading decisions (follow or ignore the AI) | |
| 4. Rate their confidence in their decisions | |
| ## Data Collected | |
| - Participant ID (anonymized) | |
| - AI tuning preferences (slider positions) | |
| - Trading decisions and response times | |
| - Confidence ratings | |
| - Portfolio outcomes | |
| ## For Researchers | |
| Data is stored in SQLite (`experiment_data.db`). Download from the Space's Files tab. | |
| Tables: | |
| - `participants`: Session-level data | |
| - `decisions`: Turn-by-turn data with all variables | |
| ## Fictional World | |
| All companies and countries are entirely fictional to prevent participants from using external knowledge. | |