QRF DIAGNOSTICS SUMMARY
============================================================
Generated: 2025-07-28 20:39:11

1. COMMON SUPPORT
----------------------------------------
Average overlap coefficient: 0.914
All overlap coefficients > 0.85: True
All SMDs < 0.25: True
All KS tests p > 0.05: True

2. VARIANCE EXPLAINED
----------------------------------------
Wages: 67%
Capital Income: 54%
Retirement Income: 71%

3. OUT-OF-SAMPLE ACCURACY
----------------------------------------
Average improvement vs hot-deck: 34.9%
Average improvement vs linear regression: 20.5%
90% prediction interval coverage: 89.4%

4. JOINT DISTRIBUTIONS
----------------------------------------
All correlation differences < 0.05: True
Average correlation difference: 0.019

============================================================
These statistics demonstrate that the QRF methodology successfully:
- Maintains strong common support between datasets
- Achieves high predictive accuracy for imputation
- Preserves joint distributions of variables
- Provides well-calibrated uncertainty estimates
