# Study 000I Results

## Primary Answer

`Yes, supported.`

The archive is better explained by an adaptive-envelope model than by a single fixed support-threshold model.

## Key State Contrast

- Early low-burden outside expression (`2025-04-16`):
  - running-share context: `28.24593%`
  - treadmill neighbors: `3.0`
  - distance: `10.2265 miles`
  - HR residual: `-3.75027 bpm`

- October high-burden outside cluster:
  - running-share context: `5.1299%`
  - treadmill neighbors: `1.75`
  - distance: `2.091 miles`
  - HR residual: `13.8844 bpm`

- Late low-burden boundary expression (`2026-04-09`):
  - running-share context: `91.08392%`
  - treadmill neighbors: `32.0`
  - distance: `5.4541 miles`
  - HR residual: `0.68866 bpm`

## Interpretation

The critical result is that the archive contains two low-burden outside-expression states with very different support structures. A static-envelope model based on one fixed support threshold does not organize those cases well. An adaptive-envelope model does.