# Study 000C Results

## Summary

The exhaustive full-data pass supports a clear result:

- the core running findings remain intact
- the ecology interpretation becomes more precise
- the later running-specialization story survives full correction
- the hybrid-cardio layer changes context, not the running-only mechanics conclusions

## Corrected yearly ecology

| year | running_count | indoor_conditioning_count | running_share_of_structured_hours | indoor_share_of_structured_hours | hybrid_indoor_cardio_count | hybrid_indoor_cardio_hours | hybrid_indoor_cardio_pct_of_indoor_hours |
|---|---|---|---|---|---|---|---|
| 2025 | 131 | 217 | 21.00 | 63.64 | 13 | 10.18 | 4.20 |
| 2026 | 120 | 19 | 84.18 | 11.38 | 5 | 3.09 | 26.34 |

The corrected ecology changes what can be said about the conditioning system:

- 2025 still reads as a Garmin-labeled conditioning-heavy year
- but only `4.20%` of those indoor-conditioning hours were GPS-bearing hybrid cardio
- 2026 has much less indoor-conditioning overall, but `26.34%` of its indoor-conditioning hours were GPS-bearing hybrid cardio

So the Garmin indoor label is imperfect in both years, but the scale and placement of that imperfection are not the same.

## Hybrid cardio inside the flagship windows

| window_label | hybrid_session_count | hybrid_total_distance_miles | hybrid_total_duration_hours | hybrid_type_counts |
|---|---|---|---|---|
| study000a_000b_early_window | 7 | 9.4458 | 4.9936 | indoor_cardio:7 |
| study000a_000b_late_window | 0 | 0 | 0 |  |
| microstudy_b_later_outdoor_phase | 5 | 4.8749 | 3.086 | indoor_cardio:5 |

This is the most important corrected-context result:

- the early flagship window contained `7` hybrid indoor-cardio sessions totaling `9.45` miles and `4.99` hours
- the late flagship window contained `0`
- the later Microstudy B phase contained `5` hybrid indoor-cardio sessions, but all of them occurred in January 2026

## Separation of later outdoor probes from the later hybrid-cardio block

| outdoor_probe_date | hr_residual_bpm | nearest_hybrid_cardio_date | nearest_hybrid_gap_days |
|---|---|---|---|
| 2025-10-02 | 12.0 | 2026-01-19 | 109 |
| 2025-10-08 | 18.06 | 2026-01-19 | 103 |
| 2025-10-09 | 12.6 | 2026-01-19 | 102 |
| 2025-10-10 | 20.59 | 2026-01-19 | 101 |
| 2026-04-09 | 0.1 | 2026-01-30 | 69 |

This check clears a major possible hole. The later outdoor probe runs are not sitting directly on top of the later hybrid-cardio block. They are `69` to `109` days away from it. That means the outdoor running probes remain interpretable as running probes rather than immediate hybrid-circuit artifacts.

## Claim stability matrix

| claim_id | claim_label | full_data_result | status | implication |
|---|---|---|---|---|
| C1 | Phase-Structured Adaptation | Retained | stable | The phase model survives full-data audit unchanged. |
| C2 | Turnover-Led Gain | Retained | stable | Cadence remains the stronger accessible lever across stabilized and higher-demand contexts. |
| C3 | Conserved Mechanics | Retained | stable | Comparatively anchored mechanics survive context expansion. |
| C4 | Later Running Stabilization Resource | Retained but narrowed | refined | The claim applies most directly to later running specialization, not to the whole conditioning ecology. |
| C5 | Mixed Internal Cost | Retained | stable | Efficiency improved while unexplained session-level burden persisted. |
| C6 | Uniformly Indoor Early Ecology | Rejected | corrected | The early flagship window included GPS-bearing hybrid outdoor conditioning hidden under indoor_cardio labels. |
| C7 | Microstudy B Directly Confounded By Hybrid Cardio | Not supported | cleared | Later outdoor probe runs are 69 to 109 days away from the later hybrid cardio block and remain interpretable as running-only probes. |

The full-data pass found only one genuinely corrected claim:

- the early conditioning ecology was not uniformly indoor

It did **not** overturn the main running findings:

- phase-structured adaptation remains
- cadence still carries `62.42%` of the cadence-stride speed gain
- `vertical_ratio_pct` still has the lowest monthly variability at `2.14%`
- later specialized outdoor cadence remains `5/5` above expected (`p = 0.03125`)
- later specialized outdoor stride remains `5/5` below expected (`p = 0.03125`)
- later specialized outdoor vertical ratio remains `5/5` above expected (`p = 0.03125`)
- later specialized outdoor vertical oscillation residual remains `-0.3111` and later specialized outdoor GCT residual remains `-19.3198`

## Strongest full-data answer

The strongest complete answer after exhausting the corrected dataset is:

`the program supports selective adaptation through corrected ecological narrowing into later stabilized running specialization, with turnover preserved more successfully than stride expression and with mixed internal cost rather than pure economy`