6
active projects
6
protocols · 2 amended, visibly
3
deviations recorded, not smoothed
4
reproducible snapshots frozen
Projects
each carries a structured research question, a protocol state and a cohort| Project | Research question | Protocol state | Cohort |
|---|---|---|---|
| Biological seed treatments · winter wheat | Do biological seed treatments change yield vs untreated in comparable UK contexts? | locked before work began · v2 | COH-BST-WW-v4 · n=23 |
| N-use efficiency across practice systems | How does NUE differ across N strategies when method and basis align? | amended · v3 · amendment visible | COH-NUE-2526 · n=19 |
| Cover crops & soil organic matter | Does overwinter cover change SOM trajectory over 3+ seasons, LOI held constant? | locked before outcome · v1 | COH-COVER-SB-v2 · n=11 |
| Reduced-N systems meta-review | Can reduced-N systems hold yield across soils, seasons and varieties? | deviation flagged · sampling date slipped | COH-RN-META · n=31 |
| Establishment methods · OSR | Establishment counts by method where counting protocol matches. | declared · v1 | COH-EST-OSR · n=9 |
| Min-till transition outcomes | Yield and fuel trajectories through transition years 1–3. | locked before work began · v1 | COH-MT-TRANS · n=14 |
Biological seed treatments · winter wheat
protocol v2 · locked 6 Oct 2023Structured research question
Population: UK winter wheat, weighbridge-measured · Intervention: biological seed treatment · Comparator: untreated, same management · Outcome metric: yield t/ha
Protocol
split-field or strip layout · reference arm required · same-management confirmed · weighbridge basis only
Cohort
COH-BST-WW-v4 · 23 included / 105 excluded, every reason named · 17 new candidates awaiting review
Protocol register — amendments and deviations stay on the record
| Event | Date | What changed | Effect |
|---|---|---|---|
| protocol locked | 6 Oct 2023 | v2 method frozen before any 2024 drilling | strongest single-record claim enabled |
| amendment | 14 Feb 2025 | yield-monitor records admitted where calibration certificate present | cohort v3 → v4 · 6 records became eligible · both versions kept |
| deviation | 2 Jun 2025 | one site sampled 11 days late (weather) | record retained, flagged · excluded from time-sensitive strata |
A method change is annotated, never hidden — trend breaks get explained, and prior cohort versions remain reproducible.
Project rules: protocols, amendments and deviations are recorded and visible; no evidence-quality score, no validity
badge, no ready-made conclusion. Reproducibility, not persuasion.