Field Measurement Protocol ZX — TOST Tool for Testing √2−1
An interactive statistical tool running entirely in the browser to test whether a specific geometric ratio in trees (such as the crown ratio (H−h)/H) is equivalent to the Silver Ratio √2−1 = tan(π/8) ≈ 0.41421356.
Statistical Methodology
TOST (Two One-Sided Tests): To test equivalence against a practical margin Δ instead of a traditional difference test.
Pre-registration Lock: Locking the hypothesis at the first tree to prevent p-hacking.
Bonferroni Correction: Correcting α/6 when no ratio is pre-specified.
Accurate t-distribution: Via logGamma + betacf to calculate CI and p-value accurately.
Power Approximation: Estimating the current statistical power and the required sample size.
Calculated Ratios
h/H — Trunk ÷ Total Height
r/H — Crown radius ÷ Height
2r/H — Full crown diameter ÷ Height
d/H — Trunk diameter ÷ Height
(H−h)/H — Crown ÷ Height (Primary target)
h/(H−h) — Trunk ÷ Crown
L/H — Shadow ÷ Height (Excluded from testing, as it depends on sun angle)
Reference Mango Sample Values
Exploratory sample n=5 mango trees (Sho'ub, Sana'a, 2026-10-14): Mean (H−h)/H = 0.41297, SD = 0.00845, SE = 0.00378, 90% CI = [0.40491, 0.42102], TOST p = 0.000801 with Δ = ±0.03. Result: Equivalence confirmed within margin, but analysis is exploratory (not pre-registered) and not generalizable as a law.
Tool Features
Input tree measurements (H, h, r, d, L) with site, species, and time.
Automatic real-time calculation of ratios.
Pre-registration lock with the ability to unlock (recording the change in history).
Filtering by tree species.
Full statistical analysis with verdict (Equivalent / Different / Inconclusive).
CSV export with pre-registration header + CSV import.
All data saved locally (localStorage) — nothing is sent to a server.
By: Eng. Abdulla Mohammed Nasser Al-Jabri — Yemen. ORCID: 0009-0003-3319-3822. License: CC BY 4.0.
🌳 Field Measurement Protocol
Hypothesis Testing: Is there a ratio in the tree ≈ √2 − 1 ≈ 0.414 ?
🔒 Hypothesis Locked since
🎯 Pre-registration Hypothesis (Before data collection)
Why lock? Choosing the ratio or margin after seeing the data opens the door to
p-hacking. Locking at the first tree makes the test honest: the chosen ratio
is tested at α=0.05 (no correction), and the rest are exploratory.
📝 New Tree
📐 Calculated Ratios
h / H (Trunk ÷ Total)—
r / H (Crown Radius ÷ Total)—
2r / H (Full Crown)—
d / H (Thickness ÷ Total)—
(H−h) / H (Crown ÷ Total)—
h / (H−h) (Trunk ÷ Crown)—
L / H (Shadow ÷ Height)—
💡 The gold card = a ratio close to 0.414 (within ±Δ of the pre-registered hypothesis)
⚠️ L/H is excluded from hypothesis testing: it depends on the sun angle, not the tree's shape.
Any vertical tree at the same angle will yield the same value — this is not a test of structure.
📋 Registered Trees0
No trees yet. Start by adding a tree.
📊 Statistical Analysis
Trees (in filter)0
Mean h/H—
SD—
Mean r/H—
SD—
√2−1 (Reference)0.4142
⏳ Waiting for sufficient data (at least 5 trees for initial analysis)
📤 Export
🔒 All data is saved on your device only — nothing is sent to any server. Completely safe.