How the calculation works
Each reference publishes LMS parameters (Box-Cox power, median, and variation) tabulated by age. We look up the parameters for the exact age — interpolating linearly between tabulated ages, exactly as the CDC documents — then convert the measurement to a z-score and a percentile with the standard normal distribution. It's the published method, implemented verbatim.
- CDC — stature-for-age, 24–240 months, standing height.
- WHO — length/height-for-age, birth–5 years; recumbent length under 24 months, standing height from 24 months (with the real ~0.7 cm step preserved).
- UK-WHO — ages 2 weeks–4 years, computed from the WHO tables the RCPCH charts use in that window.
Reading the result honestly
"Taller than X% of peers" compares against the reference population — children measured when the reference was built. It doesn't rate your child, predict the future, or diagnose anything. Percentiles are also rounded and clamped to 0.1–99.9, because beyond that the normal model implies more precision than a growth reference can support.
Want the visual? Plot the same measurement on the boys growth chart, or estimate adult height from parents' heights with the child height predictor.