. . . . "Downloaded NHANES DEMO and BMX files for the six 2007-2018 cycles, merged on SEQN, kept examined persons under 20 years with a stature measurement (recumbent length below 24 months, standing height from 24 months; n = 23551, boys 12038, girls 11513). For each sex, median stature by completed year of age gave the model-free yearly velocity v(a); a local-linear Gaussian-kernel smoother (SD 0.5 yr; 0.75 and 1.0 yr as sensitivity) on monthly medians gave the smoothed velocity curve on a monthly grid. Bootstrap over individuals (1000 resamples) gave 95% CIs for the infancy velocity, the adolescent excess over childhood velocity and the velocity at 18.5-19.5 yr. The four pre-registered sub-criteria were evaluated on boys and reported for girls." . . "Measured: 3.283333333333333 cm/yr (boys' adolescent peak yearly velocity minus mean childhood velocity, ages 4-10) (n = 12038, 95% bootstrap CI [1.80, 4.42]). Interpretation: Thompson's 'curve of growth for length or stature' and its phases are read as the first derivative (velocity, cm/yr) of median stature against age. NHANES is cross-sectional, so the population-median velocity is a smoothed version of individual velocity: the adolescent spurt is attenuated and widened relative to longitudinal curves, which makes criterion (c) conservative. 'Babyhood' is taken as the first two years, 'early boyhood' as 4-10 years, 'teens' as 10-17, 'full height' as 19-20 years. Birth length is not in NHANES, so infancy velocity is measured from the 0-11 month to the 12-23 month median (midpoints ~0.5 and ~1.5 years)." . "Claim test of thompson1917-ch03-05: supports, confidence 0.8" . . . "Planned datasets: nhanes-demo-2007-2018, nhanes-bmx-2007-2018" . "claim-test/0.1.0" . "biological" . "Refutes if (c) fails in boys, i.e. the adolescent peak does not exceed childhood velocity (bootstrap CI of the excess includes 0 or the smoothed ratio < 1.0), or if (a) fails (infancy velocity not above childhood velocity). Any other combination of failed sub-criteria (e.g. an irregular childhood plateau or incomplete arrest by 19.5 yr) gives 'mixed'." . "Birth length is absent from NHANES, so 'infancy' velocity is measured between the 0-11 and 12-23 month medians (about 14 cm/yr), which understates the true first-year velocity." . "Criterion (b) is met at the threshold: the range of boys' yearly velocities between ages 4 and 10 is exactly 2.0 cm/yr (allowed <= 2.0), driven by a pre-pubertal dip to 4.7 cm/yr at 8.5-9.5 yr. Childhood velocity in boys declines slowly (7.6 -> 4.8 cm/yr smoothed, ages 3-9) rather than being flat; 'nearly steady' holds only relative to infancy and adolescence." . "Cross-sectional data: population medians attenuate and spread the adolescent spurt; in girls the spurt (peak 10.9 yr) is so attenuated that its excess over childhood velocity (0.9 cm/yr) has a bootstrap CI including 0, so criterion (c) fails for girls even though the pattern is visible. Thompson's sentence is about boys, which is why the verdict is scored on boys." . "Deviation from pre-registration: the pre-registered 'weighted smoothing spline' derivative on monthly medians was numerically unstable (adjacent-year smoothed velocities swung between 2.7 and 13.8 cm/yr), so the smoothed velocity uses a fixed-bandwidth (SD 0.5 yr) local-linear Gaussian-kernel smoother on the same monthly medians; thresholds were not changed. Results at bandwidths 0.75 and 1.0 yr are reported and give the same sub-criterion outcomes for boys." . "Modern US population (2007-2018), not the 19th-century European boys Thompson drew on; timing of the spurt differs between populations but the sequence of phases is what was tested." . "Pre-registered, evaluated on boys (the subject of Thompson's sentence) with girls reported alongside. Velocities come from pooled NHANES 2007-2018 measured stature (recumbent length < 24 months, standing height >= 24 months) using exam age in months. Model-free yearly velocity v(a) = median(height | age in [a+1,a+2) yr) - median(height | age in [a,a+1) yr), a = 0..18; smoothed velocity = derivative of a weighted smoothing spline through monthly medians. All four must hold for boys: (a) infancy: v(0) >= 2 x mean of v(4..9); (b) childhood steadiness: coefficient of variation of the smoothed velocity over ages 4-10 < 0.15 and max-min of v(4..9) <= 2.0 cm/yr; (c) adolescent quickening: max smoothed velocity in ages 10-17 >= 1.2 x the minimum smoothed velocity in ages 6-10, and the largest v(a) for a in 10..16 exceeds mean v(4..9) by >= 1.0 cm/yr with a 95% bootstrap CI (1000 resamples of individuals) excluding 0; (d) arrest: v(18) (18.5 -> 19.5 yr) < 1.5 cm/yr and < one third of mean v(4..9)." . "Refutes if (c) fails in boys, i.e. the adolescent peak does not exceed childhood velocity (bootstrap CI of the excess includes 0 or the smoothed ratio < 1.0), or if (a) fails (infancy velocity not above childhood velocity). Any other combination of failed sub-criteria (e.g. an irregular childhood plateau or incomplete arrest by 19.5 yr) gives 'mixed'." . "0.8"^^ . . . "supports" . "The population median stature-velocity curve from birth to 20 years shows, in order: a very high velocity in infancy, a lower and nearly steady velocity through childhood, a distinct rise (adolescent spurt) in the early-to-mid teens, and a fall toward zero by the end of the teens." . "Downloaded NHANES DEMO and BMX files for the six 2007-2018 cycles, merged on SEQN, kept examined persons under 20 years with a stature measurement (recumbent length below 24 months, standing height from 24 months; n = 23551, boys 12038, girls 11513). For each sex, median stature by completed year of age gave the model-free yearly velocity v(a); a local-linear Gaussian-kernel smoother (SD 0.5 yr; 0.75 and 1.0 yr as sensitivity) on monthly medians gave the smoothed velocity curve on a monthly grid. Bootstrap over individuals (1000 resamples) gave 95% CIs for the infancy velocity, the adolescent excess over childhood velocity and the velocity at 18.5-19.5 yr. The four pre-registered sub-criteria were evaluated on boys and reported for girls." . "Pre-registered, evaluated on boys (the subject of Thompson's sentence) with girls reported alongside. Velocities come from pooled NHANES 2007-2018 measured stature (recumbent length < 24 months, standing height >= 24 months) using exam age in months. Model-free yearly velocity v(a) = median(height | age in [a+1,a+2) yr) - median(height | age in [a,a+1) yr), a = 0..18; smoothed velocity = derivative of a weighted smoothing spline through monthly medians. All four must hold for boys: (a) infancy: v(0) >= 2 x mean of v(4..9); (b) childhood steadiness: coefficient of variation of the smoothed velocity over ages 4-10 < 0.15 and max-min of v(4..9) <= 2.0 cm/yr; (c) adolescent quickening: max smoothed velocity in ages 10-17 >= 1.2 x the minimum smoothed velocity in ages 6-10, and the largest v(a) for a in 10..16 exceeds mean v(4..9) by >= 1.0 cm/yr with a 95% bootstrap CI (1000 resamples of individuals) excluding 0; (d) arrest: v(18) (18.5 -> 19.5 yr) < 1.5 cm/yr and < one third of mean v(4..9)." . "false"^^ . . "sex, age in months at exam (RIDEXAGM; RIDAGEMN for 2007-2010)" . "US Government public domain" . . "Dataset: NHANES Demographic Variables and Sample Weights (DEMO), cycles 2007-2008 to 2017-2018" . . . . . . . . "sha256:02b8e97703d04bc99769c751881bece51ac9ba68cf18a2992a45d3bdfb856147" . "sha256:78c4d4a6e5eafa9da3bbb24a90d140fef5d8c0cda932a7ff8ee609288d4fae07" . "sha256:c0b46e0345ea19404928656277c8b0d10b0cca348a9b2fe4fc3c67e8b7ee73ec" . "sha256:c9297c6c37ae8f78f29be9568fa2a03cf3b112616a39afee04030fc775a66a0d" . "sha256:eaf0525d1952626885af3e935415a1f66ad62c18698080e7354789c125af252d" . "sha256:f8f0cbb3085a323d4cde22349b164878fea1e64dbc404e65b5815c7816b547d7" . "measured standing height (BMXHT), recumbent length (BMXRECUM) and weight (BMXWT)" . "US Government public domain" . . "Dataset: NHANES Body Measures (BMX), cycles 2007-2008 to 2017-2018" . . . . . . . . "sha256:1fa42e07764037674c3f56e1e76f990e10c32245897597de3d731c57b91d869a" . "sha256:4814bfc3047ed400b9d43d285f8c3ea7c940ac6489404a9b699579715d158ec3" . "sha256:8d675e42d8826ac98714b2c3dd4c5138a5e353fb4424f7eff5e6db4a01ce838a" . "sha256:b67b50dbd12c9510b49b6348b595b9fd073c422c0c6eac182f90ac57b491a97d" . "sha256:d31da84e14212b4e58e8340598a5b8e2144fac83333563e966eb1b332e4141d6" . "sha256:fd5e9fc6e6aab0a4aee6e699f51497bbc9b62101f7f43aee924c473e38fd9442" . . "delphisci claim tester (claude-fable-5.1; cursor/claim-tests-thompson-efb7)" . . . "claude-fable-5.1" . . . . . . . . . "2026-11-08T16:24:31.891Z"^^ . . . . . "Claim test of thompson1917-ch03-05: supports, confidence 0.8" . . "RSA" . "MIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEAyFHtPB6p0mCjDozxlOY9CmO3nBm7Te3jFbXyLmM+G1tGe47Da0jYIbhkl9Oo1ZjxmCa3LCwlEqK8gj6hEIcL4ahW30JXh9HycfKVk1bugkzMEV15tJdTzAU3sGMFnbNsn43xY2LFLy45tApkRm5byEiM7oRRrooIdwoQKnP4Wzq/bfAoJ9qvW+MGkelyOTd1jbSt+KfmT7xKmDfWzAHddUlWwMX2QJOLxDaffOteTop6ZRgPiNmPJKA/S62Gac9XtSYT3oDkZ6/nr5cxxNLpOyxTf5FPISiCe11xYO/unYjY+mwnOfA8yhJIyx1l0O9I0vpOpn+E5TvyHJV+oSTrvwIDAQAB" . "reTqSziKmRlTyUJWskVTFtjjLHF17zfAVEwVyw7dF+iaQkfrNbmNu6AWmLdpFyFAvtEKLkKhLL4Gx8yy7rksmgfsynqNUWH9KLw7oFM0gta/7UTRtItfDNQTO3wHjuMzKsCFEfkAce5uqcWOCUgZXzu8qmYikZXMApOpo3tU+jfe80uVsZ2o0KJXOvLyv8Oi2RImzku17J0aCbLfFR1Cnv7lbPbrM6NX+1oVtRbra0wB5dw0vbpuI8k9i/fJLdtLbTOxN1+esJayGrzHLLQjuTO5tNNCBOz8vT3A4xJ8m808tuIEhdX4iscCJBmUcpjENL3mfni5X+6RMp4GLwcFfA==" . . .