Answers are grounded in this paper and cite their sources.
A new method of calculation of cross-sectional attained height centiles is proposed. This exploits the variance-stabilizing transformation, which has the form log(a+b height) up to age 11 years. Residuals of transformed height about the expected value for age are antilogged to give a single Gaussian distribution from which centiles can be calculated. In contrast to previous methods the standard errors of the centiles can be estimated without imprecise adjustment for smoothing. The parameters a and b of the transformation do not need precise estimation and the method can be used for population subgroups from which large samples cannot be obtained.
No takes yet. Share an insight, caveat, or question.
S Chinn (1992) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: