N. Hou (source code)
= N. Hou
{c}
\b[<N. Hou>] coauthored the human <nail (anatomy)> <growth rate> study with <S. Yaemsiri>, <M. M. Slining> and <K. He>.<Growth rate of human fingernails and toenails in healthy American young adults by S. Yaemsiri and others>[Primary byline, 2010]{ref} Its marked-nail method compared displacement over recorded follow-up intervals. The published contribution is collaborative; the byline alone does not identify which author performed a particular measurement or analysis.
A useful statistical issue is that several nails from one person do not necessarily supply independent information about population growth. Shared physiology and exposure can create <correlation> among that person's measurements. An original equal-cluster model makes the consequence explicit. Suppose there are $n$ people, each with $m$ measurements of common variance $\sigma^2$, within-person correlation $\rho$ and independence between people. Then
$$
\operatorname{Var}(\overline X)=\frac{\sigma^2}{nm}\bigl(1+(m-1)\rho\bigr).
$$
With stipulated $n=22$, $m=10$ and $\rho=0.4$, the <design effect> is $4.6$. Treating all 220 measurements as independent would understate the <standard error> by a factor $\sqrt{4.6}\approx2.14$. This is an original explanatory model, not a reconstruction of the study's analysis or number of usable nails.
The example explains why participant identity should remain attached to a biological measurement. More measurements can improve precision, but their covariance determines how much information they add. Preserving that structure separates a quantitative contribution to a <longitudinal study> from an unsupported claim of hundreds of independent people.
Back to article page
