The indentification and evaluation of measurement variability in the anthropometry of preschool children.

Measurement variability estimates for 18 different anthropometric dimensions were collected within the context of an ongoing longitudinal in vestigation of preschool Guatemalan children. Estimates of total measurement variance, intra-observer variance, and short-term intra-subject variance are presented for each variable. A simple procedure for the evaluation of measurement variance in cross-sectional and longitudinal investigations is described in which the total measurement variance is expressed as a percentage of the appropriate inter-subject variance. This statistic serves as an index of the relative reproducibility of anthropometric variables.