Reliability and validity of parental measurements of infant size.

This study evaluates the reliability and validity of parental measurements of infant size, using illustrated instructions and simple measuring tools. Following pilot tests, final methods were evaluated on a sample of 28 parents (26 mothers and 2 fathers) of infants from 1 to 6 weeks of age. Parents independently measured twice the infant's head circumference (HC), mid-upper arm circumference (MUAC), abdominal circumference (AC), and recumbent length (RL). Infants were also measured twice by a trained observer. Mean parental measures were correlated (intraclass R) with the observer criterion measures at the levels of 0.81 for RL, 0.70 for AC, 0.80 for MUAC, and 0.94 for HC. Relative differences in the means for measurements obtained by parents and the trained observer were small, ranging from 0.3% for HC and RL, to 3.8% for AC. Intraclass correlations of reliability between the two parental measurements ranged from R = 0.84 for MUAC to 0.96 for RL. Given the reliability and validity of the results, the methods tested yield measurements that are suitable for ranking individuals and for use in group-level analyses, and at least in the case of head circumference, for individual-level analyses.