Adjustment of Age-Related Height Shrinkage in Archival Data for Chinese: A Post HOC Longitudinal Survey
Stephen L. Morgan
University of Nottingham - School of Contemporary Chinese Studies; University of Nottingham Ningbo China, Faculty of Social Sciences; University of Melbourne - School of Historical Studies
September 12, 2009
Height data are a useful and concise summary measure of human welfare for historical populations in absence of conventional economical data. Most historical studies use the final attained height of adults aged between about 20-23 and 49 years on the premise that younger subjects were still growing and older subjects had begun to shrink. Data outside this range are discarded. For many studies the data lost is small and of little consequence for the study. However, where the sample includes many people older than 50 years the exclusion of these may make analysis impractical because of the resulting small sample size. Several studies have used a variety of approaches adjust height-for-age of older subjects to estimate the original attained height before next estimating the secular trend in heights. These adjustments are based on studies of the aging of European-origin populations, which may not fit the pattern observed in other human populations, such as the Chinese. In this paper I use data for nineteenth-century born Chinese immigrants to Australia whose heights were recorded repeatedly to simulate a longitudinal age-related height shrinkage study. The estimates of shrinkage are compared with estimates from other studies and applied to other archive-derived height data for Chinese to examine the reliability of adjusted height estimates in calculating secular trends in height, and in turn making inferences about their welfare.
Number of Pages in PDF File: 22
Keywords: China, human welfare, height, age-related shrinkage, measurement
JEL Classification: C23, C89, I31, N30, N35
Date posted: December 18, 2009
© 2016 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollobot1 in 2.438 seconds