Growth Study for the Traditional Society without Registration of Birth Date

研究发育过程和成人体质的比较，对研究人类的适应和进化具有重要意义。人类成长的研究对象主要是欧洲、美国和日本。然而，在传统社会中，尽管生长研究对人类变异的研究必不可少，但由于缺乏可靠的出生日期记录，生长研究做得很少。我调查了巴布亚新几内亚各种人口的人类生态研究，其中大多数缺乏可靠的出生日期登记。因此，在本研究中，我对这类传统社会的成长研究进行了以下几点考察。在传统社会中，出生日期没有正确的登记，而一个社区的成员非常清楚他们的出生顺序。根据社区所有成员的出生顺序信息，可以估计出他们的年龄，并绘制出平均生长曲线。我将年龄估计应用于Biami种群的数据，并检查了该方法的问题。提出了一种新的生长曲线拟合方法，以满足对无完整出生记录的幼年至成年数据的应用要求。引入两个逻辑函数来描述青春期前和青春期的生长曲线，从前者平滑地切换到后者，这为两次测量的Gidra儿童的身高增长提供了良好的模型。对出生日期比较清楚的马努斯人进行生长研究，虽然没有年龄问题，但平均生长曲线的估计存在样本量小的问题。我对1994年马努斯调查数据采用了平滑曲线拟合的方法。

Investigation of growth process, as well as comparison of adult body physique, is very important for the study of human adaptation and evolution. Human growth has been studied mainly on the subjects of Europe, U.S.A.and Japan. In the traditional society, however, growth study has litte done for the lack of reliable registeration of birth date, although it is indispensable for the study of human variation.I have investigated the human ecological study for various populations in Papua New Guinea, most of which lack the reliable registration of birth date. Thus, in this study, I examined the growth study on such traditional societies for the following points.Birth date is not correctly registered in the traditional societies, while the members of a community knows very well their birth order. It is possible to estimate their age from the information of birth order of all members of community and to draw the mean growth curve. I applied the age estimation to the data of the Biami population and examined the problems of the method.A new method for fitting growth curve is developed to meet the requirement of application to the data from early age to maturity without complete birth records. Two logistic functions are introduced to describe the prepubertal and adolescent growth curves, smoothly switching from the former to the latter, which provide good models of height growth in the Gidra children measured twice.Though no problem of age on the growth study of the Manus people who know their birth date very well, there exist a problem of small sample size for the estimation of mean growth curve. I adapted a method of smoothing curve fitting to the Manus survey data in 1994.