Kaarma Kh T
Arkh Anat Gistol Embriol. 1983 Sep;85(9):67-70.
The investigation has been performed in 670 young (18--20 years-old) Estonian women. They are healthy students from Tartu University having a normal menstrual cycle, or with some slight deviations in it. In every woman 56 body dimensions, including the external obstetric pelvic ones, are analysed. The measurement technique in general corresponds to the classical techniques suggested by V. V. Bunak (1941) and R. Martin (1928). The linear correlation matrix of the body dimensions statistically significant positive correlations have been revealed between all the signs studied, the body mass and length possessing, the highest correlations. When the influence of the body mass and length is eliminated, the usual correlative correlation system is destroyed. Instead of the previous system, where all the signs are statistically positively significantly connected, the only connections remain that include 46% of negative connections and 30% of statistically significant, but very weak connections (the critical value r-0,088). The factor matrix is calculated on the base of the correlation matrix by means of the main components method. For its rotation the so called method of varimax is used. Three, five and seven factors are rotated (the parts of the initial signs in the total variance are 56, 65, 72%, respectively). The factor analysis also supports the main role of the body mass and length. When the number of the factors and described by them the part of the total variance increases, only some peculiar factors connected with subsystems are added. The factor analysis demonstrates that, besides the body mass and length, there are no other equally important body dimensions which could be used for description of the woman's body structure.
该调查在670名爱沙尼亚年轻女性（18 - 20岁）中进行。她们是塔尔图大学的健康学生，月经周期正常，或有一些轻微偏差。对每位女性的56项身体尺寸进行分析，包括产科外部骨盆尺寸。测量技术总体上符合V. V. 布纳克（1941年）和R. 马丁（1928年）提出的经典技术。身体尺寸的线性相关矩阵显示，在所研究的所有体征之间存在统计学上显著的正相关，体重和身高的相关性最高。当消除体重和身高的影响时，通常的相关相关系统被破坏。取代之前所有体征在统计学上呈显著正相关的系统，仅剩下包括46%负相关和30%统计学上显著但非常弱的相关性（临界值r - 0.088）的联系。因子矩阵是基于相关矩阵通过主成分法计算得出的。对于其旋转，使用所谓的方差最大化方法。对三个、五个和七个因子进行旋转（初始体征在总方差中的占比分别为56%、65%、72%）。因子分析也支持体重和身高的主要作用。当因子数量及其所描述的总方差部分增加时，仅添加了一些与子系统相关的特殊因子。因子分析表明，除体重和身高外，没有其他同等重要的身体尺寸可用于描述女性身体结构。
