摘要
成分数据由于受其总和为1这个条件的限制而给数据分析带来不少困难。本文讨论了成分数据通过变换,用多元T^2检验、析因方差分析等方法进行均数假设检验。
There are a number of problems in statis-
tical analysis from the discipline that the sum
of compositions equal 1. In this paper, it is dis-
cusscd that, by means of transformation, T^2-test
and multiple factorial analysis of variance will
be used for mean hypothesis test of composi-
tional data. An example is introduced.
出处
《中国卫生统计》
CSCD
北大核心
1993年第6期17-20,共4页
Chinese Journal of Health Statistics