如何进行协方差矩阵的相等性的Boxs M test?

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

判别分析中需要首先检验因变量组协方差矩阵的相等性，请问如何进行协方差矩阵的相等性的Box's M test?我查了一下帮助没有相关的检验说明哈…… 在网上也看了一下还是没有发现……

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：协方差矩阵 test Boxs Est box test 矩阵 协方差 相等 Boxs

同问

Box’s M statistic is applied
M =
 
X
l
(nl − l)
!
ln

Spooled
− X
l
((nl − 1) ln |Sl|)
where nl is the sample size for the lth group, Sl is the lth group sample
covariance matrix and Spooled is the pooled sample covariance matrix
given by
Spooled =
P 1
l(nl − 1)
((n1 − 1)S1 + (n2 − 1)S2 + · · · + (ng − 1)Sg)
If the null hypothesis is true, the individual sample covariance matrices
are not expected to differ too much and, consequently, do not differ too
much from the pooled covariance matrix. Thus, Box’s M statistic will
be small.
Set
u =
 
X
l
1
nl − 1 −
P 1
l(nl − 1)
!
2p2 + 3p − 1
6(p + 1)(g − 1)

where p is the number of variables. Then
C = (1 − u)M
has an approximate 2 distribution with
v =
1
2
p(p + 1)(g − 1)
degrees of freedom. At significance level , reject H0 if C > 2
v().
Box’s 2 approximation works well if each nl exceeds 20 and if p and g
do not exceed 5.

Hi, papasuncn, thanks for your detailed explanation but can you  offer  me  an  ado file or something else using stata. I know that The test can conduct using spss, or sas macro, but  how can i  accomplish this employing stata . By the way, the display format of your reply is awful and disordered on my screen.

[此贴子已经被作者于2008-10-5 22:50:11编辑过]

同问 啊 同问啊