非参数检验——游程检验
对以下两个样本进行游程检验,minitab下stat——Nonparametrics——runs test,请问minitab是用什么原理算的?
A:13.3 14.6 13.6 17.2 14.1 10.6 15.9 14.7 14.2 18
B:14.1 15.1 9.9 14.5 17.9 16.1 16.8 15.1 13.2
MINITAB的结果如下:
Runs test for C1
Runs above and below K = 14.6789
The observed number of runs = 11
The expected number of runs = 10.4737
9 observations above K, 10 below
* N is small, so the following approximation may be invalid.
P-value = 0.803
minitab的算法是游程数法呢还是游程长度法?或则其他
A:13.3 14.6 13.6 17.2 14.1 10.6 15.9 14.7 14.2 18
B:14.1 15.1 9.9 14.5 17.9 16.1 16.8 15.1 13.2
MINITAB的结果如下:
Runs test for C1
Runs above and below K = 14.6789
The observed number of runs = 11
The expected number of runs = 10.4737
9 observations above K, 10 below
* N is small, so the following approximation may be invalid.
P-value = 0.803
minitab的算法是游程数法呢还是游程长度法?或则其他
没有找到相关结果
已邀请:
4 个回复
lwxg (威望:0)
赞同来自: Neo_944
具体计算过程如下:
1.先把A/B两列数合在一起求出平均值为14.67894737
2.然后按A B数据依次和K进行比较,小的转为-,大的转为+
结果为---+--++-+-+--++++-
3.计算游程数R为11
4.E(R)=1+2n1n2/(n1+n2)=10.47368421
5.σ(R)=sqrt(2n1n2/(n1+n2)^2/(n1+n2-1))=2.111831858
6.Z=(R-E(R))/σ(R)=0.249222393
7.对应的p值为0.803188862
不过以上的计算过程实际上是对应大样本场合(n1,n2或两者都大于20,接近正态分布)
所以本例中并不能保证结果的正确性,所以结果中加了备注。