怎样使用spss进行多元方差分析？

[rainbow]

请教好心人怎样使用spss进行多元方差分析？有以下两组数据甲：A60 B31 C3 D28 E32 F5 G5 乙：A33 B20 C5 D16 E16 F0 G4 以上是第一个问题。 还有一个问题如这样两组数据甲：A 46586 ，B 60， 乙：A 46975 ，B33能做一个什么样的检验来说有差异？？谢谢。静待佳音

 

[chrisyang]

回复: 怎样使用spss进行多元方差分析？

请教好心人怎样使用spss进行多元方差分析？有以下两组数据甲：A60 B31 C3 D28 E32 F5 G5 乙：A33 B20 C5 D16 E16 F0 G4 以上是第一个问题。 还有一个问题如这样两组数据甲：A 46586 ，B 60， 乙：A 46975 ，B33能做一个什么样的检验来说有差异？？谢谢。静待佳音

第一部分所提供的信息量不足啊！即使硬着头皮把数据输入SPSS，好像也无法进行下面的运算，因为根据提供的信息根本无法判断哪个是自变量，哪个是因变量，更不知道Rainbow要讨论的什么和什么的关系！
我的愚见而已！

 

[xiaoz]

回复: 怎样使用spss进行多元方差分析？

For question 1: the loglikelihood is 7.411 (Fisher's Exact Test score is 5.641 - as 4 cells of the contingency table has a valure less than 5, the Fisher's Exact test is more reliable), for 6 degrees of freedom: the 2-sided p level is 28.4% (measured by the LL test) or 46.6% (measured by the Fisher's Exact test) - greater than 5%. so you can say the difference in the frequencies between these 7 pairs as a whole is NOT significant statistically.

Question 2: the LL score is 8.171 for one degree of freedom: p = 0.004 (0.4%). the difference is highly significant statistically.

 

[rainbow]

回复: 怎样使用spss进行多元方差分析？

thanks for Dr.xiao . for the first question,i have made a paired sample t-test and p=0.036 which is<0.05 ,so there is difference between the the group as a whole.do u agree? now what i want to do is to make a further analysis to check which two paires(eg a=60i n group甲，a=33in group 乙 have a more distinctive difference.
i am abeginer user of spss.so please give me a more detailed analysis procedure.thanks again.

 

[hisahito]

回复: 怎样使用spss进行多元方差分析？

哈哈,问题和回答都很有意思.换一种简单的算法如何?高手快现身!

 

[rainbow]

回复: 怎样使用spss进行多元方差分析？

it is a hot question for me,i am waiting for a better and easy solusiton.