Submitted by liwenfu on Tue, 05/28/2013 - 16:20
大家好,有个问题请教下:
我在试验中收集有午睡习惯的被试每周的午睡天数,有4 5 6 7共四个选项,如果我想找到午睡习惯可能影响到的脑区(使用VBM数据),应该是使用multiple regression 还是one-way ANOVA between subjects哪?如果使用one-way ANOVA between subjects的话,contrast vector应该怎么设置才能满足我的目标那?
谢谢!
Submitted by YAN Chao-Gan on Tue, 06/04/2013 - 06:56 Permalink
Re: 这种情况应该用multiple regression 还是one-way ANOVA between subjects
根据你的实验设计,可以采用one-way ANOVA within subjects.
当然,regression可以实现所有这些model,但比较复杂。
Submitted by liwenfu on Tue, 06/04/2013 - 13:53 Permalink
Re: 这种情况应该用multiple regression 还是one-way ANOVA between subjects
哦,好的,如果用这种one-way ANOVA within subjects的话,我后面的contrast应该怎么设置那?
Submitted by YAN Chao-Gan on Tue, 06/11/2013 - 23:42 Permalink
Re: 这种情况应该用multiple regression 还是one-way ANOVA between subjects
我记得用SPM的话,会自动给你设好contrast,可以参考一下manual。
Submitted by ZHANG_RESTadmin on Mon, 07/08/2013 - 13:55 Permalink
Re: 这种情况应该用multiple regression 还是one-way ANOVA between subjects
可以用multiple regression试试看。把VBM的结果(如灰质体积)和午睡天数across subject做regression。
contrast可以设1 即可看到正相关的,-1是负相关的。
Submitted by xuqiang9042 on Thu, 10/03/2013 - 15:52 Permalink
Re: 这种情况应该用multiple regression 还是one-way ANOVA between subjects
across subject是什么意思?
Submitted by ZHANG_RESTadmin on Tue, 10/08/2013 - 21:41 Permalink
Re: 这种情况应该用multiple regression 还是one-way ANOVA between subjects
就是在被试上做相关