Submitted by chenyuchen1989 on Tue, 07/09/2013 - 14:02 老师好: 我有一个统计的问题。我有两大组,每组内有两组数据,我分别用two sample t-test计算了这两组数据的差异。现在我想计算这两组数据的差异之间的差异,应该用什么方法呢? 还是我一开始统计就应该用其他方法? 请老师指教!谢谢! Log in or register to post comments 9513 reads Submitted by ZHANG_RESTadmin on Tue, 07/09/2013 - 22:13 Permalink Re: 计算差异的差异 我觉得,可以用two way ANOVA. 差异的差异应该是interaction effect 或者,如果每组内的两组数据是被试重测数据的话,可以先做差,然后差异在两组上比较。 Log in or register to post comments Submitted by chenyuchen1989 on Tue, 07/09/2013 - 22:35 Permalink Re: 计算差异的差异 谢谢您的回答! 我两大组是健康组和病人组,然后里面分别是baseline组和follow-up组,我就是想观察健康组的变化与病人组的变化是否有差别?请问这个应该用什么方案做呢?谢谢!! Log in or register to post comments Submitted by ZHANG_RESTadmin on Tue, 07/09/2013 - 23:21 Permalink Re: 计算差异的差异 baseline组和follow up组是否是同一批人?如是的话两者都可,如不是的话可用前者。 Log in or register to post comments Submitted by YAN Chao-Gan on Sat, 07/13/2013 - 05:27 Permalink Re: 计算差异的差异 或者用SPM做ANOVA, within subject factor: baseline vs. follow-up between subject factor: health vs. control 可以参考一下SPM manual. Log in or register to post comments Submitted by chenyuchen1989 on Tue, 07/09/2013 - 23:34 Permalink Re: 计算差异的差异 是同一组人。 您说的后者:先做差,然后差异在两组上比较。在REST的statistical analysis是怎么做的? 是先做two-sample t test,然后得到两个nii,这两个nii之间怎么再做比较呢? 谢谢 Log in or register to post comments Submitted by ZHANG_RESTadmin on Wed, 07/10/2013 - 10:46 Permalink Re: 计算差异的差异 先在每个被试的baseline和followup两幅图上做差,就是后者减前者,没有统计,不需要t检验。 在REST里可以用image calculator来做。 Log in or register to post comments
Submitted by ZHANG_RESTadmin on Tue, 07/09/2013 - 22:13 Permalink Re: 计算差异的差异 我觉得,可以用two way ANOVA. 差异的差异应该是interaction effect 或者,如果每组内的两组数据是被试重测数据的话,可以先做差,然后差异在两组上比较。 Log in or register to post comments
Submitted by chenyuchen1989 on Tue, 07/09/2013 - 22:35 Permalink Re: 计算差异的差异 谢谢您的回答! 我两大组是健康组和病人组,然后里面分别是baseline组和follow-up组,我就是想观察健康组的变化与病人组的变化是否有差别?请问这个应该用什么方案做呢?谢谢!! Log in or register to post comments
Submitted by ZHANG_RESTadmin on Tue, 07/09/2013 - 23:21 Permalink Re: 计算差异的差异 baseline组和follow up组是否是同一批人?如是的话两者都可,如不是的话可用前者。 Log in or register to post comments
Submitted by YAN Chao-Gan on Sat, 07/13/2013 - 05:27 Permalink Re: 计算差异的差异 或者用SPM做ANOVA, within subject factor: baseline vs. follow-up between subject factor: health vs. control 可以参考一下SPM manual. Log in or register to post comments
Submitted by chenyuchen1989 on Tue, 07/09/2013 - 23:34 Permalink Re: 计算差异的差异 是同一组人。 您说的后者:先做差,然后差异在两组上比较。在REST的statistical analysis是怎么做的? 是先做two-sample t test,然后得到两个nii,这两个nii之间怎么再做比较呢? 谢谢 Log in or register to post comments
Submitted by ZHANG_RESTadmin on Wed, 07/10/2013 - 10:46 Permalink Re: 计算差异的差异 先在每个被试的baseline和followup两幅图上做差,就是后者减前者,没有统计,不需要t检验。 在REST里可以用image calculator来做。 Log in or register to post comments
Submitted by ZHANG_RESTadmin on Tue, 07/09/2013 - 22:13 Permalink
Re: 计算差异的差异
我觉得,可以用two way ANOVA. 差异的差异应该是interaction effect
或者,如果每组内的两组数据是被试重测数据的话,可以先做差,然后差异在两组上比较。
Submitted by chenyuchen1989 on Tue, 07/09/2013 - 22:35 Permalink
Re: 计算差异的差异
谢谢您的回答! 我两大组是健康组和病人组,然后里面分别是baseline组和follow-up组,我就是想观察健康组的变化与病人组的变化是否有差别?请问这个应该用什么方案做呢?谢谢!!
Submitted by ZHANG_RESTadmin on Tue, 07/09/2013 - 23:21 Permalink
Re: 计算差异的差异
baseline组和follow up组是否是同一批人?如是的话两者都可,如不是的话可用前者。
Submitted by YAN Chao-Gan on Sat, 07/13/2013 - 05:27 Permalink
Re: 计算差异的差异
或者用SPM做ANOVA,
within subject factor: baseline vs. follow-up
between subject factor: health vs. control
可以参考一下SPM manual.
Submitted by chenyuchen1989 on Tue, 07/09/2013 - 23:34 Permalink
Re: 计算差异的差异
是同一组人。 您说的后者:先做差,然后差异在两组上比较。在REST的statistical analysis是怎么做的? 是先做two-sample t test,然后得到两个nii,这两个nii之间怎么再做比较呢?
谢谢
Submitted by ZHANG_RESTadmin on Wed, 07/10/2013 - 10:46 Permalink
Re: 计算差异的差异
先在每个被试的baseline和followup两幅图上做差,就是后者减前者,没有统计,不需要t检验。
在REST里可以用image calculator来做。