Posted by jian on March 13, 2001 at 16:24:54:
In Reply to: Re: t-test on time series data posted by B. Douglas Ward on March 12, 2001 at 17:21:11:
Hi. thank you for your response. Now I have some naive questions.
1. what you suggest to do is to do 3dTcat with -rlt++ to concatenate the different runs within one subject first, then do FIM+. I have to do % change in FIM+ because for latter ttest accross different subjects, I need to normalize the intensity of runs of different subjects into same scale (maybe in this case, % from average is more reasonable), after this step and tlrc, I can use ttest to analyze across subjects, If I have only one data set, then I use -base1 0.0. Is my understanding correct?
2. According to the frequently asked question 41, my personal understanding is like this: in FIM+, Fit Coef and best Index is used to find the best fit between the ideal wave and real data, does they do detrend for the functional runs by defalt, togather with % change, above three is kind of intensity normalization, then people can do corrlation or other statistics, do I need to do Baseline to adjust baseline in FIM+, is my thoughts correct?
Many thanks