Dear Expert,
I found there were many outlier values in the time series data due to the particular task of naming.When I look for these outlier values by "3dtoutcount -range ", I got two column data.
Is the second column value median+3.5*MAD? what does the first column data mean? Because I want to cencor the outliers, the values in the first column which bigger than the corresponding second column values is culled.Is this reasonable?
While I use the above standard,in some case,almost 1/5 to 1/4 time points are lost. So,whether or not my design still is a block design ? And the same questin,is it reasonable?
Thanks.
