Hiya,
I did an interaction analysis with 3 groups and splitted columns covariate and get some interesting results. However when I plot the data in a spead sheet, ie the input cope vs the covariate, it looks like there are some outliers in the copes of 1 group, actually the group we are interested in. I know feat does some fancy outlier deweighting and includes lower-level variances in the calculation, so I was wondering whether I can display or extract the weighted data that feat computes to display them similarly as described above (cope vs covariate)?
Cheers,
Torsten
|