External Email - Use Caution
Dear FS List,
(apologies for reposting under different title)
I wanted to check if my GLM design for a surface based analysis is sound. I am new to FS so please don’t skip over details.
I have 3 groups of monozygotic twins: low-risk, high-risk, ill. Twin pairs may be present in the same group (e.g both are in the low-risk) or between groups (one of them in high-risk one in affected). Not all twins have their co-twin in the analysis though.
In order to account for with-pair variance correlation I was wondering if this FSGD example with 9 subjects might do it or if there might be better options.
GroupDescriptorFile 1 Title … Class LR Class HR Class ILL Variables age sex TIV Pair_no1 Pair_no2 Pair_no3 Input subject1 LR 30 M 123456 1 0 0 Input subject2 LR 30 M 123456 1 0 0 Input subject3 LR 23 F 123456 0 0 0 Input subject4 HR 20 F 123456 0 1 0 Input subject5 HR 21 F 123456 0 0 1 Input subject6 HR 22 M 123456 0 0 0 Input subject7 ILL 20 F 123456 0 1 0 Input subject8 ILL 21 F 123456 0 0 1 Input subject9 ILL 35 F 123456 0 0 0
I do have 60+ pairs so it will be a lot of variables. I know that PALM has an options to provide an exchangeability file, but FsPalm seems to perform corrections for multiple comparisons only.
Julian