2. Yes, assuming 1 is used to indicate the drug condition and 0 for the other condition. They effects do not need to be symmetric.
3. Yes - the interaction looks correct. For an interaction effect, using 1/-1 rather than 0/1 can assist with interpretation, but the p-values will be the same in both cases.
Yes - adding an interaction could reduce the ability to detect main effects (it will also increase the number of multiple comparisons), although this characteristic is not specific to the NBS and the same would hold true for any kind of regression model.
Originally posted by Max Kathofer :
Dear Andrew,
Thanks for the quick response!
- Thanks for catching the missing 1 in the permutation vector (was a copy error on my part)
- If drug contrast tests [0 1 0 0 ..] or [0 -1 0 0 ..]. the interpretation is: which network cluster is increased during drug over placebo and the other one would be which network is more active during placebo over drug right? But they do not need to yield symmetrical effects right? So if I find a cluster for contrast 1, I do not need to find a cluster for -1, right?
- The interaction between drug and high low would be coded like this I guess:
Sub1PH 0 1 0 1 0 0 ...
Sub1PL 0 0 0 1 0 0 ...
Sub1DH 1 1 1 1 0 0 ...
Sub1DL 1 0 0 1 0 0 ...
(First column main effect drug, second main effect highvlow, third column interaction drugxHigh)
The interpretation of this effect would be: is there a network that is more activated during drug and high conditions compared to everything else, right? [contrast: [0, 0, 1, 0 , 0 ..]
Last question, does the interaction reduce the power to detect main effects? E.g. normally using behavioral data I would test which regressors to include and then use the model that performs best (BIC or AIC). I usually do this since interaction terms use a lot of df and thus - if not beneficial for the glm - can result in decreased power to detect other effects right? Is there a potential with NBS in this regard as well? Can I test e.g. wether including this effect is necessary?
Thanks so much!
Cheers,
Max
Threaded View
Title | Author | Date |
---|---|---|
David de Wide | Aug 3, 2017 | |
Max Kathofer | Oct 31, 2024 | |
Andrew Zalesky | Nov 1, 2024 | |
Max Kathofer | Nov 1, 2024 | |
Andrew Zalesky | Nov 1, 2024 | |
Max Kathofer | Dec 2, 2024 | |
Andrew Zalesky | Dec 6, 2024 | |
Max Kathofer | Dec 9, 2024 | |
Andrew Zalesky | Dec 10, 2024 | |
yue zhang | Jan 11, 2020 | |
Andrew Zalesky | Aug 5, 2017 | |
Liam Nestor | Sep 27, 2019 | |
Andrew Zalesky | Sep 28, 2019 | |
Liam Nestor | Sep 28, 2019 | |
Andrew Zalesky | Sep 28, 2019 | |
Liam Nestor | Sep 30, 2019 | |
Andrew Zalesky | Oct 1, 2019 | |
Liam Nestor | Oct 1, 2019 | |
David de Wide | Aug 6, 2017 | |
Andrew Zalesky | Aug 7, 2017 | |
David de Wide | Aug 7, 2017 | |
Andrew Zalesky | Aug 7, 2017 | |
David de Wide | Aug 11, 2017 | |
Andrew Zalesky | Aug 12, 2017 | |
David de Wide | Aug 18, 2017 | |
Andrew Zalesky | Aug 19, 2017 | |