The linearity assumption surrounding a quantitative variable in common
practice may be a reasonable approximation especially when the variable
is confined within a narrow range, but can be inappropriate under some
circumstances when the variable’s effect is non-monotonic or tortuous.
As a more flexible and adaptive approach, 3dMSS adopts a principle of
learning from the data in the presence of uncertainty to dissolve the
problematic aspects of conventional polynomial fitting. It offers a powerful
analytical tool for population-level neuroimaging data analysis that involves
one or more quantitative predictors. Check out the examples in the help
document of 3dMSS. More functionality and use examples will be added
This preprint covers the underlying theory and a neuroimaging application:
Just cross-referencing some information here about 3dMSS from another thread (where @Gang was helping me out with 3dLMEr), in case someone is using 3dMSS and has the same questions:
“My [original 3dLMEr] model is for a longitudinal study with two groups and [multiple] measurements (…)”
Your current model [in 3dLMEr] assumes a linear relationship of Time. In case nonlinearity is of interest, consider 3dMSS.
“For 3dLMEr so far, my next steps are (…) 3dFWHMx and (…) 3dClustsim for cluster correction - is there a way to do this with 3dMSS?”
No, but you don’t have to adopt a harsh thresholding approach, and instead show the results in a gradation fashion.
“Would my model [in 3dLMEr, with Time=0-6: Time*Group+Age+Gender+(Time|Subj) ] translate to the following in 3dMSS?: -mrr ‘s(Time)+s(Time,by=Group)+(Age)+(Gender)’”
The above specification assumes that there are 10 or more time points. With 7 time points, try
“According to the 3dMSS [help file], I would create a pred.txt that include label, Time and Group (binarized).
Are Age and Gender okay to only be in the dataTable (not the pred.txt), and would Gender have to binarized for 3dMSS?”
All predictors are required to be specified for prediction (in the file pred.txt). Dummy coding is needed when you want to account for the interaction with >nonlinearity (e.g., Group in your case). However, Gender does not have to be dummy-coded.