Follow up to poor individual fit question

[Dina Dajani]

As a follow up to my question about nonconvergence/unstable solutions for some individuals with gimme:

If I retain this solution, should I remove subjects with nonconvergence/unstable solutions from the sim_matrix for use in perturbR?

[Katie Gates]

I would not, although for those with nonconvergence it is probably a line of zeros. 

[Dina Dajani]

Thanks Katie,

The analysis I ran was for a task-based analysis, and I noticed one of the group level paths was [Task_trial] --> [brain region]. I'm not exactly sure how to interpret that since it's not logical that brain activity in a particular region would "cause" a certain task trial to occur. Thanks for your advice!

Dina

--
You received this message because you are subscribed to a topic in the Google Groups "gimme-r" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/gimme-r/sKA5lzLIKas/unsubscribe.
To unsubscribe from this group and all its topics, send an email to gimme-r+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/gimme-r/9f5d1b86-5064-4d65-a0b0-2ea5b8e2707dn%40googlegroups.com.

[Katie Gates]

Hi Dina, 

If it was [Task_trial] --> [brain region], then this does make sense - increases in Task relate to change in brain region. The reverse would not make sense as that would suggest the brain activity dictated the task (which might make sense for bio-feedback designs, but it doesn't sound like yours is one of these). 

if it was the reverse direction, then I'd recommend entering in the Task_trial as an exogenous variable using teh "exogenous = " argument. Here you can indicate which variables (such as Task_trial) should be considered to predict other variables but not be predicted themselves. 

The arguments would be: 

...

exogenous = "Task_trial", 

...

for one variable and like this for >1 variable: 

...

exogenous = c("Task_trial1", "Task_trial2")

....

Katie