Back to PLS HelpPosted on 04/27/16 18:35:02
Number of posts: 26
Hello,
Is it a legitimate path to initially do a whole brain analysis and identify peaks and then follow up these peaks using seed based PLS? Is this double dipping?
What is the difference between a network identified through whole brain and seed based analysis?
Regards
Andrew
Replies:
Posted on 04/27/16 22:03:03
Number of posts: 394
thats a tough one to answer,Andrew. If the full-brain analysis is a task-PLS, then the questions that lead to seed PLS can be independent. What is the whole-brain analysis you are refering to?
Posted on 04/27/16 22:06:20
Number of posts: 26
The whole-brain analysis was a task-based PLS. The LV identified a couple of interesting peaks associated with two conditions within the task. The follow-up seed-based analysis of the peaks identified some group differences dependent on the condition. I'm just a little unsure of this approach.
Posted on 04/27/16 22:21:10
Number of posts: 394
The whole-brain analysis was a task-based PLS. The LV identified a couple of interesting peaks associated with two conditions within the task. The follow-up seed-based analysis of the peaks identified some group differences dependent on the condition. I'm just a little unsure of this approach.
seed based PLS is a different source of variance than task - so no worries
I'm Online
Posted on 04/28/16 08:10:36
Number of posts: 229
an example of this approach was recently published in JOCN:
http://www.ncbi.nlm.nih.gov/pubmed/25848685
Nancy
Posted on 04/28/16 08:16:36
Number of posts: 26
Thank you, that is a big help
Posted on 05/08/16 09:23:38
Number of posts: 26
seed based PLS is a different source of variance than task - so no worries
Is there a sentence or two (or a reference) I could use to satisfy the reviewers/readers that this approach is justified. Could you briefly explain the differences in variance?
Thanks for all your help.
Andrew
Posted on 05/08/16 11:33:45
Number of posts: 394
Is there a sentence or two (or a reference) I could use to satisfy the reviewers/readers that this approach is justified. Could you briefly explain the differences in variance?
Thanks for all your help.
Andrew
In addition to the paper Nancy posted there is this one:
https://www.dropbox.com/s/jpoor9ingo0cvxg/spcpathanalysis.pdf?dl=0
Login to reply to this topic.
