Back to PLS HelpPosted on 07/28/15 16:44:57
Number of posts: 9
Hi,
I'm currently working with some task and behavioral PLS results and I am unclear as to whether these results should be clusterized. Lee et al. (2012) used a cluster threshold of 100 voxels, but I haven't seen clusterizing done consistently in the literature. I think that the multivariate method of PLS may make this type of correction unnecessary, although I am unsure. Any insights are appreciated.
Thanks,
Kayleigh Ryherd
Replies:
Posted on 07/31/15 09:52:46
Number of posts: 394
Hi,
I'm currently working with some task and behavioral PLS results and I am unclear as to whether these results should be clusterized. Lee et al. (2012) used a cluster threshold of 100 voxels, but I haven't seen clusterizing done consistently in the literature. I think that the multivariate method of PLS may make this type of correction unnecessary, although I am unsure. Any insights are appreciated.
Thanks,
Kayleigh Ryherd
Clustering would be useful for summarizing the pattern within an image, but should not be considered a statistical test. That evaluation is done with the resampling methods at the whole-brain level.
Login to reply to this topic.
