When I run multiblock or behavioral PLS on my ERP dataset, I am seeing some things in the temporal correlation plot that make it seem like the program is mixing up which dimension corresponds to the subjects and which dimension corresponds to the behavioral measures. For instance, with 27 subjects, two behavioral measures, and six conditions, the temporal correlations window shows average plots for only two subjects. When I tried putting in a third behavioral measure, the window showed average plots for three subjects. For other parts of the analysis, the program is seemingly categorizing the data appropriately (e.g. it generates 12 LVs for two behavioral measures x six conditions). Any tips on how to handle this would be much appreciated.
Thanks,
Brandon
PS - I had hoped to include screenshots to demonstrate what I mean, but the VA unfortunately will only let us use the IE browser, which fails when it tries to interact with the pop-up for adding images here. Also they won't let us access image-hosting websites from VA network.
HI Bradon - can you send me some screenshots? I think what you are seeing is okay given that the temporal correlations are computed across subjects, so if you have two behaviors you will two plots per electrode and condition. With three behaviors you will have three temporal correlations plots, etc.
Baycrest is an academic health sciences centre fully affiliated with the University of Toronto
Privacy Statement - Disclaimer - © 1989-2024 BAYCREST HEALTH SCIENCE. ALL RIGHTS RESERVED