Back to PLS HelpBrain Scores and Changes in Analysis
Posted on 07/13/12 11:21:03
Number of posts: 1
Hi there
I have two questions. First, is there a way to get an output of the numerical brain score values for each participant for significant LV's? Can this also be done for specific conditions as well?
Secondly, we ran an event related PLS analysis 3 times. The first run included all conditions, the second had some conditions removed and the third had the deleted conditions re-entered. The onsets were saved from the first run in a text file which were copy and pasted into the third analysis. However, the p values, brain score plots, etc are fairly different from the original analysis. We decided to keep the datamats and session profiles from the second analysis with deleted conditions. When we recreated the datamats for the third analysis, we renamed the session profiles from the second run and saved them in another folder before editing the file. After it was saved, the conditions were added back in and the onsets copy and pasted from our text file and finally created the datamats. I'm wondering if the difference in values from the first and last runs can be due to how we recreated the datamats.
Any help would be greatly appreciated.
Replies:
Untitled Post
I'm Online
Posted on 07/13/12 11:38:24
Number of posts: 291
1. load result file. brain scores stored in b_scores variable. find significant LV, and take the corresponding column. e.g. brain_scores_LV1 = b_scores(:,1). the vector is in the order of "subject in condition in group". i.e. brain_scores_LV1(1) is grp1 cond1, subj1, and brain_scores_LV1(2) is grp1 cond1, subj2.
2. p value is calculated from bootstrap ratio, so it will be different. brain scores should be exactly the same. it could be caused by some mistake you made when delete/add back conditions. if you want me to take a look, please upload all data onto somewhere that i can have access, and post the link. please do not attach anything to email.
Brain score exports
Posted on 08/09/12 12:53:36
Number of posts: 2
Hi Jimmy
I have had some trouble exporting specific brain score values myself. I've tried to figure out your instructions but they aren't particularly clear. I am using MatLab if that makes any difference. I have loaded the result file and selected the significant LV that I would like to get the brain scores from but I don't know where to find this "b score variable" or this corresponding column you mention. Further clarification would be greatly appreciated.
Untitled Post
I'm Online
Posted on 08/09/12 13:15:05
Number of posts: 229
Hi.. here is what the variables list looks like in a typical *fMRI_result.mat file
ContrastFile 1x4 8 char
SessionProfiles 1x1 1432 cell
b_scores 15x3 180 single
boot_result 0x0 0 double
..snip..
You want the b_scores variable, as Jimmy specified. Column1 contains the brain scores for LV1. In this case, there are three conditions, with 5 subjects, so the order of the scores in the first column will be C1S1-S5, C2S1-S5, C3S1-S5. So, to continue Jimmy's instructions, in your new variable, brain_scores_LV1(6) will be Condition2,Subj1. to repeat: Subjects within Condition within Group. hope this helps nancy
ContrastFile 1x4 8 char
SessionProfiles 1x1 1432 cell
b_scores 15x3 180 single
boot_result 0x0 0 double
..snip..
You want the b_scores variable, as Jimmy specified. Column1 contains the brain scores for LV1. In this case, there are three conditions, with 5 subjects, so the order of the scores in the first column will be C1S1-S5, C2S1-S5, C3S1-S5. So, to continue Jimmy's instructions, in your new variable, brain_scores_LV1(6) will be Condition2,Subj1. to repeat: Subjects within Condition within Group. hope this helps nancy
Untitled Post
Posted on 08/09/12 14:12:52
Number of posts: 2
Yes that clarifies everything, thank you!
Login to reply to this topic.
