Brain scores
Posted on 02/27/09 04:49:43
Number of posts:
Hello,
I am wondering if the columns in b_scores stands for an LV and rows stands for subject-in-condition-in-group (g1 c1 s1, g1 c1 s2, g1 c1 s3)?
Now my question is if the values of b_scores are the average of each condition with-in a subject across the lags?in that case the average of Condition 1 with-in subject 1 across let's say 8 lags should be equal with b-scores(1,1) which is not true,can you tell me why?
My second question would be related to bring out the most dominant effects of the LVs to facilitate the interpretation.I know that you have provide something like task PLS with CI (which can also get plotted by adjusted value) to provide help for interpretation,but I am thinking to perform an ANOVA (perhaps 2 way repeated-measure ANOVA) on the brain scores (b-scores) to facilitate the intepretation.Do you think it might be fair to perform such a kind of analysis?
Thanks,
D
Untitled Post
I'm Online
jshen
Posted on 02/27/09 10:45:38
Number of posts: 291
here's the answer to your 1st question:
that's right. the columns in b_scores stands for an LV and rows stands for subject-in-condition-in-group (g1 c1 s1, g1 c1 s2, g1 c1 s3).
the values of b_scores are neither the average of each condition within a subject, or across the lags. they are the projection of your brain salience back onto your original datamat.
here's the answer to your 2nd question:
i don't think it is necessary to perform ANOVA, since PLS has used statistic information of the whole brain for multivariant analysis.
brain scores
Posted on 02/27/09 12:21:22
Number of posts:
Thank you jimmy,
so you beleive that providing brain scores with CI (with out any statistical value) figure would be the best option to report, in confiming the interpretation.I am looking for a good solution to clarify our readers that this interpretation inlines with the statistics.
Thanks again!
D