Back to PLS HelpPosted on 08/11/08 02:56:42
Number of posts: 60
hello PLS experts,
I have done PLS group analysis analysis(14 subjects) using mean centering PLS.As the result, I got two significant LV at p<0.02. LV3 showed the trend at p<0.92 which definitely cannot be considered as significant compare to LV1 and LV2(but since it shows the accessibility signal I am still interested to report it).Furthermore, LV1 and LV2 are accounted for 35.36% and 32.51% of the variances in data respectively while LV3 is accounted for 19.24% of the variances in data.(all the three LV are located at ftp://ftp.physiol.umu.se/fysiologi/out/Alireza).
Now I
cannot understand why the material-specific availability signal(reflected in
LV1) would be stronger then the material specific accessibility(reflected in
LV3)????
Just
for your clarification, the experiment consisted of three separate phase,
encoding, cue recall and recognition. During encoding subjects were presented
to pictures and sounds of easily identified objects and subjects were
instructed to memorize the sounds and pictures for the tests the following day.
during the test, for each word, subjects were supposed to indicate if they
remembered having encoded the object of the word by pressing a button. if they
remember, it means the sound or picture are accessible[two related condition:
sound(Rc+Rn+) and picture(Rc+Rn+)] but if they don't remember, ,a clue will be
provided, if they remember this time, one can say that sound or pictures were
available but not accessible(two related condition: sound(Rc-Rn+) and
picture(Rc-Rn+)) otherwise the item were forgotten (related condition:
pooled(Rc-Rn-))
The second question is related to LV2.Is there any way that one can figure out if sound(Rc+Rn+)[the blue line ] is significantly different from picture(Rc+Rn+)[the green line] ? Off course one can suggest BOLDbar plot for each region of interest in LV2 to figure out if the differences between desired conditions are significant.
Any help will be appreciate.
/Alireza
Replies:
Posted on 08/11/08 04:54:46
Number of posts: 394
Now I cannot understand why the material-specific availability signal(reflected inThe easy answer is that the data are the data. PLS does have the tendency to provide you with a more "honest" answer about whats in your data in the sense of ordering effects in terms of their strength (by the way, the percentages you quote are not % variance of the data, but rather %covariance between the data and the design - this is a very important distinction). In terms of your design, it would seem that if material was accessible, it would be a more fluid cognitive process requiring less "effort" than if it was available but not accessible.
LV1) would be stronger then the material specific accessibility(reflected in
LV3)????
Is there any way that one can figure out ifThe newest version of PLS provides a bar plot of brain scores (not temporal brain scores) with confidence intervals.
sound(Rc+Rn+)[the blue line ] is significantly different from
picture(Rc+Rn+)[the green line] ? Off course one can suggest BOLDbar plot
for each region of interest in LV2 to figure out if the differences between
desired conditions are significant.Any help will be appreciate.
Login to reply to this topic.
