Back to PLS Help
masking
Posted on 11/25/08 05:56:10
Number of posts:
Hello,
Actually,what I am trying to do, is masking the first latetnt variable (LV1) ,which was obtained from Task PLS analysis implemented on FMRI data, with encoding mask(obtain from the second level analysis using SPM using encoded data during scaning),but the problem is related to the time points. Do I need to add all the clusters for each lags(1-8) and then make a mask on the final added clusters?It will certainly change the bootstrap ratio since the final bootstrap ratio of particular cluster would be the result of summation of each bootstrap ratio of that particular cluster for each individual lag.For example if a cluster (a,b,c) has a bootstrap ration: 1,2,3,4,5,6,7,8 during each lag respectively,thes result of adding with produce the same cluster but with increased bootstrap ratio=1+2+3+4+5+6+7+8=36.
Is that a better way of doing this for masking?I know that there is an option for masking while trying to construct the datamat but that will certainly affect the LVs,so What I actually want to do is simply masking the first significant LV with encoding mask obtained from the second level analysis using SPM.
Thanks
David
Actually,what I am trying to do, is masking the first latetnt variable (LV1) ,which was obtained from Task PLS analysis implemented on FMRI data, with encoding mask(obtain from the second level analysis using SPM using encoded data during scaning),but the problem is related to the time points. Do I need to add all the clusters for each lags(1-8) and then make a mask on the final added clusters?It will certainly change the bootstrap ratio since the final bootstrap ratio of particular cluster would be the result of summation of each bootstrap ratio of that particular cluster for each individual lag.For example if a cluster (a,b,c) has a bootstrap ration: 1,2,3,4,5,6,7,8 during each lag respectively,thes result of adding with produce the same cluster but with increased bootstrap ratio=1+2+3+4+5+6+7+8=36.
Is that a better way of doing this for masking?I know that there is an option for masking while trying to construct the datamat but that will certainly affect the LVs,so What I actually want to do is simply masking the first significant LV with encoding mask obtained from the second level analysis using SPM.
Thanks
David
Replies:
Login to reply to this topic.
