External Email - Use Caution
Dear all, I was asked to provide a power calculation of my group comparison results on cortical thickness, so I extracted all clusters with significance in CT into SPSS. but I found the cluster value of CT is not significant in SPSS based on group comparison. so I got a very small effect size although the original ROI was generated by a initial threshold of p < 0.001, and corrected by monte-carlo p < 0.05, two sided. Is this possible? how can I resolve it?
best, Yali
About once a year we get an email like this saying that there are inconsistent results between SPSS and mri_glmfit. I think the problem usually boils down to having a different design matrix. Can you extract the design matrix from SPSS and compare it to Xg.dat? Or can you import Xg.dat and use it to analyze in SPSS? Or do something else to validate that they are using the same model?
On 3/23/19 11:55 PM, Yali Jiang wrote:
External Email - Use Caution
Dear all, I was asked to provide a power calculation of my group comparison results on cortical thickness, so I extracted all clusters with significance in CT into SPSS. but I found the cluster value of CT is not significant in SPSS based on group comparison. so I got a very small effect size although the original ROI was generated by a initial threshold of p < 0.001, and corrected by monte-carlo p < 0.05, two sided. Is this possible? how can I resolve it?
best, Yali
_______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
External Email - Use Caution
Greve, Thank you very much for your answer. I used QDEC for analysis in FS. The FS results were analyzed on a whole brain level by ANOVA (2*2, with one covariate ); the group comparison in SPSS was conducted in a ROI manner (the ROI is the cluster with significance from QDEC)， I am a little confused how the two design matrix could be the same? could you please give me a clue about this? If I did it wrong?
best, Yali
At 2019-03-25 23:10:33, "Greve, Douglas N.,Ph.D." DGREVE@mgh.harvard.edu wrote:
About once a year we get an email like this saying that there are inconsistent results between SPSS and mri_glmfit. I think the problem usually boils down to having a different design matrix. Can you extract the design matrix from SPSS and compare it to Xg.dat? Or can you import Xg.dat and use it to analyze in SPSS? Or do something else to validate that they are using the same model?
On 3/23/19 11:55 PM, Yali Jiang wrote:
External Email - Use Caution
Dear all, I was asked to provide a power calculation of my group comparison results on cortical thickness, so I extracted all clusters with significance in CT into SPSS. but I found the cluster value of CT is not significant in SPSS based on group comparison. so I got a very small effect size although the original ROI was generated by a initial threshold of p < 0.001, and corrected by monte-carlo p < 0.05, two sided. Is this possible? how can I resolve it?
best, Yali
_______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.eduhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Greve, Douglas N.,Ph.D. -
Yali Jiang