help
help > corr values for individual space subject
Apr 15, 2014 12:04 PM | Giannina Rita Iannotti
corr values for individual space subject
Dear Alfonso,
i took advantage from your answer to the previous posts.
I hope you or someone else could help me in the same way.
I'm performing single subject analysis in individual space. I'm analyzing the seed to voxel correlatrion.
For each seed (that I previously define on the structural wth MRIcroN) I obtain the Beta and the corr map.
I would like to know if the correlation value significant. In other words, if It's already applyed a threshold and I can trust the correlation values even if they are small, for example 0.1 or 0.4.
Thank you very much for your help.
giannarita
Originally posted by Alfonso Nieto-Castanon:
i took advantage from your answer to the previous posts.
I hope you or someone else could help me in the same way.
I'm performing single subject analysis in individual space. I'm analyzing the seed to voxel correlatrion.
For each seed (that I previously define on the structural wth MRIcroN) I obtain the Beta and the corr map.
I would like to know if the correlation value significant. In other words, if It's already applyed a threshold and I can trust the correlation values even if they are small, for example 0.1 or 0.4.
Thank you very much for your help.
giannarita
Originally posted by Alfonso Nieto-Castanon:
Dear
Khoo,
Regarding (1), in the new versions of conn the corr_* files are only generated if you select the corresponding option in the Setup->Options tab ('Create first-level seed-to-voxel r-maps'). Irrespective of this, the BETA files will always contain the z value (after Fisher transformation) if you have selected any of the 'correlation' measures in your Analysis step.
Regarding (2), yes, it is perfectly valid to enter the BETA_Subject*_Condition*_Source*.nii files in SPM for any form of second-level analyses. Also, if you are interested, you can perform within-subject inferences (such as paired T-test) in CONN directly, you simply need to define the different within-subject conditions in Setup->Conditions, and then select the combination of between-subjects and within-subjects (e.g. between-conditions) effects in your second-level analysis tab.
Hope this helps
Alfonso
Originally posted by Hui Ming Khoo:
Regarding (1), in the new versions of conn the corr_* files are only generated if you select the corresponding option in the Setup->Options tab ('Create first-level seed-to-voxel r-maps'). Irrespective of this, the BETA files will always contain the z value (after Fisher transformation) if you have selected any of the 'correlation' measures in your Analysis step.
Regarding (2), yes, it is perfectly valid to enter the BETA_Subject*_Condition*_Source*.nii files in SPM for any form of second-level analyses. Also, if you are interested, you can perform within-subject inferences (such as paired T-test) in CONN directly, you simply need to define the different within-subject conditions in Setup->Conditions, and then select the combination of between-subjects and within-subjects (e.g. between-conditions) effects in your second-level analysis tab.
Hope this helps
Alfonso
Originally posted by Hui Ming Khoo:
Dear Conn users,
I found that only BETA files was generated in the first level analysis (without corr files).
I am using the same setting as Maria (CONN.Analyses.measure is set to 1, (correlation)).
I would like to know
1)whether the content of BETA files generated is still the z value (after Fisher transformation) even the corr files were not generated,
2)is it possible to use the BETA files for within subject inference (such as paired T test) in SPM second level analysis.
Best regards,
Khoo
I found that only BETA files was generated in the first level analysis (without corr files).
I am using the same setting as Maria (CONN.Analyses.measure is set to 1, (correlation)).
I would like to know
1)whether the content of BETA files generated is still the z value (after Fisher transformation) even the corr files were not generated,
2)is it possible to use the BETA files for within subject inference (such as paired T test) in SPM second level analysis.
Best regards,
Khoo
Threaded View
Title | Author | Date |
---|---|---|
Maria Fernandez-Seara | Feb 28, 2012 | |
Hui Ming Khoo | Mar 27, 2014 | |
Alfonso Nieto-Castanon | Apr 14, 2014 | |
Giannina Rita Iannotti | Apr 15, 2014 | |
Stephen L. | Nov 30, 2017 | |
Alfonso Nieto-Castanon | Mar 1, 2012 | |
Maria Fernandez-Seara | Mar 2, 2012 | |