help
help > RE: Clarifying contrast interpretation
May 27, 2016 08:05 PM | Alfonso Nieto-Castanon - Boston University
RE: Clarifying contrast interpretation
Hi Tessa,
Regarding (1), yes, exactly, those results indicate that the average connectivity between your two DMN regions and frontal gyrus is significantly correlated with symptom scores
Regarding (2), not exactly, those results for the [1 -1] between-sources contrast indicate that there is symptoms-by-source interaction, namely that the degree of association between symptom scores and MPFC-occipitalCortex connectivity is different than the degree of association between symptom scores and RLP-occipitalCortex connectivity, but do not necessarily indicate that one of them is a positive association and the other one a negative association (e.g. if you plot symptom scores vs. MPFC-occipitalCortex connectivity, the regression line fit would have a different slope than the one obtained if you plot symptom scores vs. RLP-occipitalCortex connectivity). An alternative, and equivalent, interpretation would be that the size of the difference between MPFC-occipitalCortex connectivity and RLP-occipitalCortex connectivity is itself associated with symptom scores (e.g. subjects with higher symptom scores show larger differences between MPFC-occipitalCortex connectivity and RLP-occipitalCortex connectivity).
Hope this helps
Alfonso
Originally posted by Tessa Vuper:
Regarding (1), yes, exactly, those results indicate that the average connectivity between your two DMN regions and frontal gyrus is significantly correlated with symptom scores
Regarding (2), not exactly, those results for the [1 -1] between-sources contrast indicate that there is symptoms-by-source interaction, namely that the degree of association between symptom scores and MPFC-occipitalCortex connectivity is different than the degree of association between symptom scores and RLP-occipitalCortex connectivity, but do not necessarily indicate that one of them is a positive association and the other one a negative association (e.g. if you plot symptom scores vs. MPFC-occipitalCortex connectivity, the regression line fit would have a different slope than the one obtained if you plot symptom scores vs. RLP-occipitalCortex connectivity). An alternative, and equivalent, interpretation would be that the size of the difference between MPFC-occipitalCortex connectivity and RLP-occipitalCortex connectivity is itself associated with symptom scores (e.g. subjects with higher symptom scores show larger differences between MPFC-occipitalCortex connectivity and RLP-occipitalCortex connectivity).
Hope this helps
Alfonso
Originally posted by Tessa Vuper:
Hi all,
First, thank you for developing such a user-friendly toolbox! I have a few questions regarding interpretation of contrasts. I am interested in the effect of patient symptoms on default mode connectivity. I have selected my group 'Patients' and covariate 'Symptom Score' and entered contrast [0, 1] to test for simple main effect of symptoms. I then selected two DMN seeds (MPFC, RLP) and ran two contrasts: [0.5 0.5] and [1 -1] in a seed-voxel analysis.
1) For the [0.5 0.5] contrast, a cluster in the middle frontal gyrus was significant. Am I correct in saying that symptom score is correlated with connectivity between MPFC, RLP, and middle frontal gyrus?
2) For the [1 -1] contrast, a cluster in the lateral occipital cortex was significant. Is it correct to say that symptom score is correlated with increased connectivity between MPFC-occipital cortex and/or decreased connectivity between RLP-occipital cortex?
Thanks very much for your help,
Tessa
First, thank you for developing such a user-friendly toolbox! I have a few questions regarding interpretation of contrasts. I am interested in the effect of patient symptoms on default mode connectivity. I have selected my group 'Patients' and covariate 'Symptom Score' and entered contrast [0, 1] to test for simple main effect of symptoms. I then selected two DMN seeds (MPFC, RLP) and ran two contrasts: [0.5 0.5] and [1 -1] in a seed-voxel analysis.
1) For the [0.5 0.5] contrast, a cluster in the middle frontal gyrus was significant. Am I correct in saying that symptom score is correlated with connectivity between MPFC, RLP, and middle frontal gyrus?
2) For the [1 -1] contrast, a cluster in the lateral occipital cortex was significant. Is it correct to say that symptom score is correlated with increased connectivity between MPFC-occipital cortex and/or decreased connectivity between RLP-occipital cortex?
Thanks very much for your help,
Tessa
Threaded View
| Title | Author | Date |
|---|---|---|
| Tessa Vuper | May 27, 2016 | |
| Alfonso Nieto-Castanon | May 27, 2016 | |
| Tessa Vuper | Jun 3, 2016 | |
| Alfonso Nieto-Castanon | Jun 4, 2016 | |
| Tessa Vuper | Jun 7, 2016 | |