Show all posts by user
Dear AFNI users-
We are very pleased to announce that the new AFNI Message Board framework is up! Please join us at:
https://discuss.afni.nimh.nih.gov
Existing user accounts have been migrated, so returning users can login by requesting a password reset. New users can create accounts, as well, through a standard account creation process. Please note that these setup emails might initially go to spam folders (esp. for NIH users!), so please check those locations in the beginning.
The current Message Board discussion threads have been migrated to the new framework. The current Message Board will remain visible, but read-only, for a little while.
Sincerely,
AFNI HQ
History of AFNI updates
I suggest using 3dLMEr with something like
3dLMEr -model “Session*Velocity*Pleasantness*Oxytocin+Movement+(Oxytocin|subject)+(Pleasantness|subject)”
If you're interested in comparing Session and Velocity levels, centering within each Session and Velocity may be important for Pleasantness and Oxytocin.
by
gang
-
AFNI Message Board
> just to clarify that be the overall mean aggression (e.g., take every subject's average aggression across conditions and use that to demean?)
No, remove the mean (across subjects for each condition) from each subject. Hope this is clear.
by
gang
-
AFNI Message Board
First of all, if I understand your data structure accurately, I believe that "Aggression" is a within-subject quantitative variable because it varies within subject. Second, use 3dLMEr because 3dMVM cannot handle within-subject quantitative variable. Lastly, does Aggression correlate with Reputation to some extent? If so, you may want to center Aggression for each level of Reputation (r
by
gang
-
AFNI Message Board
Sam,
Did you run 8 separate seed-based analyses for each subject: one for each of the two seed and for each of the four waves? Also, is the output from the subject level analysis aggregated at those 8 regions or whole brain at the voxel level?
by
gang
-
AFNI Message Board
What research questions (effects of interest) are you trying to address?
by
gang
-
AFNI Message Board
Can you show the contingent table for the two factors of Reputation and Aggression? And what are your effects of interest?
by
gang
-
AFNI Message Board
> -gltLabel 1 'wk6Mwk0_int -gltCode 1 'cond : 1*tms -1*sham time : 1*wk6 -1*wk0'
This specification assesses the difference between
wk6 - wk0 under the condition tms
and
wk6 - wk0 under the condition sham
Or the difference between
tms - sham at time wk6
and
tms - sham at time wk0
by
gang
-
AFNI Message Board
3dMVM validator may help diagnose the problem:
> -glfLabel 1 ST_ANOVA_aVf_HCvPCvIED -glfCode 1 'contrasts : 1*angry -1*fear condition : 1*HC -1*IED & 1*HC -1*PC & 1*IED -1*PC'\
This specification does not make sense. By default 3dMVM would automatically provide the interaction assessment for this interaction between "condition" and "contrasts". N
by
gang
-
AFNI Message Board
> this is a within-subjects analysis of drug/placebo effects on one of our first level contrasts
Is 'contrasts' a within- or between-subject factor? Can you show the contingency table between 'Condition' and 'contrasts'?
Also, does 'Age' or 'Motion' vary within each subject?
by
gang
-
AFNI Message Board
> -bsVars 'condition*age+sex' \
> -wsVars 'contrasts'\
> -qVars 'age,motion' \
You didn't specify the variable 'motion' in the first two lines above. Maybe try the following
-bsVars 'condition*age+sex+motion' \
-wsVars 'contrasts'\
-qVars 'age,motion' \
by
gang
-
AFNI Message Board
JW,
Those three R packages are only needed for the following AFNI programs
'afex' for 3dMVM
'brms' for RBA, TRR, MBA
'metafor' for 3dICC
So, if you're not going to run any of those programs above, you can ignore the warnings. In case you want to use 3dMVM, try installing 'afex' inside R:
install.packages('afex', dependencies = N
by
gang
-
AFNI Message Board
> If I am correct, the result of the Age t is smilar to the result of 3dRegAna -tcoef. Right?
If you built the same model between 3dMVM and 3dRegAna, the result should be close to each other.
by
gang
-
AFNI Message Board
Glad that you sorted it out. I was going to say that the variable "Session" in your original script was not a within-subject factor.
by
gang
-
AFNI Message Board
The situation is more complex than I originally realized. 3dMVM needs to get a workable voxel to extract a few parameters such as degrees of freedom. It might better to go back to your original plan by making a "fake" dataset.
by
gang
-
AFNI Message Board
The following two lines
-gltLabel 2 GroupRisk -gltCode 2 'Group : 1*Singr -1*Contr Risk: ' \
-gltLabel 3 GroupAge -gltCode 3 'Group : 1*Singr -1*Contr Age: ' \
should change to
-gltLabel 2 GroupRisk -gltCode 2 'Group : 1*Singr -1*Contr Risk : ' \
-gltLabel 3 GroupAge -gltCode 3 'Group : 1*Singr -1*Contr Age : ' \
> But I am wondering if I s
by
gang
-
AFNI Message Board
It looks like lop$outInit needs to be initialized through the testing process among some voxels. Since in this case you don't have any meaningful voxels among those initial slices, add the following line
lop$outInit <- rep(0, NoBrick)
(replace NoBrick with the total number of sub-bricks in the output -- you can figure this out, right?)
before the following lines:
print(sprint
by
gang
-
AFNI Message Board
What is your AFNI version?
afni -ver
Try without the mask, and see if it works.
by
gang
-
AFNI Message Board
Erik,
I prefer a "highlight but not hide" approach to reporting results because the cluster-based adjustment tends to waste a lot of information:
If you want to adopt the cluster approach, use 3dClustSim.
by
gang
-
AFNI Message Board
Matt,
No need to worry. When there are only two levels for a within-subject factor, adding -SC would not have any impact because there is no sphericity issue with only one correlation.
by
gang
-
AFNI Message Board
You can add the following to obtain the directionality of age effecct:
-gltCode t 'Age :' \
Also, you need to add a space after Risk and Age on the following two lines:
-gltLabel 2 GroupRisk -gltCode 2 'Group : 1*Singr -1*Contr Risk:' \
-gltLabel 3 GroupAge -gltCode 3 'Group : 1*Singr -1*Contr Age:' \
by
gang
-
AFNI Message Board
Robin,
> If I my end goal is to the the stats + residuals for the bottom 10 slices. Can I make a "fake" dataset,
> containing 20 slices: first 10 are bottom slices that I'm interested in and the next 10 could be from the
> middle of the brain, just to trick MVM to think that the model/stats are correct?
Yes, that should be fine. Alternatively, in this case, you c
by
gang
-
AFNI Message Board
There are a couple of examples in the AFNI hands-on:
You can find the corresponding data and 3dLME scripts here:
by
gang
-
AFNI Message Board
> I got #14 GroupAge and #15 GroupAge t. Is it correct to put #14 on OLay and #15 on Thr?
Yes, I do recommend such specification.
by
gang
-
AFNI Message Board
> should I put the same # on the ULay and OLay options?
For F-statistic, use the same sub-brick for OLay and Thr (ULay is typically for anatomical data). For t- or z-statistic, set OLay with the effect estimates (e.g., beta) and Thr with the statistic sub-brick.
by
gang
-
AFNI Message Board
I don't really understand your question. Again, the FDR information is already embedded in the output of 3dMVM, and you don't need to do anything. When you visualize the results, the FDR value will show up when you set a threshold on the AFNI GUI.
by
gang
-
AFNI Message Board
Xiyue, the output from 3dMVM should automatically contain the information to assign the FDR values, which would show up under the threshold bar on the AFNI GUI. Thus, you do not need to run a separate 3dFDR command.
by
gang
-
AFNI Message Board
> does this mean in my case I don't have happy/sad but time1/time2?
The example script is intended to show a demo case, so change it to adapt to your specific situation.
by
gang
-
AFNI Message Board
Try something like
dMEMA -prefix ex2 \
...
-conditions happy sad \
-set happy \
...
-set sad \
...
by
gang
-
AFNI Message Board
> do I have to delete all the V trials since there are only V*no-noisy, but no V*yes-noisy trials?
Yes, that's the way to go. If you want to make any inferences regarding the V trials, set up a separate analysis.
by
gang
-
AFNI Message Board
If you have missing data, use 3dLME. The AFNI hands-on session for group analysis has a demo case (page 13):
by
gang
-
AFNI Message Board