contrast interpretation

[karyn]

HI all:

I am having a bit of a challenge understanding the meaning behind positive and negative contrasts. I want to use your Flash and Beep example you provided earlier to explain the meaning behind these intensities. If I open a SPM t map file that codes when flash is greater than beep ( 1 -1) - if I see positive activation this should be interpreted as regions where flash has more activation then beep and if I see negative activation this should be interpreted as a region where beep is less activated than flash? If I want to see regions where beep has greater activation than flash  and flash less activation then beep, then I would have to open another SPM t map file (-1 1) that codes for beep > flash. In this second case, positive activation in a region should be interpreted as beep being more active than flash and negative activation in a region should be interpreted as flash being is less active then beep?

Could you confirm or further elaborate on the meaning behind positive and negative intensities within SPM t maps and between SPM t maps?

[Xu Cui]

>if I see positive activation this should be interpreted as regions where flash has more activation then beep 

Yes

> and if I see negative activation this should be interpreted as a region where beep is less activated than flash? 

No. beep < flash is the same as flash > beep. Negative means flash < beep.

>  If I want to see regions where beep has greater activation than flash  and flash less activation then beep, then I would have to open another SPM t map file (-1 1) that codes for beep > flash.

You don't have to. The first t-test image can review both directions. SPM display only gives you one direction but xjview gives you both. The T-test image of (-1 1) is actually identical to the t-test image of (1, -1) with sign flipped.

Xu

--
You received this message because you are subscribed to the Google Groups "xjView discuss group" group.
To view this discussion on the web visit https://groups.google.com/d/msg/xjview-discuss-group/-/WJlSWPxjd6AJ.
To post to this group, send email to xjview-dis...@googlegroups.com.
To unsubscribe from this group, send email to xjview-discuss-g...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/xjview-discuss-group?hl=en.

--

----------------------------------------------------------------
Xu Cui, Ph.D
Department of Psychiatry and Behavioral Sciences
Stanford University
Blog: http://www.alivelearn.net

Profile: http://www.peaya.com/p/?cuixu
----------------------------------------------------------------

[Jonathan Beck]

Dear All,

I have a related question.  I understand that in the case of Flash - Beep, positive activation shows that F>B and negative activation shows that B>F, but that seems to assume that Flash and Beep (or more specifically, their respective betas) are both positive.  Is that true? Can that assumption be safely made?

Thank you!

Jonathan

[Xu Cui]

Jonathan,

No, it does not assume both betas are positive.

Xu

--
You received this message because you are subscribed to the Google Groups "xjView discuss group" group.

To unsubscribe from this group and stop receiving emails from it, send an email to xjview-discuss-g...@googlegroups.com.

To post to this group, send email to xjview-dis...@googlegroups.com.

Visit this group at http://groups.google.com/group/xjview-discuss-group?hl=en.

For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Jonathan Beck]

Xu,

 

Thanks for your response.  Your clarification makes complete sense, but it raised another question in my mind:

 

Deactivation subtracted from deactivation can yield a positive difference, and in that case, the positive difference does not reveal postive activation, just greater activation.  Is there a way in the xjview utility to easily view only positive values resulting from subtracting postive activation from positive activation (as opposed to positive values resulting from subtracting deactivation)?  It would help with interpretation of the results if there were.

 

Thank you for being so available to the grateful public!

 

Jonathan

[Xu Cui]

Jonathan,

> Is there a way in the xjview utility to easily view only positive values resulting from subtracting postive activation from positive activation (as opposed to positive values resulting from subtracting deactivation)? 

Unfortunately no. The t-test image you input to xjview does not contain the values of the original betas.

However, you can open beta images to see whether a region is 'positive' or 'negative'.

Best,

Xu