On Mon, May 7, 2018 at 10:24 AM, Karina Quevedo <[log in to unmask]> wrote:

Dear Toben, Manfred, and Guillaume

I have attached the job file for both batches with this email.

Please notice that the variance and covariance matrix are not the same. Batch 1 seems to have more collinearity compared to batch 2 between the columns. That is what we got when we explore the covariance structure.

Warmly,

Karina

On Mon, May 7, 2018 at 10:10 AM, Karina Quevedo <[log in to unmask]> wrote:

Dear Guillaume,
We will do the detective work today and tomorrow. In the mean time how would you recommend analyzing the data?

We do not want to loose any of the conditions (time, self, or emotion). And we want to have them all in the same model.

Warmly and more soon on those first level images,

Karina

On Mon, May 7, 2018 at 9:56 AM, Guillaume Flandin <[log in to unmask]> wrote:

Dear Karina,

From what I can see in your two second-level SPM.mat files, the GLMs
seem to be the same (same design matrix and same variance components)
so, as Volkmar and Torben pointed out, the difference must be in your
input images. Are you really sure that they are identical? For example,
are these two images from the first level identical:
RT-FMRI/3007_RT1/1stLevel/ESOM_PRE/con_0013.nii
RT-FMRI/SmO/3007/1stLevel/ESOM_PRE_SmO_Rest/con_0044.nii

I would still probably recommend to analyse the data differently at the
second level but that's something we can discuss once the detective work
is over.

Best regards,
Guillaume.

On 07/05/18 10:51, Torben Lund wrote:
> Dear Karina
>
> In the batch manager hit the save icon, and save each of the batch jobs
> if you have not already done so, with those files we can se what you
> actually specified.
>
> Best
> Torben
>
>
>> Den 6. maj 2018 kl. 19.56 skrev Karina Quevedo <[log in to unmask]
>> <mailto:[log in to unmask]>>:
>>
>> Dear Torben, Volkmar and Manfred,
>> Please see attached in a zip drive the screen shot of the matrix and
>> covariance structure and the design matrix information. SPM12 is
>> taking different decisions for the covariance structure for two sets
>> of images that we are fairly certain that we calculated in the same
>> way in parallel to check our process.
>> We are going to run the 1st level images a third time.
>>
>> Torben we know that the second level job batches have not been
>> specified in the same way already but happy to shared them with you.
>>
>> Essentially we ran the same batch (Batch 1) but uploaded a second
>> different set of 1st level contrast images for QA (which we derived
>> with the same fmri structure to begin with) and SPM took different
>> decisions automatically regarding the var-covar structure.
>>
>> If we did something wrong we just want to know what it is and why the
>> difference?
>> Warmly?
>> Karina
>>
>> On Sun, May 6, 2018 at 5:29 AM, Torben Lund <[log in to unmask]
>> <mailto:[log in to unmask]>> wrote:
>>
>> Dear Karina
>>
>> if you could share the two batch jobs then we could confirm, that
>> they have been indeed been specified the same way. If they have
>> been specified the same way I think there is a problem with SPM,
>> but I don’t really think this is the case.
>>
>>
>> Best
>> Torben
>>
>>
>>
>>> Den 5. maj 2018 kl. 21.41 skrev Karina Quevedo <[log in to unmask]
>>> <mailto:[log in to unmask]>>:
>>>
>>> Dear Volkmar, Torben and Manfred,
>>> We specify the same settings for variance in both models but we
>>> did used different images for estimating the two different
>>> models. However see the details before about how we estimated the
>>> two sets of images. (GRACIAS!)
>>>
>>> We are analyzing a complex set of data that has the following
>>> design structure,
>>>
>>> *_Pre_*: Participants see their own face and a stranger's face
>>> across 3 emotions (happy, neutral and sad)
>>> *_
>>> _*
>>> *_Intervention_*: Participants attempt to change their Bilateral
>>> Amygdala and Hippocampus (ROI) activity via a neurofeedback
>>> procedure:
>>>
>>> Condition 1: See the Self face and increase ROI.
>>> Condition 2: See the Other face and count backwards.
>>> *_
>>> _*
>>> *_Post_*: Participants see their own face and a stranger's face
>>> across 3 emotions (happy, neutral and sad).
>>>
>>> We are interested in analyzing the data *Pre versus Post*, and we
>>> used a *F**ull Factorial Design with 3 factors: Time, Face and
>>> Emotion.*
>>>
>>> The columns in the attached matrixes are as follows 1 Face a=
>>> self, b= other, 2 Emotion a=happy, b=neutral, c=sad.
>>> Column 1: Pre, 1a, 2a
>>> Column 2: Pre, 1a, 2b
>>> Column 3: Pre, 1a, 2c
>>> Column 4: Pre. 1b, 2a
>>> Column 5: Pre, 1b, 2b
>>> Column 6: Pre, 1b, 2c
>>> Column 7: Post, 1a, 2a
>>> Column 8: Post, 1a, 2b
>>> Column 9: Post, 1a, 2c
>>> Column 10: Post. 1b, 2a
>>> Column 11: Post, 1b, 2b
>>> Column 12: Post, 1b, 2c
>>>
>>> The first level analysis contrasts that we used to create these
>>> second level full factorial batches are coming from the same
>>> source in terms of pre-processing files.
>>>
>>> The only difference is that we calculated the first level
>>> contrast twice (using the same pre-processing data and initial
>>> fmri factorial specification ) and for one of the sets of first
>>> level analysis batches (_*after using the same initial fmri
>>> factorial design specification*_) we calculated more follow up t
>>> contrasts. That was all.
>>>
>>> However when we run two parallel second level analysis with those
>>> two sets of first level contrasts we are getting very different
>>> results because SPM 12 seems to be choosing to model the data
>>> differently in terms of covariance structure:
>>>
>>> *Batch 1 *in the figure we attach is the batch using images
>>> derived from 1st level batches where we calculated less t
>>> contrasts. SPM created a covariance matrix that assumes
>>> higher collinearity between column 1 an column 7 (i.e. Pre
>>> Happy Self Face, and Post Happy Self Face). It gives a value
>>> of *cosine = -0.75*.
>>>
>>> *Batch 2* in the figure we attach is the batch using images
>>> derived from 1st level batches where we calculated more t
>>> contrasts. Using the same initial Batch 1 but loading the
>>> different images, it seems that SPM 12 created a different
>>> covariance structure. *Cosine = -0.08* (more orthogonal
>>> compared to Batch 1)
>>>
>>>
>>> In addition to that, we had also looked into the SPM.mat file for
>>> both of the second level batches. We took a screenshot of the
>>> SPM.xX.W values for both batches. If you would like, we could
>>> send you the SPM.mat batches for you to look at.
>>>
>>> Our questions are:
>>> 1. Why is SPM 12 making such different decisions with regards to
>>> covariance structures?
>>>
>>> 2. Given our design which model is more appropriate Batch 1 or
>>> Batch 2?
>>>
>>> 3. We cannot run a flexible factorial design without loosing one
>>> of our conditions (in order to model participants). Is that still
>>> ok?
>>>
>>> One of our colleges is adamant about running a flexible factorial
>>> but that means that we will lose either emotion or self in the
>>> design, We think that the best model is *Batch 1* because it
>>> assumes _*more *_collinearity between the*Pre and the Post
>>> conditions*, what is your opinion?
>>>
>>> Warmly yours,
>>> Karina
>>> <design matrix information SPM_xX_W.jpg>
>>> <matrix and covariate structures.jpg>
>>>
>>>
>>> On Wed, May 2, 2018 at 4:37 AM, Volkmar Glauche
>>> <volkmar.glauche@uniklinik-freiburg.de

>>> <mailto:volkmar.glauche@uniklinik-freiburg.de>> wrote:
>>>
>>> Dear Karina,
>>>
>>> as Torben said, something seems to be different. Did you
>>> really specify the same settings for variance ((un)equal)
>>> and independence (yes/no) in both models? Did you use the
>>> same factors when specifying the models? Are both models
>>> estimated on the same set of images? Since non-sphericity
>>> estimation depends on the data, different input data may
>>> yield different non-sphericity estimates.
>>> You may also want to use "Review Design" to look at the
>>> effects of model estimation. You will find a "Design" menu in
>>> the interactive window, where you can explore the design
>>> matrix, files and factors and covariance structure. When
>>> looking at the covariance structure, a multi-page image
>>> matrix will be displayed. On the first page, you will find
>>> the estimated covariance and on the following pages the
>>> individual covariance components for which parameters have
>>> been estimated.
>>>
>>> Best,
>>> Volkmar
>>>
>>>
>>>
>>>
>>> --
>>> Karina Quevedo, Ph.D., L.P.
>>> Assistant Professor. Division of Child and Adolescent Psychiatry
>>> Room F260
>>> University of Minnesota. Department of Psychiatry
>>> 2450 Riverside Avenue, Mpls, MN 55454

>>> <https://maps.google.com/?q=2450+Riverside+Avenue,+Mpls,+MN+55454&entry=gmail&source=g>
>>> 612-273-9761 Fax: 612-273-9779
>>> [log in to unmask] <mailto:[log in to unmask]>
>>>
>>
>>
>>
>>
>> --
>> Karina Quevedo, Ph.D., L.P.
>> Assistant Professor. Division of Child and Adolescent Psychiatry
>> Room F260
>> University of Minnesota. Department of Psychiatry
>> 2450 Riverside Avenue, Mpls, MN 55454
>> 612-273-9761 Fax: 612-273-9779
>> [log in to unmask] <mailto:[log in to unmask]>
>>
>> <matrix and covariate structures.zip>
>

--
Guillaume Flandin, PhD
Wellcome Centre for Human Neuroimaging
University College London
12 Queen Square
London WC1N 3BG

--

Karina Quevedo, Ph.D., L.P.
Assistant Professor. Division of Child and Adolescent Psychiatry

Room F260

University of Minnesota. Department of Psychiatry

2450 Riverside Avenue, Mpls, MN 55454

612-273-9761 Fax: 612-273-9779

[log in to unmask]

--

Karina Quevedo, Ph.D., L.P.
Assistant Professor. Division of Child and Adolescent Psychiatry

Room F260

University of Minnesota. Department of Psychiatry

2450 Riverside Avenue, Mpls, MN 55454

612-273-9761 Fax: 612-273-9779

[log in to unmask]

Karina Quevedo, Ph.D., L.P.

Assistant Professor. Division of Child and Adolescent Psychiatry

Room F260

University of Minnesota. Department of Psychiatry

2450 Riverside Avenue, Mpls, MN 55454

612-273-9761 Fax: 612-273-9779

[log in to unmask]