subject:"\[Freesurfer\] Paired Analysis Contrasts"

Re: [Freesurfer] Paired Analysis Contrasts

2018-11-26 Thread Greve, Douglas N.,Ph.D.

answers below

On 11/21/2018 01:53 PM, srishti goel wrote:
>
> External Email - Use Caution
>
> Hi Douglas,
>
> Thank you so much for your response. We just want to be extra careful 
> about the way we are doing this and so below I have outlined the exact 
> pattern of the group descriptor files we are using along with the 
> contrast. Please let me know if you think this is correct or not. The 
> questions we are asking with this analysis are the same as I outlined 
> in the previous email.
>
> _This is our first group descriptor file:_
>
> GroupDescriptorFile 1
> Title BEIP 8 vs. 16 compare CAU (0) And FCG (1)
> Class Main
> Input    005_16.long.005    Main
> Input    005_8.long.005    Main
> Input    006_16.long.006    Main
> Input    006_8.long.006    Main
>
> _This is our second group descriptor file:_
>
> GroupDescriptorFile 1
> Title BEIP longitudinal Compare CAU (0) And FCG (1)
> Class cau
> Class fc
> Input    005_long_pair fc
> Input    006_long_pair cau
>
> The contrast to understand group differences in thickness between 8 
> and 16 yr old data sets are:
> 1 -1
>
> *Question1:* Is this correct?
Yes
> *Question 2:* If we use the below mentioned way to set up our second 
> group descriptor file, will the above contrasts still work? Or can we 
> use 0 1 contrast with the file below to get at the same thing? We 
> actually ran it both ways and the results weren't similar but I 
> specifically ask this question because we are a little worried about 
> the above 'second group descriptor file' not having a column that says 
> 'Main' and we are trying to understand the right way to do our analyses.
I think this will give you the same answer.
>
> _Example group descriptor file 2:_
> GroupDescriptorFile 1
> Title BEIP longitudinal Compare CAU (0) And FCG (1)
> Class Main
> Variables group
> Input    005_long_pair Main 1
> Input    006_long_pair Main 0
>
> Thank you so much! I appreciate all your help with this!
>
> Best,
> Srishti
> Social/Clinical Research Specialist
> Child Imaging Research and Life Experiences Lab
> University of North Carolina at Chapel Hill
> email (W): srish...@email.unc.edu 
> skype: srishti.goel12
>
>
>
> On Wed, Nov 14, 2018 at 2:50 PM srishti goel <23srishtig...@gmail.com 
> > wrote:
>
> Hi Douglas,
>
> Thank you so much for your response. We just want to be extra
> careful about the way we are doing this and so below I have
> outlined the exact pattern of the group descriptor files we are
> using along with the contrast. Please let me know if you think
> this is correct or not. The questions we are asking with this
> analysis are the same as I outlined in the previous email.
>
> _This is our first group descriptor file:_
>
> GroupDescriptorFile 1
> Title BEIP 8 vs. 16 compare CAU (0) And FCG (1)
> Class Main
> Input    005_16.long.005    Main
> Input    005_8.long.005    Main
> Input    006_16.long.006    Main
> Input    006_8.long.006    Main
>
> _This is our second group descriptor file:_
>
> GroupDescriptorFile 1
> Title BEIP longitudinal Compare CAU (0) And FCG (1)
> Class cau
> Class fc
> Input    005_long_pair fc
> Input    006_long_pair cau
>
> The contrast to understand group differences in thickness between
> 8 and 16 yr old data sets are:
> 1 -1
>
> *Question1:* Is this correct?
> *Question 2:* If we use the below mentioned way to set up our
> second group descriptor file, will the above contrasts still work?
> Or can we use 0 1 contrast with the file below to get at the same
> thing? We actually ran it both ways and the results weren't
> similar but I specifically ask this question because we are a
> little worried about the above 'second group descriptor file' not
> having a column that says 'Main' and we are trying to understand
> the right way to do our analyses.
>
> _Example group descriptor file 2:_
> GroupDescriptorFile 1
> Title BEIP longitudinal Compare CAU (0) And FCG (1)
> Class Main
> Variables group
> Input    005_long_pair Main 1
> Input    006_long_pair Main 0
>
> Thank you so much! I appreciate all your help with this!
>
> Best,
> Srishti
> Social/Clinical Research Specialist
> Child Imaging Research and Life Experiences Lab
> University of North Carolina at Chapel Hill
> email (W): srish...@email.unc.edu 
> skype: srishti.goel12
>
>
>
> On Wed, Nov 7, 2018 at 9:24 AM Greve, Douglas N.,Ph.D.
> mailto:dgr...@mgh.harvard.edu>> wrote:
>
>
>
> On 10/24/2018 09:45 PM, srishti goel wrote:
> >
> > External Email - Use Caution
> >
> > Hello,
> >
> > I am trying to perform a paired analysis for a data set that
> has
> > subjects' who belong to either group1 or group2 and have
>

Re: [Freesurfer] Paired Analysis Contrasts

2018-11-21 Thread srishti goel

External Email - Use Caution

Hi Douglas,

Thank you so much for your response. We just want to be extra careful about
the way we are doing this and so below I have outlined the exact pattern of
the group descriptor files we are using along with the contrast. Please let
me know if you think this is correct or not. The questions we are asking
with this analysis are the same as I outlined in the previous email.

*This is our first group descriptor file:*

GroupDescriptorFile 1
Title BEIP 8 vs. 16 compare CAU (0) And FCG (1)
Class Main
Input005_16.long.005Main
Input005_8.long.005Main
Input006_16.long.006Main
Input006_8.long.006Main

*This is our second group descriptor file:*

GroupDescriptorFile 1
Title BEIP longitudinal Compare CAU (0) And FCG (1)
Class cau
Class fc
Input005_long_pair fc
Input006_long_pair cau

The contrast to understand group differences in thickness between 8 and 16
yr old data sets are:
1 -1

*Question1:* Is this correct?
*Question 2:* If we use the below mentioned way to set up our second group
descriptor file, will the above contrasts still work? Or can we use 0 1
contrast with the file below to get at the same thing? We actually ran it
both ways and the results weren't similar but I specifically ask this
question because we are a little worried about the above 'second group
descriptor file' not having a column that says 'Main' and we are trying to
understand the right way to do our analyses.

*Example group descriptor file 2:*
GroupDescriptorFile 1
Title BEIP longitudinal Compare CAU (0) And FCG (1)
Class Main
Variables group
Input005_long_pair Main 1
Input006_long_pair Main 0

Thank you so much! I appreciate all your help with this!

Best,
Srishti
Social/Clinical Research Specialist
Child Imaging Research and Life Experiences Lab
University of North Carolina at Chapel Hill
email (W): srish...@email.unc.edu
skype: srishti.goel12



On Wed, Nov 14, 2018 at 2:50 PM srishti goel <23srishtig...@gmail.com>
wrote:

> Hi Douglas,
>
> Thank you so much for your response. We just want to be extra careful
> about the way we are doing this and so below I have outlined the exact
> pattern of the group descriptor files we are using along with the contrast.
> Please let me know if you think this is correct or not. The questions we
> are asking with this analysis are the same as I outlined in the previous
> email.
>
> *This is our first group descriptor file:*
>
> GroupDescriptorFile 1
> Title BEIP 8 vs. 16 compare CAU (0) And FCG (1)
> Class Main
> Input005_16.long.005Main
> Input005_8.long.005Main
> Input006_16.long.006Main
> Input006_8.long.006Main
>
> *This is our second group descriptor file:*
>
> GroupDescriptorFile 1
> Title BEIP longitudinal Compare CAU (0) And FCG (1)
> Class cau
> Class fc
> Input005_long_pair fc
> Input006_long_pair cau
>
> The contrast to understand group differences in thickness between 8 and 16
> yr old data sets are:
> 1 -1
>
> *Question1:* Is this correct?
> *Question 2:* If we use the below mentioned way to set up our second
> group descriptor file, will the above contrasts still work? Or can we use 0
> 1 contrast with the file below to get at the same thing? We actually ran it
> both ways and the results weren't similar but I specifically ask this
> question because we are a little worried about the above 'second group
> descriptor file' not having a column that says 'Main' and we are trying to
> understand the right way to do our analyses.
>
> *Example group descriptor file 2:*
> GroupDescriptorFile 1
> Title BEIP longitudinal Compare CAU (0) And FCG (1)
> Class Main
> Variables group
> Input005_long_pair Main 1
> Input006_long_pair Main 0
>
> Thank you so much! I appreciate all your help with this!
>
> Best,
> Srishti
> Social/Clinical Research Specialist
> Child Imaging Research and Life Experiences Lab
> University of North Carolina at Chapel Hill
> email (W): srish...@email.unc.edu
> skype: srishti.goel12
>
>
>
> On Wed, Nov 7, 2018 at 9:24 AM Greve, Douglas N.,Ph.D. <
> dgr...@mgh.harvard.edu> wrote:
>
>>
>>
>> On 10/24/2018 09:45 PM, srishti goel wrote:
>> >
>> > External Email - Use Caution
>> >
>> > Hello,
>> >
>> > I am trying to perform a paired analysis for a data set that has
>> > subjects' who belong to either group1 or group2 and have data at time
>> > point1 (tp1) and time point2 (tp2). I am following all the steps on
>> > the wiki but I want to clarify if the contrasts I plan to use for my
>> > analyses are appropriate.
>> >
>> > For my analyses, I only have two groups and no covariates. I am
>> > interested to know if the decrease in cortical thickness of subjects
>> > in group 1 across tp1 and tp2 is more than the decrease in cortical
>> > thickness of subjects in group 2 across tp1 and tp2. The contrast I
>> > think should be used for this is: 1 -1. Is that correct?
>> Yes, if you first take the difference between tp1 and tp2.

Re: [Freesurfer] Paired Analysis Contrasts

2018-11-14 Thread srishti goel

External Email - Use Caution

Hi Douglas,

Thank you so much for your response. We just want to be extra careful about
the way we are doing this and so below I have outlined the exact pattern of
the group descriptor files we are using along with the contrast. Please let
me know if you think this is correct or not. The questions we are asking
with this analysis are the same as I outlined in the previous email.

*This is our first group descriptor file:*

GroupDescriptorFile 1
Title BEIP 8 vs. 16 compare CAU (0) And FCG (1)
Class Main
Input005_16.long.005Main
Input005_8.long.005Main
Input006_16.long.006Main
Input006_8.long.006Main

*This is our second group descriptor file:*

GroupDescriptorFile 1
Title BEIP longitudinal Compare CAU (0) And FCG (1)
Class cau
Class fc
Input005_long_pair fc
Input006_long_pair cau

The contrast to understand group differences in thickness between 8 and 16
yr old data sets are:
1 -1

*Question1:* Is this correct?
*Question 2:* If we use the below mentioned way to set up our second group
descriptor file, will the above contrasts still work? Or can we use 0 1
contrast with the file below to get at the same thing? We actually ran it
both ways and the results weren't similar but I specifically ask this
question because we are a little worried about the above 'second group
descriptor file' not having a column that says 'Main' and we are trying to
understand the right way to do our analyses.

*Example group descriptor file 2:*
GroupDescriptorFile 1
Title BEIP longitudinal Compare CAU (0) And FCG (1)
Class Main
Variables group
Input005_long_pair Main 1
Input006_long_pair Main 0

Thank you so much! I appreciate all your help with this!

Best,
Srishti
Social/Clinical Research Specialist
Child Imaging Research and Life Experiences Lab
University of North Carolina at Chapel Hill
email (W): srish...@email.unc.edu
skype: srishti.goel12

On Wed, Nov 7, 2018 at 9:24 AM Greve, Douglas N.,Ph.D. <
dgr...@mgh.harvard.edu> wrote:

>
>
> On 10/24/2018 09:45 PM, srishti goel wrote:
> >
> > External Email - Use Caution
> >
> > Hello,
> >
> > I am trying to perform a paired analysis for a data set that has
> > subjects' who belong to either group1 or group2 and have data at time
> > point1 (tp1) and time point2 (tp2). I am following all the steps on
> > the wiki but I want to clarify if the contrasts I plan to use for my
> > analyses are appropriate.
> >
> > For my analyses, I only have two groups and no covariates. I am
> > interested to know if the decrease in cortical thickness of subjects
> > in group 1 across tp1 and tp2 is more than the decrease in cortical
> > thickness of subjects in group 2 across tp1 and tp2. The contrast I
> > think should be used for this is: 1 -1. Is that correct?
> Yes, if you first take the difference between tp1 and tp2.
> >
> > I am also interested to know if the cortical thickness decreased for
> > subjects regardless of the groups so should I use the contrast: 1 for
> > that?
> Assuming you are keeping 2 groups, then use
>   0.5 0.5
> >
> > Best,
> > Srishti
> > Social/Clinical Research Specialist
> > Child Imaging Research and Life Experiences Lab
> > University of North Carolina at Chapel Hill
> > email (W): srish...@email.unc.edu 
> > skype: srishti.goel12
> >
> >
> > ___
> > Freesurfer mailing list
> > Freesurfer@nmr.mgh.harvard.edu
> > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
>
>
> ___
> Freesurfer mailing list
> Freesurfer@nmr.mgh.harvard.edu
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Re: [Freesurfer] Paired Analysis Contrasts

2018-11-07 Thread Greve, Douglas N.,Ph.D.



On 10/24/2018 09:45 PM, srishti goel wrote:
>
> External Email - Use Caution
>
> Hello,
>
> I am trying to perform a paired analysis for a data set that has 
> subjects' who belong to either group1 or group2 and have data at time 
> point1 (tp1) and time point2 (tp2). I am following all the steps on 
> the wiki but I want to clarify if the contrasts I plan to use for my 
> analyses are appropriate.
>
> For my analyses, I only have two groups and no covariates. I am 
> interested to know if the decrease in cortical thickness of subjects 
> in group 1 across tp1 and tp2 is more than the decrease in cortical 
> thickness of subjects in group 2 across tp1 and tp2. The contrast I 
> think should be used for this is: 1 -1. Is that correct?
Yes, if you first take the difference between tp1 and tp2.
>
> I am also interested to know if the cortical thickness decreased for 
> subjects regardless of the groups so should I use the contrast: 1 for 
> that?
Assuming you are keeping 2 groups, then use
  0.5 0.5
>
> Best,
> Srishti
> Social/Clinical Research Specialist
> Child Imaging Research and Life Experiences Lab
> University of North Carolina at Chapel Hill
> email (W): srish...@email.unc.edu 
> skype: srishti.goel12
>
>
> ___
> Freesurfer mailing list
> Freesurfer@nmr.mgh.harvard.edu
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer


___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

[Freesurfer] Paired Analysis Contrasts

2018-10-24 Thread srishti goel

External Email - Use Caution

Hello,

I am trying to perform a paired analysis for a data set that has subjects'
who belong to either group1 or group2 and have data at time point1 (tp1)
and time point2 (tp2). I am following all the steps on the wiki but I want
to clarify if the contrasts I plan to use for my analyses are appropriate.

For my analyses, I only have two groups and no covariates. I am interested
to know if the decrease in cortical thickness of subjects in group 1 across
tp1 and tp2 is more than the decrease in cortical thickness of subjects in
group 2 across tp1 and tp2. The contrast I think should be used for this
is: 1 -1. Is that correct?

I am also interested to know if the cortical thickness decreased for
subjects regardless of the groups so should I use the contrast: 1 for that?

Best,
Srishti
Social/Clinical Research Specialist
Child Imaging Research and Life Experiences Lab
University of North Carolina at Chapel Hill
email (W): srish...@email.unc.edu
skype: srishti.goel12
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Re: [Freesurfer] Paired Analysis Contrasts

Re: [Freesurfer] Paired Analysis Contrasts

Re: [Freesurfer] Paired Analysis Contrasts

Re: [Freesurfer] Paired Analysis Contrasts

[Freesurfer] Paired Analysis Contrasts

5 matches

Site Navigation

Mail list logo

Footer information