subject:"Re\: \[Freesurfer\] FSFAST first level covariates"

Re: [Freesurfer] FSFAST first level covariates

2022-07-28 Thread Douglas N. Greve

Yes, see 
https://surfer.nmr.mgh.harvard.edu/fswiki/FsFastParametricModulation



On 7/25/2022 6:56 PM, Angela Fang wrote:


External Email - Use Caution

Hi Freesurfer community,

I have run participants through an event-related fMRI task in which 
subjects rate whether trait adjectives are descriptive of themselves 
or not, and afterwards asked them to rate each trait word on emotional 
valence. Is it possible to include these individual level subjective 
ratings of emotional valence as covariates in the first level contrast 
in FSFAST? If so, how?


Thanks,

Angela

---

Angela Fang, Ph.D.

Assistant Professor
Department of Psychology

University of Washington

Lab website: *MailScanner has detected a possible fraud attempt from 
"secure-web.cisco.com" claiming to be* www.uwconnectlab.com 



Pronouns: she, her, hers


___
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
The information in this e-mail is intended only for the person to whom it is 
addressed.  If you believe this e-mail was sent to you in error and the e-mail 
contains patient information, please contact the Mass General Brigham 
Compliance HelpLine at https://www.massgeneralbrigham.org/complianceline 
 .
Please note that this e-mail is not secure (encrypted).  If you do not wish to 
continue communication over unencrypted e-mail, please notify the sender of 
this message immediately.  Continuing to send or respond to e-mail after 
receiving this message means you understand and accept this risk and wish to 
continue to communicate over unencrypted e-mail.

Re: [Freesurfer] FSFAST first level covariates

2022-07-28 Thread Angela Fang

Thanks Doug. This wiki page is extremely helpful. However, my question is about 
individual subject responses. I could see how you could include a summary 
(e.g., average) value of the parametric variable across subjects in your 
“weight” column but it’s not clear to me how you could integrate individual 
subject responses to each word in the parametric modulation paradigm file? I’m 
imagining something like the FSGD file where a value is given for each subject, 
but for first-level analysis.

We have a similar design as someone else who posted a similar question 
(https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg19957.html). We 
have an event-related experiment presenting trait adjectives in terms of 
whether they describe themselves (SELF condition) or someone else (OTHER 
condition). We are interested in testing a 2x2 ANOVA to examine an interaction 
between self-relevance x emotional valence. Assuming you can’t integrate 
individual subject responses to each word in the paradigm file, would we set it 
up as follows?

“Usual” paradigm file:
0  1  2.5  1.0  SELF
2.5  0  2.5  1.0  FIXATION
5.0  1  2.5  1.0  SELF
7.5  2  2.5  1.0  OTHER

Parametric modulation paradigm file:
0  1  2.5  1.0  SELFoffset
0  2  2.5  0.8  SELFslope
0  3  2.5  1.0  VALENCEoffset
0  4  2.5  2.0  VALENCEslope

(where 0.8 reflects the percentage of time the word was endorsed as 
self-relevant and 2.0 is the average valence rating given for that word)

And then create a contrast of 2 vs 4 to test the interaction? Would testing 
contrast 1 vs 0 be a test of the main effect of self-relevance and contrast 3 
vs 0 the main effect of valence?

Thanks so much for your help!
Angela

From:  on behalf of "Douglas N. Greve" 

Reply-To: Freesurfer support list 
Date: Thursday, July 28, 2022 at 10:25 AM
To: "freesurfer@nmr.mgh.harvard.edu" 
Subject: Re: [Freesurfer] FSFAST first level covariates

Yes, see https://surfer.nmr.mgh.harvard.edu/fswiki/FsFastParametricModulation

On 7/25/2022 6:56 PM, Angela Fang wrote:

External Email - Use Caution
Hi Freesurfer community,
I have run participants through an event-related fMRI task in which subjects 
rate whether trait adjectives are descriptive of themselves or not, and 
afterwards asked them to rate each trait word on emotional valence. Is it 
possible to include these individual level subjective ratings of emotional 
valence as covariates in the first level contrast in FSFAST? If so, how?
Thanks,
Angela

---
Angela Fang, Ph.D.
Assistant Professor
Department of Psychology
University of Washington
Lab website: MailScanner has detected a possible fraud attempt from 
"secure-web.cisco.com" claiming to be 
www.uwconnectlab.com<http://secure-web.cisco.com/1EJiZPvng9zcCEb0OA91-g6prvoE6x8E2RuCN3eRqnJvVyWnOZIhlVqHUDwmp2WEJbpju-V9f5K_n7JLkG1IXMgF6ntSJr6Aa91zEsfied2HyjVTkJZCXxMUYkbVQwHF0Z6PuuEgJy0xZF4iWLQMnW1WHJWnTXmRF8Gn92JXWkNX3veeq3YElwnE8vUuXUBgXGmmFtbKV7dJKfvdquLkJq0ApBJuwyNBrhKCoiBCQx4GIbuAfZ0iAsH5aL5xid2f1julAWTaA4EyB2BlTn9hUZcuoJNQPs89B-fNI4SA4inRY7YD67lmo-MehxlSgl0xO/http%3A%2F%2Fwww.uwconnectlab.com>
Pronouns: she, her, hers




___

Freesurfer mailing list

Freesurfer@nmr.mgh.harvard.edu<mailto: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
The information in this e-mail is intended only for the person to whom it is 
addressed.  If you believe this e-mail was sent to you in error and the e-mail 
contains patient information, please contact the Mass General Brigham 
Compliance HelpLine at https://www.massgeneralbrigham.org/complianceline 
<https://www.massgeneralbrigham.org/complianceline> .
Please note that this e-mail is not secure (encrypted).  If you do not wish to 
continue communication over unencrypted e-mail, please notify the sender of 
this message immediately.  Continuing to send or respond to e-mail after 
receiving this message means you understand and accept this risk and wish to 
continue to communicate over unencrypted e-mail.

Re: [Freesurfer] FSFAST first level covariates

2022-08-01 Thread Angela Fang

External Email - Use Caution

Hi Doug,

Nevermind to my first question! I read this post 
(https://secure-web.cisco.com/1646ymi0_yM9ab72e81bZdCKw_zNbXr9RihxDaiDVPq0_Qd4EXYgDmO56zQdi9l_AyV3uyyiURXHoYWQmiu56CbMuIGdZz8EH0gbsnVrAz9KwunZAwLzh0kh-jzVwHtlbEdd1ExEJYHT7o7JtUWg2GM484JTyL0VZJymRuGRyD0ag1nQ_0BPPjQHxPCqNHEU4Y_seBsq9XsUROgyR-bX-tHVXxhshVUHgneudw6tEB2lIVYfYrL3srRbjy1QN9Bq_e3_WaNCDhkXdixnae24i41HHYwJfn3KwsmNoZ2RxLoh3SMkXXwVntAewl8PeldBY0s3UxoEPiFbDdXXuJLUjlw/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg32235.html)
 and realized that we always include a subject-specific par file in each run 
for first-level analyses.

However, I’m still confused about how to modify my paradigm file. I also need 
to model the trials of non-interest, so would it be as follows?

0 1  2.5  1.0  SelfOffset
0  2  2.5  1.0  SelfSlope (equal to 
subject’s rating of self-relevance)
0  3  2.5  1.0  ValenceOffset
0  4  2.5  3.0  ValenceSlope (equal to 
subject’s rating of valence)

2.5  0  2.5  1.0  FIXATION

5.0  1  2.5  1.0  SelfOffset
5.0  2  2.5  0  SelfSlope (equal to 
subject’s rating of self-relevance, in this case subject responded 0, or 
non-relevant)
5.0  3  2.5  1.0  ValenceOffset
5.0  4  2.5  2.0  ValenceSlope (equal to 
subject’s rating of valence)

7.5  5  2.5  1.0  OTHER

Do these contrasts look correct to you?
Self vs Fixation -a 1 -c 0 (main effect of self)
Valence vs Fixation -a 3 -c 0 (main effect of valence)
Self vs Valence -a 2 -a 4 (interaction between self x valence)

Thank you so much for your help!
Angela

From:  on behalf of Angela Fang 

Reply-To: Freesurfer support list 
Date: Thursday, July 28, 2022 at 1:02 PM
To: Freesurfer support list 
Subject: Re: [Freesurfer] FSFAST first level covariates

Thanks Doug. This wiki page is extremely helpful. However, my question is about 
individual subject responses. I could see how you could include a summary 
(e.g., average) value of the parametric variable across subjects in your 
“weight” column but it’s not clear to me how you could integrate individual 
subject responses to each word in the parametric modulation paradigm file? I’m 
imagining something like the FSGD file where a value is given for each subject, 
but for first-level analysis.

We have a similar design as someone else who posted a similar question 
(https://secure-web.cisco.com/11nFbIrJYBqRI1W_4wY-HvfdEF3GG6xLL8So8t0i9yKbcElVyl_nJoDI6XedAGY2kKd_eP-dnsWeccOw2qajd375GRCeiUjqaXv3C7vOkrGEOiSiqfcPQ9y73ROdtl0jJIGemdoYQDd3GcX-dKx6qDwBcPE_qNlqxB0ZTcsDfTwK88OkoVtftMo1zKBWSiZBV9p0GO2erUcSoXtVI-AITDr9jULRDzVL_IzxtPdtuSBrYXMASRi7ex2oKftjJjyG_HMgygf_ULhSYIsHviihCwfx4uO5_zrvh8H84AxAsv33zsFjOaYeZ826JkD3E99hxrAKW3jYr3PjfN-zNZjQLJA/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html).
 We have an event-related experiment presenting trait adjectives in terms of 
whether they describe themselves (SELF condition) or someone else (OTHER 
condition). We are interested in testing a 2x2 ANOVA to examine an interaction 
between self-relevance x emotional valence. Assuming you can’t integrate 
individual subject responses to each word in the paradigm file, would we set it 
up as follows?

“Usual” paradigm file:
0  1  2.5  1.0  SELF
2.5  0  2.5  1.0  FIXATION
5.0  1  2.5  1.0  SELF
7.5  2  2.5  1.0  OTHER

Parametric modulation paradigm file:
0  1  2.5  1.0  SELFoffset
0  2  2.5  0.8  SELFslope
0  3  2.5  1.0  VALENCEoffset
0  4  2.5  2.0  VALENCEslope

(where 0.8 reflects the percentage of time the word was endorsed as 
self-relevant and 2.0 is the average valence rating given for that word)

And then create a contrast of 2 vs 4 to test the interaction? Would testing 
contrast 1 vs 0 be a test of the main effect of self-relevance and contrast 3 
vs 0 the main effect of valence?

Thanks so much for your help!
Angela

From:  on behalf of "Douglas N. Greve" 

Reply-To: Freesurfer support list 
Date: Thursday, July 28, 2022 at 10:25 AM
To: "freesurfer@nmr.mgh.harvard.edu" 
Subject: Re: [Freesurfer] FSFAST first level covariates

Yes, see 
https://secure-web.cisco.com/1vlnv3wLgT6AWyuomHXVnJCfD3bAT8O6KYN-6kv4DVE_Kbs9JwI6WLDqHM7UN7cfJ1TP0eQKgCtR-KXf01ehJnqsV2jW5XmAXQr0QnOlGk4--dT54zncT2aoK1njMKmN9ayqCJ_tFar2vbW-JGXSkTcg6gdUPh_mngiG7m

Re: [Freesurfer] FSFAST first level covariates

2022-08-10 Thread Angela Fang

External Email - Use Caution

Hello,

Just re-sending my question below. If I have a variable with 2 levels (yes/no) 
and another variable that is continuous, based on this post 
(https://secure-web.cisco.com/1S2s08xk6_r2FFsEB5S1KdOcfq6G8ToJwyZuNFONdwgOYd87JJkB-uznJW2pelg24KQwX3lweVOmFs99TCKitjbJOqKWgEH_UW7wir5JQ113csODerDntanBrEibOdt6Mxs2QeQ5D7n69Ds6NaOSOJIbLFeMjuoaTXCkNccNydn7jvjmVd0zW2YhEXG9JtLxMNVIYt8q48ZK0sJUt8sjTP6xuCzA1pzB19MUHA078Zgygtns0YVgn1n5Sg41ZbVZ3jWciX5ZF34AejW5nWj1Z4mWO1Xyd_7RwNbKkVMPeDwG6K9W59gzBf_t0G-AzmUhxGC8zfKM0bxA9hhZv4GR2BQ/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html),
 it sounds like I should code as follows:


  1.  SelfOffset
  2.  Self-ValenceSlope (would the weight in the 4th column reflect the value 
of self multiplied by the value of valence for this participant?)
  3.  NonSelfOffset
  4.  NonSelf-ValenceSlope

If the other way of modifying the paradigm file is also acceptable to test the 
interaction (as I describe below), that would also be helpful to know.

Thanks!
Angela

From:  on behalf of Angela Fang 

Reply-To: Freesurfer support list 
Date: Monday, August 1, 2022 at 4:35 PM
To: Freesurfer support list 
Subject: Re: [Freesurfer] FSFAST first level covariates


External Email - Use Caution
Hi Doug,

Nevermind to my first question! I read this post (MailScanner has detected a 
possible fraud attempt from "secure-web.cisco.com" claiming to be 
https://secure-web.cisco.com/1CMVIrMV5SzmWczLBYJ8wrbs4I4fTd5TDfoZl4aAcMRL9BIQ-Qtf_TNzQRr7zOVT5cTiOFV8Ra3a94S103KieE3Tx520Wc4b13Figq023v6mGeLMvO4ta3onkEAhJTj2BPvgC68B6vwZKYOcU5mDuxxjAtn326ZOpm1upylv6_uFul29AAJsOxJDurtJ2p-3ktjc51QB3JOLJ_DxxHV8RQayh1MZMEkZETrsXsAYS8AzD9VTXBGYU8dxrvNBiJNcqZj1E5qRoV6I70Z6gjddHCEkR-mmClpRrGkHTBUq-0_Zgg10-XRZQFiw2xc9V9h595go0ZToz7UxuSDBebc9JAA/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg32235.html<https://secure-web.cisco.com/1646ymi0_yM9ab72e81bZdCKw_zNbXr9RihxDaiDVPq0_Qd4EXYgDmO56zQdi9l_AyV3uyyiURXHoYWQmiu56CbMuIGdZz8EH0gbsnVrAz9KwunZAwLzh0kh-jzVwHtlbEdd1ExEJYHT7o7JtUWg2GM484JTyL0VZJymRuGRyD0ag1nQ_0BPPjQHxPCqNHEU4Y_seBsq9XsUROgyR-bX-tHVXxhshVUHgneudw6tEB2lIVYfYrL3srRbjy1QN9Bq_e3_WaNCDhkXdixnae24i41HHYwJfn3KwsmNoZ2RxLoh3SMkXXwVntAewl8PeldBY0s3UxoEPiFbDdXXuJLUjlw/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg32235.html>)
 and realized that we always include a subject-specific par file in each run 
for first-level analyses.

However, I’m still confused about how to modify my paradigm file. I also need 
to model the trials of non-interest, so would it be as follows?

0 1  2.5  1.0  SelfOffset
0  2  2.5  1.0  SelfSlope (equal to 
subject’s rating of self-relevance)
0  3  2.5  1.0  ValenceOffset
0  4  2.5  3.0  ValenceSlope (equal to 
subject’s rating of valence)

2.5  0  2.5  1.0  FIXATION

5.0  1  2.5  1.0  SelfOffset
5.0  2  2.5  0  SelfSlope (equal to 
subject’s rating of self-relevance, in this case subject responded 0, or 
non-relevant)
5.0  3  2.5  1.0  ValenceOffset
5.0  4  2.5  2.0  ValenceSlope (equal to 
subject’s rating of valence)

7.5  5  2.5  1.0  OTHER

Do these contrasts look correct to you?
Self vs Fixation -a 1 -c 0 (main effect of self)
Valence vs Fixation -a 3 -c 0 (main effect of valence)
Self vs Valence -a 2 -a 4 (interaction between self x valence)

Thank you so much for your help!
Angela

From:  on behalf of Angela Fang 

Reply-To: Freesurfer support list 
Date: Thursday, July 28, 2022 at 1:02 PM
To: Freesurfer support list 
Subject: Re: [Freesurfer] FSFAST first level covariates

Thanks Doug. This wiki page is extremely helpful. However, my question is about 
individual subject responses. I could see how you could include a summary 
(e.g., average) value of the parametric variable across subjects in your 
“weight” column but it’s not clear to me how you could integrate individual 
subject responses to each word in the parametric modulation paradigm file? I’m 
imagining something like the FSGD file where a value is given for each subject, 
but for first-level analysis.

We have a similar design as someone else who posted a similar question 
(MailScanner has detected a possible fraud attempt from "secure-web.cisco.com" 
claiming to be 
https://secure-web.cisco.com/1S2s08xk6_r2FFsEB5S1KdOcfq6G8ToJwyZuNFONdwgOYd87JJkB-uznJW2pelg24KQwX3lweVOmFs99TCKitjbJOqKWgEH_UW7wir5JQ113csODerDntanBrEibOdt6Mxs2QeQ5D7n69Ds6NaOSOJIbLFeMjuoaTXCkNccNydn7jvjmVd0zW2YhEXG9JtLxMNVIYt8q48ZK0sJUt8sjTP6xuCzA1pzB19MUHA078Zgygtns0YVgn1n5Sg41ZbVZ3jWciX5ZF34AejW5nWj1Z4mWO1Xyd_7R

Re: [Freesurfer] FSFAST first level covariates

2022-08-14 Thread Douglas N. Greve

Those look like they are the same coding scheme. What is different? You 
can only have one offset. The Self vs Valence -a 2 -a 4 is not testing 
for an interaction. If you want an interaction you have to create a new 
variable which is SelfRating*ValenceRating.



On 8/10/2022 2:38 PM, Angela Fang wrote:


External Email - Use Caution

Hello,

Just re-sending my question below. If I have a variable with 2 levels 
(yes/no) and another variable that is continuous, based on this post 
(*MailScanner has detected a possible fraud attempt from 
"secure-web.cisco.com" claiming to be* 
https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg19957.html 
<https://secure-web.cisco.com/1S2s08xk6_r2FFsEB5S1KdOcfq6G8ToJwyZuNFONdwgOYd87JJkB-uznJW2pelg24KQwX3lweVOmFs99TCKitjbJOqKWgEH_UW7wir5JQ113csODerDntanBrEibOdt6Mxs2QeQ5D7n69Ds6NaOSOJIbLFeMjuoaTXCkNccNydn7jvjmVd0zW2YhEXG9JtLxMNVIYt8q48ZK0sJUt8sjTP6xuCzA1pzB19MUHA078Zgygtns0YVgn1n5Sg41ZbVZ3jWciX5ZF34AejW5nWj1Z4mWO1Xyd_7RwNbKkVMPeDwG6K9W59gzBf_t0G-AzmUhxGC8zfKM0bxA9hhZv4GR2BQ/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html>), 
it sounds like I should code as follows:


 1. SelfOffset
 2. Self-ValenceSlope (would the weight in the 4^th column reflect the
value of self multiplied by the value of valence for this
participant?)
 3. NonSelfOffset
 4. NonSelf-ValenceSlope

If the other way of modifying the paradigm file is also acceptable to 
test the interaction (as I describe below), that would also be helpful 
to know.


Thanks!

Angela

*From: * on behalf of Angela 
Fang 

*Reply-To: *Freesurfer support list 
*Date: *Monday, August 1, 2022 at 4:35 PM
*To: *Freesurfer support list 
*Subject: *Re: [Freesurfer] FSFAST first level covariates

*External Email - Use Caution *

Hi Doug,

Nevermind to my first question! I read this post (*MailScanner has 
detected a possible fraud attempt from "secure-web.cisco.com" claiming 
to be* 
https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg32235.html 
<https://secure-web.cisco.com/1646ymi0_yM9ab72e81bZdCKw_zNbXr9RihxDaiDVPq0_Qd4EXYgDmO56zQdi9l_AyV3uyyiURXHoYWQmiu56CbMuIGdZz8EH0gbsnVrAz9KwunZAwLzh0kh-jzVwHtlbEdd1ExEJYHT7o7JtUWg2GM484JTyL0VZJymRuGRyD0ag1nQ_0BPPjQHxPCqNHEU4Y_seBsq9XsUROgyR-bX-tHVXxhshVUHgneudw6tEB2lIVYfYrL3srRbjy1QN9Bq_e3_WaNCDhkXdixnae24i41HHYwJfn3KwsmNoZ2RxLoh3SMkXXwVntAewl8PeldBY0s3UxoEPiFbDdXXuJLUjlw/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg32235.html>) 
and realized that we always include a subject-specific par file in 
each run for first-level analyses.


However, I’m still confused about how to modify my paradigm file. I 
also need to model the trials of non-interest, so would it be as follows?


0 1  2.5  1.0  SelfOffset

0 2  2.5  1.0  SelfSlope (equal to 
subject’s rating of self-relevance)


0 3  2.5  1.0  ValenceOffset

0 4  2.5  3.0  ValenceSlope (equal to 
subject’s rating of valence)


2.5 0  2.5  1.0  FIXATION

5.0 1  2.5  1.0  SelfOffset

5.0 2  2.5  0  SelfSlope (equal to 
subject’s rating of self-relevance, in this case subject responded 0, 
or non-relevant)


5.0 3  2.5  1.0  ValenceOffset

5.0 4  2.5  2.0  ValenceSlope (equal to 
subject’s rating of valence)


7.5 5  2.5  1.0  OTHER

Do these contrasts look correct to you?

Self vs Fixation -a 1 -c 0 (main effect of self)

Valence vs Fixation -a 3 -c 0 (main effect of valence)

Self vs Valence -a 2 -a 4 (interaction between self x valence)

Thank you so much for your help!

Angela

*From: * on behalf of Angela 
Fang 

*Reply-To: *Freesurfer support list 
*Date: *Thursday, July 28, 2022 at 1:02 PM
*To: *Freesurfer support list 
*Subject: *Re: [Freesurfer] FSFAST first level covariates

Thanks Doug. This wiki page is extremely helpful. However, my question 
is about individual subject responses. I could see how you could 
include a summary (e.g., average) value of the parametric variable 
across subjects in your “weight” column but it’s not clear to me how 
you could integrate individual subject responses to each word in the 
parametric modulation paradigm file? I’m imagining something like the 
FSGD file where a value is given for each subject, but for first-level 
analysis.


We have a similar design as someone else who posted a similar question 
(*MailScanner has detected a possible fraud attempt from 
"secure-web.cisco.com" claiming to be* 
https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg19957.html 
<https://secure-web.cisco.com/11nFbIrJYBqRI1W_4wY-HvfdEF3GG6xLL8So8t0i9yKbcElVyl_nJoDI6XedAGY2kKd_eP-dnsWeccOw2qajd375GRCeiUjqaXv3C7vOkrGEOiSiqfcPQ9y73ROdtl0jJIGemdoYQDd3GcX-dKx6qDwBcPE_qNlqxB0Z

Re: [Freesurfer] FSFAST first level covariates

2022-08-14 Thread Angela Fang

External Email - Use Caution

The two coding schemes are different because the second one does include the 
self*valence variable you’re talking about, whereas the first one doesn’t. I 
only included the 2nd offset because you suggested to someone else to include 
it (see 
https://secure-web.cisco.com/1UvSz7SvjZnlE1QfkdS-VBc1GzXeZMYqUcsOqE3dGoo43anjoKIpvfs4NnozgTlCx23dB10wC_oFWTi8Zyazh1v1oufO7QQBf9hJanKAiwbu0cr4NfMvGMSOSaaOt5nSATHi-J-55MTqcCUhjz8_rRM1YuYWhtxzDVrlNJ5mD3QlEmdQlhRlYoneii_5mWjAZB1gcbpR_0Zl1nUaCy9BfmpcQRNLpIdfE1NMjg7OnqHOX5jAdPz1gGqxVTKgstTqx8RhxmTkDYOQPLc6hC3by-Atu2VfBnloD3GbZOzG04LjV1Of0uYaB6pk6oSsZFULTsifcwPSiwh1m9gKSg5lD7Q/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html).
 If we don’t need it, would it just be 2 conditions, as follows?


  1.  SelfOffset
  2.  Self*ValenceSlope

But then I’m not clear how to get the main effect of valence (brain regions 
that scale with increasing emotion valence, while holding self-relevance 
constant)?

From:  on behalf of "Douglas N. Greve" 

Reply-To: Freesurfer support list 
Date: Sunday, August 14, 2022 at 1:37 PM
To: "freesurfer@nmr.mgh.harvard.edu" 
Subject: Re: [Freesurfer] FSFAST first level covariates

Those look like they are the same coding scheme. What is different? You can 
only have one offset. The Self vs Valence -a 2 -a 4 is not testing for an 
interaction. If you want an interaction you have to create a new variable which 
is SelfRating*ValenceRating.

On 8/10/2022 2:38 PM, Angela Fang wrote:

External Email - Use Caution
Hello,

Just re-sending my question below. If I have a variable with 2 levels (yes/no) 
and another variable that is continuous, based on this post (MailScanner has 
detected a possible fraud attempt from "secure-web.cisco.com" claiming to be 
https://secure-web.cisco.com/1UvSz7SvjZnlE1QfkdS-VBc1GzXeZMYqUcsOqE3dGoo43anjoKIpvfs4NnozgTlCx23dB10wC_oFWTi8Zyazh1v1oufO7QQBf9hJanKAiwbu0cr4NfMvGMSOSaaOt5nSATHi-J-55MTqcCUhjz8_rRM1YuYWhtxzDVrlNJ5mD3QlEmdQlhRlYoneii_5mWjAZB1gcbpR_0Zl1nUaCy9BfmpcQRNLpIdfE1NMjg7OnqHOX5jAdPz1gGqxVTKgstTqx8RhxmTkDYOQPLc6hC3by-Atu2VfBnloD3GbZOzG04LjV1Of0uYaB6pk6oSsZFULTsifcwPSiwh1m9gKSg5lD7Q/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html<https://secure-web.cisco.com/1S2s08xk6_r2FFsEB5S1KdOcfq6G8ToJwyZuNFONdwgOYd87JJkB-uznJW2pelg24KQwX3lweVOmFs99TCKitjbJOqKWgEH_UW7wir5JQ113csODerDntanBrEibOdt6Mxs2QeQ5D7n69Ds6NaOSOJIbLFeMjuoaTXCkNccNydn7jvjmVd0zW2YhEXG9JtLxMNVIYt8q48ZK0sJUt8sjTP6xuCzA1pzB19MUHA078Zgygtns0YVgn1n5Sg41ZbVZ3jWciX5ZF34AejW5nWj1Z4mWO1Xyd_7RwNbKkVMPeDwG6K9W59gzBf_t0G-AzmUhxGC8zfKM0bxA9hhZv4GR2BQ/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html>),
 it sounds like I should code as follows:


  1.  SelfOffset
  2.  Self-ValenceSlope (would the weight in the 4th column reflect the value 
of self multiplied by the value of valence for this participant?)
  3.  NonSelfOffset
  4.  NonSelf-ValenceSlope

If the other way of modifying the paradigm file is also acceptable to test the 
interaction (as I describe below), that would also be helpful to know.

Thanks!
Angela

From: 
<mailto:freesurfer-boun...@nmr.mgh.harvard.edu>
 on behalf of Angela Fang <mailto:angf...@uw.edu>
Reply-To: Freesurfer support list 
<mailto:freesurfer@nmr.mgh.harvard.edu>
Date: Monday, August 1, 2022 at 4:35 PM
To: Freesurfer support list 
<mailto:freesurfer@nmr.mgh.harvard.edu>
Subject: Re: [Freesurfer] FSFAST first level covariates


External Email - Use Caution
Hi Doug,

Nevermind to my first question! I read this post (MailScanner has detected a 
possible fraud attempt from "secure-web.cisco.com" claiming to be 
https://secure-web.cisco.com/1HVMVTjRAiW9L4gY48Veh_kfiv1-IcKlH19YGnZA4uiivgCNnRBDiv1ha5bxC1DRvtBdo4S9DEUaBOOp996Bdoi48IxhMuUmcx4E11Q99LMPWXjV0ecDedB5lM7q-j6FaDPW_yfxmKxeLQI69yz9F7iGzCVSPjYO7_j48Y_0GQwwlNxpi1PogvndydGw96sgQmHDfsqftIDgpdKoQR5mq560WyTAUgVDQNiSRTqxYnUgrm6S315JtmMFs8DczARMB5Y3Dj19BwS9yGVTRpaOJSQTsTswFIgqA2CnTKjgtiUWGy8TCI2QjTXdaWfOdW83XCXkMMOwu1kghjraT1mnSqw/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg32235.html<https://secure-web.cisco.com/1646ymi0_yM9ab72e81bZdCKw_zNbXr9RihxDaiDVPq0_Qd4EXYgDmO56zQdi9l_AyV3uyyiURXHoYWQmiu56CbMuIGdZz8EH0gbsnVrAz9KwunZAwLzh0kh-jzVwHtlbEdd1ExEJYHT7o7JtUWg2GM484JTyL0VZJymRuGRyD0ag1nQ_0BPPjQHxPCqNHEU4Y_seBsq9XsUROgyR-bX-tHVXxhshVUHgneudw6tEB2lIVYfYrL3srRbjy1QN9Bq_e3_WaNCDhkXdixnae24i41HHYwJfn3KwsmNoZ2RxLoh3SMkXXwVntAewl8PeldBY0s3UxoEPiFbDdXXuJLUjlw/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg32235.html>)
 and realized that we always include a subject-specific par file in each run 
for first-level analyses.

However, I’m still confused about how to modify my paradigm file. I also need 
to model the trials of non-interest, so would it be as follows?

0 1  2.5

Re: [Freesurfer] FSFAST first level covariates

2022-08-28 Thread Douglas N. Greve

1. Offset
2. Self
3. Valence
4. Self*Valence
I've never tried the interaction (self*valence). You might have to
demean before computing the product

On 8/14/2022 4:58 PM, Angela Fang wrote:

External Email - Use Caution

The two coding schemes are different because the second one does
include the self*valence variable you’re talking about, whereas the
first one doesn’t. I only included the 2^nd offset because you
suggested to someone else to include it (see *MailScanner has detected
a possible fraud attempt from "secure-web.cisco.com" claiming to be*
https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg19957.html
<https://secure-web.cisco.com/1UvSz7SvjZnlE1QfkdS-VBc1GzXeZMYqUcsOqE3dGoo43anjoKIpvfs4NnozgTlCx23dB10wC_oFWTi8Zyazh1v1oufO7QQBf9hJanKAiwbu0cr4NfMvGMSOSaaOt5nSATHi-J-55MTqcCUhjz8_rRM1YuYWhtxzDVrlNJ5mD3QlEmdQlhRlYoneii_5mWjAZB1gcbpR_0Zl1nUaCy9BfmpcQRNLpIdfE1NMjg7OnqHOX5jAdPz1gGqxVTKgstTqx8RhxmTkDYOQPLc6hC3by-Atu2VfBnloD3GbZOzG04LjV1Of0uYaB6pk6oSsZFULTsifcwPSiwh1m9gKSg5lD7Q/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html>).
If we don’t need it, would it just be 2 conditions, as follows?

1. SelfOffset
2. Self*ValenceSlope

But then I’m not clear how to get the main effect of valence (brain
regions that scale with increasing emotion valence, while holding
self-relevance constant)?

*From: * on behalf of "Douglas
N. Greve"

*Reply-To: *Freesurfer support list
*Date: *Sunday, August 14, 2022 at 1:37 PM
*To: *"freesurfer@nmr.mgh.harvard.edu"
*Subject: *Re: [Freesurfer] FSFAST first level covariates

Those look like they are the same coding scheme. What is different?
You can only have one offset. The Self vs Valence -a 2 -a 4 is not
testing for an interaction. If you want an interaction you have to
create a new variable which is SelfRating*ValenceRating.

On 8/10/2022 2:38 PM, Angela Fang wrote:

*External Email - Use Caution *

Hello,

Just re-sending my question below. If I have a variable with 2
levels (yes/no) and another variable that is continuous, based on
this post (*MailScanner has detected a possible fraud attempt from
"secure-web.cisco.com" claiming to be*
https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg19957.html

<https://secure-web.cisco.com/1S2s08xk6_r2FFsEB5S1KdOcfq6G8ToJwyZuNFONdwgOYd87JJkB-uznJW2pelg24KQwX3lweVOmFs99TCKitjbJOqKWgEH_UW7wir5JQ113csODerDntanBrEibOdt6Mxs2QeQ5D7n69Ds6NaOSOJIbLFeMjuoaTXCkNccNydn7jvjmVd0zW2YhEXG9JtLxMNVIYt8q48ZK0sJUt8sjTP6xuCzA1pzB19MUHA078Zgygtns0YVgn1n5Sg41ZbVZ3jWciX5ZF34AejW5nWj1Z4mWO1Xyd_7RwNbKkVMPeDwG6K9W59gzBf_t0G-AzmUhxGC8zfKM0bxA9hhZv4GR2BQ/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html>),
it sounds like I should code as follows:

1. SelfOffset
2. Self-ValenceSlope (would the weight in the 4^th column reflect
the value of self multiplied by the value of valence for this
participant?)
3. NonSelfOffset
4. NonSelf-ValenceSlope

If the other way of modifying the paradigm file is also acceptable
to test the interaction (as I describe below), that would also be
helpful to know.

Thanks!

Angela

*From: *
<mailto:freesurfer-boun...@nmr.mgh.harvard.edu> on behalf of
Angela Fang <mailto:angf...@uw.edu>
*Reply-To: *Freesurfer support list

<mailto:freesurfer@nmr.mgh.harvard.edu>
*Date: *Monday, August 1, 2022 at 4:35 PM
*To: *Freesurfer support list
<mailto:freesurfer@nmr.mgh.harvard.edu>
*Subject: *Re: [Freesurfer] FSFAST first level covariates

*External Email - Use Caution *

Hi Doug,

Nevermind to my first question! I read this post (*MailScanner has
detected a possible fraud attempt from "secure-web.cisco.com"
claiming to be*
https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg32235.html

<https://secure-web.cisco.com/1646ymi0_yM9ab72e81bZdCKw_zNbXr9RihxDaiDVPq0_Qd4EXYgDmO56zQdi9l_AyV3uyyiURXHoYWQmiu56CbMuIGdZz8EH0gbsnVrAz9KwunZAwLzh0kh-jzVwHtlbEdd1ExEJYHT7o7JtUWg2GM484JTyL0VZJymRuGRyD0ag1nQ_0BPPjQHxPCqNHEU4Y_seBsq9XsUROgyR-bX-tHVXxhshVUHgneudw6tEB2lIVYfYrL3srRbjy1QN9Bq_e3_WaNCDhkXdixnae24i41HHYwJfn3KwsmNoZ2RxLoh3SMkXXwVntAewl8PeldBY0s3UxoEPiFbDdXXuJLUjlw/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg32235.html>)
and realized that we always include a subject-specific par file in
each run for first-level analyses.

However, I’m still confused about how to modify my paradigm file.
I also need to model the trials of non-interest, so would it be as

Re: [Freesurfer] FSFAST first level covariates

2022-08-29 Thread Angela Fang

External Email - Use Caution

Great, thanks so much Doug- this setup makes sense to me. However, isn’t it a 
problem that the brain regions associated with the contrast for -a 3 and the 
contrast for -a 4 would be the same for trials that subjects rate as 1 
(self-relevant) vs 0 (not self-relevant)? Could we code it as 2 (self-relevant) 
and 1 (not self-relevant)?

Many thanks!!

From:  on behalf of "Douglas N. Greve" 

Reply-To: Freesurfer support list 
Date: Sunday, August 28, 2022 at 1:07 PM
To: "freesurfer@nmr.mgh.harvard.edu" 
Subject: Re: [Freesurfer] FSFAST first level covariates

I was under the impression that Self and Valence were ratings from the same 
event (in that mail archive, they were different events and so needed different 
offsets). If Self and Valence are from the same event, then you would have 
something like
1. Offset
2. Self
3. Valence
4. Self*Valence
I've never tried the interaction (self*valence). You might have to demean 
before computing the product
On 8/14/2022 4:58 PM, Angela Fang wrote:

External Email - Use Caution
The two coding schemes are different because the second one does include the 
self*valence variable you’re talking about, whereas the first one doesn’t. I 
only included the 2nd offset because you suggested to someone else to include 
it (see MailScanner has detected a possible fraud attempt from 
"secure-web.cisco.com" claiming to be 
https://secure-web.cisco.com/1sZZR9kggc35UyK7lSAda3dE6il9II_AD7YGowE7cmcitllK2WkMkXZrYMKNHWVfSSVhSBTpzhl3tRvJs-7MjxeemfsGXeS63jstCurLf6ur3TzjDLZaSY3nEp6mFhrKVYKnRkv7Wdxkn-BI3_cR0HvigETBgmiDZpQpzNHRDgc-7DYuwysbhd26aCuwOQ6YPKSgA5EhF0RwgNj-Y5XAyBeBe3b0QJ7Xgsxvk-wejbxLU2hQx3CIxif6YZC-0J5u3IdWQ1xP3WxhoSZ_z3MEylWxOEDyLyIWLX2dGKMd6-xvTML2huIRERSgZF-zI8ZNUPUzt2Klohmq9IfNQbdNZVw/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html<https://secure-web.cisco.com/1UvSz7SvjZnlE1QfkdS-VBc1GzXeZMYqUcsOqE3dGoo43anjoKIpvfs4NnozgTlCx23dB10wC_oFWTi8Zyazh1v1oufO7QQBf9hJanKAiwbu0cr4NfMvGMSOSaaOt5nSATHi-J-55MTqcCUhjz8_rRM1YuYWhtxzDVrlNJ5mD3QlEmdQlhRlYoneii_5mWjAZB1gcbpR_0Zl1nUaCy9BfmpcQRNLpIdfE1NMjg7OnqHOX5jAdPz1gGqxVTKgstTqx8RhxmTkDYOQPLc6hC3by-Atu2VfBnloD3GbZOzG04LjV1Of0uYaB6pk6oSsZFULTsifcwPSiwh1m9gKSg5lD7Q/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html>).
 If we don’t need it, would it just be 2 conditions, as follows?


  1.  SelfOffset
  2.  Self*ValenceSlope

But then I’m not clear how to get the main effect of valence (brain regions 
that scale with increasing emotion valence, while holding self-relevance 
constant)?

From: 
<mailto:freesurfer-boun...@nmr.mgh.harvard.edu>
 on behalf of "Douglas N. Greve" 
<mailto:dgr...@mgh.harvard.edu>
Reply-To: Freesurfer support list 
<mailto:freesurfer@nmr.mgh.harvard.edu>
Date: Sunday, August 14, 2022 at 1:37 PM
To: "freesurfer@nmr.mgh.harvard.edu"<mailto:freesurfer@nmr.mgh.harvard.edu> 
<mailto:freesurfer@nmr.mgh.harvard.edu>
Subject: Re: [Freesurfer] FSFAST first level covariates

Those look like they are the same coding scheme. What is different? You can 
only have one offset. The Self vs Valence -a 2 -a 4 is not testing for an 
interaction. If you want an interaction you have to create a new variable which 
is SelfRating*ValenceRating.

On 8/10/2022 2:38 PM, Angela Fang wrote:

External Email - Use Caution
Hello,

Just re-sending my question below. If I have a variable with 2 levels (yes/no) 
and another variable that is continuous, based on this post (MailScanner has 
detected a possible fraud attempt from "secure-web.cisco.com" claiming to be 
https://secure-web.cisco.com/1sZZR9kggc35UyK7lSAda3dE6il9II_AD7YGowE7cmcitllK2WkMkXZrYMKNHWVfSSVhSBTpzhl3tRvJs-7MjxeemfsGXeS63jstCurLf6ur3TzjDLZaSY3nEp6mFhrKVYKnRkv7Wdxkn-BI3_cR0HvigETBgmiDZpQpzNHRDgc-7DYuwysbhd26aCuwOQ6YPKSgA5EhF0RwgNj-Y5XAyBeBe3b0QJ7Xgsxvk-wejbxLU2hQx3CIxif6YZC-0J5u3IdWQ1xP3WxhoSZ_z3MEylWxOEDyLyIWLX2dGKMd6-xvTML2huIRERSgZF-zI8ZNUPUzt2Klohmq9IfNQbdNZVw/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html<https://secure-web.cisco.com/1S2s08xk6_r2FFsEB5S1KdOcfq6G8ToJwyZuNFONdwgOYd87JJkB-uznJW2pelg24KQwX3lweVOmFs99TCKitjbJOqKWgEH_UW7wir5JQ113csODerDntanBrEibOdt6Mxs2QeQ5D7n69Ds6NaOSOJIbLFeMjuoaTXCkNccNydn7jvjmVd0zW2YhEXG9JtLxMNVIYt8q48ZK0sJUt8sjTP6xuCzA1pzB19MUHA078Zgygtns0YVgn1n5Sg41ZbVZ3jWciX5ZF34AejW5nWj1Z4mWO1Xyd_7RwNbKkVMPeDwG6K9W59gzBf_t0G-AzmUhxGC8zfKM0bxA9hhZv4GR2BQ/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html>),
 it sounds like I should code as follows:


  1.  SelfOffset
  2.  Self-ValenceSlope (would the weight in the 4th column reflect the value 
of self multiplied by the value of valence for this participant?)
  3.  NonSelfOffset
  4.  NonSelf-ValenceSlope

If the other way of modifying the paradigm file is also acceptable to test the 
interaction

Re: [Freesurfer] FSFAST first level covariates

2022-08-30 Thread Douglas N. Greve

Oh, I think I see now. In realty, each trial is either self or not self
and then you have a separate valence measure, so your coding should be

1. Self-offset
2. SelfValence
3. NoSelf-offset
4. NoSelfValence

Where 1+2 are used for a self event and 3+4 are used for a noself event.
If you want to test for the difference between self and noself valence
slope, then you would just use -a 2 -c 4

On 8/29/2022 12:08 PM, Angela Fang wrote:

External Email - Use Caution

Great, thanks so much Doug- this setup makes sense to me. However,
isn’t it a problem that the brain regions associated with the contrast
for -a 3 and the contrast for -a 4 would be the same for trials that
subjects rate as 1 (self-relevant) vs 0 (not self-relevant)? Could we
code it as 2 (self-relevant) and 1 (not self-relevant)?

Many thanks!!

*From: * on behalf of "Douglas
N. Greve"

*Reply-To: *Freesurfer support list
*Date: *Sunday, August 28, 2022 at 1:07 PM
*To: *"freesurfer@nmr.mgh.harvard.edu"
*Subject: *Re: [Freesurfer] FSFAST first level covariates

1. Offset
2. Self
3. Valence
4. Self*Valence
I've never tried the interaction (self*valence). You might have to
demean before computing the product

On 8/14/2022 4:58 PM, Angela Fang wrote:

*External Email - Use Caution *

<https://secure-web.cisco.com/1UvSz7SvjZnlE1QfkdS-VBc1GzXeZMYqUcsOqE3dGoo43anjoKIpvfs4NnozgTlCx23dB10wC_oFWTi8Zyazh1v1oufO7QQBf9hJanKAiwbu0cr4NfMvGMSOSaaOt5nSATHi-J-55MTqcCUhjz8_rRM1YuYWhtxzDVrlNJ5mD3QlEmdQlhRlYoneii_5mWjAZB1gcbpR_0Zl1nUaCy9BfmpcQRNLpIdfE1NMjg7OnqHOX5jAdPz1gGqxVTKgstTqx8RhxmTkDYOQPLc6hC3by-Atu2VfBnloD3GbZOzG04LjV1Of0uYaB6pk6oSsZFULTsifcwPSiwh1m9gKSg5lD7Q/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html>).
If we don’t need it, would it just be 2 conditions, as follows?

1. SelfOffset
2. Self*ValenceSlope

But then I’m not clear how to get the main effect of valence
(brain regions that scale with increasing emotion valence, while
holding self-relevance constant)?

*From: *
<mailto:freesurfer-boun...@nmr.mgh.harvard.edu> on behalf of
"Douglas N. Greve"
<mailto:dgr...@mgh.harvard.edu>
*Reply-To: *Freesurfer support list

<mailto:freesurfer@nmr.mgh.harvard.edu>
*Date: *Sunday, August 14, 2022 at 1:37 PM
*To: *"freesurfer@nmr.mgh.harvard.edu"
<mailto:freesurfer@nmr.mgh.harvard.edu>

<mailto:freesurfer@nmr.mgh.harvard.edu>
*Subject: *Re: [Freesurfer] FSFAST first level covariates

Those look like they are the same coding scheme. What is
different? You can only have one offset. The Self vs Valence -a 2
-a 4 is not testing for an interaction. If you want an interaction
you have to create a new variable which is SelfRating*ValenceRating.

On 8/10/2022 2:38 PM, Angela Fang wrote:

*External Email - Use Caution *

Hello,

Just re-sending my question below. If I have a variable with 2
levels (yes/no) and another variable that is continuous, based
on this post (*MailScanner has detected a possible fraud
attempt from "secure-web.cisco.com" claiming to be*

https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg19957.html

1. SelfOffset
2. Self-ValenceSlope (would the weight in the 4^th column
reflect the value of self multiplied by the value of
valence for this participant?)
3. NonSelfOffset
4. NonSelf-ValenceSlope

If the other way of modifying the paradigm file is also
acceptable to test the interaction (as I describe below), that
would

Re: [Freesurfer] FSFAST first level covariates

2022-08-31 Thread Angela Fang

External Email - Use Caution

That makes sense! And if we wanted to identify brain regions associated with 
valence (across self/non-self) that is unique from condition 2, can we add a 
5th condition? Maybe we can code self as 2 and non-self as 1 so that condition 
2 (self*valence) would be different from condition 5 (valence ratings only)?

Incredibly helpful, Doug- thank you so much!

From:  on behalf of "Douglas N. Greve" 

Reply-To: Freesurfer support list 
Date: Tuesday, August 30, 2022 at 7:38 AM
To: "freesurfer@nmr.mgh.harvard.edu" 
Subject: Re: [Freesurfer] FSFAST first level covariates

Oh, I think I see now. In realty, each trial is either self or not self and 
then you have a separate valence measure, so your coding should be
1. Self-offset
2. SelfValence
3. NoSelf-offset
4. NoSelfValence

Where 1+2 are used for a self event and 3+4 are used for a noself event. If you 
want to test for the difference between self and noself valence slope, then you 
would just use -a 2 -c 4

On 8/29/2022 12:08 PM, Angela Fang wrote:

External Email - Use Caution
Great, thanks so much Doug- this setup makes sense to me. However, isn’t it a 
problem that the brain regions associated with the contrast for -a 3 and the 
contrast for -a 4 would be the same for trials that subjects rate as 1 
(self-relevant) vs 0 (not self-relevant)? Could we code it as 2 (self-relevant) 
and 1 (not self-relevant)?

Many thanks!!

From: 
<mailto:freesurfer-boun...@nmr.mgh.harvard.edu>
 on behalf of "Douglas N. Greve" 
<mailto:dgr...@mgh.harvard.edu>
Reply-To: Freesurfer support list 
<mailto:freesurfer@nmr.mgh.harvard.edu>
Date: Sunday, August 28, 2022 at 1:07 PM
To: "freesurfer@nmr.mgh.harvard.edu"<mailto:freesurfer@nmr.mgh.harvard.edu> 
<mailto:freesurfer@nmr.mgh.harvard.edu>
Subject: Re: [Freesurfer] FSFAST first level covariates

I was under the impression that Self and Valence were ratings from the same 
event (in that mail archive, they were different events and so needed different 
offsets). If Self and Valence are from the same event, then you would have 
something like
1. Offset
2. Self
3. Valence
4. Self*Valence
I've never tried the interaction (self*valence). You might have to demean 
before computing the product
On 8/14/2022 4:58 PM, Angela Fang wrote:

External Email - Use Caution
The two coding schemes are different because the second one does include the 
self*valence variable you’re talking about, whereas the first one doesn’t. I 
only included the 2nd offset because you suggested to someone else to include 
it (see MailScanner has detected a possible fraud attempt from 
"secure-web.cisco.com" claiming to be 
https://secure-web.cisco.com/1axFb8Uk3EidN7RU3PxCMFcOtbS8C3cmNULxUDXhNgudqc2v-5LUyAYUUUC6m-FcwrPwdVKy2Y7n8DAVcyRmfV5k4NWvno0aemc07RzaIinRQ8-6_FFiXeiHdOzhbARiwJyY4xdCGKBBbKJGuolxWhUQ5ZaDv2NtX0TJ7BmyhFTGOFw4S9sp6o7il_xApboosWIOBOJb1zh7hBwMks3eFrd04OkImYWLahgGjjS7Zs4BjsQy1dYPiBCXK_vv2cHadYnHraA451p-pjerGjuvaFs9cp4jm03dcia0ilLlFFYK5n74UApp-Da_ShPd4EN_o7az4F96Vwm7OtA1A2eemJw/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html<https://secure-web.cisco.com/1UvSz7SvjZnlE1QfkdS-VBc1GzXeZMYqUcsOqE3dGoo43anjoKIpvfs4NnozgTlCx23dB10wC_oFWTi8Zyazh1v1oufO7QQBf9hJanKAiwbu0cr4NfMvGMSOSaaOt5nSATHi-J-55MTqcCUhjz8_rRM1YuYWhtxzDVrlNJ5mD3QlEmdQlhRlYoneii_5mWjAZB1gcbpR_0Zl1nUaCy9BfmpcQRNLpIdfE1NMjg7OnqHOX5jAdPz1gGqxVTKgstTqx8RhxmTkDYOQPLc6hC3by-Atu2VfBnloD3GbZOzG04LjV1Of0uYaB6pk6oSsZFULTsifcwPSiwh1m9gKSg5lD7Q/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html>).
 If we don’t need it, would it just be 2 conditions, as follows?


  1.  SelfOffset
  2.  Self*ValenceSlope

But then I’m not clear how to get the main effect of valence (brain regions 
that scale with increasing emotion valence, while holding self-relevance 
constant)?

From: 
<mailto:freesurfer-boun...@nmr.mgh.harvard.edu>
 on behalf of "Douglas N. Greve" 
<mailto:dgr...@mgh.harvard.edu>
Reply-To: Freesurfer support list 
<mailto:freesurfer@nmr.mgh.harvard.edu>
Date: Sunday, August 14, 2022 at 1:37 PM
To: "freesurfer@nmr.mgh.harvard.edu"<mailto:freesurfer@nmr.mgh.harvard.edu> 
<mailto:freesurfer@nmr.mgh.harvard.edu>
Subject: Re: [Freesurfer] FSFAST first level covariates

Those look like they are the same coding scheme. What is different? You can 
only have one offset. The Self vs Valence -a 2 -a 4 is not testing for an 
interaction. If you want an interaction you have to create a new variable which 
is SelfRating*ValenceRating.

On 8/10/2022 2:38 PM, Angela Fang wrote:

External Email - Use Caution
Hello,

Just re-sending my question below. If I have a variable with 2 levels (yes/no) 
and another variable that is continuous, based on this post (MailScanner has 
detected a possible fraud a

Re: [Freesurfer] FSFAST first level covariates

2022-08-31 Thread Douglas N. Greve

sorry, you lost me there ...

On 8/31/2022 12:39 PM, Angela Fang wrote:

External Email - Use Caution

That makes sense! And if we wanted to identify brain regions
associated with valence (across self/non-self) that is unique from
condition 2, can we add a 5^th condition? Maybe we can code self as 2
and non-self as 1 so that condition 2 (self*valence) would be
different from condition 5 (valence ratings only)?

Incredibly helpful, Doug- thank you so much!

*From: * on behalf of "Douglas
N. Greve"

*Reply-To: *Freesurfer support list
*Date: *Tuesday, August 30, 2022 at 7:38 AM
*To: *"freesurfer@nmr.mgh.harvard.edu"
*Subject: *Re: [Freesurfer] FSFAST first level covariates

Oh, I think I see now. In realty, each trial is either self or not
self and then you have a separate valence measure, so your coding
should be

1. Self-offset
2. SelfValence
3. NoSelf-offset
4. NoSelfValence

Where 1+2 are used for a self event and 3+4 are used for a noself
event. If you want to test for the difference between self and noself
valence slope, then you would just use -a 2 -c 4

On 8/29/2022 12:08 PM, Angela Fang wrote:

*External Email - Use Caution *

Great, thanks so much Doug- this setup makes sense to me. However,
isn’t it a problem that the brain regions associated with the
contrast for -a 3 and the contrast for -a 4 would be the same for
trials that subjects rate as 1 (self-relevant) vs 0 (not
self-relevant)? Could we code it as 2 (self-relevant) and 1 (not
self-relevant)?

Many thanks!!

*From: *
<mailto:freesurfer-boun...@nmr.mgh.harvard.edu> on behalf of
"Douglas N. Greve"
<mailto:dgr...@mgh.harvard.edu>
*Reply-To: *Freesurfer support list

<mailto:freesurfer@nmr.mgh.harvard.edu>
*Date: *Sunday, August 28, 2022 at 1:07 PM
*To: *"freesurfer@nmr.mgh.harvard.edu"
<mailto:freesurfer@nmr.mgh.harvard.edu>

<mailto:freesurfer@nmr.mgh.harvard.edu>
*Subject: *Re: [Freesurfer] FSFAST first level covariates

I was under the impression that Self and Valence were ratings from
the same event (in that mail archive, they were different events
and so needed different offsets). If Self and Valence are from the
same event, then you would have something like
1. Offset
2. Self
3. Valence
4. Self*Valence
I've never tried the interaction (self*valence). You might have to
demean before computing the product

On 8/14/2022 4:58 PM, Angela Fang wrote:

*External Email - Use Caution *

https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg19957.html

1. SelfOffset
2. Self*ValenceSlope

But then I’m not clear how to get the main effect of valence
(brain regions that scale with increasing emotion valence,
while holding self-relevance constant)?

*From: *
<mailto:freesurfer-boun...@nmr.mgh.harvard.edu> on behalf of
"Douglas N. Greve"
<mailto:dgr...@mgh.harvard.edu>
*Reply-To: *Freesurfer support list

<mailto:freesurfer@nmr.mgh.harvard.edu>
*Date: *Sunday, August 14, 2022 at 1:37 PM
*To: *"freesurfer@nmr.mgh.harvard.edu"
<mailto:freesurfer@nmr.mgh.harvard.edu>

<mailto:freesurfer@nmr.mgh.harvard.edu>
*Subject: *Re: [Freesurfer] FSFAST first level covariates

Those look like they are the same coding scheme. What is
different? You can only have one offset. The Self vs Valence
-a 2 -a 4 is not testing for an interaction. If you want an
interaction you have to create a new variable which is
SelfRating*ValenceRating.

On 8/10/2022 2:38 PM, Angela Fang wrote:

*External Email - Use Caution *

Hello,

Just re-sending my question below. If I have a variable
with 2 le

Re: [Freesurfer] FSFAST first level covariates

2022-08-31 Thread Angela Fang

External Email - Use Caution

Sorry if I was unclear. The analysis we have set up will allow me to test for 
the difference between self and non self valence slope, but I’m also interested 
in testing for the slope of valence alone (identifying the brain regions 
associated with increasing valence ratings). Can I add another condition for 
valence, and if so, how do we make the weights different from the self*valence 
condition (because self is always coded 1 so the weights for self*valence and 
valence only will be identical)?


On Aug 31, 2022, at 8:08 PM, Douglas N. Greve  wrote:

 sorry, you lost me there ...

On 8/31/2022 12:39 PM, Angela Fang wrote:

External Email - Use Caution
That makes sense! And if we wanted to identify brain regions associated with 
valence (across self/non-self) that is unique from condition 2, can we add a 
5th condition? Maybe we can code self as 2 and non-self as 1 so that condition 
2 (self*valence) would be different from condition 5 (valence ratings only)?

Incredibly helpful, Doug- thank you so much!

From: 
<mailto:freesurfer-boun...@nmr.mgh.harvard.edu>
 on behalf of "Douglas N. Greve" 
<mailto:dgr...@mgh.harvard.edu>
Reply-To: Freesurfer support list 
<mailto:freesurfer@nmr.mgh.harvard.edu>
Date: Tuesday, August 30, 2022 at 7:38 AM
To: "freesurfer@nmr.mgh.harvard.edu"<mailto:freesurfer@nmr.mgh.harvard.edu> 
<mailto:freesurfer@nmr.mgh.harvard.edu>
Subject: Re: [Freesurfer] FSFAST first level covariates

Oh, I think I see now. In realty, each trial is either self or not self and 
then you have a separate valence measure, so your coding should be
1. Self-offset
2. SelfValence
3. NoSelf-offset
4. NoSelfValence

Where 1+2 are used for a self event and 3+4 are used for a noself event. If you 
want to test for the difference between self and noself valence slope, then you 
would just use -a 2 -c 4

On 8/29/2022 12:08 PM, Angela Fang wrote:

External Email - Use Caution
Great, thanks so much Doug- this setup makes sense to me. However, isn’t it a 
problem that the brain regions associated with the contrast for -a 3 and the 
contrast for -a 4 would be the same for trials that subjects rate as 1 
(self-relevant) vs 0 (not self-relevant)? Could we code it as 2 (self-relevant) 
and 1 (not self-relevant)?

Many thanks!!

From: 
<mailto:freesurfer-boun...@nmr.mgh.harvard.edu>
 on behalf of "Douglas N. Greve" 
<mailto:dgr...@mgh.harvard.edu>
Reply-To: Freesurfer support list 
<mailto:freesurfer@nmr.mgh.harvard.edu>
Date: Sunday, August 28, 2022 at 1:07 PM
To: "freesurfer@nmr.mgh.harvard.edu"<mailto:freesurfer@nmr.mgh.harvard.edu> 
<mailto:freesurfer@nmr.mgh.harvard.edu>
Subject: Re: [Freesurfer] FSFAST first level covariates

I was under the impression that Self and Valence were ratings from the same 
event (in that mail archive, they were different events and so needed different 
offsets). If Self and Valence are from the same event, then you would have 
something like
1. Offset
2. Self
3. Valence
4. Self*Valence
I've never tried the interaction (self*valence). You might have to demean 
before computing the product
On 8/14/2022 4:58 PM, Angela Fang wrote:

External Email - Use Caution
The two coding schemes are different because the second one does include the 
self*valence variable you’re talking about, whereas the first one doesn’t. I 
only included the 2nd offset because you suggested to someone else to include 
it (see MailScanner has detected a possible fraud attempt from 
"secure-web.cisco.com" claiming to be 
https://secure-web.cisco.com/1ibJu9DNCEQ80IS8XnYQe64G3qUwBk_mJpSfJiHpiFjdObD9fykxEUJll6wFW1-iyN02rfgtRiXTnU4tJAcadUeslBYPf-5CRhCzFKS5taH5e0tcr9yHQucLQZtMPwSNjLSKr-EMGQl4s3bOljQ7SzM7l6ZQZBC6x74q-pCqkQlgW50Mo8uKIDwwDfPnF04ok_kWyqNSL0h4dA3bGXpjQGWY4qb8REA8F3zJTfPieNKjhTBe92G4KDFXX89v_EwUnQw-1GiIs15MK4BUf5eS1LmzTtShQ99QPfHfFSABruvyX4IXN1a1oyCVeBCtRlgya/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg19957.html
 
[secure-web.cisco.com]<https://secure-web.cisco.com/1n9NaxxOQGYBn2vf8MJ5Y3HsNwtbaKNPmxR056SHywqcZ1M5Pr3EmrcpD5GO7Bz4VWqz0_QYUglcetQzxo3XfZo52OzpC1h-OFx-iNmhdpBsBWhahZ6gq3wTQ-7drXxoKE3PW6bJvfRmwtp6ttZtwroM5e7FgTaNR0-rYAe7OULwrxMTvFRHcwUT9ZBRYZA_hbtEv0-G0ytcUWEDBddCevIKtYVhpgVwLLS0Exk_XdFbnAvvT1GZTg2WfFucz-ljQluQXurRZBZ6pXaLlq0MOQtFA2RGTM46p_WEB8ZrlOwRc4pLoR-O7EzXnBEQFv2P0/https%3A%2F%2Furldefense.com%2Fv3%2F__https%3A%2F%2Fsecure-web.cisco.com%2F1UvSz7SvjZnlE1QfkdS-VBc1GzXeZMYqUcsOqE3dGoo43anjoKIpvfs4NnozgTlCx23dB10wC_oFWTi8Zyazh1v1oufO7QQBf9hJanKAiwbu0cr4NfMvGMSOSaaOt5nSATHi-J-55MTqcCUhjz8_rRM1YuYWhtxzDVrlNJ5mD3QlEmdQlhRlYoneii_5mWjAZB1gcbpR_0Zl1nUaCy9BfmpcQRNLpIdfE1NMjg7OnqHOX5jAdPz1gGqxVTKgstTqx8RhxmTkDYOQPLc6hC3by-Atu2VfBnloD3GbZOzG04LjV1Of0uYaB6pk6oSsZFULTsifcwPSiwh1m9gKSg5lD7Q%2Fhttps*3A*2F*2Fwww.mail-archive.com*2Ffreesurfer*40nmr.mgh.harvard.edu*2Fmsg19957.html__;JSUlJSUl!!

Re: [Freesurfer] FSFAST first level covariates

Re: [Freesurfer] FSFAST first level covariates

Re: [Freesurfer] FSFAST first level covariates

Re: [Freesurfer] FSFAST first level covariates

Re: [Freesurfer] FSFAST first level covariates

Re: [Freesurfer] FSFAST first level covariates

Re: [Freesurfer] FSFAST first level covariates

Re: [Freesurfer] FSFAST first level covariates

Re: [Freesurfer] FSFAST first level covariates

Re: [Freesurfer] FSFAST first level covariates

Re: [Freesurfer] FSFAST first level covariates

Re: [Freesurfer] FSFAST first level covariates

12 matches

Site Navigation

Mail list logo

Footer information