Repeated predictor variable analysis

classic Classic list List threaded Threaded
8 messages Options
Reply | Threaded
Open this post in threaded view
|

Repeated predictor variable analysis

Nkem Ntonghanwah
good day all, 
I am trying to figure out the appropriate model following the problem.
Say Participants are enrolled in a study and baseline covariates measured along with a predictor variable (but not the outcome variable). At Time 2, 3,4 (say),  The predictor variable is again measured. The outcome variable is then measured at future time 5 (say). Interest is to determine at which time point is best (in terms of predictive power) to measure the predictor.
Below is an example of data. COVAR1, COVAR2, and Predt1 are measured at the same time while PredT2 is measured some times later - say after 5 weeks while Response is measured again sometimes after PredT2. Time difference between measurements is not fixed.
 
Reference articles would also be appreciated.
 Thanks
Forcheh
id,COVAR1,COVAR2, PredT1,PredT2,Response
1,22,1,90,36,75
2,28,1,49,49,90
3,25,1,62,16,91
4,23,0,22,66,69
5,32,1,66,87,85
6,24,0,77,18,65
7,31,1,48,76,86
8,31,0,70,80,64
9,26,1,17,59,78
10,22,1,66,56,92
11,30,0,4,44,69
12,24,1,17,86,62
13,30,0,55,59,58
14,30,1,96,58,84
15,26,1,22,33,99
16,30,0,14,47,58


===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
Reply | Threaded
Open this post in threaded view
|

Re: Repeated predictor variable analysis

Ryan Black
Setting aside covariates, consider fitting a random coefficient model [using the MIXED procedure], with time incorporated as a fixed effect term and a random effect term, and the predictor variable treated as y:

Level 1 Equation:

y = B0J + B1J*time + eij

Level 2 Equations:

B0J = Gamma00 + u0j
B1J = Gamma10 + u1j

Full Equation:

y = Gamma00 +
    + Gamma10*time
    + ( u1j*time + u0j + eij )

Next, regress your actual y on u0j and u1j, perhaps allowing u0j and u1j to covary [using AMOS]. 

This approach will allow you to evaluate the extent to which: 

(1) your predictor is related to y at baseline, u0j effect, (assuming you code time such that ‘0’ is BL);

(2) change in your predictor (over time) is related to your y, u1j effect. 

Note: You could change the coding of time such that the effect of U0j is associated with a different time point. 

Ryan

Sent from my iPhone

On Aug 9, 2020, at 8:32 PM, Nkem Ntonghanwah <[hidden email]> wrote:


good day all, 
I am trying to figure out the appropriate model following the problem.
Say Participants are enrolled in a study and baseline covariates measured along with a predictor variable (but not the outcome variable). At Time 2, 3,4 (say),  The predictor variable is again measured. The outcome variable is then measured at future time 5 (say). Interest is to determine at which time point is best (in terms of predictive power) to measure the predictor.
Below is an example of data. COVAR1, COVAR2, and Predt1 are measured at the same time while PredT2 is measured some times later - say after 5 weeks while Response is measured again sometimes after PredT2. Time difference between measurements is not fixed.
 
Reference articles would also be appreciated.
 Thanks
Forcheh
id,COVAR1,COVAR2, PredT1,PredT2,Response
1,22,1,90,36,75
2,28,1,49,49,90
3,25,1,62,16,91
4,23,0,22,66,69
5,32,1,66,87,85
6,24,0,77,18,65
7,31,1,48,76,86
8,31,0,70,80,64
9,26,1,17,59,78
10,22,1,66,56,92
11,30,0,4,44,69
12,24,1,17,86,62
13,30,0,55,59,58
14,30,1,96,58,84
15,26,1,22,33,99
16,30,0,14,47,58


===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
Reply | Threaded
Open this post in threaded view
|

Re: Repeated predictor variable analysis

Jon Peck
Why not just use all the predictor time point values in ordinary regression?  There aren't that many of them.

On Sun, Aug 9, 2020 at 7:36 PM Ryan Black <[hidden email]> wrote:
Setting aside covariates, consider fitting a random coefficient model [using the MIXED procedure], with time incorporated as a fixed effect term and a random effect term, and the predictor variable treated as y:

Level 1 Equation:

y = B0J + B1J*time + eij

Level 2 Equations:

B0J = Gamma00 + u0j
B1J = Gamma10 + u1j

Full Equation:

y = Gamma00 +
    + Gamma10*time
    + ( u1j*time + u0j + eij )

Next, regress your actual y on u0j and u1j, perhaps allowing u0j and u1j to covary [using AMOS]. 

This approach will allow you to evaluate the extent to which: 

(1) your predictor is related to y at baseline, u0j effect, (assuming you code time such that ‘0’ is BL);

(2) change in your predictor (over time) is related to your y, u1j effect. 

Note: You could change the coding of time such that the effect of U0j is associated with a different time point. 

Ryan

Sent from my iPhone

On Aug 9, 2020, at 8:32 PM, Nkem Ntonghanwah <[hidden email]> wrote:


good day all, 
I am trying to figure out the appropriate model following the problem.
Say Participants are enrolled in a study and baseline covariates measured along with a predictor variable (but not the outcome variable). At Time 2, 3,4 (say),  The predictor variable is again measured. The outcome variable is then measured at future time 5 (say). Interest is to determine at which time point is best (in terms of predictive power) to measure the predictor.
Below is an example of data. COVAR1, COVAR2, and Predt1 are measured at the same time while PredT2 is measured some times later - say after 5 weeks while Response is measured again sometimes after PredT2. Time difference between measurements is not fixed.
 
Reference articles would also be appreciated.
 Thanks
Forcheh
id,COVAR1,COVAR2, PredT1,PredT2,Response
1,22,1,90,36,75
2,28,1,49,49,90
3,25,1,62,16,91
4,23,0,22,66,69
5,32,1,66,87,85
6,24,0,77,18,65
7,31,1,48,76,86
8,31,0,70,80,64
9,26,1,17,59,78
10,22,1,66,56,92
11,30,0,4,44,69
12,24,1,17,86,62
13,30,0,55,59,58
14,30,1,96,58,84
15,26,1,22,33,99
16,30,0,14,47,58


===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD


--
Jon K Peck
[hidden email]

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
Reply | Threaded
Open this post in threaded view
|

Re: Repeated predictor variable analysis

Ryan Black
In reply to this post by Nkem Ntonghanwah
I assumed there were more than three time points based on “(say)”. If there are only three time points, what Jon suggests seems reasonable to me. OTOH, if there are many time points, I’d consider an alternative approach such as the one in my initial reply.

Sent from my iPhone

On Aug 9, 2020, at 9:50 PM, Jon Peck <[hidden email]> wrote:


Why not just use all the predictor time point values in ordinary regression?  There aren't that many of them.

On Sun, Aug 9, 2020 at 7:36 PM Ryan Black <[hidden email]> wrote:
Setting aside covariates, consider fitting a random coefficient model [using the MIXED procedure], with time incorporated as a fixed effect term and a random effect term, and the predictor variable treated as y:

Level 1 Equation:

y = B0J + B1J*time + eij

Level 2 Equations:

B0J = Gamma00 + u0j
B1J = Gamma10 + u1j

Full Equation:

y = Gamma00 +
    + Gamma10*time
    + ( u1j*time + u0j + eij )

Next, regress your actual y on u0j and u1j, perhaps allowing u0j and u1j to covary [using AMOS]. 

This approach will allow you to evaluate the extent to which: 

(1) your predictor is related to y at baseline, u0j effect, (assuming you code time such that ‘0’ is BL);

(2) change in your predictor (over time) is related to your y, u1j effect. 

Note: You could change the coding of time such that the effect of U0j is associated with a different time point. 

Ryan

Sent from my iPhone

On Aug 9, 2020, at 8:32 PM, Nkem Ntonghanwah <[hidden email]> wrote:


good day all, 
I am trying to figure out the appropriate model following the problem.
Say Participants are enrolled in a study and baseline covariates measured along with a predictor variable (but not the outcome variable). At Time 2, 3,4 (say),  The predictor variable is again measured. The outcome variable is then measured at future time 5 (say). Interest is to determine at which time point is best (in terms of predictive power) to measure the predictor.
Below is an example of data. COVAR1, COVAR2, and Predt1 are measured at the same time while PredT2 is measured some times later - say after 5 weeks while Response is measured again sometimes after PredT2. Time difference between measurements is not fixed.
 
Reference articles would also be appreciated.
 Thanks
Forcheh
id,COVAR1,COVAR2, PredT1,PredT2,Response
1,22,1,90,36,75
2,28,1,49,49,90
3,25,1,62,16,91
4,23,0,22,66,69
5,32,1,66,87,85
6,24,0,77,18,65
7,31,1,48,76,86
8,31,0,70,80,64
9,26,1,17,59,78
10,22,1,66,56,92
11,30,0,4,44,69
12,24,1,17,86,62
13,30,0,55,59,58
14,30,1,96,58,84
15,26,1,22,33,99
16,30,0,14,47,58


===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD


--
Jon K Peck
[hidden email]

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
Reply | Threaded
Open this post in threaded view
|

Re: Repeated predictor variable analysis

Maguin, Eugene

Thank all of you for your replies. After working with someone who has done this before, I (finally) understood how to set-up and do the analyses. The data source is an ecological momemtary assessment (EMA) of a behavior. For an eight week period, each person receives a text survey in the morning and a  series of four surveys during the rest of the day. This is done every day. The morning surveys are dataset 1 and the rest-of-the-day surveys are dataset 2. Both datasets are processed to yield a weekly compliance proportion for each dataset. The analysis problem is build a multilevel (ML) model for each dataset and compares the two models. Those of you that use mplus will recognize this model. I was trying to implement that parallel process growth curve model in an ML setting. After getting help from the aforementioned person, this is the syntax statement.

 

Mixed y with int slope/fixed int slope int*slope/print solution/

Random intercept | subjects(sid) cotype(id).

 

The datasets were added together and sorted by sid, tod (morning or rest-of-day), week. Week was recoded to model a linear curve for each dataset, i.e., 0-7. Int was coded as 0 for all weeks in the morning dataset and as 1 for all weeks in the rest-of-day dataset.

 

Gene Maguin

 

 

 

 

 

 

 

 

 

 

 

 

 

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Ryan Black
Sent: Sunday, August 09, 2020 10:32 PM
To: [hidden email]
Subject: Re: Repeated predictor variable analysis

 

I assumed there were more than three time points based on “(say)”. If there are only three time points, what Jon suggests seems reasonable to me. OTOH, if there are many time points, I’d consider an alternative approach such as the one in my initial reply.

 

Sent from my iPhone



On Aug 9, 2020, at 9:50 PM, Jon Peck <[hidden email]> wrote:



Why not just use all the predictor time point values in ordinary regression?  There aren't that many of them.

 

On Sun, Aug 9, 2020 at 7:36 PM Ryan Black <[hidden email]> wrote:

Setting aside covariates, consider fitting a random coefficient model [using the MIXED procedure], with time incorporated as a fixed effect term and a random effect term, and the predictor variable treated as y:

Level 1 Equation:

y = B0J + B1J*time + eij

Level 2 Equations:

B0J = Gamma00 + u0j
B1J = Gamma10 + u1j

Full Equation:

y = Gamma00 +
    + Gamma10*time
    + ( u1j*time + u0j + eij )

Next, regress your actual y on u0j and u1j, perhaps allowing u0j and u1j to covary [using AMOS]. 

 

This approach will allow you to evaluate the extent to which: 

 

(1) your predictor is related to y at baseline, u0j effect, (assuming you code time such that ‘0’ is BL);

 

(2) change in your predictor (over time) is related to your y, u1j effect. 

 

Note: You could change the coding of time such that the effect of U0j is associated with a different time point. 

 

Ryan

 

Sent from my iPhone



On Aug 9, 2020, at 8:32 PM, Nkem Ntonghanwah <[hidden email]> wrote:



good day all, 

I am trying to figure out the appropriate model following the problem.

Say Participants are enrolled in a study and baseline covariates measured along with a predictor variable (but not the outcome variable). At Time 2, 3,4 (say),  The predictor variable is again measured. The outcome variable is then measured at future time 5 (say). Interest is to determine at which time point is best (in terms of predictive power) to measure the predictor.

Below is an example of data. COVAR1, COVAR2, and Predt1 are measured at the same time while PredT2 is measured some times later - say after 5 weeks while Response is measured again sometimes after PredT2. Time difference between measurements is not fixed.

 

Reference articles would also be appreciated.

 Thanks

Forcheh

id,COVAR1,COVAR2, PredT1,PredT2,Response

1,22,1,90,36,75

2,28,1,49,49,90

3,25,1,62,16,91

4,23,0,22,66,69

5,32,1,66,87,85

6,24,0,77,18,65

7,31,1,48,76,86

8,31,0,70,80,64

9,26,1,17,59,78

10,22,1,66,56,92

11,30,0,4,44,69

12,24,1,17,86,62

13,30,0,55,59,58

14,30,1,96,58,84

15,26,1,22,33,99

16,30,0,14,47,58

 

 

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD


 

--

Jon K Peck
[hidden email]

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
Reply | Threaded
Open this post in threaded view
|

Re: Repeated predictor variable analysis

Bruce Weaver
Administrator
Hi Gene.  Your post confuses me.  Did you intend to post it to this thread?

http://spssx-discussion.1045642.n5.nabble.com/parallel-process-model-in-mixed-td5739473.html

If so, I can move it there (in the Nabble archive).  Thanks for clarifying.  

Cheers,
Bruce



Maguin, Eugene wrote

> Thank all of you for your replies. After working with someone who has done
> this before, I (finally) understood how to set-up and do the analyses. The
> data source is an ecological momemtary assessment (EMA) of a behavior. For
> an eight week period, each person receives a text survey in the morning
> and a  series of four surveys during the rest of the day. This is done
> every day. The morning surveys are dataset 1 and the rest-of-the-day
> surveys are dataset 2. Both datasets are processed to yield a weekly
> compliance proportion for each dataset. The analysis problem is build a
> multilevel (ML) model for each dataset and compares the two models. Those
> of you that use mplus will recognize this model. I was trying to implement
> that parallel process growth curve model in an ML setting. After getting
> help from the aforementioned person, this is the syntax statement.
>
> Mixed y with int slope/fixed int slope int*slope/print solution/
> Random intercept | subjects(sid) cotype(id).
>
> The datasets were added together and sorted by sid, tod (morning or
> rest-of-day), week. Week was recoded to model a linear curve for each
> dataset, i.e., 0-7. Int was coded as 0 for all weeks in the morning
> dataset and as 1 for all weeks in the rest-of-day dataset.
>
> Gene Maguin
>
>
>
>
>
>
>
>
>
>
>
>
>
> From: SPSSX(r) Discussion [mailto:

> SPSSX-L@.UGA

> ] On Behalf Of Ryan Black
> Sent: Sunday, August 09, 2020 10:32 PM
> To:

> SPSSX-L@.UGA

> Subject: Re: Repeated predictor variable analysis
>
> I assumed there were more than three time points based on “(say)”. If
> there are only three time points, what Jon suggests seems reasonable to
> me. OTOH, if there are many time points, I’d consider an alternative
> approach such as the one in my initial reply.
>
> Sent from my iPhone
>
>
> On Aug 9, 2020, at 9:50 PM, Jon Peck &lt;

> jkpeck@

> &lt;mailto:

> jkpeck@

> &gt;> wrote:
> 
> Why not just use all the predictor time point values in ordinary
> regression?  There aren't that many of them.
>
> On Sun, Aug 9, 2020 at 7:36 PM Ryan Black &lt;

> ryan.andrew.black@

> &lt;mailto:

> ryan.andrew.black@

> &gt;> wrote:
> Setting aside covariates, consider fitting a random coefficient model
> [using the MIXED procedure], with time incorporated as a fixed effect term
> and a random effect term, and the predictor variable treated as y:
>
> Level 1 Equation:
>
> y = B0J + B1J*time + eij
>
> Level 2 Equations:
>
> B0J = Gamma00 + u0j
> B1J = Gamma10 + u1j
>
> Full Equation:
>
> y = Gamma00 +
>     + Gamma10*time
>     + ( u1j*time + u0j + eij )
> Next, regress your actual y on u0j and u1j, perhaps allowing u0j and u1j
> to covary [using AMOS].
>
> This approach will allow you to evaluate the extent to which:
>
> (1) your predictor is related to y at baseline, u0j effect, (assuming you
> code time such that ‘0’ is BL);
>
> (2) change in your predictor (over time) is related to your y, u1j effect.
>
> Note: You could change the coding of time such that the effect of U0j is
> associated with a different time point.
>
> Ryan
>
> Sent from my iPhone
>
>
> On Aug 9, 2020, at 8:32 PM, Nkem Ntonghanwah &lt;

> nkemntongha@

> &lt;mailto:

> nkemntongha@

> &gt;> wrote:
> 
> good day all,
> I am trying to figure out the appropriate model following the problem.
> Say Participants are enrolled in a study and baseline covariates measured
> along with a predictor variable (but not the outcome variable). At Time 2,
> 3,4 (say),  The predictor variable is again measured. The outcome variable
> is then measured at future time 5 (say). Interest is to determine at which
> time point is best (in terms of predictive power) to measure the
> predictor.
> Below is an example of data. COVAR1, COVAR2, and Predt1 are measured at
> the same time while PredT2 is measured some times later - say after 5
> weeks while Response is measured again sometimes after PredT2. Time
> difference between measurements is not fixed.
>
> Reference articles would also be appreciated.
>  Thanks
> Forcheh
> id,COVAR1,COVAR2, PredT1,PredT2,Response
> 1,22,1,90,36,75
> 2,28,1,49,49,90
> 3,25,1,62,16,91
> 4,23,0,22,66,69
> 5,32,1,66,87,85
> 6,24,0,77,18,65
> 7,31,1,48,76,86
> 8,31,0,70,80,64
> 9,26,1,17,59,78
> 10,22,1,66,56,92
> 11,30,0,4,44,69
> 12,24,1,17,86,62
> 13,30,0,55,59,58
> 14,30,1,96,58,84
> 15,26,1,22,33,99
> 16,30,0,14,47,58
>
>
> ===================== To manage your subscription to SPSSX-L, send a
> message to

> LISTSERV@.UGA

> &lt;mailto:

> LISTSERV@.UGA

> &gt; (not to SPSSX-L), with no body text except the command. To leave the
> list, send the command SIGNOFF SPSSX-L For a list of commands to manage
> subscriptions, send the command INFO REFCARD
> ===================== To manage your subscription to SPSSX-L, send a
> message to

> LISTSERV@.UGA

> &lt;mailto:

> LISTSERV@.UGA

> &gt; (not to SPSSX-L), with no body text except the command. To leave the
> list, send the command SIGNOFF SPSSX-L For a list of commands to manage
> subscriptions, send the command INFO REFCARD
>
>
> --
> Jon K Peck

> jkpeck@

> &lt;mailto:

> jkpeck@

> &gt;
> ===================== To manage your subscription to SPSSX-L, send a
> message to

> LISTSERV@.UGA

> &lt;mailto:

> LISTSERV@.UGA

> &gt; (not to SPSSX-L), with no body text except the command. To leave the
> list, send the command SIGNOFF SPSSX-L For a list of commands to manage
> subscriptions, send the command INFO REFCARD
>
> =====================
> To manage your subscription to SPSSX-L, send a message to

> LISTSERV@.UGA

>  (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD





-----
--
Bruce Weaver
[hidden email]
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.

--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
--
Bruce Weaver
bweaver@lakeheadu.ca
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.
Reply | Threaded
Open this post in threaded view
|

Re: Repeated predictor variable analysis

Nkem Ntonghanwah
Many thanks Jon.

I shall follow this very promising lead. I have not yet installed the R extensions but will do so asap.
Will surely report my success once done.

Regards
Forcheh

On Mon, Aug 10, 2020 at 5:23 PM Bruce Weaver <[hidden email]> wrote:
Hi Gene.  Your post confuses me.  Did you intend to post it to this thread?

http://spssx-discussion.1045642.n5.nabble.com/parallel-process-model-in-mixed-td5739473.html

If so, I can move it there (in the Nabble archive).  Thanks for clarifying. 

Cheers,
Bruce



Maguin, Eugene wrote
> Thank all of you for your replies. After working with someone who has done
> this before, I (finally) understood how to set-up and do the analyses. The
> data source is an ecological momemtary assessment (EMA) of a behavior. For
> an eight week period, each person receives a text survey in the morning
> and a  series of four surveys during the rest of the day. This is done
> every day. The morning surveys are dataset 1 and the rest-of-the-day
> surveys are dataset 2. Both datasets are processed to yield a weekly
> compliance proportion for each dataset. The analysis problem is build a
> multilevel (ML) model for each dataset and compares the two models. Those
> of you that use mplus will recognize this model. I was trying to implement
> that parallel process growth curve model in an ML setting. After getting
> help from the aforementioned person, this is the syntax statement.
>
> Mixed y with int slope/fixed int slope int*slope/print solution/
> Random intercept | subjects(sid) cotype(id).
>
> The datasets were added together and sorted by sid, tod (morning or
> rest-of-day), week. Week was recoded to model a linear curve for each
> dataset, i.e., 0-7. Int was coded as 0 for all weeks in the morning
> dataset and as 1 for all weeks in the rest-of-day dataset.
>
> Gene Maguin
>
>
>
>
>
>
>
>
>
>
>
>
>
> From: SPSSX(r) Discussion [mailto:

> SPSSX-L@.UGA

> ] On Behalf Of Ryan Black
> Sent: Sunday, August 09, 2020 10:32 PM
> To:

> SPSSX-L@.UGA

> Subject: Re: Repeated predictor variable analysis
>
> I assumed there were more than three time points based on “(say)”. If
> there are only three time points, what Jon suggests seems reasonable to
> me. OTOH, if there are many time points, I’d consider an alternative
> approach such as the one in my initial reply.
>
> Sent from my iPhone
>
>
> On Aug 9, 2020, at 9:50 PM, Jon Peck &lt;

> jkpeck@

> &lt;mailto:

> jkpeck@

> &gt;> wrote:
> 
> Why not just use all the predictor time point values in ordinary
> regression?  There aren't that many of them.
>
> On Sun, Aug 9, 2020 at 7:36 PM Ryan Black &lt;

> ryan.andrew.black@

> &lt;mailto:

> ryan.andrew.black@

> &gt;> wrote:
> Setting aside covariates, consider fitting a random coefficient model
> [using the MIXED procedure], with time incorporated as a fixed effect term
> and a random effect term, and the predictor variable treated as y:
>
> Level 1 Equation:
>
> y = B0J + B1J*time + eij
>
> Level 2 Equations:
>
> B0J = Gamma00 + u0j
> B1J = Gamma10 + u1j
>
> Full Equation:
>
> y = Gamma00 +
>     + Gamma10*time
>     + ( u1j*time + u0j + eij )
> Next, regress your actual y on u0j and u1j, perhaps allowing u0j and u1j
> to covary [using AMOS].
>
> This approach will allow you to evaluate the extent to which:
>
> (1) your predictor is related to y at baseline, u0j effect, (assuming you
> code time such that ‘0’ is BL);
>
> (2) change in your predictor (over time) is related to your y, u1j effect.
>
> Note: You could change the coding of time such that the effect of U0j is
> associated with a different time point.
>
> Ryan
>
> Sent from my iPhone
>
>
> On Aug 9, 2020, at 8:32 PM, Nkem Ntonghanwah &lt;

> nkemntongha@

> &lt;mailto:

> nkemntongha@

> &gt;> wrote:
> 
> good day all,
> I am trying to figure out the appropriate model following the problem.
> Say Participants are enrolled in a study and baseline covariates measured
> along with a predictor variable (but not the outcome variable). At Time 2,
> 3,4 (say),  The predictor variable is again measured. The outcome variable
> is then measured at future time 5 (say). Interest is to determine at which
> time point is best (in terms of predictive power) to measure the
> predictor.
> Below is an example of data. COVAR1, COVAR2, and Predt1 are measured at
> the same time while PredT2 is measured some times later - say after 5
> weeks while Response is measured again sometimes after PredT2. Time
> difference between measurements is not fixed.
>
> Reference articles would also be appreciated.
>  Thanks
> Forcheh
> id,COVAR1,COVAR2, PredT1,PredT2,Response
> 1,22,1,90,36,75
> 2,28,1,49,49,90
> 3,25,1,62,16,91
> 4,23,0,22,66,69
> 5,32,1,66,87,85
> 6,24,0,77,18,65
> 7,31,1,48,76,86
> 8,31,0,70,80,64
> 9,26,1,17,59,78
> 10,22,1,66,56,92
> 11,30,0,4,44,69
> 12,24,1,17,86,62
> 13,30,0,55,59,58
> 14,30,1,96,58,84
> 15,26,1,22,33,99
> 16,30,0,14,47,58
>
>
> ===================== To manage your subscription to SPSSX-L, send a
> message to

> LISTSERV@.UGA

> &lt;mailto:

> LISTSERV@.UGA

> &gt; (not to SPSSX-L), with no body text except the command. To leave the
> list, send the command SIGNOFF SPSSX-L For a list of commands to manage
> subscriptions, send the command INFO REFCARD
> ===================== To manage your subscription to SPSSX-L, send a
> message to

> LISTSERV@.UGA

> &lt;mailto:

> LISTSERV@.UGA

> &gt; (not to SPSSX-L), with no body text except the command. To leave the
> list, send the command SIGNOFF SPSSX-L For a list of commands to manage
> subscriptions, send the command INFO REFCARD
>
>
> --
> Jon K Peck

> jkpeck@

> &lt;mailto:

> jkpeck@

> &gt;
> ===================== To manage your subscription to SPSSX-L, send a
> message to

> LISTSERV@.UGA

> &lt;mailto:

> LISTSERV@.UGA

> &gt; (not to SPSSX-L), with no body text except the command. To leave the
> list, send the command SIGNOFF SPSSX-L For a list of commands to manage
> subscriptions, send the command INFO REFCARD
>
> =====================
> To manage your subscription to SPSSX-L, send a message to

> LISTSERV@.UGA

>  (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD





-----
--
Bruce Weaver
[hidden email]
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.

--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
Reply | Threaded
Open this post in threaded view
|

Re: Repeated predictor variable analysis

vidita
I am   Field Maintenance Technician
<https://www.fieldengineer.com/blogs/how-can-a-field-maintenance-technician-help>
at field engineer which is an on-demand marketplace that connects businesses
and freelance field engineers in the telecom industry.





--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD