Login  Register

How to model interaction between time of the day (on many different days) and group effect in linear mixed model

classic Classic list List threaded Threaded
1 message Options Options
Embed post
Permalink
Reply | Threaded
Open this post in threaded view
| More
Print post
Permalink

How to model interaction between time of the day (on many different days) and group effect in linear mixed model

jdube
This post was updated on Jul 20, 2019; 5:17pm.
Hello all,

Just new to this forum/ML, which seems great to encourage collaboration
between SPSS users (and between expert-non experts).

For my self, I am starting with linear mixed models and trying to specify
the syntax for the model I have in mind. I'm not sure if I'm syntax is
correct and if it the best for my model. Any help would be appreciated. I am
using SPSS 25 on Windows.

As for the model, I have around 20 animals (in 2 age groups) with measures
of brain activity across 12h periods (12 time-points) for a variable number of days
(between 8 and 31) per animal. I am interested in modelling an interaction
between age and hour (I expect age effect at specific hours but not others).
In my understanding of things, I have two random factors: 1) the days where
the recording took place and 2) the animals which participated. As for my
fixed factor, I see 1) time of the day and 2) age group. Time of the day is
a repeated measures but do not seem random to me as these points are
deliberately fixed.

My retained syntax is below. I am using an autoregressive covariance
structure as I expect high correlation between contiguous data point in the
time domain.

MIXED Variable BY Agegroup Day Hour
  /CRITERIA=CIN(95) MXITER(100) MXSTEP(10) SCORING(1)
SINGULAR(0.000000000001) HCONVERGE(0,
    ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE)
  /FIXED=Agegroup Hour Agegroup*Hour | SSTYPE(3)
  /METHOD=REML
  /RANDOM=INTERCEPT Day | SUBJECT(animal) COVTYPE(VC) SOLUTION
  /REPEATED=Hour*Day | SUBJECT(animal) COVTYPE(ARH1).

Any thoughts regarding the model (choices of random and fixed factors) and
syntax would be appreciated!
Thank you all !



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

=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@LISTSERV.UGA.EDU (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