Mixed models with repeated meassures

[karenyearwood]

Hi everyone

I eish someone can help me with this one

I have a sample of patients who went through bariatric surgery and their weight at time 1,2,3,4 but with lots of missing data

so I opted for mixed model with repeated meassures using the following syntax

DATASET ACTIVATE DataSet1.
MIXED peso BY time
  /CRITERIA=CIN(95) MXITER(100) MXSTEP(10) SCORING(1) SINGULAR(0.000000000001) HCONVERGE(0,
    ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE)
  /FIXED=time | SSTYPE(3)
  /METHOD=REML
  /PRINT=SOLUTION.

and then

MIXED peso BY time
  /CRITERIA=CIN(95) MXITER(100) MXSTEP(10) SCORING(1) SINGULAR(0.000000000001) HCONVERGE(0,
    ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE)
  /FIXED=time | SSTYPE(3)
  /METHOD=REML
  /PRINT=R SOLUTION TESTCOV
  /REPEATED=time | SUBJECT(code) COVTYPE(UN).

I wanted to confirm if it is the correct syntax and then ask how can I aseses the variable of number of nutrition appointments if I want to know if this is a significant predictor of weight loss.

Just on a linear regression of time 4 by time 1 and number of appointments? or any suggestions

thanks

Karen

[Bruce Weaver]

I don't have a lot of time to study your situation carefully right now, but from a quick glance, I think you might want to consider exploring covariance structures other than UN.  The material at the following links may be helpful.

http://jonathantemplin.com/files/anova/ersh8310f09/ersh8310f09_lecture14.pdf
http://www.ats.ucla.edu/stat/spss/faq/repeat_covar.htm
http://www.youtube.com/watch?v=LKT4brBV-jo

I haven't viewed all of the last one.  He does not exit the dialogs via PASTE, but he does appear to explore using various covariance structures.

HTH.

karenyearwood wrote

Hi everyone

I eish someone can help me with this one

I have a sample of patients who went through bariatric surgery and their weight at time 1,2,3,4 but with lots of missing data

so I opted for mixed model with repeated meassures using the following syntax

DATASET ACTIVATE DataSet1.
MIXED peso BY time
  /CRITERIA=CIN(95) MXITER(100) MXSTEP(10) SCORING(1) SINGULAR(0.000000000001) HCONVERGE(0,
    ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE)
  /FIXED=time | SSTYPE(3)
  /METHOD=REML
  /PRINT=SOLUTION.

and then

MIXED peso BY time
  /CRITERIA=CIN(95) MXITER(100) MXSTEP(10) SCORING(1) SINGULAR(0.000000000001) HCONVERGE(0,
    ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE)
  /FIXED=time | SSTYPE(3)
  /METHOD=REML
  /PRINT=R SOLUTION TESTCOV
  /REPEATED=time | SUBJECT(code) COVTYPE(UN).

I wanted to confirm if it is the correct syntax and then ask how can I aseses the variable of number of nutrition appointments if I want to know if this is a significant predictor of weight loss.

Just on a linear regression of time 4 by time 1 and number of appointments? or any suggestions

thanks

Karen

[Alex Reutter]

Hi Karen,

You may also want to look at: http://pic.dhe.ibm.com/infocenter/spssstat/v22r0m0/topic/com.ibm.spss.statistics.cs/spss/tutorials/mixed_diet_intro.htm

Alex

[Ryan]

You might find a message I posted a while back helpful:

 

http://listserv.uga.edu/cgi-bin/wa?A2=ind1101&L=spssx-l&P=R53088

 

Sorry but no time to respond further right now.

 

Ryan

On Thu, Mar 20, 2014 at 4:42 AM, karenyearwood <[hidden email]> wrote:

Hi everyone

I eish someone can help me with this one

I have a sample of patients who went through bariatric surgery and their
weight at time 1,2,3,4 but with lots of missing data

so I opted for mixed model with repeated meassures using the following
syntax

DATASET ACTIVATE DataSet1.
MIXED peso BY time
  /CRITERIA=CIN(95) MXITER(100) MXSTEP(10) SCORING(1)
SINGULAR(0.000000000001) HCONVERGE(0,
    ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE)
  /FIXED=time | SSTYPE(3)
  /METHOD=REML
  /PRINT=SOLUTION.

and then

MIXED peso BY time
  /CRITERIA=CIN(95) MXITER(100) MXSTEP(10) SCORING(1)
SINGULAR(0.000000000001) HCONVERGE(0,
    ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE)
  /FIXED=time | SSTYPE(3)
  /METHOD=REML
  /PRINT=R SOLUTION TESTCOV
  /REPEATED=time | SUBJECT(code) COVTYPE(UN).

I wanted to confirm if it is the correct syntax and then ask how can I
aseses the variable of number of nutrition appointments if I want to know if
this is a significant predictor of weight loss.

Just on a linear regression of time 4 by time 1 and number of appointments?
or any suggestions

thanks

Karen






--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Mixed-models-with-repeated-meassures-tp5724942.html
Sent from the SPSSX Discussion mailing list archive at 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