SPSS Statistics

 View Only

 Help with Linear Mixed Model

Nicholas Cheng's profile image
Nicholas Cheng posted Thu July 17, 2025 04:19 PM

Hi, 

I'm currently in the process of running a linear mixed model in SPSS. Our study design is a repeated measures crossover design with 20 participants, 4 conditions and 5 trials (no between factors). We also wanted to add sex as a covariate and we wanted to add participant ID as a random effect. Our syntax can be found below: 

MIXED RANKL BY ParticipantID Trial Time WITH Sex

/CRITERIA=DFMETHOD(SATTERTHWAITE) CIN(95) MXITER(100) MXSTEP(10) SCORING(1) SINGULAR(0.000000000001) HCONVERGE(0.00000001, RELATIVE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0, ABSOLUTE)

/FIXED=Trial Time Sex Trial*Time Trial*Sex Time*Sex Trial*Time*Sex | SSTYPE(3)

/METHOD=REML

/RANDOM=ParticipantID | COVTYPE(VC) 

/REPEATED=Trial*Time | SUBJECT(ParticipantID) COVTYPE(CS)

/EMMEANS=TABLES(Trial) COMPARE ADJ(LSD)

/EMMEANS=TABLES(Time) COMPARE ADJ(LSD)

/EMMEANS=TABLES(Trial*Time) .

Is this correct?

Nick

Emanuel Festino's profile image
Emanuel Festino

I think the syntax is ok, but do you have problem with "the final Hessian matrix is not positive"? Because i have a similar dataset, and i don't know if it is right to keep these results