Latent 3-way interaction xwith error PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 Tim Manfred Götz posted on Monday, September 17, 2018 - 4:26 am
Dear Colleagues

i try to estimate a three-way interaction model with mplus by using the xwith command. Unfortunately I always get the same error message for the third two-way interaction (INT3) and I do not know how to fix this problem. Anyone has some thoughts what might be the problem in here? Sample size is 8000, fiml used for missing data treatment.


Analysis:
TYPE = RANDOM;
ALGORITHM = INTEGRATION;
ESTIMATOR = MLR;
Model:
TECH BY tech_1_t1 tech_2_t1 tech_3_t1;
SUSE BY suse_1_t1 suse_2_t1 suse_3_t1 ;
SOCC BY soc_1_t1 soc_2_t1 soc_3_t1;
INT1 | TECH XWITH SUSE;
INT2 | TECH XWITH SOCC;
INT3 | SUSE XWITH SOCC;
INT4 | TECH XWITH INT3;
wa_lt_t2 ON TECH SUSE SOCC INT1 INT2 INT3 INT4; "covariates excluded in here"
Output:
TECH1 TECH5 ;



ERROR MESSAGE:
"THE STANDARD ERRORS OF THE MODEL PARAMETER ESTIMATES MAY NOT BE TRUSTWORTHY FOR SOME PARAMETERS DUE TO A NON-POSITIVE DEFINITE FIRST-ORDER DERIVATIVE PRODUCT MATRIX. THIS MAY BE DUE TO THE STARTING VALUES BUT MAY ALSO BE AN INDICATION OF MODEL NONIDENTIFICATION. THE CONDITION NUMBER IS -0.315D-17. PROBLEM INVOLVING THE FOLLOWING PARAMETER: Parameter 33, WA_LT_T2 ON INT3"


Parameter 33 is the betacoefficient of wa_lt_t2 on INT3.

I am thankful for any advice..
Best regards,
Tim
 Bengt O. Muthen posted on Monday, September 17, 2018 - 2:15 pm
Please send output and data to Support along with your license number.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: