Bayesian estimation and standardized ... PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 Erica Kamphorst posted on Friday, April 14, 2017 - 2:00 am
I am using Bayesian estimation, for the following model:
define:
DCxCT= ComDis * ComTm;
center ComTm ComDis (grandmean);

analysis:
ESTIMATOR IS BAYES;

FBITERATIONS IS 5000;
THIN is 50;
STVALUES IS ML;

model:
BE1 by B11 B12 B15 B16 B17;
EE1 by E11 E12 E13 E14 E15 E16;
EE15 with EE16;


BE1 ON Dur AGTm AGSm ComTm ComSm AgDis AgEnt AgTCT ComDis
ComEnt ComTCT DCxCT;
EE1 ON Dur AGTm AGSm ComTm ComSm AgDis AgEnt AgTCT ComDis
ComEnt ComTCT DCxCT;



I have two questions concerning the results:
- is it true that standardised results are not available with Bayesian estiamtion? If so, is there an alternative method to compare the influence of different predictors?
- my model fit seems rather good (posterior predictve p-value of .37; CI with zero almost in the middel); however none of the individual regression parameters are significant. How should I interpret this seemingly contradictory result? Or have I misspecified some part of the input?
 Bengt O. Muthen posted on Friday, April 14, 2017 - 3:47 pm
MPlus 7.4 gives standardized results with
Bayes.

You can get a good fit if you have categorical outcomes, or a small sample, or low correlations.
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: