|
|
 |
| Fit statistics for SEM for MIRT with ... |
 
|
| Message/Author |
|
|
| Chris Han posted on Tuesday, June 12, 2012 - 8:43 am
|
|
|
Hello,
I am estimating multidimensional IRT parameters based on data sets that have a lot of missing responses.
The item (intercepts/slopes) and person (factor scores) parameters were estimated very accurately without any problem.
The only problem was that Mplus did not provide any useful fit statistics like chi-square values and RMSEA in the output (attached below). My question is that how can I make Mplus compute those fit statics (as it does in most of the other analyses).
Thanks,
Chris
==Mplus OUTPUT FILE==
...
THE MODEL ESTIMATION TERMINATED NORMALLY
THE CHI-SQUARE TEST CANNOT BE COMPUTED BECAUSE THE FREQUENCY TABLE FOR THE
LATENT CLASS INDICATOR MODEL PART IS TOO LARGE.
MODEL FIT INFORMATION
Number of Free Parameters 281
Loglikelihood
H0 Value
...
Information Criteria
Akaike (AIC)
...
==Mplus OUTPUT FILE==
==INPUT FILE==
DATA:
FILE IS C:\1.DAT;
FORMAT IS (I8,2X,100I1);
VARIABLE:
NAMES ARE ID u1-u100;
CATEGORICAL ARE u1-u100;
MISSING IS BLANK;
ANALYSIS:
ESTIMATOR IS MLR;
LINK IS LOGIT;
MODEL:
...
SAVEDATA:
SAVE = FSCORES;
FILE IS 1.sco;
OUTPUT: STDYX TECH4 TECH8;
==INPUT FILE== |
|
|
|
|
| With categorical factor indicators and maximum likelihood estimation, means, variances, and covariances are not sufficient statistics for model estimation. Because of this, chi-square and related fit statistics are not available. Nested models can be compared using -2 times the loglikelihood difference which is distributed as chi-square. |
|
|
| Back to top |
|
|
