mimic sem use

[zara ndri]

Hi, i try to estimate a mimic model with these data after computing i got this at the end The LR test of model vs. saturated is not reported because the fitted
model is not full rank. issue with maximum likelihood. I wish to know how to deal with this. I put below my data and the code.

the code : sem ( SE -> gdp elec m2pib ) ( SE <- infl2 tax2 ouvpib ), iterate(10) latent (SE)

the data :
_group _type gdp elec m2pib infl2 ouvpib tax2
1 1
1 2 224120 1779 25.0499 26.583 57.8894 4.100e+18
1 3 224120 18235.4 1112.7 -957.406 -1683.8 9.500e+11
1 4 18235.4 1779 106.23 -87.9688 -83.1208 7.600e+10
1 5 1112.7 106.23 25.0499 -2.28771 -.987539 4.500e+09
1 6 -957.406 -87.9688 -2.28771 26.583 2.07095 -3.500e+09
1 7 -1683.8 -83.1208 -.987539 2.07095 57.8894 -8.400e+09
1 8 9.500e+11 7.600e+10 4.500e+09 -3.500e+09 -8.400e+09 4.100e+18

the boarding result

Endogenous variables

Measurement: gdp elec m2pib
Latent: SE

Exogenous variables

Observed: infl2 tax2 ouvpib

Fitting target model:

Iteration 0: log likelihood = -12423.647
Iteration 1: log likelihood = -12423.418
Iteration 2: log likelihood = -12423.417

Structural equation model Number of obs = 300
Estimation method = ml
Log likelihood = -12423.417

( 1) [gdp]SE = 1
------------------------------------------------------------------------------
| OIM
| Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
Structural |
SE |
infl2 | -4.958594 .5024914 -9.87 0.000 -5.943459 -3.973729
tax2 | 2.40e-07 1.57e-09 153.13 0.000 2.37e-07 2.43e-07
ouvpib | 6.064368 .4376418 13.86 0.000 5.206605 6.92213
-------------+----------------------------------------------------------------
Measurement |
gdp |
SE | 1 (constrained)
-----------+----------------------------------------------------------------
elec |
SE | .0816146 .0021036 38.80 0.000 .0774917 .0857376
-----------+----------------------------------------------------------------
m2pib |
SE | .0049517 .0005412 9.15 0.000 .003891 .0060124
-------------+----------------------------------------------------------------
var(e.gdp)| 591.2152 546.0812 96.72208 3613.812
var(e.elec)| 289.1315 24.01109 245.701 340.2388
var(e.m2pib)| 19.5039 1.594131 16.61688 22.89252
var(e.SE)| 1149.009 547.2842 451.7391 2922.53
------------------------------------------------------------------------------
Note: The LR test of model vs. saturated is not reported because the fitted
model is not full rank.

[Rattiya Lippe]

Hallo Zara Ndri,

I have started to run the MIMIC model using Stata and found your post as of today. I also experienced the same problem and would like to ask whether you found any further solutions.

Thank you,