Gsem interpretation

[Chiara Tasselli]

Good evening,
I am performing the GSEM model of the following sintax:

gsem (SUBJ -> intrins, ) (SUBJ -> Extr, ) (SUBJ -> CosT, ) (SUBJ -> Belonging) (SUBJ -> AbilitySC, ) (SUBJ -> Abilitygap, )  ///
(female Stereotype 1.female#c.Stereotype -> SUBJ, ) (SUBJ -> Hum, family(gaussian, lcensored(0)) link(identity))(SUBJ -> Ecomath, family(gaussian, lcensored(0)) link(identity) ) (SUBJ -> Stem, family(gaussian, lcensored(0)) link(identity))  (abilitygap -> Hum, family(gaussian, lcensored(0)) link(identity)) (abilitygap -> Ecomath, family(gaussian, lcensored(0)) link(identity)) (abilitygap -> Stem, family(gaussian, lcensored(0)) link(identity))  ///
(M1[school] -> Ecomath, family(gaussian, lcensored(0)) link(identity)) (M1[school] -> Hum, family(gaussian, lcensored(0)) link(identity) ) (M1[school] -> Stem, family(gaussian, lcensored(0)) link(identity))   if time == 0 , ///
covstruct(_lexogenous, diagonal) latent(SUBJ M1 ) nocapslatent

and this is what I get:

 
Log likelihood = -5504.9157

( 1) [Hum]M1[school] = 1

( 2) [Hum]SUBJ = 1




Coef.
Std. Err. z
P>z
[95% Conf.
Interval]




intrins

SUBJ .1935517
.0315363 6.14
0.000
.1317416
.2553618

_cons -.0956396
.0610386 -1.57
0.117
-.2152729
.0239938




Extr

SUBJ .1147982
.0238522 4.81
0.000
.0680489
.1615476

_cons 2.605941
.0530425 49.13
0.000
2.50198
2.709903




CosT

SUBJ -.1312346
.0263191 -4.99
0.000
-.1828192
-.07965

_cons 3.10876
.0543323 57.22
0.000
3.002271
3.215249




Belonging

SUBJ .1044975
.0241242 4.33
0.000
.0572149
.1517801

_cons 3.534031
.0516605 68.41
0.000
3.432779
3.635284




AbilitySC

SUBJ .195256
.0321747 6.07
0.000
.1321948
.2583172

_cons 3.15322
.0613774 51.37
0.000
3.032923
3.273517




Abilitygap

SUBJ .3168933
.0354582 8.94
0.000
.2473966
.3863901

_cons -.3568302
.0793652 -4.50
0.000
-.5123831
-.2012774




Hum

abilitygap -2.402295
.383102 -6.27
0.000
-3.153162
-1.651429

M1[school] 1
(constrained)

SUBJ 1
(constrained)

_cons .2856486
.4672304 0.61
0.541
-.6301062
1.201403




Ecomath

abilitygap -2.504259
.4187047 -5.98
0.000
-3.324906
-1.683613

M1[school] .4628181
.1106909 4.18
0.000
.2458679
.6797682

SUBJ 1.207185
.1039694 11.61
0.000
1.003408
1.410961

_cons -.0857572
.3947757 -0.22
0.828
-.8595034
.6879891




Stem

abilitygap -2.013367
.3579137 -5.63
0.000
-2.714865
-1.311869

M1[school] .6351652
.1046598 6.07
0.000
.4300358
.8402947

SUBJ .9870099
.0893207 11.05
0.000
.8119445
1.162075

_cons -.0352007
.3731809 -0.09
0.925
-.7666219
.6962204




SUBJ

female 1.023342
.3656752 2.80
0.005
.3066324
1.740053

Stereotype .4980086
.2418418 2.06
0.039
.0240073
.9720098

female#c.Stereotype

F -1.053142
.3101316 -3.40
0.001
-1.660989
-.4452955




var(M1[school]) .7269493
.4823096
.1980418
2.668403

var(e.SUBJ) 7.165785
2.024763
4.118589
12.46749




var(e.intrins) .7206302
.0493199
.6301679
.8240786

var(e.Extr) .9109222
.0615937
.7978582
1.040008

var(e.CosT) .8784698
.0593297
.7695534
1.002801

var(e.Belonging) .9054402
.0612125
.7930746
1.033726

var(e.AbilitySC) .7239465
.0495701
.633028
.8279231

var(e.Abilitygap) .2489258
.0535419
.163299
.3794517

var(e.Hum) 1.431831
.1371872
1.186687
1.727617

var(e.Ecomath) .5157694
.1262512
.3192247
.8333253

var(e.Stem) 1.009848
.113371
.810393
1.258394

My question is: does Gsem still have a measurement and structural part as a SEM?
Because in my idea the variables: intrins, Extr, CosT, Belonging, AbilitySC are indicators of the latent variable SUBJ , meanwhile the latent variable SUBJ affect Abilitygap, Ecomath, Stem, Hum.

My question is: does Stata read them in the same way as mine or it took everything as an indicator of SUBJ?

Many thanks in advance for your time