! Rater Bias Model
! CBC data in US twins

G1: model parameters
Data Calc NGroups=3
Begin Matrices;
X Lower 1 1 Free    ! genetic path, a, to latent phenotype
Y Lower 1 1 Free    ! shared env path, c, to latent phenotype
Z Lower 1 1 Free    ! non-shared env path, e, to latent phenotype
B Full 4 2          ! rater bias paths (betas)
F Diag 4 4          ! residual specific variance ratings (Rm, Rf)
S Diag 2 2          ! path, alpha, from latent phenotype to 
!                     paternal rating, when the path from the 
!                     latent phenotype to the maternal rating
!                     is set at 1.0
I Iden 2 2          ! an identity matrix for later use
End Matrices;
Specify B
4 0                 ! parameter number 4 is the maternal bias path
4 0
0 5                 ! parameter number 5 is the paternal bias path
0 5 
Specify F 
6 6 7 7             ! this specifies a residual (error) for the 
!                     maternal ratings (parameter number 6), and
!                     a residual error for the paternal ratings
!                     (parameter number 7) 
Specify S 15 15     ! parameter number 15 is alpha
Begin Algebra;
A= X*X' ;           ! these square the paths to give  
C= Y*Y' ;           ! the corresponding variance 
E= Z*Z' ;           ! components
R= B*B' ;           !
J= F*F' ;           !
End Algebra;
End

G2: MZ boys twin pairs
Data NInput_vars=4 NObservations=96
Labels morg_t1 morg_t2 farg_t1 farg_t2
CMatrix File=uscbcmzm.cov
Matrices= Group 1
Covariances R+ J+ ((I_S)*(A+C+E | A+C _
                          A+C   | A+C+E )*(I_S)') /
!
!
! In terms of the parameters in the model given in Figure 16.1, 
! page 316 in Neale and Cardon, or Figure 11.1, page 173 in Neale and Maes,
! or in Figure 1, page 297, in 
! Hewitt et al (1992) Analysis of parental ratings of children's
! behavior using LISREL, Behavior Genetics, 22:3, 293-317, 
! this covariance matrix is:
!
!
!
!
!
!
!
!
!
!
Option RSidual
End 

G3: DZ boys twin pairs
Data NInput_vars=4 NObservations=102
Labels morg_t1 morg_t2 farg_t1 farg_t2
CMatrix File=uscbcdzm.cov
Matrices= Group 1
H Full 1 1
Covariances R+ J+ ((I_S)*(A+C+E | H@A+C _
                          H@A+C | A+C+E )*(I_S)') /
! This covariance matrix is the same as that
! for the MZs, except that for the off diagonal 
! components involving a2, the a2 terms are 
! multiplied by 0.5
Matrix H .5
Option RSidual 
Option NDecimals = 4 Iterations = 200
End