FIGURE 4

Structural model. Where is the vector relating to exogenous latent variables, the vector relating to endogenous latent variables, x the vector relating to manifest variables of exogenous latent variables , y the vector relating to manifest variables of endogenous latent variables , the vector of measurement error of x, the vector of measurement error of y, the parameter of directional relation between a latent variable and its indicators, the vector of errors in the equations (disturbance terms) corresponding to endogenous latent variables , the parameter of directional relation between two endogenous latent variables, the parameter of directional relation between an endogenous and exogenous latent variable, the parameter of non-directional relation (covariance) between two latent variables, the (m n) matrix of structural coefficients between exogenous latent variables and endogenous latent variables, Â the (m n) matrix of structural coefficients between endogenous latent variables, the (n n) matrix of variances and co-variances between exogenous latent variables, the (m 1) matrix of residual variances and co-variances of the endogenous latent variables, _x the (p m) matrix of factor loading for the x variables, _y the (q n) matrix of factor loading for the y variables, the (p p) matrix of measurement error variances and co-variances of the x variables, the (q q) matrix of measurement error variances and co-variances of the y variables, the (q p) matrix of measurement error co-variances between the x variables and the y variables, m the number of endogenous latent variables , n the number of exogenous latent variables , p the number of the x variables, q the number of the y variables.