|
|
Interpretation of statistical model |
|
Message/Author |
|
|
Hello, on page 11 of the technical appendix (version 3) and in the 1984 Psychometrika paper by Muthen, the structural system is defined as eta=alpha+Beta*eta+gamma*x+zeta Why is eta on both sides of the equation? Thank you, Chris |
|
|
Because we want to do regressions between the different eta variables. Beta picks up those regression coefficients. The diagonal of the Beta matrix consists of zeros, so no eta variable is regressed on itself. |
|
Back to top |
|
|