|
|
Hi,
Would be grateful if anyone could help me with this:
I have specified a three factor model in which two items loaded onto each of the three proposed latent variables. The measurement model contained no double loading indicators, all measurement error was presumed to be uncorrelated and the latent factors were allowed to correlate in order to measure the multidimensionality (three-factor) of collective identity.
I want to use the MIMIC approach to look at the effect of age and religious group (3 groups) onto the dimensions and have used the following syntax:
Sem.TextOutput ( )
Sem.Standardized ()
Sem.Smc
Sem.Mods (10)
Sem.BeginGroup ("C:\cfareligion.sav")
Sem.AStructure ("Q2FeelReli = (1) Feel + (1) e1")
Sem.AStructure ("Q2FeelRelib = Feel + (1) e2")
Sem.AStructure ("Q2ImpReli = (1) Import + (1) e3")
Sem.AStructure ("Q2ImpRelib = Import+ (1) e4")
Sem.AStructure ("Q2AttaReli = (1) Attach + (1) e5")
Sem.AStructure ("Q2AttaRelib = Attach + (1) e6")
Sem.AStructure ("Feel = Religion2 + (1) D1")
Sem.AStructure ("Import = Religion2+ (1) D2")
Sem.AStructure ("Attach = Religion2 + (1) D3")
Sem.AStructure ("D1 <-> D2")
Sem.AStructure ("D1 <-> D3")
Sem.AStructure ("D2 <-> D3")
Sem.FitModel ()
Finally
Sem.Dispose ()
End Try
End Sub
My question is do i need to also add the age variable and look at that simultaneously or do i replace the 'religion2' by age and then perform two tests. In other words are all the variables (age, religion (two dummy code) and interaction) looked at simultaneously? In this case I have a problem of identification.
Would be grateful for an answer!
Best wishes,
Caroline
|
Locked
