Indirect effects with multiple group analysis

[Amanda]

Hi everyone,

I am running a multiple groups analysis with five groups. I have already figured out which groups should be allowed to vary, and which groups must be constrained. My question is how I can run indirect effects now. The problem is that I have already asked the program to constrain (or not) the groups. The way I have been taught to calculate indirect effects is to essentially multiple a letter by the path you want and then use the argument ab := a*c. This is problematic with a multiple groups analysis because I already am multiplying the constrained (c) function with letters on the path. 

E.g. With the Multiple groups analysis I have:

#Direct Effects

          SWL ~ c(a,a,a,a,a)*SELF_ESTEEM + c(o,o,p,q,o)*FRS_IC + c(z,z,d,d,z)*FRS_EC + dsex + PARENT_EDUCATION

#Mediators

          SELF_ESTEEM ~ c(k,k,k,k,k)*FRS_IC + c(t,u,u,v,v)**FRS_EC 

I need the path from Self_Esteem to SWL to be a ( SWL ~ a*SELF_ESTEEM) and the path from FRS_IC to Self_Esteem to be b ( SELF_ESTEEM ~ b*FRS_IC) and then 

ab : = a*b.

Does anyone know how to get around this problem of having the constraint multipled and labeling the paths? I want the indirect effects for all groups to be run. Any help would be appreciated. 

[Terrence Jorgensen]

I want the indirect effects for all groups to be run

Then just define the indirect effect for each group using the parameters for that groups.  The fact that some parameters in the calculation will overlap because of the constraints is irrelevant to how an indirect effect is calculated.  For the first, second, and third groups, for example, the indirect effect of FRS_EC on SWL via SELF_ESTEEM would be:

indEC_g1 := t*o
indEC_g2 := u*o
indEC_g3 := u*p

Terrence D. Jorgensen

Postdoctoral Researcher, Methods and Statistics

Research Institute for Child Development and Education, the University of Amsterdam

UvA web page: http://www.uva.nl/profile/t.d.jorgensen