SEM with three IVs, one mediation, and two DVs

[Mehrnoush Sarafan]

Hi,

I'm relatively new to R and the lavaan package so excuse me if some of the questions sound very basic. I'm trying to fit a structural equation model to my experimental data. I have three IVs (all categorical with two levels:0,1), some control variables - age, gender, work experience, industry- (all categorical), one mediator (measured using three items on 7-point Likert scale) and two dependent variables (both measured using three items on 7-point Likert scale).

I need to do conformity factor analysis for the three measured variables as well as mediation analysis to fit the model. Here is my syntax for only one of the DVs and without considering control variables. I was wondering whether this is the right way of do the analysis? how I can add my control variables? and is it possible to have two DVs?

SEM.model.buffering <- '

#measurement model 

Blame =~ Q1a + Q1b + Q1c

Buffering =~ Q2a + Q2b + Q2c

#direct effects

Buffering ~ s*Severity + r*Relational_Norms + c*Controllability + b*Blame

#mediator

Blame ~ sb*Severity + rb*Relational_Norms + cb* Controllability

# indirect effect

sbb := sb*b

rbb:= rb*b

cbb:= cb*b

# total effect

total := s + (sb*b) + r + (rb*b) + c + (cb*b)'

fit <- sem(SEM.model.buffering, data=Data)

[Terrence Jorgensen]

I was wondering whether this is the right way of do the analysis?

The syntax looks correct, but the "total" effect doesn't make any sense.  There are 3 IVs, so you need to calculate 3 separate total effects. 

total_s := s + sbb 

total_r := r + rbb

total_c := c + cbb

how I can add my control variables?

The same way you add any other predictor.

and is it possible to have two DVs?

Yes, just include it in the model syntax.

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

[Mehrnoush Sarafan]

Thanks!It’s been really helpful.

--
You received this message because you are subscribed to a topic in the Google Groups "lavaan" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/lavaan/eek8WiakdrE/unsubscribe.
To unsubscribe from this group and all its topics, send an email to lavaan+un...@googlegroups.com.
To post to this group, send email to lav...@googlegroups.com.
Visit this group at https://groups.google.com/group/lavaan.
For more options, visit https://groups.google.com/d/optout.