Please enable JavaScript.
Coggle requires JavaScript to display documents.
SEM 2 (CFA (Measurement model (how latent variables are measured with…
SEM 2
CFA
Measurement model
how latent variables are measured with indicator variables
Theta-delta matrix
loadings of observed varon delta var
Lambda-x
loadings of observed variables on latent factors
confirmatory factor analyses
Assumes underlying latent variables that can explain variance
Error variances are left free
Designs
Cross sectional
all variables measured at the same time
Cross lagged
causal effect of variable X at time 1 on variable Y at time 2
Instantaneous
present condition leads to outcome
Global fit
RMSEA
Devided by DF
parsimony
values
<0.05 good
0.1 bad
Accounts for N
ECVI
Accounts for
free parameters
expected value of chi if cross validated with different sample
AIC
ECVI*(N-1)
Should be low
Standardized residuals
look at values >2
specify error correlation
Measurement equations
hypothesized
Equality constraints
some paths are equal
structural equations
gives extra degrees of freedom
more parsimony