Extension of lavaan code for RI-CLPM (Mund & Nestler, 2017)

[zrc2...@gmail.com]

I am trying to run Random Intercept Cross-Lagged Panel Model (RI-CLPM) using the Lavaan syntax Mund & Nestler (2017) provided in their paper.

If I could get some help extending this code to add a third variable (mediator) to the two variable model, I would find it tremendously helpful. 

Below is Mund & Nestler's syntax, to which I attempted to add a third variable. All edits are highlighted in yellow, specific questions highlighted in green.

Thank you in advance for any help (and credit to the authors of this paper)

Figure of two-variable RI-CLPM

#### ++++++++++++++++++++++++++++++++++++++++++++++++++++++

####     Random-Intercept Cross-Lagged Panel Model     ####

#### ++++++++++++++++++++++++++++++++++++++++++++++++++++++

 

## load package

library(lavaan)

## load data

data <- read.table("./data/Fakedata_RI-CLPM_R.dat", header = T)

 

## Define model for lavaan

riclpm <- '

# Define intercept factors

ix =~ 1*x1+1*x2+1*x3+1*x4

iy =~ 1*y1+1*y2+1*y3+1*y4

iz =~ 1*z1+1*z2+1*z3+1*z4

# Define phantom latent variables

etax1 =~ 1*x1

etax2 =~ 1*x2

etax3 =~ 1*x3

etax4 =~ 1*x4

 

etay1 =~ 1*y1

etay2 =~ 1*y2

etay3 =~ 1*y3

etay4 =~ 1*y4

 

etaz1 =~ 1*z1

etaz2 =~ 1*z2

etaz3 =~ 1*z3

etaz4 =~ 1*z4

 

 

# Autoregressive effects

etax2 ~ a1*etax1

etax3 ~ a1*etax2

etax4 ~ a1*etax3

 

etay2 ~ a2*etay1

etay3 ~ a2*etay2

etay4 ~ a2*etay3

 

etaz2 ~ a2*etaz1

etaz3 ~ a2*etaz2

etaz4 ~ a2*etaz3

 

# Crosslagged effects

etay2 ~ c1*etax1

etay3 ~ c1*etax2

etay4 ~ c1*etax3

 

etax2 ~ c2*etay1

etax3 ~ c2*etay2

etax4 ~ c2*etay3

 

etaz2 ~ c1*etax1

etaz3 ~ c1*etax2

etaz4 ~ c1*etax3

 

etax2 ~ c2*etaz1

etax3 ~ c2*etaz2

etax4 ~ c2*etaz3

 

etay2 ~ c1*etaz1

etay3 ~ c1*etaz2

etay4 ~ c1*etaz3

 

etaz2 ~ c2*etay1

etaz3 ~ c2*etay2

etaz4 ~ c2*etay3

 

# Some further constraints on the variance structure

# 1. Set error variances of the observed variables to zero

x1 ~~ 0*x1

x2 ~~ 0*x2

x3 ~~ 0*x3

x4 ~~ 0*x4

 

y1 ~~ 0*y1

y2 ~~ 0*y2

y3 ~~ 0*y3

y4 ~~ 0*y4

 

z1 ~~ 0*z1

z2 ~~ 0*z2

z3 ~~ 0*z3

z4 ~~ 0*z4

 

# 2. Let lavaan estimate the variance of the latent variables

etax1 ~~ varx1*etax1

etax2 ~~ varx2*etax2

etax3 ~~ varx3*etax3

etax4 ~~ varx4*etax4

 

etay1 ~~ vary1*etay1

etay2 ~~ vary2*etay2

etay3 ~~ vary3*etay3

etay4 ~~ vary4*etay4

 

etaz1 ~~ varz1*etaz1

etaz2 ~~ varz2*etaz2

etaz3 ~~ varz3*etaz3

etaz4 ~~ varz4*etaz4

 

 

 

 

 

# 3. We also want estimates of the intercept factor variances and an

#    estimate of their covariance

ix ~~ varix*ix

iy ~~ variy*iy

iz ~~ variz*iz

ix ~~ covi*iy

ix ~~ covi*iz

iy ~~ covi*iz

 

# 4. We have to define that the covariance between the intercepts and

#    the latents of the first time point are zero

etax1 ~~ 0*ix

etay1 ~~ 0*ix

etaZ1 ~~ 0*ix

 

etax1 ~~ 0*iy

etay1 ~~ 0*iy

etaZ1 ~~ 0*iy

 

etax1 ~~ 0*iz

etay1 ~~ 0*iz

etaZ1 ~~ 0*iz

 

# 5. Finally, we estimate the covariance between the latents of x and y

#    of the first time point, the second time-point and so on. note that

#    for the second to fourth time point the correlation is constrained to

#    the same value

etax1 ~~ cov1*etay1

etax2 ~~ e1*etay2

etax3 ~~ e1*etay3

etax4 ~~ e1*etay4

 

etax1 ~~ cov1*etaz1

etax2 ~~ e1*etaz2

etax3 ~~ e1*etaz3

etax4 ~~ e1*etaz4

 

etay1 ~~ cov1*etaz1

etay2 ~~ e1*etaz2

etay3 ~~ e1*etaz3

etay4 ~~ e1*etaz4

 

# The model also contains a mean structure and we have to define some

# constraints for this part of the model. the assumption is that we

# only want estimates of the mean of the intercept factors. all other means

# are defined to be zero:

x1 ~ 0*1

x2 ~ 0*1

x3 ~ 0*1

x4 ~ 0*1

y1 ~ 0*1

y2 ~ 0*1

y3 ~ 0*1

y4 ~ 0*1

z1 ~ 0*1

z2 ~ 0*1

z3 ~ 0*1

z4 ~ 0*1

 

etax1 ~ 0*1

etax2 ~ 0*1

etax3 ~ 0*1

etax4 ~ 0*1

 

etay1 ~ 0*1

etay2 ~ 0*1

etay3 ~ 0*1

etay4 ~ 0*1

 

etaz1 ~ 0*1

etaz2 ~ 0*1

etaz3 ~ 0*1

etaz4 ~ 0*1

ix ~ 1

iy ~ 1

iz ~ 1

 

## define correlations (NOT MODIFIED—QUESTION)

cori := covi / (sqrt(varix) * sqrt(variy))

cor1 := cov1 / (sqrt(varx1) * sqrt(vary1))

cort2 := e1 / (sqrt(varx2) * sqrt(vary2))

cort3 := e1 / (sqrt(varx3) * sqrt(vary4))

cort4 := e1 / (sqrt(varx4) * sqrt(vary4))

'

 

fit <- sem(riclpm, data = data)

summary(fit, fit.measures = T)

[zrc2...@gmail.com]

Apologies for another post but would really find any feedback very helpful. Questions are highlighted in green. Thank you again.

[Christopher D Desjardins]

Hi,

What are your questions? I do see the highlighted green areas. What specifically are your questions?

--
You received this message because you are subscribed to the Google Groups "lavaan" group.
To unsubscribe from this group and stop receiving emails from it, 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.

[zrc2...@gmail.com]

I'm sorry that my question wasn't clear. I am extending the code that was written for 2 variables to 3 and wasn't sure that the modifications I made in highlighted green areas were accurate. 

[Christopher David Desjardins]

I see. Everything looks fine to me.

[zrc2...@gmail.com]

Wonderful. Thank you!

[Andrej Skoko]

Did it work out at the end? I tried the same, but somehow lavaan didn't recognise varz...