1
library(lavaan)

#model

fig2.5.data <- '
1
.5 1
.1 .1 1
.2 .3 .2 1
'

fig2.5.cor <- getCov(fig2.5.data, names=c(c("X1", "X2", "X3", "X4")))

fig2.5.model <- '
# latent variables
F1 =~ NA*X1 + a*X1 + b*X2
F2 =~ NA*X3 + c*X3 + d*X4

# variances/covariances

# fix variance to 1
F1 ~~ 1*F1

# fix variance to 1
F2 ~~ 1*F2
F1 ~~ e*F2

# label residual variances
X1 ~~ w*X1
X2 ~~ x*X2
X3 ~~ y*X3
X4 ~~ z*X4

It's an extract from the textbook I'm reading at the moment.

I understand everything else but "NA*" parts.

What are they?

I've encountered many NAs when using R, such as when there are missing data, but this one seems to be totally incomprehensible to me.

Any help would be much appreciated!

No Ru
  • 37
  • 3

1 Answers1

2

From lavaan's tutorial, page 12:

To force this factor loading to be free, we pre-multiply it with NA, as a hint to lavaan that the value of this parameter is ‘missing’ and therefore still unknown.

Ramiro Magno
  • 3,085
  • 15
  • 30