The following model specification works when testing for first order factors, but when I attempt to test for a second order factor by adding the last 4 lines in the model, I get the error message below:
model.cfa.ru <- specifyModel()
sRU1 <- sRU, NA, 1
sRU2 <- sRU, lam12
sRU3 <- sRU, lam13
sRU4 <- sRU, lam14
sRU5 <- sRU, lam15
sRU6 <- sRU, lam16
sRU <-> sRU, mak1
sRU1 <-> sRU2, cors1
sRU5 <-> sRU3, cors2
sRU6 <-> sRU2, cors3
sRU6 <-> sRU1, cors4
pRU1 <- pRU, sfsf
pRU2 <- pRU, lam22
pRU3 <- pRU, lam23
pRU4 <- pRU, lam24
pRU5 <- pRU, lam25
pRU6 <- pRU, lam26
pRU <-> pRU, mak2
pRU5 <-> pRU3, corp1
pRU4 <-> pRU6, corp2
pRU5 <-> pRU6, corp3
pRU6 <-> pRU2, corp4
rRU1 <- rRU, sdfsdf
rRU2 <- rRU, lam32
rRU3 <- rRU, lam33
rRU4 <- rRU, lam34
rRU5 <- rRU, lam35
rRU6 <- rRU, lam36
rRU7 <- rRU, lam37
rRU8 <- rRU, lam38
rRU <-> rRU, mak3
rRU3 <-> rRU1, corr1
rRU7 <-> rRU4, corr2
rRU8 <-> rRU7, corr3
rRU8 <-> rRU2, corr4
rRU5 <-> rRU2, corr5
rRU7 <-> rRU5, corr6
rRU8 <-> rRU5, corr7
sRU <- sof, lok1
pRU <- sof, lok2
rRU <- sof, lok3
sof <-> sof, lok4
cfa.ru <- sem(model.cfa.ru, data=ru)
Error in w_mat %*% p_deriv_mat %*% invMat : requires numeric/complex matrix/vector arguments In addition: Warning message: In eval(expr, envir, enclos) : Negative parameter variances. Model may be underidentified.
I've done a number of modifications to the model specification with no success. I'm wondering if I'm missing something? Thank you!
