Please Help I am new to R initial model-implied matrix (Sigma) is not positive definite; check your model and/or starting parameters in group 2.

Ok, first sorry If I sound dumb, I am new at this.

I am trying to run a measurement invariance test with 2 groups in a CFA. I have 2 Factors NV and VO. Whenever, I try to fit the model with 2 groups, I get the following error "lavaan ERROR: initial model-implied matrix (Sigma) is not positive definite; check your model and/or starting parameters in group 2."

I have no idea what this means. Here is my code (I correlated all of the residual covariances based on MI, I tried running it without these and encountered the same error):

MITcon_SRO_W3 <-'

NV =~ NA* NV1 + NV2 + NV3 + NV4 + NV5 + NV6 + NV7 + NV8 + NV9 + NV10 + NV11 + NV12 + NV13 + NV15 + NV16 + NV17 + NV18 + NV19 + NV20 + NV21

VO =~ NA * VO1 + VO2 + VO3 + VO4 + VO5 + VO6 + VO7 + VO8 + VO9 + VO10 + VO11

NV ~~ 1* NV

VO ~~ 1* VO

NV1 ~~ NV2 + NV3 + NV5 + NV6 + NV7 + NV10 + NV17 + NV18 + NV20 + VO2 + VO6

NV2 ~~ NV4 + NV5 + NV8 + NV12 + NV19 + NV21 + VO2 + VO3 + VO4

NV3 ~~ NV12 + NV13 + NV16 + NV19 + NV21 + VO1 + VO2 + VO3 + VO5 + VO8 + VO9

NV4 ~~ NV5 + NV7 + NV10 + NV11 + NV13 + NV16 + NV17 + NV19 + NV20 + VO3 + VO7 + VO10

NV5 ~~ NV10 + NV11 + VO4 + VO7

NV6 ~~ NV8 + NV10 + NV11 + NV12 + NV16 + NV19 + VO1 + VO2 + VO3 + VO8 + VO9

NV7 ~~ NV3 + NV10 + NV11 + NV12 + NV13 + NV17 + NV19 + NV20 + VO4 + VO8

NV10 ~~ NV8 + NV11 + NV12 + NV13 + NV15 + VO3 + VO4 + VO9 + VO10

NV11 ~~ NV13 + NV19 + VO3 + VO4 + VO7 + VO10

NV12 ~~ NV9 + NV15 + NV19 + NV20 + VO3 + VO5 + VO6

NV13 ~~ NV19 + NV16 + VO4 + VO5 + VO6 + VO7

NV15 ~~ NV21 + VO3 + VO4 + VO7 + VO8

NV16 ~~ NV17 + NV18 + NV20 + VO5 + VO7 + VO8

NV17 ~~ NV12 + NV18 + NV19 + NV20 + NV21 + VO3 + VO4 + VO7 + VO8 + VO 11

NV18 ~~ NV8 + NV12 + VO2 + VO3 + VO4 + VO7 + VO8

NV19 ~~ NV9 + NV21 + VO1 + VO6 + VO7 +VO8 + VO11

NV21 ~~ NV9 + NV20 + VO2 + VO4

VO1 ~~ NV16 + VO2 + VO4 + VO5 + VO6 + NV8 + VO10

VO2 ~~ NV8 + VO3 +VO6 + VO9

VO3 ~~ VO11

VO4 ~~ NV9 + VO7 + VO11

VO5 ~~ VO2 + VO6 + VO7 + VO10

VO7 ~~ VO3 + VO8 + VO11 + NV20

VO8 ~~ VO3 + VO4 + VO11

VO9 ~~ VO6 + VO10

VO10 ~~ VO2 + VO3 + VO6 + VO11'

fitMITcon_SRO_W3 <- cfa(MITcon_SRO_W3, data = ProjectData, missing = "fiml", group = "HI_Bi_B21")

summary(fitMITcon_SRO_W3,standardized=TRUE, fit.measures=TRUE)

Hi, and welcome!

Please see the FAQ: What's a reproducible example (`reprex`) and how do I do one? Using a reprex, complete with representative data will attract quicker and more answers.

For this question, it's particularly important: 1) it's impossible to reverse-engineer the problem without the data and all the libraries and 2) the lavaan questions don't attract a lot of answers, so there may be few in the community who can tell at a glance what may be wrong.

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

Thank you! I will do this.

1 Like