SEM convergence not achieved

[Maryam Bidgoli]

I am working on SEM, however I got the error that "convergence not achieved". This is my code

sem (Trust1 -> q0020_0001, ) (Trust1 -> q0020_0002, ) (Trust1 -> q0020_0003, ) (Trust1 -> Attitude, ) (Trust1 -> vaccine, ) (Trust2 -> q0020_0009, ) (Trust2 -> q0020_0010, ) (Trust2 -> q0020_0011, ) (Trust2 -> Attitude, ) (Trust2@1 -> vaccine, ) (Attitude -> attitud1, ) (Attitude -> attitud2, ) (Attitude -> attitud3, ) (Attitude -> attitud5, ) (Attitude -> vaccine, ) (gender_1 -> Attitude, ) (gender_1 -> vaccine, ) (race_2 -> Attitude, ) (race_2 -> vaccine, ) (race_3 -> Attitude, ) (race_3 -> vaccine, ) (race_4 -> Attitude, ) (race_4 -> vaccine, ) (ms_2 -> Attitude, ) (ms_2 -> vaccine, ) (ms_3 -> Attitude, ) (ms_3 -> vaccine, ) (income -> Attitude, ) (income -> vaccine, ) (edu -> Attitude, ) (edu -> vaccine, ), covstruct(_lexogenous, diagonal) cov(_lexogenous*_oexogenous@0) difficult iterate(300) latent(Trust1 Trust2 Attitude ) cov( Trust1*Trust2) nocapslatent

Can anyone help me with this? Much appreciated.

[Erik Ruzek]

You have a lot going on here. In general, I recommend breaking up a complex model into pieces and then running those pieces to make sure you aren't getting estimation problems.

Start with the latent variable/measurement model pieces:

(Trust1 -> q0020_0001 q0020_0002 q0020_0003) (Trust2 -> q0020_0009 q0020_0010 q0020_0011) ///
(Attitude -> attitud1 attitud2 attitud3 attitud5)

Can you fit just the measurement model?

Next, layer in the paths from the two trust latent variables to the attitude latent variable:

(Trust1 -> Attitude) (Trust2 -> Attitude)

Before moving onto the rest of the model, see if you can estimate the combined measurement model + paths between the latent variables.

Almost everything else seems to be predictors of latent variables (the race variables, income, edu, and ms variables). The exception is vaccine. That plays some role that I can't quite understand. Can you elaborate on that?

Also, what is this about?

(Trust2@1 -> vaccine)