Mplus and Lavaan

[merm.m...@gmail.com]

Hi Yves,

I've noticed the estimated covariance marix or correlation matrix from the data that eventually gets modeled is not the same between lavaan and Mplus.  Am I doing something wrong?

Thanks.

[merm.m...@gmail.com]

Please note, this is even true when I choose mimic=Mplus and the estimator is the same.  I do get the same parameter estimates from lavaan as Mplus although the estimated covariance/correlation matrix is off from each other.

[yrosseel]

On 06/21/2013 03:31 PM, merm.m...@gmail.com wrote:
> Please note, this is even true when I choose mimic=Mplus and the
> estimator is the same. I do get the same parameter estimates from
> lavaan as Mplus although the estimated covariance/correlation matrix is
> off from each other.

I don't think this is the case. Why do you think that? An example would
help.

Yves.

[merm.m...@gmail.com]

fitted(model) give me the model implied covariance matrix?

Thanks and sorry for all the questions.

On Friday, June 21, 2013 6:29:17 AM UTC-7, merm.m...@gmail.com wrote:

[yrosseel]

On 06/22/2013 05:26 PM, merm.m...@gmail.com wrote:
> fitted(model) give me the model implied covariance matrix?

Yes. To see the sample-based covariance, use

inspect(fit, "sampstat")

Yves.