Tobit-type CFA

[M. van der Plaat]

Hello all,

I am wondering whether Lavaan supports Tobit-type CFAs/SEMS. The dataset I use includes variables that are left censored and want to use this data to estimate a CFA-model.

To my knowledge, a way of modeling such data is to use some kind of Tobit model. I have looked online, and found some papers referring to Tobit-CFAs (see the source below for some examples). Looking through the documentation, I can't find how to implement such models with Lavaan. My question therefore is: can I estimate a Tobit-CFA with Lavaan? If not, is there another way of accounting for left-censored data.

Kind regards,

Mark

----

Zhou, X., Liu, X., 2009. The Monte Carlo EM method for estimating multivariate tobit latent variable models. Journal of Statistical Computation and Simulation 79, 1095–1107.

Costa, D., Lachos, V., Bazan, J., Azevedo, C., 2014. Estimation methods for multivariate Tobit confirmatory factor analysis. Computational Statistics & Data Analysis 79, 248–260.

Castro, L., Costa, D., Prates, M., Lachos, V., 2015. Likelihood-based inference for Tobit confirmatory factor analysis using the multivariate student-t distribution. Statistics and Computing 25, 1163–1183.

[Yves Rosseel]

On 11/21/22 13:22, M. van der Plaat wrote:
> I am wondering whether Lavaan supports Tobit-type CFAs/SEMS.

I am afraid not. Never needed this before. The technology exists. It has
just never been implemented in lavaan.

Not sure about alternatives.

Yves.