Stable specification search in structural equation model with latent variables

TL;DR

This paper extends the stable specification search method to latent variables in structural equation models, demonstrating improved performance over existing methods on simulated and real-world data.

Contribution

The study introduces S3C-Latent, a novel extension of stable specification search for latent variables, enhancing causal discovery in SEMs.

Findings

S3C-Latent outperforms PC-MIMBuild on simulated data.

Results on real-world ADHD and mental ability data align with prior research.

S3C-Latent provides more stable and parsimonious causal structures.

Abstract

In our previous study, we introduced stable specification search for cross-sectional data (S3C). It is an exploratory causal method that combines stability selection concept and multi-objective optimization to search for stable and parsimonious causal structures across the entire range of model complexities. In this study, we extended S3C to S3C-Latent, to model causal relations between latent variables. We evaluated S3C-Latent on simulated data and compared the results to those of PC-MIMBuild, an extension of the PC algorithm, the state-of-the-art causal discovery method. The comparison showed that S3C-Latent achieved better performance. We also applied S3C-Latent to real-world data of children with attention deficit/hyperactivity disorder and data about measuring mental abilities among pupils. The results are consistent with those of previous studies.