arXiv subjects

Holger Brandt

Publications and source records attributed to Holger Brandt.

6 recordsLinked to original sources

arXiv2026

Representation Learning for Semiparametric Causal Mediation Analysis under No Essential Heterogeneity

We propose a two-stage estimator for structural mediation parameters that combines deep representation learning with G-estimation under the "no essential heterogeneity" (NEH) assumption. We call the method UNIT. In the first stage,TARNet estimates the heterogeneous effect of a randomized treatment on a mediator by learning a shared covariate representation across treatment arms.The resulting conditional average treatment effect (CATE) estimate provides a plug-in approximation to the heterogeneity-dependent component of the weight function entering the G-estimating equation of Zheng and Zhou (2015), which identifies the structural parameters even in the presence of unmeasured mediator-outcome confounding. We show that more accurate first-stage representation learning can yield a more informative plug-in weight and thereby improve the precision of the structural parameter estimator. In simulations with non-Gaussian covariates and nonlinear mediator effects, TARNet weights reduce the Stage-2 standard error of the mediation coefficient by a factor of $1.45$ to $1.51$ (median across replications, $n \ge 2000$) relative to the classical approach, at no cost to bias or coverage.

stat.ML↗

arXiv2026

RAPSEM: Identifying Latent Mediators Without Sequential Ignorability via a Rank-Preserving Structural Equation Model

Standard structural equation models (SEMs) are often used to identify latent mediators. However, valid inference typically relies on the strong, frequently violated Sequential Ignorability assumption. We introduce the Rank-Preserving Structural Equation Model (RAPSEM), which increases robustness through G-estimation while maintaining the measurement model's integrity through a two-stage method of moments (2SMM) for factor score corrections. RAPSEM replaces the no unmeasured mediator-outcome confounding with the weaker no unobserved effect modification assumption. By leveraging treatment randomization, RAPSEM achieves identification in a manner equivalent to instrumental variable estimation through structurally emerging instruments. Specifically, identification relies on treatment-covariate interactions that influence the mediator but have no direct effect on the outcome, allowing researchers to utilize natural heterogeneity in treatment response as a testable source of identification. We provide a robustness assessment for the core identifying assumption and establish the consistency and asymptotic normality of the resulting estimator. Simulation studies demonstrate that RAPSEM remains unbiased under unobserved confounding, whereas standard SEM yields biased results. RAPSEM achieves reasonable power for sample sizes above 500, depending on the strength of the structural instruments. The method is implemented in the accompanying rapsem R package, and its practical utility is illustrated through an empirical example from educational research. The code is available at https://github.com/PsychometricsMZ/RAPSEM.

stat.ME↗

arXiv2026

Dynamic Latent Class Structural Equation Modeling: A Hands-On Tutorial for Modeling Intensive Longitudinal Data

In this tutorial, we provide a hands-on guideline on how to implement complex Dynamic Latent Class Structural Equation Models (DLCSEM) in the Bayesian software JAGS. We provide building blocks starting with simple Confirmatory Factor and Time Series analysis, and then extend these blocks to Multilevel Models and Dynamic Structural Equation Models (DSEM). Subsequently, we introduce Hidden Markov Switching Models (HMSM) and demonstrate their integration with DSEM to yield DLCSEM. Leading through the tutorial is an example from clinical psychology using data on a generalized anxiety treatment that includes scales on anxiety symptoms and the Working Alliance Inventory that measures alliance between therapists and patients. Within each block, we provide an overview, specific hypotheses we want to test, the resulting model and its implementation, as well as an interpretation of the results. The aim of this tutorial is to provide a step-by-step guide for applied researchers that enables them to use this flexible DLCSEM framework for their own analyses.

stat.ME↗

arXiv2025

Frequentist forecasting in regime-switching models with extended Hamilton filter

Psychological change processes, such as university student dropout in math, often exhibit discrete latent state transitions and can be studied using regime-switching models with intensive longitudinal data (ILD). Recently, regime-switching state-space (RSSS) models have been extended to allow for latent variables and their autoregressive effects. Despite this progress, estimation methods for handling both intra-individual changes and inter-individual differences as predictors of regime-switches need further exploration. Specifically, there's a need for frequentist estimation methods in dynamic latent variable frameworks that allow real-time inferences and forecasts of latent or observed variables during ongoing data collection. Building on Chow and Zhang's (2013) extended Kim filter, we introduce a first frequentist filter for RSSS models which allows hidden Markov(-switching) models to depend on both latent within- and between-individual characteristics. As a counterpart of Kelava et al.'s (2022) Bayesian forecasting filter for nonlinear dynamic latent class structural equation models (NDLC-SEM), our proposed method is the first frequentist approach within this general class of models. In an empirical study, the filter is applied to forecast emotions and behavior related to student dropout in math. Parameter recovery and prediction of regime and dynamic latent variables are evaluated through simulation study.

stat.ME↗

arXiv2025

Advantages and limitations in the use of transfer learning for individual treatment effects in causal machine learning

Generalizing causal knowledge across diverse environments is challenging, especially when estimates from large-scale datasets must be applied to smaller or systematically different contexts, where external validity is critical. Model-based estimators of individual treatment effects (ITE) from machine learning require large sample sizes, limiting their applicability in domains such as behavioral sciences with smaller datasets. We demonstrate how estimation of ITEs with Treatment Agnostic Representation Networks (TARNet; Shalit et al., 2017) can be improved by leveraging knowledge from source datasets and adapting it to new settings via transfer learning (TL-TARNet; Aloui et al., 2023). In simulations that vary source and sample sizes and consider both randomized and non-randomized intervention target settings, the transfer-learning extension TL-TARNet improves upon standard TARNet, reducing ITE error and attenuating bias when a large unbiased source is available and target samples are small. In an empirical application using the India Human Development Survey (IHDS-II), we estimate the effect of mothers' firewood collection time on children's weekly study time; transfer learning pulls the target mean ITEs toward the source ITE estimate, reducing bias in the estimates obtained without transfer. These results suggest that transfer learning for causal models can improve the estimation of ITE in small samples.

stat.ML↗

arXiv2021

Forecasting intra-individual changes of affective states taking into account inter-individual differences using intensive longitudinal data from a university student drop out study in math

The longitudinal process that leads to university student drop out in STEM subjects can be described by referring to a) inter-individual differences (e.g., cognitive abilities) as well as b) intra-individual changes (e.g., affective states), c) (unobserved) heterogeneity of trajectories, and d) time-dependent variables. Large dynamic latent variable model frameworks for intensive longitudinal data (ILD) have been proposed which are (partially) capable of simultaneously separating the complex data structures (e.g., DLCA; Asparouhov, Hamaker, & Muthén, 2017; DSEM; Asparouhov, Hamaker, & Muthén, 2018; NDLC-SEM, Kelava & Brandt, 2019). From a methodological perspective, forecasting in dynamic frameworks allowing for real-time inferences on latent or observed variables based on ongoing data collection has not been an extensive research topic. From a practical perspective, there has been no empirical study on student drop out in math that integrates ILD, dynamic frameworks, and forecasting of critical states of the individuals allowing for real-time interventions. In this paper, we show how Bayesian forecasting of multivariate intra-individual variables and time-dependent class membership of individuals (affective states) can be performed in these dynamic frameworks. To illustrate our approach, we use an empirical example where we apply forecasting methodology to ILD from a large university student drop out study in math with multivariate observations collected over 50 measurement occasions from multiple students (N = 122). More specifically, we forecast emotions and behavior related to drop out. This allows us to model (i) just-in-time interventions, (ii) detection of heterogeneity in trajectories, and (iii) prediction of emerging dynamic states (e.g. critical stress levels or pre-decisional states).

stat.ME↗