Statistical Inference on Latent Structures
Research
Our research focuses on the development, evaluation, and application of advanced statistical methods designed to uncover, measure, and draw valid inferences about unobserved phenomena. By bridging the gap between abstract psychometric theory and empirical practice, we aim to provide researchers with the rigorous methodological tools necessary to test complex theories, account for measurement error, and ensure robust scientific conclusions.
Methodological Developments in Structural Equation Modeling
At the core of our methodological work is the continuous development and evaluation of Structural Equation Modeling as a flexible framework for multivariate data analysis. We investigate and refine estimation techniques, evaluate model fit criteria, and address common violations of distributional assumptions. Our goal is to enhance the statistical power, accuracy, and overall robustness of parameter estimation in complex models.
Selected Publiations:
- Schamberger, T., Schuberth, F., Henseler, J. (2026). Moderated Mediation with Composites: The Composite Moderated Structural Equations Approach, Behavior Research Methods, 58:104, https://doi.org/10.3758/s13428-025-02930-w
- Schuberth, F., Schamberger, T., Kemény, I., Henseler, J. (2025). The Sum Score Model: Specifying and Testing Equally Weighted Composites Using Structural Equation Modeling. Psychometrika. 90(1): 358-383. https://doi.org/10.1017/psy.2024.5
- Schuberth, F., Schamberger, T., Hensler, J. (2024). More Powerful ParameterTests? No, Rather Biased Parameter Estimates. Some Reflections on Path Analysis With Weighted Composites. Behavior Research Methods, 56: 4205-4215, https://doi.org/10.3758/s13428-023-02256-5
- Schamberger, T., Schuberth, F., Henseler, J. (2023). Confirmatory Composite Analysis in Human Development Research, International Journal of Behavioral Development 47(1): 89-100. https://doi.org/10.1177/01650254221117506
Modeling Different Types of Latent Structures
A central focus of our research lies in the critical distinction between different types of latent structures. We analyze the statistical properties, estimation challenges, and substantive implications of reflective measurement models (where indicators are viewed as manifestations caused by the latent construct) versus composite models (where indicators jointly form and constitute the construct). A primary objective of our work is to overcome traditional methodological divides and enable researchers to investigate both types of latent structures equally within a unified modeling framework.
Projects:
FWO (Research Foundation Flanders) program “Senior research project fundamental research”: „Structural equation modeling with both latent variables and composites” with Prof. Dr. Yves Rosseel (Ghent University), 01.01.2026 - 31.12.2029
Selected Publications:
- Bauer, J., Mayer, A., Fuchs, C., and Schamberger, T. (2025). Misspecification in Structural Equation Modeling: The Choice of Latent Variables, Causal-Formative Constructs or Composites (preprint: https://doi.org/10.48550/arXiv.2507.21998)
- Schamberger, T., Schuberth, F., Henseler, J., Rosseel, Y. (2025). Factor- and Composite-Based Structural Equation Modeling - A New Approach to Incorporate Composites in the Traditional SEM Framework (preprint: https://doi.org/10.48550/arXiv.2508.06112)
- Henseler, J., Xi, Y., Schamberger, T., Hancock, G.R., Schuberth, F. (2025). Specifying Composites in Structural Equation Modeling: Traditional, Recent and New Approaches (preprint, https://doi.org/10.48550/arXiv.2509.01292)
Interdisciplinary Applications
Methodological developments only unfold their full value when applied to real-world empirical questions. Therefore, a key component of our work involves the interdisciplinary application of the developed methods across diverse fields such as health sciences, sports science, and psychology. By closely collaborating with domain-specific researchers, we help translate complex theoretical frameworks into rigorous empirical designs, ensuring that sophisticated statistical tools generate meaningful and reliable insights in practice.
Selected Publications:
- Adam, T.*, Bauer, J.*, Deutscher, C.*, Fuchs, C.*, Schamberger, T.*, Winkelmann, D.* (2026). Measuring Football Fever Through Wearable Technology, Scientific Reports, 16: 3866, https://doi.org/10.1038/s41598-026-36182-1
- Alaze, A., Hagen, D., Schamberger, T., Razum, O., Miani, C. (2025). Empirical Validation of a Conceptual Framework Linking Gendered Pathways to Adolescent Mental Health Using Cross- Sectional Data from Ethiopia