Single Case Experimental Designs (SCEDs) are important tools in personalized interventions in health and behavior research and useful for studying treatment effects for small populations (e.g., individuals with rare diseases). Until recently, there has been little work on developing mediation methods for these designs. Recently, we received a grant from Institute of Education Sciences (IES) titled “Methods for synthesizing causal mediation effects in single-case experimental design data”. This project will explore these different techniques for estimating causal mediation effects and aggregating causal mediation effects across multiple N-of-1 studies.  

Relevant publications: 

• Valente, M. J., Pang, J., Rijnhart, J. J.M., Ferron, J., & Miočević, M. (2025). Estimating mediation effects in ABAB reversal designs. Evaluation & the Health Professions, 48(1), 129-142. 
• Valente, M. J., Rijnhart, J. J.M., & Miočević, M. (2023). Causal mediation effects in single case experimental designs. Psychological Methods, 28(2), 488-506. 

Often intervention theories predict that the intervention will operate differently for different subgroups or that the intervention must be varied to achieve similar effects in a specific population resulting in heterogeneity of treatment and mediated effects. Moderation, or effect modification, analysis is an important step in conducting subgroup analysis. Effect modification (moderation) is straightforward when the effect modifier is a single categorical variable (e.g., gender) but becomes complicated as soon as there are multiple effect modifiers of mixed variable types (categorical and continuous). In this project we are developing Monte Carlo simulation methods to account for highly heterogenous mediated effects and provide interpretable subgroup effects.  

Relevant publications: 

• Valente, M. J., Pang, J., & Cao, B. (2025). Understanding Who Benefits the Most from Interventions: Implications for Baseline Target Moderated Mediation Analysis with Multiple Moderators. Prevention Science, 26, 149-160. 
• Valente, M. J., Rijnhart, J. J. M., & Gonzalez, O. (2025). A novel approach to estimate moderated treatment effects and moderated mediated effects with continuous moderators. Psychological methods, 30(1) 1 - 15. 

The development of methods for causal mediation analysis based on the potential outcomes framework and counterfactual framework has been an important advancement in the causal inference literature. To make these causal mediation analysis methods accessible for a wide audience, various software programs have emerged over the past two decades. In 2020, we published a paper with an overview of causal mediation software programs available at that time. Since 2020, more causal mediation software programs have been developed. In this project, we aim to map all causal mediation software programs available to date and provide resources to help applied researchers decide which program is the best fit for their project. 

Relevant publications: 

• Schuster, N. A., Twisk, J. W. R., Heymans, M. W., & Rijnhart, J. J. M. (2022). Causal Mediation Analysis with a Binary Mediator: The Influence of the Estimation Approach and Causal Contrast. Structural Equation Modeling: A Multidisciplinary Journal, 1-10. 
• Valente, M. J., Rijnhart, J. J. M., Smyth, H. L., Muniz, F. B., & MacKinnon, D. P. (2020). Causal mediation programs in R, M plus, SAS, SPSS, and Stata. Structural equation modeling: a multidisciplinary journal, 27(6), 975-984. 

Systematic reviews and meta-analyses are often considered high-quality sources of evidence. However, effect estimates in systematic reviews of observational studies tend to exhibit high heterogeneity, which limits their usefulness for information interventions and policies. In this project, we are using insights from the causal inference literature to help improve the quality of evidence gained from systematic reviews of observational studies. 

Relevant publications: 

• Rijnhart, J. J.M., Rabbers, A., Rizzuto, S., Howard, A. M., & Valente, M. J. (2025). An umbrella review reveals that control variables are rarely considered as a source of heterogeneity in systematic reviews of observational studies. Journal of Clinical Epidemiology, 111826. 
• Rijnhart, J. J.M., Rabbers, A., & Rizzuto, S. (2024). Also long overdue: consideration of collider bias in guidelines and tools for systematic reviews and meta-analyses of observational studies. International Journal of Epidemiology, 53(6), dyae147. 

Dementia is more prevalent among certain subpopulations compared to others. In this project, we aim to study the mechanisms that are responsible for these differences in dementia prevalence across subpopulations. Funded by the Alzheimer’s Association, we are currently studying whether differences in educational attainment and occupations between women and men explain why women are at higher risk of dementia.  

Relevant publications: 

• Rijnhart, J. J., Bailey, R. J., Agbodo, J., Agrawal, V., Rodriguez-Olmo, V. M., & Salemi, J. L. (2025). Leveraging Mediation Analysis as a Tool to Study Mechanisms Underlying Health Inequities. Annals of Epidemiology. 
• Van Zutphen E.M., Rijnhart J. J. M., Rhebergen, D., Muller, M., Huisman, M., Beekman, A., Kok, A. A. K., & Appelman, Y. (2021). Do cardiovascular risk factors and cardiovascular disease explain sex differences in cognitive functioning in old age? Journal of Alzheimer’s Disease, 1-13.