Methodology researcher Prof. Dr. Marie-Ann Sengewald is the first recipient of the LIfBi Research Award. She was honored for her publication on causal impact analysis with latent variables and categorical indicators using the EffectLiteR package. Her work focuses on how large data sets can be used to reliably determine whether a particular factor—such as private tuition-actually has an effect on learning performance. The award, which is being presented for the first time in 2025, will be given annually to employees of the institute who have made outstanding contributions in the areas of research or infrastructure.
Marie-Ann Sengewald’s work was selected by LIfBi’s Scientific Advisory Board for the award. The reasoning emphasizes that the article is situated within the institute’s core research areas and stands out for its methodological excellence, broad applicability, and high scientific as well as socio-political relevance—particularly regarding questions on the impact of private tuition. The tutorial-like structure, including provided R code and software implementation that foster open and collaborative research, was also rated especially positively.
In her method-oriented work, the psychologist uses NEPS data to demonstrate an advanced statistical approach to estimating causal effects with non-experimental observational data, taking into account measurement errors that are always present. To this end, Sengewald compared different modeling strategies, as these differ in their handling of measurement errors and the accuracy of causal effect estimates. Using an application example on the “effects of private tuition,” she also showed that latent variables, i.e., statistical estimates of abilities that cannot be measured directly, provide more accurate results than conventional ability values.
For the award winner, receiving the LIfBi Research Award is an incentive to translate scientific findings from methodological research into practical applications. “What drives me is the conviction that methodological precision and social relevance are not mutually exclusive. Research unfolds its potential when statistical models provide answers to questions that are of actual importance to families, schools, and politics.”
During her time at LIfBi in Department 3, ‘“Research Data Center, Method Development,” Marie-Ann Sengewald focused not only on replication research but also on causal inference in models of latent variables. Since September 2025, she has held the position of W1 Professor of Psychological Methods and Diagnostics at Friedrich-Alexander-Universität Erlangen-Nürnberg.
Original publication:
Sengewald, M.-A., & Mayer, A. (2024). Causal effect analysis in nonrandomized data with latent variables and categorical indicators: The implementation and benefits of EffectLiteR. Psychological Methods, 29(2), 287–307. https://doi.org/10.1037/met0000489