B.G. Langenberg

My research interests include statistical methods longitudinal and experimental designs in the field of psychology, medicine and life sciences, such as structural equation models, mixed models and multilevel models, growth curve models, and analysis of variance. I also have a strong interest in methods and algorithms for model selection and data mining.

Main Research Topics:

• Structural equation modeling, for instance:
  • growth curve models
  • cross-lagged designs
  • longitudinal designs
  • time-series (dynamic) structural equations models (DSEM)
  • Bayesian structural equation models (BSEM)
• Linear mixed models + Regression + ANOVA
• Algorithms for subgroup discovery in large data sets
  • In clinical trials, we are oftentimes interested in finding combinations of covariates that describe subgroups that benefit particularly well from a treatment. For instance, consider patients with depressive symptoms after a diagnosis of a severe medical condition. We may find that participants with high family support benefit particularly well from cognitive behavioral therapy. Data mining algorithms can be used to efficiently search through a large number of covariates (e.g., family support). Identifying those covariates can aid in increasing the effectiveness of a treatment by also focussing on the covariates during the treatment.