The Department of Statistics (Faculty 5) is responsible for profile projects in the area of method development. Postdocs develop flexible quantitative methods with improved predictive accuracy for large data sets (large-scale studies; panel data; digital behavioral traces) and small case numbers (single-case/few-case research designs; heterogeneous (intervention) effects in small subgroups, repeated measures in ABAB designs and within-person RCTs). In addition, there are techniques that combine these areas and thus can quantify causal effects of agile interventions. Through methodological research, participation in data (re)analysis and prognostic modeling, as well as study design, the department is the scientific liaison for the FAIR profile area.

In the Department of Rehabilitation Sciences, the focus of Agile PAIR lies on analysing data relating to a computer-based intervention for children with learning disorders (large-scale data). Further, we conduct small-sample studies pertaining to individually adaptive prevention and intervention measures relating to mathematics and vision. Here, a core concern in Agile PAIR is the data-driven development of a tablet-based, adaptive intervention for primary school-aged children with mathematical difficulties taking into account their visual capabilities. Additional measures will consider increased screen time and its potential influence on school performance as well as the development of vision.

The Dortmund Data Science Center (DoDSc) is an interdisciplinary research center that integrates long-standing expertises in handling large, high-dimensional data and in Bayesian statistics. These methods enable an efficient processing of large data streams and distributed data, and statistical analysis in presence of small case numbers.

The DoDSc provides method development for FAIR in the area of advanced data processing: for instance, scalable sketching and subsampling techniques provide interpretable aggregates of behavioral traces adaptively and in real time, and complex data streams of agile interventions can be analyzed using streamed statistical models.

In addition, the DoDSc is responsible within FAIR for the conception, organization, and implementation of modular qualification programs in the fields of quantitative methodological training for handling large data sets or small case numbers within social science oriented research questions.