Development of R User Interfaces for the Bayesian Functional Mixed Model

The goal of this project is to develop open and user-friendly interface packages in the statistical environment R for the Bayesian functional mixed model (FMM), developed by Jeffrey S. Morris (Department of Biostatistics, The University of Texas MD Anderson Cancer Center). The FMM provides methods for the statistical analysis of a wide variety of general curve data, including one-dimensional (e.g. time-series data) as well as two- or three-dimensional curves (e.g. two-dimensional image data such as spectrograms), and also offers methods for more specific types of curve data, such as multi-channel electroencephalographic measurements (EEG). In contrast to canonical approaches to the analysis of curve data, the FMM makes use of the full amount of information contained in the data, including the curve-internal correlational structure, and allows to accommodate a wide range of experimental designs due to its implementation as a mixed effects model. The R packages developed in this project will make the FMM core software accessible for researchers from different fields in which curve data play a pivotal role, by providing user-friendly, intuitive and freely available FMM-interfaces. These will allow to visualize a model's underlying curve data, specify a FMM and work flexibly with its output with a minimal amount of specialized knowledge and coding. Building upon existing R code and functionality, we will incorporate a number of extensions of the core FMM software, provide methods for handling and visualizing big data sets and numerous other features. A special focus will be on integrating specialized FMM methods for data types which are widely used in cognitive neuropsychology (e.g. EEG data).