Kerby Shedden, PhD
- Professor, Statistics
- Professor, Biostatistics
Dr. Shedden is interested in methods, software, and theory for statistical analysis of many forms of complex data. Through his work directing the statistical consulting center (CSCAR) he collaborates with UM investigators in a wide variety of areas. He is also actively involved in efforts to build and improve statistical software in multiple language environments.
- PhD, UCLA, 1999
- BS, University of Michigan, 1994
Quantile and expectile analysis of clustered, longitudinal, and spatial data, regularized projection pursuit, mediation analysis.
Shedden's work on the epigenetics of human growth and development is based around a longitudinal cohort study in Mali, and focuses on intergenerational transmission of signals influencing growth of a child in response to environmental exposures of the parents.
Shedden's work on the spatial morphometrics of human tissues occurs in the context of studies of chronic kidney disease. The goal is to use statistical methods to analyze images of human kidney tissue, aiming to characterize patterns of normal and atypical glomerular distribution related to renal impairment associated with aging, or driven by disease processes.
Shedden's work on signal analysis of EEG data is motivated by collaborations with neurologists seeking to better delineate the region of origin of epileptic seizures, based on propagation of high-frequency oscillations and other features of neuronal signalling captured by modern high-resolution sensors.
Shedden's work on sensors and physical activity is based on a project looking at dyadic heartrate coupling in patient/caregiver dyads, utilizing novel methods of covariance analysis to identify shared and unique features of heart rate signals captured by the new generation of wearable heartrate monitors.
Areas of Expertise: Biostatistics