Dr. William Cumberland's interests are in the areas of finite population sampling and biomedical applications of stochastic processes. His research in sampling falls under the heading of model based estimation, and it shows that relying solely on the sampling plan can lead to very misleading inferences. Through the use of models, better estimates of standard errors have been developed. Cumberland's interest in sampling has led to collaborative work in health services research in the UCLA Center for Health Policy Research, concentrating on the analysis of National Center for Health Sciences and Current Population Survey data in the study of health insurance. His work in stochastic processes has led to models for progression of CD4 counts in HIV. Cumberland has also been active in biostatistical applications in tumor immunology, especially in AIDS research.