The projects in the Center span a broad range of experimental paradigms, ranging from in vitro preparations and animal studies to clinical studies on humans. The statistical requirements are corresponding broad, and the statistical core plans to consult as needed across this broad range of experiments. In general, the core will provide guidance regarding experiments which involve standard statistical techniques and well as those that involve statistical innovation. At the planning stage, we will consult on appropriate experimental design, effective methods of data collection and storage, and using power analysis to determine adequate sample sizes. At the analysis stage, we will provide guidance in performing both exploratory as well as final analysis of the data, including locating outliers, testing assumptions, using parametric data transformations or nonparametric methods as appropriate, advising on multiple comparison techniques, and curve fitting (both linear and nonlinear). Many of the experiments involving standard designs will use t tests and one or two way ANOVA. It is anticipated that innovative statistical techniques will be required with some of the projects, particularly with projects 6 and 2. Dr. Iyengar is already consulting with Dr. Lewis on an experiment that calls for a nonstandard statistical approach. The experiment involves comparing the density of pyramidal neurons from the prefrontal cortex of schizophrenic and control subjects. In addition to consulting on planning experiments, and analyzing data, the statistical core will conduct workshops on statistical matters relevant to some aspect of the Center's mission. Possible topics include the analysis of longitudinal data, multiple comparisons, and the mathematical modeling of ion channels.
Showing the most recent 10 out of 382 publications
