Recipient Organization
OREGON STATE UNIVERSITY
(N/A)
CORVALLIS,OR 97331
Performing Department
STATISTICS
Non Technical Summary
Ill-fitted experimental or study designs can often waste both time and money and may lead to incorrect or misleading conclusions. Inefficient data analysis methods can require large sample sizes to achieve sufficient statistical power. Appropriate statistical procedures and methods change according to the project at hand. Drs. Gitelman and Qu will remain available to researchers and graduate students in the College of Agricultural Science for statistical consultation and assistance.
Animal Health Component
50%
Research Effort Categories
Basic
25%
Applied
50%
Developmental
25%
Goals / Objectives
The objectives of the Statistical Consulting project are three-fold: To assist researchers in the Agriculture Sciences with study design and data analytic methods, thereby helping to maximize information gain and relevant scientific content from efficient data collection. To contribute to the development of new statistical tools and methodology for use in the Agriculture Sciences, and to assist in the application of existing statistical tools and methodology to problems encountered by the Agriculture Sciences. To serve on Masters of Science and PhD committees for students in the College of Agricultural Sciences, thereby assisting in the educational and research pursuits of students at Oregon State University. Any one of these three objectives may result in scientific publications on w hich Drs. Gitelman and Qu are collaborating authors.
Project Methods
Dr. Gitelman's work: Dr. Gitelman has worked extensively with Dr. Bruce McCune (faculty, Botany and Plant Pathology); Dr. Mina McDaniel (faculty, Food Science and Technology); and Dr. Dan Ball (faculty, Weed Science - Pendleton Experiment Station). Each of these collaborations has resulted in publications in press or in preparation for journals in the applied areas of Ecology, Sensory Science and Weed Science. In addition, the work with Dr. Dan Ball has resulted in a submission to the Journal of Agricultural Biological and Environmental Statistics (JABES) in which Dr. Gitelman presents a Bayesian hierarchical change-point model. Detailed descriptions of the statistical assistance provided to each of these researchers is given below. Dr. Gitelman's contribution was in providing a framework in which a Bayes Factor could be computed to compare the SpOcc model results to other, parametric model results. The Bayes Factor is a quantity that is similar to a Likelihood Ratio
Statistic, but more flexible in that it does not require that one model be nested within the other one. The collaborative work with Dr. Mina McDaniel has involved several students' masters work. In particular, Dr. Gitelman has assisted in the analysis of multivariate sensory data, from which it can be difficult to detect meaningful signals since there is quite a bit of noise in these data. Dr. Qu's work: Dr. Qu has worked with Ajoy Velayudhan and John Selker (faculty members) and Ling Zhang (Ph.D. student) in the Bioengineering Department on numerical estimation of multicomponent adsorption isotherms in preparative chromatography. They have published a joint paper in Journal of Chromatography, and continue their collaboration. Dr. Qu has also worked with researchers (John Jaeger) in Animal Science on predicting the time of parturition in spring calving beef cows.