"Everything should be made as simple as possible, but not simpler." - Albert Einstein

Seminar Announcement

Stratification Design Assessment for Waterfowl Production Surveys on Tetlin National Wildlife Refuge

Jonathan Greenberg, STAT-MS Candidate, Colorado State University

Tuesday, May 10, 2011

10:00 a.m., room 008, Statistics Bldg

ABSTRACT

The Tetlin National Wildlife Refuge in Alaska is a primary breeding ground for a number of species of migratory waterfowl.  Accurate and precise estimation of the total waterfowl population is difficult and expensive.  In this paper, I will explain the current estimator being used and then I will explore the effectiveness of five alternative stratified estimators.  Each estimator will be assessed on the basis of bias, standard deviations, and root median squared error across 19 different scenarios developed using simulation testing techniques.  In the end, I find that there is no superior estimator overall, but different estimators excel in different scenarios.  The Penalized Post-Stratification and Markov Post-Stratification estimators are superior under Baseline scenarios and scenarios with steep, crossing abundance curves.  The Baseline scenario is superior under scenarios with non-crossing abundance curves.  Based on the results from this project, I will make suggestions to Tetlin wildlife biologists for improving the estimation of total waterfowl population.  

Advisory Committee:

Dr. Geof Givens, Advisor, Statistics Department, CSU
Dr. Jennifer Hoeting, Committee Member, Statistics Department, CSU
Dr. Dan Cooley, Committee Member, Statistics Department, CSU
Dr. Stephen Jablonsky, Outside Member, Accounting Department, CSU