Date of Award
2001
Level of Access Assigned by Author
Open-Access Thesis
Degree Name
Master of Arts (MA)
Department
Mathematics
Advisor
William A. Halteman
Second Committee Member
Robert D. Franzosa
Third Committee Member
Sundar Subramanian
Abstract
Many discrete response variables have counts as possible outcomes. Poisson regression has been recognized as an important tool for analyzing count data. This technique includes the simple Poisson generalized linear model and mixtures of independent Poisson models as special cases. Generalized linear models have been found useful in many statistical analysis. Count data analyzed under such models often exhibit overdispersion. In many practical circumstances the restriction that the mean and variance are equal is not realistic. Especially, when there is overdispersion in the data, a conditional negative binomial mixed model, given some random effects, could be an attractive alternative. This paper focuses on the data analysis using mixed Poisson regressions and mixed Negative Binomial regressions. The motivation comes from attempts to analyze habitat use from the snow tracking data.
Recommended Citation
Jung, Jungah, "Using generalized linear models with a mixed random component to analyze count data" (2001). Electronic Theses and Dissertations. 409.
https://digitalcommons.library.umaine.edu/etd/409