| Dissertation |
Thesis (M.Sc.) --NUI, 2016 at Department of Statistics, UCC. |
| Summary |
Chapter one of this thesis features a dynamic PET study with data gathered from a single patient. This study aimed to minimise the length of the scan time. Two compartmental models were used to model kinetic processes. Parameter estimates for the flux, flow, volume of blood and volume of distribution were formed. Simulations were generated and parameter estimates extracted. Single time points were removed from the study and parameter estimates recalculated. Parameter estimates were compared to original estimates using measures of bias, MSE and variance. It was found that scan time could be reduced to 25 minutes. Chapter two features a study introducing competing risks to the SCORE project. This study aimed to develop a system of risk estimation for clinical practice in Europe. A Fine and Gray adjusted cox regression model was used to form coefficients of the hazard of the subdistribution. A cox proportional hazards model was used to form the cause specific hazards coefficients. Schoenfeld plots and Schoenfeld-type plots were used to assess the proportionality of the covariates. Little difference existed between the coefficients of the hazard of the subdistribution and the cause specific hazards coefficients. Risk charts based on the Breslow estimator were considered. Chapter three involves a study of the Teaching and Learning instrument included in the Student Experience Survey. Data was gathered from previous iterations of the survey. This study aimed to establish subgroups of students based on their responses to the Teaching and Learning instrument using multiple group latent class analysis. A secondary aim was to establish if being a first year student was a good predictor of latent class membership. It was found that seven latent classes existed and had the same meaning across all time-points. Being a first year was found to be a good predictor of latent class membership. |
| Subject |
Statistics.
|
|
Medical statistics.
|
|
Educational statistics.
|
| Collection |
Theses Masters (Research)
|
|
Theses Statistics Department
|
| Description |
167 pages ; 30 cm. |
|
