This paper introduces a framework of techniques that separately have in the past been developed and used on survival data but are now proposed to be separate parts of a sequence of stages in a much larger modelling process. By integrating these methods to¬-gether, it is proposed that the survival data undergoes a much more thorough investiga¬tion and as a result yields a greater in-depth reporting of the survival distributions, and the covariates that best characterise the survival distribution behaviours in a manner that retains the simple straightforward way in which these methods are clear and easy to use. The framework is applied to a specific data set to study the Length of Studies (LoS) of students at a Greek university enrolled on courses until the event of interest occurs, that is, the depar¬tu¬re from the course due to graduation, dropout or extensive period of study time. Results conclude that the approaches applied in this framework offer a more in-depth insight into student behaviour highlighting also the most influential characteristics.
Kalamatianou, A., Marshall, A., Zenga, M. (2014). Coxian Phase-type Distributions, Survival Trees and the Gini Index Collaborating together to Provide an in-Depth Knowledge of the Distribution of the Length of University Studies. In C. H. Skiadas (a cura di), Theoretical and Applied Issues in Statistics and Demography (pp. 103-120). ISAST.
Coxian Phase-type Distributions, Survival Trees and the Gini Index Collaborating together to Provide an in-Depth Knowledge of the Distribution of the Length of University Studies
ZENGA, MARIANGELAUltimo
2014
Abstract
This paper introduces a framework of techniques that separately have in the past been developed and used on survival data but are now proposed to be separate parts of a sequence of stages in a much larger modelling process. By integrating these methods to¬-gether, it is proposed that the survival data undergoes a much more thorough investiga¬tion and as a result yields a greater in-depth reporting of the survival distributions, and the covariates that best characterise the survival distribution behaviours in a manner that retains the simple straightforward way in which these methods are clear and easy to use. The framework is applied to a specific data set to study the Length of Studies (LoS) of students at a Greek university enrolled on courses until the event of interest occurs, that is, the depar¬tu¬re from the course due to graduation, dropout or extensive period of study time. Results conclude that the approaches applied in this framework offer a more in-depth insight into student behaviour highlighting also the most influential characteristics.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.