Non-traditional Student Withdrawal: a Grounded Bayesian Vector Auto Regression Framework

1 Chris Laing, Southampton Institute, Technology Faculty, Southampton, SO14 0YN, UK, [email protected] 2 Alan Robinson, Southampton Institute, Technology Faculty, Southampton, SO14 0YN, UK, [email protected] Abstract  Previously it has been proposed that explanations of non-traditional withdrawal might be defined by the underlying characteristics of the teaching and learning environment, especially on how a student’s perceptions and expectations of that environment impact on their decision to withdraw. An ethnographic study using Grounded Theory was used to capture these underlying characteristics. The results of that study provided an explanation of the teaching and learning environment as a function of student beliefs, staff-student actions, and institutional intentions. A follow-up longitudinal study is now being undertaken. The aim of this study is, (a) refine the grounded analysis, and (b) model the grounded teaching and learning environment within a Bayesian Vector Auto Regression (BVAR) framework.