Modeling the longevity of completion of higher degree studies at Makerere University

Abstract

Student persistence in graduate programs is widely regarded as an important yet challenging aspect of higher education in the literature. Research into this phenomenon is particularly problematic when (i) a considerable number of students have not completed their studies at the time of data collection, (ii) low enrollment and/or completion figures are observed, (iii) a normality assumption of completion time is made, and (iv) a detailed assessment of non-completion is required. A time-to-event approach in a Cox model, which uses enrollment time as censored for students who have not completed their studies at the time of the study, and makes no assumption regarding the distribution of completion time, solves the first three problems. A multinomial logistic allowing for at-least three outcomes of PhD candidature deals with the fourth problem. The aptness of implementing these approaches is illustrated using administrative data of all 295 PhD students at Makerere University in the 2000 to 2005 enrollment cohorts. The total elapsed time from first enrollment to submission of a final dissertation or thesis copy was adopted as a measure of completion time. A median completion time of 5.0 years for the 89 enrollees who had completed their studies by November 2010 points to a delayed completion of PhDs at Makerere. A 15% (n = 44) completion estimate, based on the five-year maximum permissible period of PhD training stipulated in the University Human Resource Manual, suggests a low timely completion rate for the University’s doctoral students. The rate of completion was higher for younger students, those registered in science disciplines, international students, and those from earlier enrollment periods (i.e., 2000 to 2002). Extended candidature beyond five years, rather than withdrawal, was more likely among the financially aided students and those from science disciplines. The results revealed that the dynamics of completion are not entirely unique to Makerere but relate with experiences reported among other universities elsewhere. However, the observed associations, modeled by a range of candidates, candidature, and institutional variables including discipline area, suggest the need for establishing measures to promote progress in doctoral studies at early stages of commencement as well as throughout the course of candidature.