Postgraduate student satisfaction: a multilevel analysis of PTES data

Student satisfaction has received growing attention in Higher Education systems in recent years, and are increasingly used for internal and external accountability in the sector. This does leave us with questions on the extent to which variance in student satisfaction can be explained by Higher Education Institution (HEI) and course attended rather than by individual student characteristics; and on what factors may predict student satisfaction. In this study we used the Postgraduate Taught Experience Survey (PTES) to look at these questions in a sample of approximate 70000 UK postgraduate students from 100 HEI’s. Firstly, confirmatory factor analyses were used to test the validity of the proposed multidimensional structure of student satisfaction in PTES. Then multilevel models were used to look at variance at three levels (HEI, course and student), and the relationships between institutional and student factors collected in PTES and student satisfaction. Two years of data (2014 and 2015) were used. Findings suggest that over 90% of variance is explained at the student level for all dimensions of student satisfaction in both years, with variance at the HEI level being particularly low. Individual student characteristics explain more variance than institutional characteristics, but only some (such as BME status) are significant, and explained variance is low. These findings suggest using student satisfaction as an accountability measure in HE may be highly problematic