Multilevel modelling of learning gains: The impact of module particulars on students' learning in Higher Education

In the UK, the introduction of the Teaching Excellence Framework (TEF) has increased interest in appropriate and valid measurement approaches of learning gains in Higher Education. Usually learning gains are measured using pre-post testing, but this study examines whether academic performance can be effectively used as proxy to estimate students' learning progress. Academic performance of 21,192 online learners from two major faculties was retrieved from university database. A three-level growth-curve model was estimated and results showed that 16% to 46% of variance in students' initial academic performance, and 51% to 77% of variance in their subsequent learning gains was due to them studying at a particular module. In addition, the results illustrate that students who studied in modules with initial high student achievements exhibited lower learning gains than students learning in modules with low initial student achievements. The importance of assessment and learning design for learning gains are outlined.