A multi-level longitudinal analysis of 80,000 online learners: Affective-Behaviour-Cognition models of learning gains

Conference of the EARLI SIG 17

Qualitative and Quantitative Approaches to Research on Learning and Instruction 17-19 August 2016 Maastricht University, Maastricht, the Netherlands.

Authors: Jekaterina Rogaten, Bart Rienties, Denise Whitelock, Simon J. Cross and Allison Littlejohn.

Open University, UK

One of the challenges facing higher education is understanding what counts for an excellent educational outcome. Historically academic performance was a variable of choice for measuring ‘excellence’ in education, but more recently a concept of learning gain, which can be defined as change in knowledge, skills and personal development across time (e.g., Andrews et al., 2011; Boyas et al., 2012) gained momentum. Educational research also mainly looked at cognitive gain largely ignoring affective changes (attitude) and behaviour (Tempelaar et al., 2015a). Current research aims to address this gap by developing and testing an Affective-Behaviour-Cognition model of learning gains using longitudinal multilevel modelling. The learner-generated affective-behaviour-cognition data was retrieved from university database for 80,000+ undergraduate students who started their degree in autumn 2013/14. The preliminary multilevel modelling revealed that cognitive and behaviour learning gains are well explained by the hypothesised Affective-Behaviour-Cognition model, whereas the more complex affective learning gains model needs further refinement. The main strength of this research is that approach used is a practical and scalable solution that could be used by teachers, learners, higher education institutions and the sector as a whole in facilitating students’ learning gains by further improving and personalising provision of higher education.

Aims of the ABC model of learning

An evidence-based approach of learning gains can be applied across Higher Educational Institutions (HEIs) would be of value to students, institutions and national organisations.

Most studies of learning gains have focussed solely on cognitive learning gains (Bowman, 2010; Liu, 2009; McGrath et al., 2015). In line with well-established educational psychology principles and recent learning analytics approaches, which enable greater insight to be achieved from large data sets, we propose an Affective-Behaviour-Cognition (ABC) model of learning, to broaden the concept of learning gain, and – more importantly – to develop, test, implement and evaluate a range of measurements for learning gains at each of the ABC levels.

A Higher Education Challenge

One of the challenges facing higher education is in understanding what counts for an excellent educational outcome, how students’ learning can be measured effectively, and how these measurements might be used to guide current investments and inform future developments.

While there is a substantial body of research examining learning gains in the USA and in the Netherlands (Bowman, 2010; Pascarella, Blaich, Martin, & Hanson, 2011; Tempelaar, Rienties, & Giesbers, 2015a), recent review of the learning gain literature by McGrath, Guerin, Harte, Frearson, and Manville (2015) indicated that approaches to measuring  learning gains are in their infancy in higher education in England.