In this chapter, the use of Structural Equation Modelling (SEM) in the analysis of the effects of curricula designs on student learning and experience in higher education is explored. The origins, fundamental assumptions, metaphysical and ontological commitments of SEM, are explicated. This is followed by an exemplification of the method by use of a case study. The chapter includes an argument for the adoption of a realist account of latent variables on the basis that the constructs they represent are in principle manipulable, even though experimental manipulation is not typically a feature of research on curricula in higher education. The chapter concludes with the application of these ideas to the scholarship of learning and teaching (SoLT). It is asserted that only by the adoption of a realist perspective on the data collected in the pursuit of SoLT goals can these goals be adequately attained.
TopIntroduction
The study of higher education students’ experiences and learning outcomes is a substantive field in its own right, but it has many variants, in both method and purpose. For instance students have been surveyed or observed for a range of purposes, including to address the following themes, goals and issues: quality and nature of teachers’ practices (Chau & Hocevar, 1994; for an example and a review of the literature see Marsh & Roche, 1994); assurance of learning or the outcomes thereof (Barrie, 2006; Barrie, Hughes, Smith, & Thomson, 2009; Biggs & Collis, 1982; Griffin, Coates, McInnes, & James, 2003; Wilson, Lizzio, & Ramsden, 1997); and students’ approaches to studying (Entwistle, Hanley, & Hounsell, 1979; Entwistle & Ramsden, 1983). In this chapter, the use of Structural Equation Modeling (SEM) for the exploration of the impact of aspects of curriculum design on student learning and satisfaction is explored. Along the way, there will be an exploration of the philosophical commitments that such an approach implies.