“…Despite the aforementioned early statistical validation of the tripartite model (Breckler, 1984;Ostrom, 1969), studies measuring attitudes have consistently and increasingly assumed unidimensionality, through the use of unidimensional scaling such as Likert scales (e.g., (Baum and Nelson, 2007;Gjersing et al, 2010;Macknight and Powell, 2001;Reuben et al, 1998;Wånggren et al, 2013). In other words, research typically produces an average or summated score, along a single scale, despite the statistical evidence that the scores represent different components of an attitude (Ostrom, 1969).…”