In the area of educational media research, the dismissal of media comparison approaches led to a rise of investigations concerned with “media attributes” (Wetzel, Radtke , & Stern, 1994). Educational media researchers (e.g., Clark, 1994; Kozma , 1994) now urge investigators to consider those variables that cluster around “media” (e.g., speed of presentation, familiarity, editing style, clarity of images, topic). Additionally, researchers need to examine those associated with “method” (e.g., instructor behaviour, repeated viewings, length of exposure, motivational attitudes) as a way to account for differences in performance . In a similar vein, Tatsuki (1993) claims that CALL research has also suffered because of significant flaws that include exceedingly small sample sizes, a lack of control groups, a tendency to overgeneralize, and a failure to operationalize key variables. As with Dunkel (1991), Tatsuki (1993) called on researchers to abandon comparative designs in favor of more “basic research into how learners learn language and how specific media affect language learning” (p. 24). Fortunately, such advice has been largely heeded and researchers now look more closely at how the interactions of computers, learners, and instructors influence the process of language learning . In regard to the examination of learner behaviours, or strategies, CALL researchers need to explore the framework of “constructively responsive” readers set out by Pressley and Afflerbach (1995). This perspective, based on the underpinnings of cognitive constructivism (for an overview, see Driscoll, 2000), regards comprehends as flexible, concerned with main ideas, and, most importantly, responsive to the presentation of textual resources as they attempt to build a coherent macro-structure.