2003; Moreno, 2006; Moreno and Mayer, 1999). Although Mayer and colleagues’ cognitive model of multimedia learning has not been expressly tested and developed in informal environments, it may be of interest to the field. Researchers have developed this work through a series of laboratory-based studies in which they use computer technology to test the influence of (1) modality (text versus audio) and (2) spatial and temporal contiguity of particular media elements (e.g., the spatial and temporal proximity of text and a related video animation in a computer environment) on cognition. Through a series of controlled laboratory experiments primarily with college student volunteers, researchers have established that, when words and images are represented contiguously in time and/or space, the effectiveness of multimedia instruction increases, influencing recall (Mayer, 1989; Mayer, Steinhoff, Bower, and Mars, 1995) and transfer to novel problems (Mayer, 1997). Mayer and Anderson (1991, 1992) have also established that, in a laboratory setting, concurrent presentation of textual and graphic information (e.g., explanation of how a tire pump works and relevant imagery) is more conducive to recall and transfer than presentation of the same information in a series (e.g., text then graphic or graphics then text). They have also found that students presented with auditory verbal materials plus animations recalled more, solved problems better, and were better able to match the visual and verbal elements than those who learn with on-screen text plus animations. These ideas may be fruitful for design and further testing in informal environments for science learning.
There is also evidence that elements of certain media may help to focus learners on important issues. This feature can be useful for designed settings, in which opportunities to engage participants are often narrow and fleeting. For example, motion pictures use various cinematic techniques, such as panning and zooming, to help learners attend to filmed details that might otherwise go unnoticed during casual viewing (Salomon, 1994). Similar claims have been made about the interactive properties that computational media afford. For example, well-designed computer games and simulations have been touted by several researchers (Baillie and Percoco, 2001; Gee, 2007; Greenfield, 1984; Papert, 1980) as ideal spaces for learning science for a number of reasons: they enable learners to customize the learning environment; they situate learning in a more authentic context; they provide direct experiences and interaction with intangible, abstract, ideal, complex, or otherwise unavailable scientific phenomena; and they engage users in collaborative, active, and problem-based learning.
Perhaps more important is Robert Kozma’s rebuttal to Clark’s argument, especially in light of the strands discussed throughout this volume. Kozma (1991) argued that one could study learning occurred when the same message was presented in different media formats. But he emphasized the importance of theories of distributed cognition (Salomon, 1993) that describe how individual cognition is developed through interactions with peers and