Against the background of Moreno's “cognitive-affective theory of learning with media” (CATLM) (Moreno, 2006), three papers on cognitive and affective processes in learning with multimedia are discussed in this commentary. The papers provide valuable insights in how cognitive processing and learning results can be affected by constructs such as “situational interest”, “positive emotions”, or “confusion”, and they suggest questions for further research in this field.
For a long time, research on technology-based learning and instruction has primarily been focusing on cognitive aspects. For example, the “Cognitive Theory of Multimedia Learning” (CTML) (Mayer, 2005) or the “Integrative Model of Text and Picture Comprehension” (Schnotz, 2005) have, together with “Cognitive Load Theory” (Sweller, 1988), widely been used to inspire research in the field of multimedia learning. From this research we know a lot about learning with combinations of verbal and pictorial information, and based on these theories and models we are able to describe, explain, and predict effects of different formats and designs of multimedia instruction. Affective aspects, however, such as emotion and motivation, although having a long tradition in research on learning and instruction, have largely been neglected in research on technology-based learning and instruction. Against this background, Moreno (2006) proposed a “Cognitive-Affective Theory of Learning with Media” (CATLM), which expands Mayer's CTML by including the concept of “affective mediation”, that is, the idea that motivational factors may mediate learning by affecting the cognitive engagement of the learners. The papers by Magner et al., 2013 and Plass et al., 2013, and D'Mello, Lehman, Pekrun & Graesser (2013), to be discussed in the present commentary, follow this line of reasoning and provide valuable insights in how learning with multimedia is affected by emotional and motivational factors.
The three papers discussed so far follow Moreno's (Moreno, 2006) line of reasoning that learning with multimedia is affected by emotional and motivational factors. They provide valuable insights in how this is achieved with a focus on “situational interest” as a motivational construct (Magner and colleagues) as well as on “positive emotions” (Plass and colleagues) and “confusion” (D'Mello and collegues) as emotional constructs. All three papers applied sophisticated statistical analyses based on experimental approaches, that is, differences in motivation and emotion were induced in learners by varying the design of learning environments. Thus hypotheses about cause and effects could appropriately be investigated which is a major advantage compared to correlational studies. In line with Moreno's CATLM, all authors had mediation hypotheses in mind: Experimentally induced differences in motivation/emotion lead to differences in the cognitive processing of learning materials which in turn leads to differences in learning results. Unfortunately however, presumably due to the unavailability of appropriate measures of cognitive processing, in none of the three papers the crucial link between motivation/emotion and cognitive processing was studied directly. So far, the effects of motivation/emotion on learning can, at least for the present papers, only be interpreted in terms of modified cognitive processing. Future research should try to develop and use independent measures of cognitive processing in order to be able to test the mediation hypothesis.
Motivation and emotion, as studied in the present papers, are state constructs that may occur in a given learning situation. For further research, an interesting question would be to ask whether the occurrence of specific motivations/emotions as states in a given learning situation is perhaps modified by individual differences in overarching (non-cognitive) trait constructs like, for example, neuroticism, openness, conscientiousness, anxiety, frustration tolerance, learned helplessness etc. It could even be the case that the whole mediation process from motivation/emotion over cognitive processing to learning results might be moderated by specific traits. There is a growing body of research on cognitive trait constructs that moderate the effectiveness of multimedia design features for learning like, for example, prior knowledge ( Magner et al., 2013), spatial ability (e.g., Höffler, 2010 and Höffler and Leutner, 2011) or learning preferences (e.g., Plass, Chun, Mayer, & Leutner, 1998). So far however, research on non-cognitive trait constructs that might moderate the effectiveness of multimedia design features, however, seems to be at its very beginning. The present papers provide a valuable step in this direction.
