UCLA

Multimedia learning refers to learning from a combination of words and images. In the present dissertation, a multimedia lesson is defined as an animated, narrated educational video that depicts a scientific process--a format of instructional material becoming increasingly common in online, hybrid, and traditional classrooms. The overarching goal of the present research was to investigate how to optimize learning from multimedia lessons using two related theories of multimedia learning (the cognitive theory of multimedia learning [CTML], Mayer, 2009; the cognitive-affective theory of learning with multimedia [CATLM], Moreno, 2006) as guiding frameworks. A secondary goal was to explore students' metacognitive beliefs regarding the efficacy of various types of multimedia presentation conditions for learning.

In one body of research, we attempted to strike a balance between the positive and negative effects of including seductive details in a multimedia presentation (i.e., the tendency of tangentially related facts to engage interest but at the cost of impairing learning) by varying the level of redundancy between the narration and on-screen text. While inclusion of identical on-screen text has been shown to impair multimedia learning, we previously found that inclusion of a certain level of non-identical on-screen text could actually increase learning (Yue, Bjork, & Bjork, 2013). We utilized this finding by presenting seductive details with identical on-screen text and key points with non-identical on-screen text, thereby inducing shallow processing of the former information and deeper processing of the latter information. Our results indicate that such a strategy leads to an optimal presentation scheme--that is, one in which seductive details can be presented during the lesson to engage students' interest but without a cost to their learning of key information.

In another body of research, we focused on optimizing processing during restudy, rather than during initial encoding. After watching a multimedia lesson for the first time, participants practiced retrieving either the auditory modality while experiencing the visual modality, or vice versa. Although not leading to greater learning than re-presenting the intact lesson, this type of retrieval practice did result in equivalent learning, indicating that students can engage in such practice without harm to their performance on a later exam when needed (e.g., should they only have access to an instructor's slides but not the accompanying narration from a lecture).

In a third body of research, we explored potential indirect benefits of testing on learning from multimedia lessons. After experiencing a lesson, students either read half the facts from the lesson or took a test on those facts, after which they studied either the same lesson again or an entirely different multimedia lesson. Taking the intervening test, as compared to restudying facts, led students to make more effective use of a second study opportunity (as reflected in superior final test performance) whether the multimedia lesson presented in that second study opportunity was the same or different from the first. That is, the test-taking experience generalized to better study of even totally different material. Interestingly, however, it was the students who studied facts--not those who took a test on those facts--who had the more unrealistic sense of confidence about how well they would do on the final test.

In summary, the present research offers new insights--of both a theoretical and applied nature--for our understanding of multimedia learning. Our findings reveal new methods for inducing deep processing of key information and, importantly, ones that can be easily implemented in the construction and delivery of multimedia instructional materials for a variety of educational domains.