EFFECTS OF SEGMENTING, SIGNALING, AND WEEDING ON LEARNING FROM EDUCATIONAL VIDEO By MOHAMED IBRAHIM Bachelor of Arts in Archeology Cairo University Cairo, Egypt 1984 Master of Arts in Political Science Oklahoma State University Stillwater, Oklahoma 1997 Submitted to the Faculty of the Graduate College of the Oklahoma State University in partial fulfillment of the requirements for the Degree of DOCTOR OF PHILOSOPHY May, 2011 EFFECTS OF SEGMENTING, SIGNALING, AND WEEDING ON LEARNING FROM EDUCATIONAL VIDEO Dissertation Approved: Dr. Pasha Antonenko Dissertation Adviser Dr. Jesse Mendez Dr. Carmen Greenwood Outside Committee Member Dr.
Payton Dean of the Graduate College ii ACKNOWLEDGMENTS Thank you to my esteemed committee: Dr. Pasha Antonenko, Dr. John Curry, Dr. Jesse Mendez, and Dr.
Your passion for teaching and enthusiasm for interesting research is inspiring. Thank you to my wife for her love and support: I simply could not have done it without you. iv TABLE OF CONTENTS Chapter Page I.1 Statement of the Problem .6 Definition of Terms. REVIEW OF LITERATURE.12 The Benefits of Using Video in Education .12 The Challenges of Using Video in Education.14 Multimedia Design Theories and Principles .28 Summary and Implications for the Design of Educational Video .31 Segmentation and Cognition .31 Signaling and Cognition .32 Weeding and Cognition.43 v Chapter Page IV.47 Multicollinearity, Singularity and Linearity .52 Scope and Limitations.72 vi LIST OF TABLES Table Page 1 Segment titles .40 2 Concepts signaled in the SSW video .41 3 Descriptive statistics for the dependent measures .45 4 Descriptive statistics for the participants’ demographics .46 5 Normality levels for each dependent variable .47 6 Correlations among dependent variables and prior knowledge .48 7 Univariate analyses of the effects of SSW on the dependent variables .50 vii LIST OF FIGURES Figure Page 1 Educational uses of video .13 2 Problems associated with learning from audiovisuals .40 5 Topics that make up the Insect Body Parts segment.42 6 A diagram used for signaling and segment summary .42 viii CHAPTER I INTRODUCTION The use of educational video has increased over the past decade.
In 2009, it became the third most popular genre for learning and reached 38% of adult Internet users (Purcell, 2010). Empirical research on the use of dynamic audiovisual learning materials in education demonstrates that learners not only prefer instructional video over text, but are also more likely to gain deeper conceptual understanding of the content from video than from words alone (Baggett, 1984; Mayer, 2002, 2003; Mayer & Moreno, 2002). In many learning contexts, knowledge acquisition is better achieved through presenting materials in formats optimized to use both the visual and auditory sensory channels at the same time (Mayer, 2001). Content presented in video is also more memorable than text-based instruction (Jonassen, Peck, & Wilson, 1999).
A major assumption underlying this empirical work is that humans can construct a mental representation of the semantic meaning from either auditory or visual information alone, but when instruction is presented in both formats, each source provides complementary information that is relevant to learning (Baggett, 1984). At the same time, other empirical evidence suggests that video, like other dynamic and complex audiovisuals, may be no better than a series of equivalent content static images because dynamic visuals is difficult for students to perceive and understand, and may interfere with successful learning (Catrambone & Seay, 2002; Hegarty, Kriz, & Cate, 2003; Hegarty, Narayanan, & Freitas, 2002; Mayer, 2005; Tversky, Morrison, & Betrancourt, 2002). 1 The perceived difficulty of the learning materials may also be increased, in particular for novice students, because they do not possess adequate knowledge to discriminate relevant from irrelevant information (Bromage & Mayer, 1981; Graesser, 1981) and are often distracted by focusing on non-essential features of presentations at the expense of more important information (Lowe, 1999, 2003). Statement of the Problem Video requires high levels of cognitive processing to synthesize the visual and auditory streams of information and to extract the semantics of the message (Homer, Plass, & Blake, 2008).
This increased processing increases the learner’s cognitive load, especially when students are novices in the knowledge domain and lack appropriate prior knowledge to guide their attention (Moreno, 2004; Sweller, 1999). Therefore, a key problem in using video as an instructional device is how to direct learners’ attention to relevant information and decrease cognitive load, creating conditions for the learners’ cognitive system to meet the processing demands that are needed to organize and integrate knowledge from a stream of visual and auditory information. More specifically, cognitive researchers have identified three major challenges in using audiovisuals in instruction: (1) the transitory nature of the dynamic materials, (2) the difficulty of focusing students’ attention on essential information in the complex and fast stream of visual and verbal information, and (3) the inclusion of extraneous content that competes with the essential information for limited cognitive resources (e., Ayres & Paas, 2007; Lowe, 1999, 2003; Tversky, et al. Theoretical Framework In an attempt to overcome the challenges associated with processing information from multimedia materials, such as video, cognitive scientists have developed a number of theories to explain learning from materials rich in media and have proposed design principles to manage learners’ cognitive load and to enhance knowledge acquisition.
Cognitive Theory of Multimedia 2 Learning (CTML) ( Mayer, 2001) and Cognitive Load Theory (CLT) (Sweller, Van Merrienboer, & Paas, 1998) help explain and predict learning from educational multimedia. Both theories were tested in multimedia instructional environments (Moreno, 2006) and are based on assumptions regarding the relationship between cognition and learning from dual representation information formats. Five of these assumptions are particularly relevant to learning from video. First, the cognitive architecture assumption postulates that the human mind consists of an unlimited, long-term memory (LTM) in which all prior knowledge is stored and a limited working memory (WM) in which new information is processed.
Second, the dual-channel assumption proposes that WM has two channels for visual/pictorial and auditory/verbal processing and that the two channels are structurally and functionally distinct (Clark & Paivio, 1991). Third, the limited capacity assumption states that each channel has limited capacity for information that can be processed at one time (Baddeley, 1986; Baddeley & Logie, 1999). Fourth, the active processing assumption explains that humans actively engage in the cognitive processes to select relevant verbal and non-verbal information from the learning materials, organize the selected information into cognitive structures, and integrate these cognitive structures with the existing knowledge to construct a new (or update an old) mental representation (Mayer, 1996a). Finally, the cognitive load assumption maintains that during learning, humans are typically exposed to three types of cognitive load that compete for the limited resources of WM: (1) intrinsic load is the cognitive processing required to comprehend content, (2) extraneous load is caused by ineffective formats of content presentation, and (3) germane load, which is beneficial to learning, enables learners to engage in deeper cognitive processing of the to-be-learned material (Sweller, et al.
According to CTML and CLT, integrating complex learning material into LTM may burden the limited cognitive resources of the learner. In the case of learning from video, the human cognitive system can process only small portions of the large amounts of visual and auditory stimuli received. Unlike processing printed text, learners in formal educational contexts typically do not have the 3 opportunity to stop the video presentation and reflect on what was learned and identify potential gaps in their knowledge. Thus, information processing in this situation frequently requires longer and more intense periods of cognitive and metacognitive activity.
Regardless of the amount of information presented in each sensory channel, the learner’s WM will accept, process, and send to LTM only a limited number of information units (Attneave, 1954; Jacobson, 1950, 1951). Thus, working memory requires direct prompting to accept, process, and send to the long-term storage only the most crucial information (Clark, Nguyen, & Sweller, 2006). Empirical research informed by CTML and CLT suggested a number of prescriptive principles to help multimedia designers create learning materials that are better aligned with human cognitive architecture. These design principles can be categorized into two groups.
The first group comprises strategies aimed at reducing extraneous cognitive load (i., processing that is not related to the instructional goal) and increasing germane load (i., processing that results in deeper learning). These strategies include adding cues to signal the main ideas (called signaling) and eliminating the unnecessary content from learning materials (called weeding). In signaling, the presentation’s main ideas are summarized and highlighted to aid learners in selecting relevant information and organizing it into coherent mental representations. In weeding, non-essential content is eliminated in order to allow students to engage in processing only the essential content.
The second group of design principles is aimed at managing intrinsic cognitive load (i., essential processing related to the learning goal), such as dividing the presentation into small units, called segmentation. With segmentation, learning material is broken up into several segments of information to help students process one cluster of related information elements before moving to the next one. Purpose Statement Prior research on multimedia learning demonstrates that when applied individually segmentation, signaling, and weeding (SSW) can effectively decrease learners’ self-reported mental 4 effort (e., Mayer, 2001; Mayer, Mathias, & Wetzell, 2002; Moreno & Mayer, 1999; Pollock, Chandler, & Sweller, 2002) and improve knowledge acquisition (Mayer & Chandler, 2001; Mayer & Moreno, 2003; Mayer, Moreno, Boire, & Vagge, 1999). For example, with segmentation, learners are able to process pre-structured information and maintain the cognitive capacity necessary to understand the learning content, which results in improved transfer of knowledge (Mayer & Chandler, 2001).
In signaled multimedia presentations, learners can build a mental outline of the presentation, which improves both the retention and transfer of knowledge (Mautone & Mayer, 2001). Similarly, applying weeding to multimedia learning materials was found to reduce extraneous cognitive load and improve learners’ transfer of knowledge (Mayer, Heiser, & Lonn, 2001). While numerous studies applied the segmentation, signaling, and weeding principles to the design of animations, hypermedia, and educational games (Mautone & Mayer, 2001; Mayer & Chandler, 2001; Mayer, et al., 2001; Moreno & Mayer, 2000), little research has examined the effects of these design principles in the context of educational video. Moreno (2007) analyzed the effect of directing attention to relevant information with signaling and segmentation (SS) in dynamic audiovisuals.
In this study, instructional video and animation were designed using the SS principles and compared to video and animation designed without SS. The findings showed that, while the non- SS group outperformed the SS group on the retention of conceptual information, the SS group performed better on the test of knowledge transfer and reported lower levels of cognitive load. Research Questions The present study builds on prior research in two important ways. First, the study examines how the segmentation, signaling, and weeding design principles in educational video affect students’ cognitive load and learning outcomes as compared to students learning from a non-SSW version of the same video.
The second contribution is to outline a theoretical and empirical basis for the domain of educational video design. Many of the design techniques that are used in educational video today 5 reflect the subjective perceptions of ―what works best‖ acquired through the designer’s personal experience and what is considered best practices in the field, rather than empirical evidence (Najjar, 1996; Wetzel, Radtke, & Stern, 1994). Another challenge in educational video design has been to identify information presentation techniques that facilitate higher-order learning, such as transfer of knowledge and structural knowledge acquisition (Gerjets, Scheiter, & Catrambone, 2004). Enhancing knowledge transfer is particularly important because successful instruction should not focus exclusively on the retention of knowledge but should also encourage creative applications of newly acquired knowledge in novel situations (Sternberg & Mio, 2009).
Specifically, this study was guided by these three research questions: 1. Will the SSW intervention affect the perceived learning difficulty of novice learners in the context of educational video? 2. Will the SSW intervention affect retention of knowledge, knowledge transfer, and structural knowledge acquisition for novice learners in the context of educational video? 3.