The Split-Attention Principle in Multimedia Learning
Total Page:16
File Type:pdf, Size:1020Kb
CHAPTER 8 The Split-Attention Principle in Multimedia Learning Paul Ayres John Sweller University of New South Wales Abstract Definition of Split-Attention The split-attention principle states that Instructional split-attention occurs when when designing instruction, including mul learners are required to split their atten timedia instruction, it is important to tion between and mentally integrate sev avoid formats that require learners to split eral sources of physically or temporally dis their attention between, and mentally in tegrate, multiple sources of information. parate information, where each source of Instead, materials should be formatted so information is essential for understanding that disparate sources of information are the material. Cognitive load is increased by physically and temporally integrated thus the need to mentally integrate the multi obviating the need for learners to en ple sources of information. This increase in gage in mental integration. By eliminat extraneous cognitive load (see chapter 2) is ing the need to mentally integrate multiple likely to have a negative impact on learn sources of information, extraneous work ing compared to conditions where the infor ing memory load is reduced, freeing re mation has been restructured to eliminate sources for learning. This chapter provides the need to split attention. Restructuring the theoretical rationale, based on cog occurs by physically or temporally inte nitive load theory, for the split-attention principle, describes the major experiments grating disparate sources of information to that establish the validity of the princi eliminate the need for mental integration. ple, and indicates the instructional de The split-attention effect occurs when learn sign implications when dealing with multi ers studying integrated information outper media materials. form learners studying the same information 135 THE SPLIT-ATTENTION PRINCIPLE IN MULTIMEDIA LEARNING 136 THE CAMBRIDGE HANDBOOK OF MULTIMEDIA LEARNING A F B have to locate the given information (the pIe sources of information is likely to be a ~------~----------~ parallel lines and the two angles) on the di major source of extraneous cognitive load. agram. If the learner can write this informa The order in which the solution steps are tion on the diagram then split-attention only calculated, are also marked on the diagram occurs once at this stage. However, to fol and indicated by the numbers 1 and 2. As low the two steps to solution, learners have a consequence of this physical integration, to mentally integrate these steps with spe the need for mental integration is reduced cific angles and geometrical configurations and extraneous cognitive load is kept to in the diagram. This requirement to split at a minimum. tention between the diagram and text fol lowed by mental integration is a classic ex ample of split-attention. If the learner is a Basic Research Into the novice, possessing few developed schemas in Split-Attention Effect Task: In the above diagram find the value of angle BFE when this domain, it might be expected that sub stantial cognitive resources will need to be The initial research into the split-attention line AB is parallel to line DC devoted to splitting attention between the effect was conducted by Tarmizi and Sweller disparate sources of information and men (1988) who investigated the effectiveness Angle BCD = 11 0° tally integrating them. of worked examples on learning geometry. To avoid split-attention, researchers have Prior to this study, worked examples (see chapter 1 5) had proven to be highly effec Angle BEF = 50° successfully employed the strategy of phys ically integrating the various sources of in tive for learning algebra (Cooper & Sweller, formation. Figure 8.2 demonstrates how the 1987; Sweller & Cooper, 1985) and in Solution: two parts (diagram and text) of the worked other mathematical domains (Zhu & Simon, example in Figure 8.1 have been integrated. 1987). However, in their initial experi Angle FBE = 180 -110 = 70° (Co-interior angles between parallel lines sum to 180°) First, the given information on angles and ments, Tarrnizi and Sweller found that nei parallel lines are drawn on the diagram. Par ther worked examples nor guided solutions (highly directed but not a full worked exam Angle BFE = 180 - 50 -70 (Angles in a triangle sum to 180) allel lines are represented by the universal symbol of the two arrows, and the two an pIe) enhanced performance compared with gles HOO and 50° are marked. Second, the conventional problem-solving strategies. =60°. two steps to solution are written on the dia The failure of worked examples in ge Figure 8.1. Split-attention in a geometry worked example. gram at the precise location where the values ometry was initially perplexing. However for the angles are calculated thus eliminat Tarmizi and Sweller reasoned that the ing the need for the learner to keep refo format of the worked examples, a dia presented in split-attention format. The using both words and pictures" (Mayer 2001, cusing attention from diagram to text and gram followed by the solution steps (Fig split-attention principle flows from the split p. 1) it can be seen that split-attention will vice versa. Searching for referents in multi- ure 8.1 provides an example), must increase attention effect. It states that when pre frequently occur using multimedia as there senting disparate sources of information that will always be at least two sources of infor must be mentally integrated in order for the mation involved. A F B information to be understood, those sources Figure 8.1 demonstrates an example of of.infc:>rmation should be presented in inte materials that include a requirement to split 1. Angle FB = 180 - 110 = 70° grated format. (Co-jnte or angles between attention in the mathematical domain of ge paralle ines sum to 180°) ometry. In Figure 8.1 the diagram is sepa rated from the solution that explains how Examples of the Split-Attention Effect the task (find Angle BFE) is completed. Neither source of information makes sense The different sources of information that without the other. The diagram provides no cause split-attention vary. For example, the solution information and the solution in sources can be text and text, or text and formation is unintelligible without the dia mathematical equations, or different forms gram. To understand this worked example, of multimedia. Using Mayer's definition of a learner will be forced to integrate many D C multimedia as "the presentation of materials pieces of information. Initially; learners will Figure 8.2. Integrated worked example of a geometry problem. 13 8 THE CAMBRIDGE HANDBOOK OF MULTIMEDIA LEARNING THE SPLIT-ATTENTION PRINCIPLE IN MULTIMEDIA LEARNING cognitive load. In particular, they argued Additional evidence of the split-attention that the source of increased cognitive load effect and how it could be avoided was pro was the requirement for learners to men vided by Ward and Sweller (1990) in the area a) tally integrate the two sources of informa of physics. Using mechanics problems based tion (diagram and textual solutions) that on the formulae associated with constant A car moves from rest to a speed of 40mls in 20 seconds. What is the had split their attention. They further hy acceleration, Ward and Sweller found that pothesised that if diagrams and texts were worked examples compared poorly with a acceleration of the car? integrated, then split-attention would be problem-solving strategy. Following on from u = 0 mis, v = 40 mis, t = 20 s avoided leading to effective worked ex the earlier research of Tarmizi and Sweller, amples. In the final two experiments of Ward and Sweller reasoned that the worked v=u+at this study, Tarmizi and Sweller successfully examples were structured using a format showed that learners who studied integrated that promoted split-attention. Figure 8.p v-u=at worked examples performed at a superior depicts a worked example in dynamics fol level (fewer errors and quicker solution lowing the traditional textbook format. The a=(v-u)/t times) compared to learners who followed a problem statement and the initial given conventional problem-solving strategy dur states are presented first followed by the ap a = (40-0)/20 ing acquisition. propriate formulae and the solution steps. Researchers sought to test these findings Using this structure, the learner has to men in other domains following the identifica tally integrate the problem statement, the tion of the split-attention effect and the initial givens, the formula, and the solu b) success of integrated material in the ge tion steps at various points. In contrast Fig A car moves from rest (u = 0) to a speed (v = 40mls) in 20 seconds (t = 20): ometry domain. Sweller, Chandler, Tierney, ure 8.3 b demonstrates how this informa and Cooper (1990) reproduced the Tarmizi tion can be integrated physically to reduce v = u + at, v - u = at, a = (v - u)/t, a = (40-0)/20, a = 2 mls2• What is the and Sweller findings using coordinate ge split-attention. The key to this integration ometry materials. Again, traditionally struc is to place the algebraic variables (e.g., v) acceleration of the car? tured worked examples in this domain have immediately next to their numeric values used a diagram associated with solution to reduce search for the appropriate refer steps next to or under the diagram. Using ents and to complete the algebraic manipu Figure 8.3. Split-attention and integrated dynamics problem. this format, as indicated previously, learn lation and substitution before the question is ers are forced to search for the location stated to reduce the problem-solving search split-attention effect could be avoided, re and Sweller (1991) was superior to the in of referents in either the diagram or the associated with a problem goal.