The Expertise Reversal Effect for Sequential Presentation in ...

JOURNAL OF

Journal of Sport & Exercise Psychology, 2013, 35, 260-269 © 2013 Human Kinetics, Inc.

SPORT EXERCISE PSYCHOLOGY

Official Journal of NASPSPA

www.JSEP-Journal.com ORIGINAL RESEARCH

The Expertise Reversal Effect for Sequential Presentation in Dynamic Soccer Visualizations Aïmen Khacharem,1,2 Bachir Zoudji,2,3 Slava Kalyuga,4 and Hubert Ripoll1,2 1Aix-Marseille

University; 2Information and System Sciences Laboratory, UMR-CNRS 6168, France; of Valenciennes and Hainaut Cambrésis; 4University of New South Wales

3University

Cognitive load perspective was used as a theoretical framework to investigate effects of expertise and type of presentation of interacting elements of information in learning from dynamic visualizations. Soccer players (N = 48) were required to complete a recall reconstruction test and to rate their invested mental effort after studying a concurrent or sequential presentation of the elements of play. The results provided evidence for an expertise reversal effect. For novice players, the sequential presentation produced better learning outcomes. In contrast, expert players performed better after studying the concurrent presentation. The findings suggest that the effectiveness of different visual presentation formats depend on levels of learner expertise. Keywords: animation, learning, expertise reversal, cognitive load, soccer, sequential presentation Static diagrams have long been used by team-sport coaches to illustrate the functioning of a playing system and describe the movements of players in the field through the use of schematic elements such as cross and arrow. In recent years, with the growth in graphic technologies, such visualizations have become dynamic, more realistic, and displaced from traditional media like blackboard to the computer. From an intuitive perspective, dynamic visualizations (e.g., animations and videos) are believed to have an enormous potential for improving players’ tactical skills because they can deliver explicitly the information concerning the spatiotemporal changes over time (e.g., motion, acceleration, and trajectory). Thereby, they may compensate the inability of learners to mentally infer these changes from static visualizations. It can also be assumed that recall of information related to team sports could be enhanced when subjects learn in dynamic conditions, which present a set of contextual features (e.g., change-related information) analogous to those experienced in realistic game conditions (the Aïmen Khacharem is with the Faculty of Sport Sciences, AixMarseille University, France, and with the Information and System Sciences Laboratory, UMR-CNRS 6168, France. Bachir Zoudji is with the Information and System Sciences Laboratory, UMR-CNRS 6168, France, and with the Faculty of Sport Sciences and Management, University of Valenciennes and Hainaut Cambrésis, France. Slava Kalyuga is with the School of Education, University of New South Wales, Sydney, NSW, Australia. Hubert Ripoll is with the Faculty of Sport Sciences, Aix-Marseille University, France, and with the Information and System Sciences Laboratory, UMR-CNRS 6168, France.

260

encoding specificity principle; e.g., Tulving & Thomson, 1973). According to the congruency principle (Tversky, Morrison, & Bétrancourt, 2002), dynamic visualizations are considered the most suited and natural instructional means when they are used to portray a content characterized by a high level of dynamism because they can convey this dynamism in a direct and explicit way (Scheiter & Gerjets, 2010). Despite the presumed advantages of dynamic visualizations, some studies have shown that the temporal constraints of animations may impose a high cognitive load on the learners’ working memory (WM) and actually decrease performance when compared with learning from static pictures (e.g., Hegarty, Kriz, & Cate, 2003; Mayer, Hegarty, Mayer, & Campbell, 2005; Tversky et al., 2002). A number of researchers have suggested that certain design techniques may be employed to overcome these difficulties. These include, for example, altering the animations’ presentation speed (e.g., De Koning, Tabbers, Rikers, & Paas, 2011; Fischer, Lowe, & Schwan, 2008; Fischer & Schwan, 2010), using attentional cueing (e.g., De Koning, Tabbers, Rikers, & Paas, 2007), and segmentation (e.g., Spanjers, Wouters, Van Gog, & Van Merriënboer, 2011; Spanjers, Van Gog, Wouters, & Van Merriënboer, 2012). Since such techniques are based on theory and research in cognitive psychology, they have been referred to as cognitive principle design (Hegarty & Kriz, 2007). In the current study, we suggest a novel instructional technique called the sequential presentation method, which involves displaying parts of the dynamic visualization in a successive way rather than in a simultaneous way. Even though this type of presentation has been frequently

The Expertise Reversal Effect for Sequential Presentation  261

used with static visualizations such as statistical formulas (e.g., Kester, Kirschner, & Van Merriënboer, 2004) and diagrams (e.g., Bétrancourt, Bisseret, & Faure, 2001; Bétrancourt, Dillenbourg, & Montarnal, 2003; Jamet & Arguel, 2008; Jamet, Gavota, & Quaireau, 2008), there have hardly been any investigations with dynamic visualizations. Furthermore, reviews and empirical studies of the expertise reversal effect (e.g., Kalyuga, 2007; Kalyuga, Ayres, Chandler, & Sweller, 2003; Lee & Kalyuga, 2011) have demonstrated that instructional design methods that have a positive effect for experts might have no, or even a negative effect for novices. This study investigated whether an expertise reversal effect would occur for the sequential presentation method used with animated soccer visualizations.

Learning From Dynamic Visualizations and Cognitive Load To derive meaning from a complex dynamic system such as a soccer-playing system, players must develop several types of knowledge (Hegarty & Just, 1993; Narayanan & Hegarty, 1998). Firstly, it involves knowledge of the components of the system (i.e., teammates, opponents, and ball) and how they are placed and organized in the field, that is, the configuration of the system. Secondly, it includes knowledge of how the players move and affect each other’s motions, that is, the behavior of the system. Finally, it involves understanding of the function of the playing system and what it is designed to do. Effective integration of these types of knowledge allows players to construct an internal representation called mental model that depicts precisely the playing system (Kriz & Hegarty, 2007). North and colleagues (e.g., North, Ward, Ericsson, & Williams, 2011; North, Williams, Hodges, Ward, & Ericsson, 2009; Williams, Hodges, North, & Barton, 2006) extended this knowledge and showed, using point light, that the effective perception of dynamic soccer scenes requires players to process the structural, relational information between display features (e.g., movements and relative positions of the players) rather than identifying superficial, low-level surface features (e.g., background information, isolated distinctive actions). However, research within cognitive load theory (Sweller, Van Merriënboer, & Paas, 1998; Paas, Renkl, & Sweller, 2003, 2004; Sweller, Ayres, & Kalyuga, 2011) indicates that animations are often too complex to be easily perceived and processed in WM (cf. the apprehension principle in Tversky et al., 2002). According to this theory, capacity and duration limitations of WM as a major component of human cognitive architecture may critically influence the effectiveness of learning new information from various instructional procedures and presentation formats. In particular, animations may impose the following two types of cognitive load that can seriously impede learning. The extraneous cognitive load is a type of load that is not actually required for achieving learning goals but is imposed due to the design of information presentations.

With animations, it could be caused by the fleeting nature of such information presentations (transient information effect; e.g., Sweller et al., 2011) imposing significant WM demands. When dealing with animated information, learners are required to temporally hold in WM the current information (A) while attending to the succeeding information (B). If information (B) is processed before the information (A) is committed to long-term memory (LTM), information (B) may interfere with information (A). This phenomenon, called interference, or retroactive inhibition, could overwhelm WM (Mayer, DeLeeuw, & Ayres, 2007). For example, it has been known that storage of dynamic soccer information (e.g., relative movement of the players) in WM suffers as a function of retroactive interference (Khacharem, Zoudji, & Ripoll, 2013; North & Williams, 2008) Furthermore, the extraneous cognitive load might also be due to the high perceptual (attentional) demands, since the learners’ visual attention could be spatially distributed over the different elements moving at once (split-attention effect, De Koning, Tabbers, Rikers, & Paas, 2009). As a result, some of the presented elements will receive learners’ attention whereas others could be overlooked. Lowe (1999) demonstrated that learners, especially novices, found it difficult to detect and integrate the relevant information conveyed by dynamic visualizations (weather maps), because their visual attention was often diverted by salient features, such as fast movements, that were not necessarily relevant for understanding the content. The intrinsic cognitive load is determined by interactivity of the elements of information that are essential for learning. An animation that contains a large number of interacting elements is considered as more difficult to learn from than animation with a smaller number of interacting elements. Low interactivity animations mostly consist of single elements that can be learned in isolation, whereas in high interactivity animations, individual elements can only be well understood in relation to other elements (Sweller, 1994; Sweller et al., 1998). Intrinsic load is also influenced by learners’ prior knowledge because, as a result of their experience within a domain, learners can develop extensive cognitive schemas in LTM that enable them to encapsulate several elements into larger chunks or templates that are treated as single units in WM. Thus, the same animation might generate less intrinsic load for learners with higher levels of prior knowledge than for learners with lower levels of prior knowledge (Khacharem, Spanjers, Zoudji, Kalyuga, & Ripoll, 2013; Spanjers et al., 2011).

Sequential Presentation Method and Levels of Learner Expertise In some circumstances, the level of element interactivity of an instructional presentation may be too high to be accurately understood by the learners. In this case, it may be helpful to first present the material as isolated elements that could be processed serially rather than concurrently.

262  Khacharem et al.

By presenting the elements of complex information sequentially, the learners could perceive lower intrinsic cognitive load because the element interactivity of the material would be artificially reduced (Van Merriënboer & Ayres, 2005). The added value of the sequential presentation is widely attested in learning from static pictures/diagrams presented alone (e.g., Bétrancourt & Tversky, in press; Bétrancourt et al., 2001) or associated with either written text (e.g., Bétrancourt et al., 2003; Wright, Hull, & Black, 1990) or spoken commentary (Jamet & Arguel, 2008). For example, using a geographic map as learning material, Bétrancourt and Tversky (in press; Experiment 3) showed that a meaningful sequential presentation of the map’s components allowed faster processing and more accurate recall performance than the static display of the map. Bétrancourt et al. (2003) found that learners, who studied a sequential presentation of financial concepts, achieved higher performance scores than learners who saw the same content in the simultaneous static presentation. According to these studies, the sequential presentation format may be beneficial as it reduces the cognitive-perceptual demands by providing learners with less information to be processed in WM concurrently and facilitating the integration of information in LTM. In addition, displaying sequentially the components of a system in a defined order can be seen as a form of temporal cueing, thus facilitating the building of wellstructured and organized knowledge in LTM (Jamet, Gavota, & Quaireau, 2008). In other words, participants who during the learning phase studied visualizations sequentially would in the test phase retrieve corresponding items in the order in which they appeared (Taylor & Tversky, 1992). Bétrancourt et al. (2001) used a map of a library to compare the effects of sequential and simultaneous static presentation formats. The findings indicated that, although the sequential presentation did not affect learning outcomes, the order of recalling of the map was consistent with the type of sequential presentation used. However, a number of research studies have shown that the effectiveness of instructional materials is a function of the expertise of the learner. There have been interactions observed between the effectiveness of different instructional formats and levels of learner expertise. The information presentation that is effective for a novice could be ineffective for an expert and vice versa. This interaction is referred to as the expertise reversal effect (e.g., Kalyuga et al., 2003; Khacharem, Zoudji, Kalyuga, & Ripoll, 2013). For example, Blayney, Kalyuga, and Sweller (2010) demonstrated that the effect of sequential presentation in accounting training was influenced by levels of learner expertise. Learners with different levels of expertise viewed either isolated elements presented in a sequential form (e.g., using several less complex intermediate working formulas in spreadsheet-based calculations) or in a concurrent form (using a single complex formula). The results indicated that for novice learners, the sequential presentation was superior to the

concurrent presentation. But for more experienced learners, this method did not improve learning because these learners had to cross-reference the sequence of presented simplified elements with their available knowledge of more complex structures, thus using up additional cognitive resources. The sequential presentation technique has only been investigated with static information. It is yet unknown how this technique would affect learning from complex dynamic presentations involving many interacting elements. This study aims to address this issue by assessing the impact of the sequential presentation technique in learning from animated soccer scenes, and to investigate whether the effectiveness of this instructional design would depend on levels of player expertise. Two forms of presentations were tested: concurrent and sequential. The soccer game activity was chosen to investigate these issues given its dynamic nature and the complex interaction occurring simultaneously between numerous elements (North et al., 2011). On the basis of the expertise reversal effect, it could be predicted that expert players would benefit more from the concurrent presentation (i.e., they would obtain better recall test scores, need a lower number of repetitions, and invest less mental effort when learning) as their superior knowledge base and domain-specific memory skills acquired through deliberate practice would allow them to deal with the constraints of the concurrent animation. These players were expected to process easily the interacting elements moving at once (i.e., concurrently), since they were already matched to an internal semantic template or schema held in LTM. In contrast, novices would benefit more from the sequential presentation (i.e., perform better, need a lower number of repetitions, and invest less mental effort) because many interacting elements occurring at once may exceed their WM capacity.

Method Participants A total of 48 male players served as participants in the experiment. The 24 experts (Mage = 24.2, SD = 1.8) were all semiprofessional soccer players. They had been playing soccer for an average of 13.2 years (SD = 2.2) and trained or played for an average of 11.0 hr (SD = 3.5) per week. The 24 novice players (Mage = 23.5, SD = 2.7) knew the rules of the game, but they had never played soccer or any other team ball sport in a club. This requirement was applied to avoid the effects of transfer across sports (e.g., Smeeton, Ward, & Williams, 2004). None of the participants had ever taken part in the present type of experiment. Participants who used to play video games related to team ball sports (e.g., Fifa, Pro-Evolution Soccer, or any equivalent games) were excluded. The participants volunteered to participate and provided informed consent. Approval for this project was granted by the local ethics committee.

The Expertise Reversal Effect for Sequential Presentation  263

Apparatus The experiment was conducted using a HP Pavilion dv6 computer. The stimuli were presented on a 250 cm × 200 cm screen from a video projection system (Sony VPL EX120 XGA) placed at a distance of 3 m. The display on the screen was 200 cm × 160 cm, with a 45° viewing angle.

Materials Two different animated soccer sequences were designed using SimulFoot three-dimensional software. (SimulFoot was developed by researchers from the SimGraph team of the Information Science and Systems Laboratory (LSIS) at Aix-Marseille University.) Each sequence showed the development of a clearly structured counterattack. The structured sequences contained coherent, organized counterattack scenes that put players in conditions likely to be experienced during competition matches. To guarantee that the presented information was a realistic depiction of a soccer playing system and was matched to specific offensive tactical components, the sequences were developed in collaboration with three experienced soccer coaches. Firstly, together with an expert coach, we created two coherent counterattack sequences that involved six players who carried out a rapid tactical combination of play composed of five passes toward the opponent’s court before a shot at goal was taken. During each pass, each player should move in relation to the ball and the teammates’ positions to offer an appropriate solution to the ball carrier. Each pass corresponded to a new stage

Figure 1 — Schematic diagram of the concurrent presentation.

made up of multiple offensive actions carried out by the players (e.g., moving, position switching, overlapping). The two sequences of 31.5 s and 32.3 s in length were captured from a sideline camera in an elevated position (25°), such that the entire field and all players were visible. This type of presentation allows a concurrent display of the components of the situation (see Figure 1). Secondly, we generated a sequential presentation from each animation (i.e., concurrent presentation). In this version, each stage of play was divided into a set of meaningful chunks composed of two, three, or four players that transited successively and separately from one stage to another (see Figure 2). Finally, another two experienced soccer coaches independently rated on a 9-point Likert scale each sequence in relation to (a) whether the pattern of play corresponded to a structured counterattack soccer scene (0 = very unstructured; 5 = very structured) and (b) whether the chunks presented in each sequence were representative of a meaningful (typical) grouping of players (0 = very nonrepresentative; 5 = very representative). All sequences were rated 4 or above and thus were kept for the experimentation. Note that all coaches (including the first author) were licensed by the French Football Federation and had satisfactory experience at a semiprofessional level (M = 8.42 years, SD = 1.74).

Procedure The experimental sessions were conducted individually and involved a study phase and a test phase. The experimental sessions lasted no more than 45 min. Before being exposed to the experimental materials, participants were

264  Khacharem et al.

Figure 2 — Schematic diagram of the sequential presentation.

given a printed outline of the different phases of the study, in which they were informed about the general procedure. The presentation order of the two scenes (i.e., concurrent and sequential) was counterbalanced within subjects. Immediately after having watched either a concurrent or sequential presentation (study phase) players were asked to rate the mental effort invested in studying the scene of play and perform a recall-reconstruction test (test phase). Study Phase.  In this phase, the participant was asked to memorize the soccer scenes as accurately and as rapidly as possible. For each presentation, the projection of the scene began as soon as the learner clicked on the “Ready!” button. Participants could not stop the display, but when it was over, they could see it again as many times as desired by clicking on the “Repeat” button (the number of repetitions was registered by the system). When the subjects felt they had learned the scene of play, they had to click on the “End” button. Test Phase.  The test phase occurred immediately

following the study phase of each presentation type. The participant was asked to perform two successive tasks, as follows. (i) Evaluate the mental effort invested in studying the learning material (“Please indicate how much mental effort did you invest in studying the animation?”) on a 9-point subjective rating scale, ranging from (1) very, very low mental effort to (9) very, very high mental effort (Paas, 1992; Paas, Tuovinen, Tabbers, & Van Gerven, 2003). This self-report measure is widely accepted as a valid and noninvasive method of estimating cognitive load (e.g., Paas, 1992; Paas, Tuovinen, Tabbers, & Van Gerven, 2003; Paas & Van Merriënboer, 1993, 1994; Paas, Van Merriënboer, & Adam, 1994). (ii) Reconstruct the scene

of play by drawing it on a sheet of paper that contained an empty soccer field, with the same viewing angle as in the experimental conditions (see Figure 3). The soccer field was divided into five parts. For each part (stage of play), participants were instructed to reconstruct as accurately as possible the position of the six players (with different numbers) in relation to the ball which had been already placed in its correct position. In each part, they had to reproduce (a) the position of the ball carrier, (b) players located in front of the ball (on the left, in the middle, or on the right), and (c) players located behind the ball (on the left, in the middle, or on the right). One point was awarded for each correct placement of a player in the field; no point was awarded for each incorrect placement of a player in the field. The scores could, therefore, range from 0 to 30. An individual naive regarding the hypothesis of the project assured this processes of coding.

Data Analyses A mixed-design ANOVA was used to analyze the data. The between-participants factor was Expertise (expert vs. novice) and the within-participants factor was Condition (concurrent vs. sequential). For all analyses throughout this study, we used p < .05 as the criterion for significance, partial eta squared (ηp2) values are provided as a measure of effect size for all main effects and interactions, and Cohen’s mean standardized differences (d) for post hoc comparisons. Partial eta squared values of .01, .06, and .14 and Cohen’s d values of .20, .50, and .80 represent small, moderate, and large effect sizes, respectively. The dependent variables were recall performance (0–30), number of repetitions needed to memorize the

The Expertise Reversal Effect for Sequential Presentation  265

Figure 3 — Schematic presentation of the empty soccer play.

scene (unlimited), mental effort invested in studying the scene (1–9), and instructional efficiency (score). Instructional efficiency was calculated using Paas and Van Merriënboer’s (1993) computational approach. Efficiency = ( Z Recall – Z Mental Effort ) / 2 This measure combines standardized z-scores for the test performance scores with standardized z-scores for the ratings of mental effort associated with studying the materials. According to this definition of relative instructional efficiency, a lower rating of effort combined with higher performance scores would provide evidence of a

more efficient learning condition. The denominator √2 is used to make the graphical interpretation of the formula more straightforward by representing the mental effort and performance z-scores in a cross of axes. The relative efficiency at any point on the diagram can be measured as the distance from this point to the line of zero efficiency (Z Recall = Z Mental Effort).

Results Descriptive statistics for recall accuracy, mental effort, number of repetitions, and learning efficiency for novice and expert participants are presented in Table 1.

Table 1  Means Scores (SD) Concerning Recall Accuracy, Mental Effort, Number of Repetitions, and Learning Efficiency for Each Experimental Condition and Level of Expertise Concurrent

Sequential

Recall Accuracy

26.79 (1.02)

26.45 (1.31)

Mental Effort

3.50 (0.93)

4.37 (0.57)

Number of Repetitions

4.25 (0.67)

4.70 (1.08)

Learning Efficiency

1.32 (0.55)

0.73 (0.55)

Recall Accuracy

21.79 (0.93)

23.29 (1.82)

Mental Effort

6.04 (1.00)

4.95 (0.69)

Number of Repetitions

6.58 (0.97)

6.29 (0.85)

Learning Efficiency

–1.55 (0.55)

–0.50 (0.60)

Expert

Novice

266  Khacharem et al.

Recall Accuracy ANOVA revealed a significant main effect of expertise in recall performance, F(1,46) = 226.79, p < .001, ηp2 = .83. Expert participants (M = 26.62, SD = 0.24) were more accurate than novice counterparts (M = 22.54, SD = 1.06). There was also a significant main effect of condition, F(1,46) = 4.72, p = .035, ηp2 = .09, indicating that participants performed better in the sequential presentation condition (M = 24.87, SD = 2.23) than in the concurrent presentations condition (M = 24.29, SD = 3.53). There was a signification Expertise × Condition interaction, F(1,46) = 11.64, p = .001, ηp2 = .20. The analyses of simple effects revealed a significant difference for the novices, p = .001, d = 1.04, and a nonsignificant difference for the experts, ns, d = 0.28. Thus, whereas the novice participants performed better in the sequential than in the concurrent condition, the expert participants performed at the same level regardless of the condition.

Mental Effort There was a significant main effect of expertise, F(1,46) = 91.95, p < .001, ηp2 = .66, demonstrating that the novice participants (M = 5.49, SD = 0.77) had to invest more mental effort than the expert participants (M = 3.93, SD = 0.61). Although there was no main effect of condition, F(1,46) = 37.17, p = .54, ηp2 = .008, there was a significant expertise × condition interaction, F(1,46) = 32.84, p < .001, ηp2 = .41. The analyses of simple effects revealed significant differences for both the novices (p < .001, d = 1.27) and experts (p = .004, d = 0.16). Whereas the novices invested more mental effort in the concurrent condition than in the sequential condition, the experts invested more mental effort in the sequential condition than in the concurrent condition.

Number of Repetitions ANOVA indicated a significant main effect of expertise, F(1,46) = 106.40, p < .001, ηp2 = .69. The novice participants (M = 6.43, SD = 0.20) needed a significantly larger number of repetitions than the expert counterparts (M = 4.47, SD = 0.31). There was no significant effect of condition, F(1,46) = .21, p = .320, ηp2 = .004; however, there was a significant interaction between expertise and condition, F(1,46) = 4.25, p = .04, ηp2 = .008. The analyses of simple effects following the significant interaction found no significant differences in the number of repetitions between the two conditions either for novices (ns, d = –0.31) or for experts (ns, d = 0.32).

Learning Efficiency The data for this analysis are illustrated in Figure 4. The results showed a significant main effect of expertise, F(1,46) = 363.76, p < .001, ηp2 = .88, indicating that learning was more efficient for the expert participants (M = 1.02, SD = 0.41) than for the novice counterparts (M = –1.02, SD = 0.74). Although there was no significant main

Figure 4 — Learning Efficiency per experimental group, based on Mental Effort and Recall performance. The diagram is a representation after Paas and Van Merriënboer (1993), where Ex = Expert, No = Novice, Con = Concurrent, and Seq = Sequential.

effect for condition, F(1,46) = 3.31, p = .07, ηp2 = .006, there was a significant interaction between expertise and condition, F(1,46) = 44.24, p < .001, ηp2 = .49. Following the significant interaction, the analyses of simple effects revealed a significant difference between the two conditions for both the novices (p < .001, d = 1.82) and experts (p < .01, d = 1.07). Whereas for the novices, the sequential presentation was more efficient than the concurrent presentation, for the experts, the concurrent presentation was more efficient than the sequential presentation.

Discussion The main goal of this study was to investigate the effects of concurrent versus sequential presentation in learning from animated soccer scenes depending on players’ level of expertise. The results supported the predicted interaction between the presentation formats and levels of player expertise (i.e., expertise reversal effect), with the best format of instructional animation being dependent on the levels of learner expertise. The sequential presentation was more beneficial for novice players (i.e., they attained higher recall performance, invested less mental load, and needed the same number of repetitions). This result is consistent with the previous research using static visualizations that indicated advantages of sequential display over complete display (e.g., Jamet & Arguel, 2008). From the perspective of cognitive load theory, by displaying sequentially the elements of the play, the high intrinsic cognitive load linked to the complexity of the animation was avoided since

The Expertise Reversal Effect for Sequential Presentation  267

the visual elements were gradually (serially) processed and integrated, resulting in better learning outcomes. In addition, the sequential presentation highlighted different chunks, as well as their display order on the screen, which could facilitate their structuration, categorization, and organization in LTM (Jamet et al., 2008). Finally, we can assume that the animated elements that formed a chunk might stand out against the rest of the display due to their movements (i.e., transition from one stage to another) and thus cause a dynamic contrast (Schnotz & Lowe, 2008) by automatically attracting players’ visual attention (cf. Hillstrom & Chai, 2006; Hillstrom & Yantis, 1994). This could make them more aware of particular changes that occur at a macro level in the animation. However, which of these explications is more plausible or whether it is their combination that makes sequential presentation more effective for novice players is an open question for future research to address. On the other hand, expert players benefited primarily from the concurrent presentation (i.e., they attained the same recall performance and needed the same number of repetition, but invested less mental effort) because their available knowledge structures allowed them to manage many concurrent interacting elements of information. However, in the sequential presentation, they would need to cross-reference and reconcile their knowledge structures with instructional details that were for them redundant and superfluous. Additional extraneous (unnecessary) cognitive load may be required for such processing, thus increasing WM demands and reducing relative learning outcomes. These findings extend previous research demonstrating that the expertise reversal effect for the sequential presentation method could not only occur when using static presentations (e.g., Blayney et al., 2010), but also when using animations. The expertise reversal effect demonstrated in the current study fits well with some studies conducted within the frameworks established by the theories of skill acquisition and automaticity (e.g., Anderson, 1987; Fitts & Posner, 1967). According to these studies, training in conditions that require explicit cognitive control and monitoring of the execution of complex sensorimotor skills (e.g., golf putting, soccer dribbling, and baseball batting) could enhance performance of novices but hinder performance of expert counterparts (Beilock, Bertenthal, McCoy, & Carr, 2004; Gray, 2004). In the initial learning phase (cognitive phase), learning and performance of novices are based mostly on declarative knowledge, and the skill execution is supported by a set of unintegrated control structures that are held in WM and attended in a step-by-step manner (Fitts & Posner, 1967). However, with deliberate and repeated practice, experts develop more proceduralized and automated knowledge that operates outside of WM (autonomous phase). As a result, attention to execution at the step-by-step level is thought to disrupt the automated functioning of such skills (Beilock & Carr, 2001). Some questions remain unanswered and require further research. This study focused on learning from

animated soccer scenes presented alone without verbal commentaries that could potentially direct learner’s attention to crucial elements and/or to describe the individual and collective actions in the field. It is unclear whether the current results would be replicated with animations that are accompanied by text or verbal instructions. In addition, this study provided evidence of an expertise reversal effect for the sequential presentation method in learning from soccer visualizations. Naturally, one wonders how well our results generalize to learning materials that are both more and less intrinsically complex, and transfer to other team sports activities or unrelated areas. Finally, in this study, the learner performance was measured using a recall reconstruction test. It would be interesting to investigate whether the sequential presentation method would also improve the learning of tactical skills measured by players’ problem-solving abilities in dealing with novel game situations. Since such skills are likely related to the integrated, concurrent vision of the game situations, using an ordered set of combined sequential and concurrent presentation formats could possibly be the best instructional approach to be used with novice learners (similar to isolated-interacting elements effect in cognitive load theory; Sweller et al., 2011). On a practical level, these results suggest caution in the use of instructional animations that depict the functioning of a playing system. For novice players, it is more judicious to use animations that sequentially display the components of the play. However, with the increase of expertise in the domain, concurrent animations are recommended to further enhance the constructed schemas. In conclusion, these results provide strong evidence that processing and integrating a visual representation depends not only on characteristics of the visual display, but also on characteristics of learners, especially on their level of expertise. Accordingly, tailoring visual displays to players with differing levels of expertise should be a crucial part in the work of both soccer coaches and multimedia designers.

References Anderson, J.R. (1987). Skill acquisition: Compilation of weakmethod problem solutions. Psychological Review, 94, 192–210. doi:10.1037/0033-295X.94.2.192 Beilock, S.L., & Carr, T.H. (2001). On the fragility of skilled performance: What governs choking under pressure? Journal of Experimental Psychology. General, 130, 701–725. PubMed doi:10.1037/0096-3445.130.4.701 Beilock, S.L., Bertenthal, B.I., McCoy, A.M., & Carr, T.H. (2004). Haste does not always make waste: Expertise, direction of attention, and speed versus accuracy in performing sensorimotor skills. Psychonomic Bulletin & Review, 11, 373–379. PubMed doi:10.3758/BF03196585 Bétrancourt, M., & Tversky, B. (in press).Simple animations for organizing diagrams. International Journal of HumanComputer Studies. Bétrancourt, M., Bisseret, A., & Faure, A. (2001). Sequential display of pictures and its effect on mental representa-

268  Khacharem et al.

tions. In J-F. Rouet, J.J. Levonen, & A. Biardeau (Eds.), Multimedia learning: Cognitive and instructional issues (pp. 112–118). Amsterdam: Elsevier Science. Bétrancourt, M., Dillenbourg, P., & Montarnal, C. (2003). Computer technologies in powerful learning environments: the case of using animated and interactive graphics for teaching financial concepts. In E. De Corte, L. Verschaffel, N. Entwistle, & J. Van Merrienböer (Eds.), Unrevealing basic components and dimensions of powerful learning environments (pp. 143–157). Oxford, UK: Elsevier Science. Blayney, P., Kalyuga, S., & Sweller, J. (2010). Interactions between the isolated-interactive elements effect and levels of learner expertise: experimental evidence from an accountancy class. Instructional Science, 38, 277–287. doi:10.1007/s11251-009-9105-x De Koning, B.B., Tabbers, H.K., Rikers, R.M.J.P., & Paas, F. (2011). Attention cueing in an instructional animation: The role of presentation speed. Computers in Human Behavior, 27, 41–45. doi:10.1016/j.chb.2010.05.010 De Koning, B.B., Tabbers, H.K., Rikers, R.M.J.P., & Paas, F. (2007). Attention cueing as a means to enhance learning from an animation. Applied Cognitive Psychology, 21, 731–746. doi:10.1002/acp.1346 De Koning, B.B., Tabbers, H.K., Rikers, R.M.J.P., & Paas, F. (2009). Towards a framework for attention cueing in instructional animations: Guidelines for research and design. Educational Psychology Review, 21, 113–140. doi:10.1007/s10648-009-9098-7 Fischer, S., & Schwan, S. (2010). Comprehending Animations. Effects of Spatial Cueing versus Temporal Scaling. Learning and Instruction, 20, 465–475. doi:10.1016/j. learninstruc.2009.05.005 Fischer, S., Lowe, R.K., & Schwan, S. (2008). Effects of Presentation Speed of a Dynamic Visualization on the Understanding of a Mechanical System. Applied Cognitive Psychology, 22, 1126–1141. doi:10.1002/acp.1426 Fitts, P.M., & Posner, M.I. (1967). Human performance. Belmont: Brooks/Cole. Gray, R. (2004). Attending to the execution of a complex sensorimotor skill: Expertise differences, choking, and slumps. Journal of Experimental Psychology. Applied, 10, 42–54. PubMed doi:10.1037/1076-898X.10.1.42 Hegarty, M., & Kriz, S. (2007). Effects of knowledge and spatial ability on learning from animation. In R. Lowe & W. Schnotz (Eds.), Learning from Animation (pp. 3–29). Cambridge University Press. Hegarty, M., & Just, M.A. (1993). Constructing mental models of machines from text an diagrams. Journal of Memory and Language, 32, 717–742. doi:10.1006/jmla.1993.1036 Hegarty, M., Kriz, S., & Cate, C. (2003). The roles of mental animations and external animations in understanding mechanical systems. Cognition and Instruction, 21, 209–249. doi:10.1207/s1532690xci2104_1 Hillstrom, A.P., & Chai, Y-C. (2006). Factors that guide or disrupt attentive visual processing. Computers in Human Behavior, 22, 648–656. doi:10.1016/j.chb.2005.12.003 Hillstrom, A.P., & Yantis, S. (1994). Visual motion and attentional capture. Perception & Psychophysics, 55, 399–411. PubMed doi:10.3758/BF03205298

Jamet, E., & Arguel, A. (2008). La compréhension d’un document technique multimédia peut-elle être améliorée par une présentation séquentielle de son contenu? Le Travail Humain, 71, 253–270. doi:10.3917/th.713.0253 Jamet, E., Gavota, M., & Quaireau, C. (2008). Attention guiding in multimedia learning. Learning and Instruction, 18, 135–145. doi:10.1016/j.learninstruc.2007.01.011 Kalyuga, S. (2007). Expertise reversal effect and its implications for learner-tailored instruction. Educational Psychology Review, 19, 509–539. doi:10.1007/s10648-007-9054-3 Kalyuga, S., Ayres, P., Chandler, P., & Sweller, J. (2003). The expertise reversal effect. Educational Psychologist, 38, 23–31. doi:10.1207/S15326985EP3801_4 Kester, L., Kirschner, P.A., & Van Merriënboer, J.J.G. (2004). Timing of information presentation in learning statistics. Instructional Science, 32, 233–252. doi:10.1023/ B:TRUC.0000024191.27560.e3 Khacharem, A., Spanjers, I.A.E., Zoudji, B., Kalyuga, S., & Ripoll, H. (2013). Using segmentation to support the learning from animated soccer scenes: An effect of prior knowledge, Psychology of Sport and Exercise, 14, 154–160. doi:10.1016/j.psychsport.2012.10.006 Khacharem, A. Zoudji, B., & Ripoll, H. (2013). Effect of presentation format and expertise on attacking-drill memorization in soccer. Journal of Applied Sport Psychology, doi:10.10 80/10413200.2012.718312. Khacharem, A., Zoudji, B., Kalyuga, S., & Ripoll, H. (2013). Developing tactical skills through the use of static and dynamic soccer visualizations: An expert–nonexpert differences investigation. Journal of Applied Sport Psychology. doi:10.1080/10413200.2012.735342. Kriz, S., & Hegarty, M. (2007). Top-down and bottom-up influences on learning from animations. International Journal of Human-Computer Studies, 65, 911–930. doi:10.1016/j. ijhcs.2007.06.005 Lee, C.H., & Kalyuga, S. (2011). Effectiveness of on-screen pinyin in learning Chinese: An expertise reversal for multimedia redundancy effect. Computers in Human Behavior, 27, 11–15. doi:10.1016/j.chb.2010.05.005 Lowe, R.K. (1999). Extracting information from an animation during complex visual learning. European Journal of Psychology of Education, 14, 225–244. doi:10.1007/ BF03172967 Mayer, R.E., DeLeeuw, K.E., & Ayres, P. (2007). Creating retroactive and proactive Interference in Multimedia Learning. Applied Cognitive Psychology, 21, 795–809. doi:10.1002/acp.1350 Mayer, R.E., Hegarty, M., Mayer, S.Y., & Campbell, J. (2005). When Static Media Promote Active Learning: Annotated Illustrations Versus Narrated Animations in Multimedia Instruction. Journal of Experimental Psychology. Applied, 11, 256–265. PubMed doi:10.1037/1076-898X.11.4.256 Narayanan, N.H., & Hegarty, M. (1998). On designing comprehensible hypermedia manuals. International Journal of HumanComputer Studies, 48, 267–301. doi:10.1006/ijhc.1997.0169 North, J.S., & Williams, A.M. (2008). The critical time period for information extraction when identifying patterns of play in soccer. Research Quarterly for Exercise and Sport, 79(2), 268–273. PubMed

The Expertise Reversal Effect for Sequential Presentation  269

North, J.S., Williams, A.M., Hodges, N.J., Ward, P., & Ericsson, K.A. (2009). Perceiving patterns in dynamic action sequences: Investigating the processes underpinning stimulus recognition and anticipation skill. Applied Cognitive Psychology, 23, 878–894. doi:10.1002/acp.1581 North, J., Ward, P., Ericsson, K.A., & Williams, A.M. (2011). Mechanisms underlying skilled anticipation and recognition in a dynamic and temporally constrained domain. Memory (Hove, England), 19, 155–168. PubMed doi:10. 1080/09658211.2010.541466 Paas, F. (1992). Training strategies for attaining transfer of problem-solving skill in statistics: A cognitive-load approach. Journal of Educational Psychology, 84, 429– 434. doi:10.1037/0022-0663.84.4.429 Paas, F., & Van Merriënboer, J.J.G. (1993). The efficiency of instructional conditions: An approach to combine mentaleffort and performance measures. Human Factors, 35, 737–743 doi:10.1177/001872089303500412. Paas, F., & Van Merriënboer, J.J.G. (1994). Variability of worked examples and transfer of geometrical problemsolving skills: A cognitive-load approach. Journal of Educational Psychology, 86, 122–133. doi:10.1037/00220663.86.1.122 Paas, F., Renkl, A., & Sweller, J. (2003). Cognitive load theory and instructional design: Recent developments. Educational Psychologist, 38, 1–4. doi:10.1207/ S15326985EP3801_1 Paas, F., Renkl, A., & Sweller, J. (2004). Cognitive load theory: Instructional implications of the interaction between information. Instructional Science, 32, 1–8. doi:10.1023/ B:TRUC.0000021806.17516.d0 Paas, F., Tuovinen, J.E., Tabbers, H.K., & Van Gerven, P.W.M. (2003). Cognitive load measurement as a means to advance cognitive load theory. Educational Psychologist, 38, 63–71. doi:10.1207/S15326985EP3801_8 Paas, F., Van Merriënboer, J.J.G., & Adam, J.J. (1994). Measurement of cognitive load in instructional research. Perceptual and Motor Skills, 79, 419–430. PubMed doi:10.2466/ pms.1994.79.1.419 Scheiter, K., & Gerjets, P. (2010). Cognitive and socio-motivational aspects in learning with animations: There is more to it than ‘do they aid learning or not’. Instructional Science, 38, 435–440. doi:10.1007/s11251-009-9118-5 Schnotz, W., & Lowe, R.K. (2008). A unified view of learning from animated and static graphics. In R.K. Lowe & W. Schnotz (Eds.), Learning with animation: Research and design implications. New York: Cambridge University Press.

Smeeton, N.J., Ward, P., & Williams, A.M. (2004). Do pattern recognition skills transfer across sports? A preliminary analysis. Journal of Sports Sciences, 22, 205–213. PubMed doi:10.1080/02640410310001641494 Spanjers, I.A.E., Van Gog, T., Wouters, P., & Van Merriënboer, J.J.G. (2012). Explaining the segmentation effect in learning from animations: The role of pausing and temporal cueing. Computers & Education, 59, 274–280. doi:10.1016/j.compedu.2011.12.024 Spanjers, I.A.E., Wouters, P., Van Gog, T., & Van Merriënboer, J.J.G. (2011). An expertise reversal effect of segmentation in learning from animated worked-out examples. Computers in Human Behavior, 27, 46–52. doi:10.1016/j. chb.2010.05.011 Sweller, J. (1994). Cognitive load theory, learning difficulty, and instructional design. Learning and Instruction, 4, 295–312. doi:10.1016/0959-4752(94)90003-5 Sweller, J., Ayres, P., & Kalyuga, S. (2011). Cognitive load theory. New York: Springer. Sweller, J., Van Merriënboer, J.J.G., & Paas, F. (1998). Cognitive architecture and instructional design. Educational Psychology Review, 10, 251–296. doi:10.1023/A:1022193728205 Taylor, H.A., & Tversky, B. (1992). Descriptions and depictions of environments. Memory & Cognition, 20, 483–496. PubMed doi:10.3758/BF03199581 Tulving, E., & Thomson, D.M. (1973). Encoding specificity and retrieval processes in episodic memory. Psychological Review, 80, 352–373. doi:10.1037/h0020071 Tversky, B., Morrison, J.B., & Bétrancourt, M. (2002). Animation: Can it facilitate? International Journal of Human-Computer Studies, 57, 247–262. doi:10.1006/ ijhc.2002.1017 Van Merriënboer, J.J.G., & Ayres, P. (2005). Research on cognitive load theory and its design implications for e-learning. Educational Technology Research and Development, 53, 5–13. doi:10.1007/BF02504793 Williams, A.M., Hodges, N.J., North, J.S., & Barton, G. (2006). Perceiving patterns of play in dynamic sport tasks: Identifying the essential information underlying skilled performance. Perception, 35, 317–332. PubMed doi:10.1068/p5310 Wright, P., Hull, A., & Black, D. (1990). Integrating diagrams and text. The Technical Writing Teacher, 17, 244–254. Manuscript submitted: August 6, 2012 Revision accepted: January 12, 2013