Volitional aspects of multimedia learning.

Research on multimedia learning has produced a vast body of findings which, however, are not yet being integrated into a comprehensive framework of reference. For a considerable time, cognitive centered approaches have dominated the literature. Although motivational variables are now being taken into account, there is still a large gap in regard to an adequate representation of motivation in multimedia learning. This is an important concern given the various challenges and obstacles, such as navigational problems, distractions, and cognitive overload, that learners have to face due to the very nature of hypermedia. A promising area of theory that can help concerning this matter is represented by volition, an old concept in the study of human motivation and action (James, 1902), which has been reestablished within recent developments in psychology, such as the theory of action control (Kuhl, 1984). In this article, a volitional framework to supplement the mostly cognitively-based research on multimedia learning is introduced to serve as a basis for critically reviewing and reinterpreting current research findings. In particular, the volitional framework is applied to common phenomena in multimedia such as lost in hyperspace, cognitive overload, and seductive details together with other obstacles to persistence and learning. In addition, several future directions in research on both theory and practical strategies based on the application of volitional strategies in multimedia learning are provided.

**********

Peter, a 35-year-old product export supervisor in a large multinational corporation, is eligible for promotion, but it requires completion of a three-month intensive course. Usually this training is provided as a workshop at the company's south headquarters. Since costs have been rising considerably, general management decided to reduce course delivery costs and avoid having people away from work by introducing a self-directed multimedia-based course on CD-Rom with a user guide and Internet-interactions. Employees like Peter are now expected to complete this self-directed course on their own time. After three months they will receive a final exam, which will be an important factor in determining the promotion decision. At first, Peter is highly motivated to do the course. During the first days he works quite diligently and makes considerable progress. Then, in the second week, his motivation starts to wane because he has to fulfill other obligations. He tries to maintain his motivation to work through the CD, but other interests in his life are becoming stronger. As a result, Peter decreases his study time. The extensive freedom he has to decide, when, where, and how long he wants to work with the CD and Internet compared to being required to attend a class is apparently contributing to his vanishing motivation. Even though he had a strong desire to master the course because of the valuable benefits of the promotion, right now he is not able to protect his intention against competing tendencies.

This overall phenomenon of problems with self-regulation in self-directed multimedia learning is accompanied by some inherent aspects of multimedia that can contribute to decreased motivation. For example, assume that a while after Peter starts exploring the CD with its Internet links he realizes that he does not know where he actually is or where he has come from and he has gotten confused about what he was actually looking for. Feelings of disorientation triggered by being "lost in hyperspace" arise. Moreover, some features of the CD are relatively complicated and thus require a lot of Peter's working memory. Even though the CD was distributed with the purpose to facilitate learning it also contributes to a higher amount of cognitive load (Sweller, 1994) on the learner compared to a traditional instructor-led workshop, because Peter has to manage his own learning. Another problematic issue pertains to attractive stimuli embedded in multimedia environments like Peter's CD. Since technical improvements now allow producers to easily create sophisticated video-clips and animations, they tend to be more widely used even though they might not contribute directly to the learning objectives. For instance, Peter discovers a really intriguing animation about a prototype of a futuristic sports car, which he enjoys very much and therefore watches it several times. However, when it is time for him to complete his exercises, he realizes that he wasted time on this irrelevant animation which actually interfered with his learning goals. Distracting but interesting features such as these can be referred to as "seductive details" and they can have a detrimental effect on motivation and learning.

NEED FOR AN INTEGRATED CONCEPTION OF MULTIMEDIA LEARNING AND VOLITION

As this introductory example illustrates, self-directed multimedia learning differs substantially from traditional instructor-led learning settings. In an instructor-led environment, even if it incorporates multimedia and Internet access, information tends to be conveyed to the learners in the way that is intended by the instructor (instructor control). In contrast, self-directed multimedia learning environments are characterized by the ability to make information available in a nonlinear fashion (using nodes and hotlinks) and by random access (that is, learners can decide the order in which they access information). Thereby, the learner is given full control over his own instruction; he is responsible for the learning process and outcomes and related benefits (leaner control).

However, as illustrated by the example of Peter, there are problems inherent in multimedia environments and relatively little knowledge about them is offered in the literature. This problem is amplified by inconsistent results of experimental studies as revealed by recent meta-analyses (e.g., Dillon & Gabbard, 1998). For instance, learner control, which is regarded as one of the major advantages of hypermedia learning, showed limited and inconsistent effects since it is heavily dependent on learners' abilities; that is, some learners took advantage of the opportunity for learner control and some did not.

Another aspect of these problems, and of overall research on multimedia learning, is that while current research is making progress in gaining a better understanding of the cognitive information processing aspects (Mayer, 2001), but very little attention has been paid to motivation (Astleitner & Wiesner, 2004; Keller, 2004), and in particular major variables such as volition have been neglected. Volition refers to one's capability of maintaining attention and effort toward goals in spite of possible distractions due to waning motivation or competing goals. It represents an old psychological concept which can be traced back to the work of William James (1902) but was not incorporated into most contemporary theories of motivation. However, this has been changing and volitional theory, such as Kuhl's (1984) action control theory, which has been validated in settings other than multimedia learning can also be incorporated into this context.

In this regard, the purpose of this article is to describe how modern concepts of volition such as action control can provide an additional dimension to bring theory into multimedia research, and how it can help explain more effectively certain phenomena that have been observed for a considerable time to occur. In addition to that, helpful suggestions as to how learners can shield their motivation against competing distractions and therefore enable them to take full advantage of hypermedia-based learning environments are provided.

The remainder of this article begins with a brief overview of recent research. It is followed by descriptions of the construct of volition and the theory of action control is introduced. Then, a comprehensive volitional framework for analyzing and re-interpreting research findings in the area of multimedia learning is presented. Moreover, some practical implications to overcome current problems using action control strategies as well as related concepts are provided. Finally, some directions for future research are outlined.

STATUS OF AND GAPS IN MULTIMEDIA LEARNING THEORY

Learning with multimedia-, hypermedia-, or hypertext-based learning environments, terms which will be used interchangeably throughout this article, combined with e-learning have gained tremendous attention and are accompanied by huge expectations. One of the key features of multimedia that can benefit learning is its flexibility. It potentially allows the user to access any piece of information in whatever manner and in whatever sequence he/she wishes. Therefore, access to information is entirely interactive (that is, nonsequential) and is guided by the needs of the user (learner control). But unfortunately this is not an advantage in all situations and for all learners, especially learners with low abilities (Bannert, 2004), or learners who get confused in the multimedia environment because they do not know how to manage the large amount of freedom they have. In particular, learners' skills to regulate their learning process is crucial. As a result, normal ability students working in a multimedia environment might score less on performance tests than do lower ability students who are working in a more structured learning context. However, using action control strategies to manage fluctuating motivation can help prepare people to take full advantage of hypermedia programs. But, before exploring these possibilities, the discussion of the status of multimedia research will be continued.

On the one hand, a close look into the research literature on multimedia reveals a strong orientation towards technical aspects such as screen design (Lee & Boling, 1999), or navigation (Lawless & Brown, 1997). Findings in this area indicate the helpfulness of technical features (Chen, Fan, & Macredie, in press). For example, participants who were given a graphical overview in a hierarchically structured hypertext experienced less disorientation than participants without such an overview.

However, this area of research tends not to be grounded in any kind of theory such as communications, learning, or motivation theories that could help provide explanations of the results and serve as a basis for generalizable principles. For example, variations in student persistence and navigational patterns could be investigated more systematically by introducing theoretically-based mediating variables such as student's volitional skills.

On the other hand, there is a theory of multimedia learning (Mayer, 2001), but it is comprised almost exclusively of cognitive variables. For example, Mayer's formulation of traditional information processing theory assumes that successful multimedia learning depends on the degree to which a person selects words and pictures from multimedia presentations, organizes sounds and images into coherent verbal and pictorial models, and integrates these models with prior knowledge that is stored in the long-term memory. By applying these principles, Mayer (2003) assumed that they will lead to what he called "the promise of multimedia learning" which is that "students can learn more deeply from well-designed multimedia messages consisting of words and pictures than from more traditional modes of communication involving words alone" (p. 125). Consequently, this perception refers to a rather deterministic learning process: as long as the instructional designer did a good job; that is, using both words and pictures, successful learning will take place!

But what about possible mediating variables? What about learner's expectations or values, two classical and well-established factors of human motivation (McClelland, 1985)? What about the influence of whether the learner believes he/she is able to solve the problem, which is referred to as self-efficacy and considered a powerful predictor of learning outcomes (Bandura, 1997)? What about the effects of student's goals? There is a vast amount of findings pertaining to the consequences of (a) how people set their goals (Locke & Latham, 2002), and (b) goal orientation which can be, among other things, intrinsic or extrinsic (Deci & Ryan, 1985) as well as egocentric or task-centric. Questions like these lie outside the boundaries of Mayer's theory and therefore provide considerable opportunities for expanded areas of investigation into learning and motivation in multimedia environments.

A more systematic conception of motivation and multimedia learning is provided by Astleitner and Wiesner (2004) whose framework incorporated motivation based on well-established theoretical approaches such as goal-setting. Moreover, they referred to the theory of action control by Kuhl (1985) in which volition is defined as a mediating factor that "energizes the maintenance and enactment of intended actions" (Kuhl, p. 90) and therefore goes beyond motivation. However, they do not systematically distinguish between motivation and volition in their framework, and they do not fully incorporate the elements of action control theory. Nevertheless, the approach by Astleitner and Wiesner (2004) is promising as it contributes to a more comprehensive understanding of multimedia learning, especially a theory-based conception of motivation. Also, it can serve as an indicator that current research is trying to expand its focus.

In summary, research efforts have tried to incorporate cognitive and motivational variables into the process of multimedia-based learning. Yet, none of the preceding approaches has explicitly included the notion of volition. However, this concept has been reemerging as an important variable in the context of studies of self-regulation, which refers to the tendency of students to be self-motivated and to possess volitional strategies (Pintrich, 1999). Furthermore, Pintrich and Garcia (1994) pointed out that the influence of volition becomes even more important for college students "who, when you talk to them, are very motivated and concerned about doing well, but often have a very difficult time enacting their intentions, given all the internal and external distractions they confront in college life" (p. 126f). In a similar vein, Wolters (2003) highlighted the usefulness of volition in explaining how students manage distractions or other problems that might interfere with the timely completion of academic work.

Since multimedia learning requires a large amount of self-regulated activities such as dealing with distractions, it seems promising to integrate volitional aspects into both theoretical considerations and empirical examinations. To do so, the concept of volition is presented in the context of action control theory, which is the theory that underlies much of the recent research on self-regulation (Corno, 2001), covers circumstances as previously described, and provides powerful tools to overcome decreased motivation.

VOLITION AND THE THEORY OF ACTION CONTROL

Why does Peter fail to keep working through the CD even he knows that he has to do that in order to get promoted? Speaking from an educational perspective: What could be done to help Peter stay on task in his self-directed course of study?

The theory of action control, developed by German psychologist Julius Kuhl (1984) specifically addresses these questions. Moreover, it provides a comprehensive framework based on an information-processing perspective including motivation, emotion, cognition, and volition.

Volitional approaches such as action control theory contain specific control strategies which serve as a support for motivational tendencies by warding off competitive tendencies. Kuhl (1984) postulated six strategies and assumed that as soon as an action tendency attains the status of a current intention (by commitment), the motivational maintenance system (MMS) is activated, and it gets access to the full repertoire of action control processes, which are:

1. Selective attention: also called the "protective function of volition" (Kuhl, p. 125): it shields the current intention against competing action tendencies.

2. Encoding control: facilitates the protective function of volition by selectively encoding those features of incoming stimulus that are related to the current intention.

3. Emotion control: control of emotional states according to the current intention.

4. Motivation control: selectively processing expectancy- and/or value-related information that supports the current intention.

5. Environment control: control of the environment (e.g., making social commitments) to protect the current intention.

6. Parsimonious information processing: using stop rules to terminate the generation of more and more information regarding action alternatives.

It should be noted that processes of action control underlie virtually any kind of activity: "It is assumed that even the enactment of seemingly simple intentions such as reading a letter, opening a window requires a certain amount of self-regulatory control" (Kuhl, 1985, p. 102). Consequently, "if the MMS is inactive, an activated plan cannot be maintained for very long nor can it gain access to the execution system" (Kuhl, 1987, p. 284); that is, the plan cannot be performed because the person has not activated any resource of energy for the activity. Accordingly, action control is highly relevant for educational settings (Corno, 2001; Kuhl, 2000) which require self-regulatory behavior. In this regard, a considerable number of theories and concepts have been established (Zimmerman, 2001) in recent years.

As already pointed out, self-directed multimedia learning requires the learner to take responsibility for his learning process, especially in regard to setting and adhering to a schedule of study and managing the sequence of instruction. Hence, this article proposes a volitional framework based on modern conceptions such as action control theory to reanalyze major aspects of multimedia learning (navigational patterns, seductive details) and in particular some of its most characteristic problems (serendipity effect, lost in hyperspace, and cognitive overload). Moreover, it suggests strategies to facilitate volitional skills. The goal of this approach is to inspire research towards disregarded aspects of human motivation and volition in the growing area of multimedia learning.

VOLITIONAL EXPLANATIONS OF COMMON OBSTACLES IN MULTIMEDIA LEARNING

As aforementioned, research on hypermedia learning has grown tremendously over the past two decades leading to a vast body of findings that are largely unconnected and without a theoretical basis in human motivation and learning. This section presents a compilation of these findings and explains how they can be reinterpreted in terms of volitional categories.

Serendipity Effect and Embedded Digression Problem

The "serendipity effect" (Kuhlen, 1991) describes a phenomenon that can take place in hypermedia environments when the user comes across unforeseen, albeit interesting items of information by chance. The serendipity effect occurs if the current intention can not be supported by the MMS any longer because a newly activated intention resulting from an unexpected opportunity such as an interesting looking link becomes stronger and hence gets access to the execution system (Kuhl, 1987). As a result, the new intention will be supported and protected. To illustrate this volitional mechanism, assume that Peter has been working online for two hours or so on one of his multimedia-based lessons and suddenly, "out of the blue," discovers an animated feature containing interesting tips about how to plan an automobile road trip, which he is planning to do in two months. Even though this feature is not closely connected to his initial intention (current motivational tendency) he keeps on exploring this serendipitous opportunity for a considerable time. Apparently his initial intention was actually rather weak, maybe because of a low commitment. Therefore, the MMS is no longer able to support the intention by warding off distractions like the animated feature.

There is another effect, closely related to the serendipity effect, which is called the "embedded digression problem" (Foss, 1989). It is known to happen when users pursue digressive paths and lose their initial place or forget to return to it. In terms of volition, a motivational tendency such as trying to complete a multimedia learning task is endangered by competing tendencies that can be activated while browsing the environment (for example, the appearance of a link to an online game). This is analogous to the serendipity effect inasmuch as it might cause a superficial learning process and like it, there has been no theoretical approaches to explain why a person does or does not maintain action toward accomplishing the original goal, but action control theory does offer an explanation in terms of competing intentions.

To illustrate, several of the six action control strategies (Kuhl, 1987) can help overcome distractions due to the serendipity effect or the embedded digression problem. For example, one tactic is to guide users to focus only on relevant pieces of instruction (selective attention/active attentional selectivity), which also "inhibits the processing of information supporting competing tendencies" (Kuhl, p. 287). Another possibility is to enhance current motivation by warding off irrelevant materials (motivation control). And still another is to consider what will happen if one fails to accomplish the task (incentive escalation). In Peter's case, he might need to re-energize his motivational base by using these tactics to visualize the consequences if he fails the final exam because he wasted too much time on rather irrelevant features in the learning environment. When conflicts such as these occur, a person allows the original intention to slip away even though it is important. The appeal of the current distraction overtakes the original intention unless the person actively re-establishes it.

To sum up, volition theory can be used to explain and provide guidance for both the serendipity effect and the embedded digression problem. It can also be effective in providing a theoretical explanation for other characteristics of people in multimedia learning environment, such as navigational patterns which were empirically derived.

Navigational Patterns (cluster analysis)

The examination of different navigational patterns represents a large subfield of current multimedia research. A knowledge of patterns, or habits, may help predict learning behavior and outcomes. In this vein, Barab, Bowdish, and Lawless (1997) explored navigational paths of 66 undergraduate students while working with a multimedia environment. They discovered four different types of users whom the researchers labeled as: (a) model users, (b) disenchanted volunteers, (c) feature explorers, and (d) cyber cartographers.

This approach to tracking learner behavior makes it possible to take a look "inside" the learning process, but the resulting observations are purely descriptive; they do not explain why the learners behaved as they did. However, volitional concepts can be applied to provide a theoretically-based explanation. From this perspective, disenchanted volunteers seem to be the most interesting. These users were apparently sufficiently motivated to begin navigating the kiosk program very effectively (Barab et al., 1997, p. 34). Then, their motivation began to wane, which in turn led to the termination of the action (exploring the kiosk). Among all groups, disenchanted volunteers spent the second least amount of time using the kiosk and their information-retrieval scores were the lowest of all groups. Furthermore, they showed a relatively low degree of self-efficacy as compared to other users.

According to Kuhl (1984, p. 119) self-efficacy is one of the major determinants affecting the motivational basis of a current intention; that is, the higher the perceived self-efficacy is, the higher is the person's ability to control the current intention and protect it against other intentions. Disenchanted volunteers can therefore be characterized as users lacking in volitional skills which impedes successful usage of the kiosk. To prevent these negative effects, instructions reinforcing self-efficacy could be helpful. For instance, Ghoshal and Bruch (2003) reported that managers often employed strategies to foster self-efficacy when they were facing problems and obstacles. They tried to recall earlier experiences analogus to the current situation in which the problem was solved successfully.

Seductive Details

The phenomenon of seductive details which arises when learners get distracted by interesting yet irrelevant aspects of a stimulus environment first appeared in research on text comprehension. Within this line of research, Garner, Gillingham, and White (1989) labeled the result of being distracted by added interesting but unimportant text segments as "seductive details." Their study also showed that learning for the main points in the text declined in the presence of seductive details. In subsequent years scientists tried to fathom the reasons for the effects of seductive details which have led to an increased research effort. However, consistent patterns have not been revealed (Thalheimer, 2004), which may be caused by a missing comprehensive theoretical framework that includes major variables such as volition.

In particular volitional skills, which enable the person to stay on task seem to be crucial when dealing with seductive details. First of all, adult learners such as Peter who have many obligations in their daily work often find it difficult to maintain a habit of logging on regularly to work on their self-managed multimedia lessons. When they do log on, they find themselves in a learning environment that consists of several interrelated steps; that is, choose a goal, plan and implement their actions, control and monitor their progress, and self-evaluate their outcomes and progress. Current studies are showing that students oftentimes have difficulties fulfilling these tasks efficiently (Bannert, 2004), and the addition of seductive details instead of motivating but relevant details (Keller, 1983) make their task even more difficult. Current research on seductive details (e.g., Mayer, Heiser, & Lonn, 2001) does not assess volitional or self-regulatory skills, but it can be postulated that participants who have detrimental effects from seductive details apparently fail to control their attention towards the important points of the instruction.

Furthermore, students may not able to terminate their information processing at an appropriate time; that is, they do not try to determine when they have obtained sufficient knowledge for the task. This dysfunctional learning behavior can be explained by Kuhl's (1984) concept of "parsimonious information processing" which states: "Whenever the actor believes that further processing of information bearing on potential action alternatives may jeopardize the execution of the current intention, the continuation of [this] process ... should be brought to a halt" (p. 127). Following this line of argument, seductive details would appear to provoke further yet unnecessary information processing, which in turn leads to a shift of the focus. The learner loses interest in his/her initial task because of interfering seductive aspects. Therefore, he/she starts exploring new aspects in the environment.

To help people prevail over these behaviors, strategies of self-monitoring or attention control seems promising. For example, Peter can try to focus on the important information regardless of the type of media it is being presented in: "I'll try to regard a specific media (for example animation, video clip) only if I assume that it will bring me closer to the promotion!" However, since animations or video clips sometimes might pretend to be important and try to capture the interest of the learner by using attractive curiosity-arousing stimuli, self-instructional strategies like this one could also help people become aware of this danger. As observed in a recent study on learning behavior in online environments (Rapp & Dresel, 2003), learners tend to click on every clickable object regardless of its potential usefulness. Therefore, becoming aware of the distraction potential of multimedia objects constitutes an important step towards improved multimedia learning.

In summary, although research on the seductive detail effect has been discovering some of the underlying cognitive principles (Mayer et al., 2001), so far it does not provide any consistent theoretical explanation (Thalheimer, 2004). Applying a volitional frame of reference can help to bridge the gap by taking further moderating variables such as learner's volitional skills into consideration.

Lost in Hyperspace

From the very beginning when hypermedia systems emerged, feelings of disorientation which arise when the user is lost in the environment have become familiar, yet still remain unresolved. In general, two forms of being lost can be distinguished: (a) users do not know where they are in the network, and (b) users do not know how to get to some other place that they know (or think) exists in the network.

One reason disorientation in hypermedia environments might come about is insufficiently supportative navigation design. Nielsen (2000) postulated three fundamental questions which navigational design should address: (a) where am I?, (b) where have I been?, and (c) where can I go?. Embedded navigational aids that meet these criteria seem to be helpful in overcoming disorientation. In a small experimental study (N=20), Danielson (2002) examined the effects of a constantly visible site map for relatively small sites and local views of five real-world web sites (e.g., TopGolf, http://www.topgolf.co.uk). He offered the site map to the experimental group whereas participants in the control group could not make use of the site map. The experiment contained tasks such as fact-finding missions (e.g., "About how long does it take for an individual person to play three games of TopGolf"?). Results indicated that, the site map participants made less use of the "back" button, deserted fewer information-seeking tasks, and better estimated the amount of distinct pages they had visited while navigating than the control group participants.

However, albeit those findings, it can be questioned whether lost in hyperspace is the result of design or psychological problems. However, it can be assumed that it is a complex problem and should therefore be approached in a multi-disciplinary manner. This allows researchers to draw upon and integrate knowledge and findings in seemingly diverse disciplines such as Cognitive Psychology, Artificial Intelligence, and Software Engineering. Yet, these approaches, as Mayer's Theory of Multimedia Learning, focus solely on cognitive aspects, apart from the design issues, and abridge other elements of the complex mechanism of human learning. Therefore, in keeping with the general thesis of this article, it is postulated to integrate volitional aspects in the study of the lost in hyperspace phenomenon.

In particular, the action control strategy "selective attention" can help overcome feelings of disorientation. Kuhl (1984, p. 125) defined this as a "means of selectively strengthening the activation and processing of information that supports the current intention." With regard to hypermedia learning, attention control/selective attention can be employed by trying to focus only on task-relevant material. Before exploring any content, users can ask themselves, "Is this information necessary for solving the task?" or, "Do I really have to visit this feature?"

Recently, Gollwitzer (1999) introduced the concept of implementation intentions as a contribution to modern theory of volition. Based on the proven possibilities for helping people stay on task this concept can be used as an attention control strategy. By anticipating possible distractions which might occur while learning with a multimedia environment, such as animations or video clips, and linking them to specific responses (for example, "Whenever a video clip pops up, I will ignore it"), people are able to protect their intentions in a relatively straightforward and powerful way. Research using the delay of gratification paradigm indicated that implementation intentions can support peoples' intentions and help ward off distraction. However, studies which specifically address aspects of hypermedia learning such as the lost in hyperspace phenomena are still missing.

Cognitive (Over-)Load

Similar to the lost in hyperspace phenomena, cognitive overload appeared along with early innovations of hypermedia systems. Based on the theoretical foundations derived from the work of Sweller (1994), cognitive load theory (CLT) assumes a limited working memory which is connected to a virtually unlimited long-term memory. When working with a hypermedia system, the user is faced with three types of cognitive load:

    1. Intrinsic cognitive load: load caused by the complexity and the
       structure of the material.
    2. Germane cognitive load: load induced by the learner's effort to
       process and comprehend the provided information (effective
       cognitive load).
    3. Extraneous cognitive load: this load is a form of overhead that
       does not contribute to an understanding of the material
       (ineffective cognitive load). For example "instructional
       techniques that require students to engage in activities that are
       not directed at schema acquisition or automation" (Sweller, p.
       299).

Given the constant technical improvements leading to highly sophisticated hypermedia environments, the amount of cognitive load increases, too. Learning in such an environment requires the user to take responsibility for what to read and the sequence for reading it; hence, it augments the amount of extraneous cognitive load and decreases the amount of germane cognitive load. This is one of the problems in learning settings such as Peter's where one has to work with a multimedia environment on one's own, that is, without an immediate supervision by a teacher. As a result, less working memory is available to process the actual learning content. Therefore, it is not surprising that some learners in multimedia environments obtain worse learning outcomes compared to traditional settings such as classroom (Bannert, 2004).

For a considerable time, several attempts have been undertaken to reduce extraneous and foster germane cognitive load; however, these approaches have considered only cognitive oriented strategies such as, among other things, worked-out examples and goal-free activities. However, recent critics say this is a conceptual shortcoming of CLT. For example, Gerjets and Scheiter (2003) alluded to the so-called "one-to-one mapping" (p. 34) between specific patterns of instructional design (e.g., the design of a multimedia kiosk) and resulting outcomes (e.g., the score in an information-retrieval task) without accounting for different learner activities that might mediate between the instructional design and subsequent learning outcomes. For that reason, they postulate goal-orientations and processing strategies as potential mediating variables. Their conception is promising inasmuch as it points beyond the prevalent cognitive orientation of CLT.

In another approach to improving CLT theory, Valcke (2002) introduced metacognitive load as a fourth type of cognitive load. By doing so, Valcke also took leaner activities, which are beyond cognition into account. It is assumed, especially, that monitoring plays a significant role and influences different parts of the learning process such as the selection and organization of sensory information to working memory. Since monitoring activities refer to learner's effort to process and comprehend the information, Valcke incorporated metacognitive load under the category of germane cognitive load.

Both Gerjets and Scheiter (2003) and Valcke (2002) provided new and inspiring insights into CLT as they refer to specific learner activities which are beyond the mere processing of information based on principles of Cognitive Psychology. Nevertheless, volitional aspects are not being considered in those conceptions. Although it has a close relationship to the concept of metacognition, volition goes beyond it as it is based on a broader theoretical framework, which also includes aspects of regulation of motivation and emotion. As pointed out by Wolters (2003), regulation of motivation is conceptually similar to regulation of cognition (that is metacognition) but differs with regard to its target and purpose:

    Strategies for the regulation of cognition are meant to affect
    students' use of cognitive learning strategies or how students are
    processing, constructing, or understanding the material with which
    they are interacting. In contrast, strategies for regulating
    motivation ultimately are meant to affect students' willingness to
    process information, to construct meaning, or to continue working.
    (p. 192)

Consequently, focusing only on metacognitive activities such as monitoring is too limiting. Moreover, some other basic psychological variables such as emotion can interfere with metacognitive strategies and confine their effects. Therefore, the volitional framework which incorporates motivation and emotion broadens the spectrum of possible interventions regarding cognitive load.

Another volitional approach, which can inspire CLT research is the concept of implementation intentions (Gollwitzer, 1999). Using field studies as well as laboratory settings, it has been shown that implementation intentions have the potential to decrease extraneous cognitive load and to increase germane cognitive load (Gollwitzer, Fujita, & Oettingen, 2004). Following Valcke's (2002) conception, metacognitive load, which is induced by activities such as planning consumes part of the germane load and therefore reduces the available amount of working memory. Consequently, it should be kept at a minimum. Implementation intentions as a volitional strategy analogous to metacognitive activities satisfy these requirements perfectly. They have proven to be very simple which, for instance, allows people to use them even in highly demanding contexts. Evidence for this is provided in a series of studies using a dual-task paradigm (Gollwitzer et al.). Therefore, it can be concluded that: "Apparently, the immediacy of responding induced by implementation intentions is also efficient in the sense that is does not require much in the way of cognitive resources" (Gollwitzer et al., p. 213). Nevertheless, research has not yet investigated the effectiveness of implementation intentions in multimedia learning environments in particular dealing with aspects of cognitive load.

SUMMARY

It has been shown how fruitful and helpful volitional considerations can be in explaining some of the major problematic issues in research on multimedia learning. These include the serendipity effect/"embedded digression problem," navigational patterns, lost in hyperspace, seductive details, and cognitive overload. Furthermore, feasible strategies to overcome those problems were provided and several practical implications for instructional designers are offered below. However, in addition to providing a useful added dimension to previous and current areas of research, volitional considerations can also be useful in regard to supporting emerging and future areas of research.

FUTURE DIRECTIONS

In this section some directions for future research in multimedia learning will be briefly discussed to obtain a more comprehensive theoretical foundation.

Validating the Theoretical Framework

The proposed volitional framework is to be seen as an initial step towards a systematic theoretical foundation based on the extension of action control theory to the volitional aspects in multimedia learning. Although action control theory has made one of the major contributions to modern conception of volition, its applications have been primarily in areas of behavioral change and, to a lesser degree, in classroom education settings. As has been advocated in this article, there are many ways in which additional research based on action control theory and its applications can become more and more elaborated and sophisticated through continued refinements, especially to adequately represent multimedia-based learning.

One first step for volitional-centered conceptions might be to attempt to capture distinctive volitional activities of students while learning with multimedia. Analogous to Mayer's cognitive approach, this might reveal basic volitional thoughts and behaviors. For example, qualitative research programs using methods such as thinking aloud protocols or as log-file analysis can help to identify certain volitional components such as dealing with distractions or other behaviors that exemplify various action control strategies. The goal of this approach is to "look inside" the learning process and to document different patterns of covert volitional activities and to relate them to outcomes such as persistence and achievement. Subsequent statistical calculations (for example cluster analysis) can be used to explore different profiles of learners.

The findings of such qualitative research programs should be integrated into a comprehensive volitional framework, which would be a next step towards a more comprehensive and integrated volitional conception of multimedia learning.

Developing and Fostering Volitional of Strategies

Besides contributing to the theoretical foundation, an area of equal importance pertains to the development of volitional strategies. From an educational point of view, volitional strategies are standing in the center of the theory of action control and there have been numerous applications of them (Corno, 2001), but mostly they have been in settings other than multimedia.

Multimedia research should therefore try to adapt action control strategies. The research program initiated by Corno as well as the volitional skill training program (VEP; Trawick & Corno, 1995) can serve as helpful guidelines, as can Keller's (1983) ARCS model, which provides a validated, systematic approach to diagnosing learners' motivational and volitional problems and prescribing strategies according to the four dimensions of attention, relevance, confidence, and satisfaction. Each of these categories subsumes a set of psychological constructs that can be clustered based on their shared attributes. For example, one might test the effects of having instructors in online settings moderate an opening discussion about advantages and disadvantages of online learning among students using a discussion board. Each student could post his/her experiences or expectations. Following that, the instructor might focus more specifically on motivational issues by, for example, posing the question, "Why do you think motivation is important for online learning?". This strategy would help ensure that students perceive the instruction to be personally relevant. Also she/he should pick possible obstacles and challenges out as central themes for online learning settings ("What obstacles and challenges do you expect to encounter in an online learning course?") and thus make students become aware of it ("How would you deal with that?"). Finally, students could post their own strategies to overcome those problems and discuss them. These postings should be available for every student throughout the entire course. This method would presume that there are serious volitional challenges to the learners in this setting, and using a motivation/volitional design process that incorporates learner motivational analysis (Keller, 2004) would be important.

Moreover, research in the area of distance education offers some hints for practical and useful strategies that can be used in support of learners' volition. One of these is the technique called "motivational messages" as developed by Visser and Keller (1990). This technique proved to be both effective and efficient in different environments. For instance, Visser and Keller reported of an intervention using simple written statements, which were distributed to the learner at the moment when he or she was most likely to benefit from some externally provided motivation. It is important to understand that these messages were not the type of generic motivational "boosters" that one finds on posters and calendars. They were formulated as a result of analyzing the motivational characteristics of the audience at various key points in the course and preparing messages that were targeted to the needs of the learners. Results of qualitative and quantitative analysis indicate that the motivational messages had positive effects on student attitude and performance.

Since technical improvements are constantly proceeding in regard to resources to support learning and the delivery of instruction, new methods for producing motivational messages are now available which are far beyond the simple written statements that were distributed by hand (Visser & Keller, 1990). There has been a rapid development of technology called "animated pedagogical agents," which has been derived from intelligent tutoring systems and enables the developer to convey messages in different manners such as animation, video clip, or audio (e.g., Atkinson, 2002). These agents could be an effective means of conveying motivational and volitional messages to learners when most needed. Even learners who seem to be motivated adequately as Peter by means of rewards are likely to benefit by additional support.

A closely related yet not so sophisticated concept as animated pedagogical agents is that of prompting. This technique presents the learner with a pop-up window throughout a multimedia learning environment. The prompt can include such things as scaffolding tactics to assist learning, but can also be used to motivate the learner (for example a cheer up message such as, "Don't worry, try once more!"). Another motivational/volitional application refers to the regulation of the learning process; for example, prompting students to make them aware of problems while learning. Results of a recent study (Stark, Tyroller, & Krause, 2004) indicated positive effects. Students who were prompted to give reasons for their selections in the course of learning with examples and problem-solving tasks performed significantly and substantially better in a knowledge test than their colleagues who were not prompted.

Taken together, theoretical developments and technical innovations greatly enrich volitional research by providing opportunities to investigate and apply action-control strategies in the context of multimedia learning. The possibilities described here are only a small sampling of the kinds of inquiry that are underway or could be initiated.

CONCLUSION

The present article introduces a volitional framework to expand the current theoretical models of multimedia learning, which are mostly cognitive in nature and, which have not yet achieved a coherent and consistent theoretical foundation. Moreover, since recent meta-analyses of the effectiveness of multimedia learning environments detected mixed results this article suggests that further important variables, such as motivation and volition, should be added to the current research models. Although recent approaches such as the Integrated Model of Multimedia Learning and Motivation by Astleitner and Wiesner (2004) have moved forward in a promising direction, there is still a considerable gap regarding motivation and volition. This article provides a sharper distinction between these two concepts, and presents the construct of volition in a manner that illustrates its utility in helping close the gap within multimedia research. Specifically, based on the explanations contained in this article, it is believed that volitional concepts, such as those contained in action control theory and self-regulation theory can be incorporated into the theoretical models, can be used to provide more theoretically-based explanations for several observed phenomena in multimedia learning, and can be used to guide some areas of future areas of research on theoretical assumptions and applications in practice. Based on this approach, perhaps it will be possible to build a more coherent and integrated body of knowledge on multimedia learning.

References

Astleitner, H., & Wiesner, C. (2004). An integrated model of multimedia learning and motivation. Journal of Educational Multimedia and Hypermedia, 13(1), 3-21.

Atkinson, R. K. (2002). Optimizing learning from examples using animated pedagogical agents. Journal of Educational Psychology, 94(2), 416-427.

Bandura, A. (1997). Self-efficacy. The exercise of control. New York: Freeman.

Bannert, M. (2004). Designing metacognitive support for hypermedia learning. In H. M. Niegemann, D. Leutner, & R. Brunken (Eds.), Instructional design for multimedia learning (pp. 19-30). Munster, Germany: Waxmann.

Barab, S. A., Bowdish, B. E., & Lawless, K. A. (1997). Hypermedia navigation: Profiles of hypermedia users. Educational Technology Research & Development, 45(3), 23-41.

Chen, S. Y., Fan, J. P., & Macredie, R. D. (in press). Navigation in hypermedia learning systems: Experts vs. novices. Computers in Human Behavior.

Corno, L. (2001). Volitional aspects of self-regulated learning. In B. J. Zimmerman & D. H. Schunk (Eds.), Self-regulated learning and academic achievement. Theoretical perspectives (pp. 191-226). Mahwah, N.J.: Lawrence Erlbaum.

Danielson, D. R. (2002). Web navigation and the behavioral effects of constantly visible site maps. Interacting with Computers, 14(5), 601-618.

Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. New York: Plenum Press.

Dillon, A., & Gabbard, R. (1998). Hypermedia as an educational technology: A review of the quantitative research literature on learner comprehension, control, and style. Review of Educational Research, 68(3), 322-349.

Foss, C. L. (1989). Tools for reading and browsing hypertext. Information Processing and Management, 25(4), 407-418.

Garner, R., Gillingham, M. G., & White, C. S. (1989). Effects of "seductive details" on macroprocessing and microprocessing in adults and children. Cognition and Instruction, 6(1), 41-57.

Gerjets, P., & Scheiter, K. (2003). Goal configurations and processing strategies as moderators between instructional design and cognitive load: Evidence from hypertext-based instruction. Educational Psychologist, 38(1), 33-41.

Ghoshal, S., & Bruch, H. (2003). Going beyond motivation to the power of volition. MIT Sloan Management Review, 44(3), 51-57.

Gollwitzer, P. M. (1999). Implementation intentions. Strong effects of simple plans. American Psychologist, 54(7), 493-503.

Gollwitzer, P. M., Fujita, K., & Oettingen, G. (2004). Planning and the implementation of goals. In R. F. Baumeister & K. D. Vohs (Eds.), Handbook of self-regulation. Research, theory, and applications (pp. 211-228). New York: Guilford Press.

James, W. (1902). The principles of psychology. New York: Holt.

Keller, J. M. (1983). Motivational design of instruction. In C. M. Reigeluth (Ed.), Instructional-design theories and models: An overview of their current status (pp. 383-434). Hillsdale, N.J.: Lawrence Erlbaum.

Keller, J. M. (2004). A predictive model of motivation, volition, and multimedia learning, Proceedings of the International Symposium and Conference, Educational Media in Schools (pp. 9-19): Osaka, Japan: Kansai University.

Kuhl, J. (1984). Volitional aspects of achievement motivation and learned helplessness: Toward a comprehensive theory of action control. In B. A. Maher & W. B. Maher (Eds.), Progress in experimental personality research (pp. 101-171). Orlando, FL: Academic Press.

Kuhl, J. (1985). Volitional mediators of cognitive-behavior-consistency; self-regulatory processes and action versus state orientation. In J. Kuhl & J. Beckmann (Eds.), Action control. From cognition to behavior (pp. 101-128). Berlin: Springer.

Kuhl, J. (1987). Action control: The maintenance of motivational states. In F. Halisch & J. Kuhl (Eds.), Motivation, intention and volition (pp. 279-291). Berlin: Springer.

Kuhl, J. (2000). The volitional basis of personality systems interaction theory: Applications in learning and treatment contexts. International Journal of Educational Research, 33(7-8), 665-703.

Kuhlen, R. (1991). Hypertext: Ein nicht-lineares Medium zwischen Buch und Wissensbank. Berlin: Springer.

Lawless, K. A., & Brown, S. W. (1997). Multimedia environments: Issues of learner control and navigation. Instructional Science, 25(2), 117-131.

Lee, S. H., & Boling, E. (1999). Screen design guidelines for motivation in interactive multimedia instruction: A survey and framework for designers. Educational Technology, 39(3), 19-26.

Locke, E. A., & Latham, G. P. (2002). Building a practically useful theory of goal setting and task motivation. American Psychologist, 57(9), 705-717.

Mayer, R. E. (2001). Multimedia learning. Cambridge, UK: Cambridge University Press.

Mayer, R. E. (2003). The promise of multimedia learning: Using the same instructional design methods across different media. Learning and Instruction, 13(2), 125-139.

Mayer, R. E., Heiser, J., & Lonn, S. (2001). Cognitive constraints on multimedia learning: When presenting more material results in less understanding. Journal of Educational Psychology, 93(1), 187-198.

McClelland, D. C. (1985). Human motivation. Glenview, IL.: Scott, Foresman.

Nielsen, J. (2000). Designing web usability. Indianapolis, IN: New Riders.

Pintrich, P. R. (1999). Taking control of research on volitional control: Challenges for future theory and research. Learning & Individual Differences, 11(3), 335-351.

Pintrich, P. R., & Garcia, T. (1994). Self-regulated learning in college students: Knowledge, strategies, and motivation. In C. E. Weinstein (Ed.), Student motivation, cognition, and learning (pp. 113-134). Hilsdale, NJ: Lawrence Erlbaum.

Rapp, A. M., & Dresel, M. (2003). Man klickt alles durch ohne zu iiberlegen--Vorgehensweisen von Studenten beim Lernen mit neuen Medien. Paper presented at the 9. Tagung der Fachgruppe Padagogische Psychologie der Deutschen Gesellschaft fur Psychologie, Bielefeld, Germany.

Stark, R., Tyroller, M., & Krause, U.-M. (2004). Effects of a metacognitive prompting procedure in the context of a computer-based learning environment: Practical relevance and explanation by metacognitive and motivational processes. In P. Gerjets, P. A. Kirschner, J. Elen, & R. Joiner (Eds.), Instructional design for effective and enjoyable computer-supported learning. Proceedings of the first joint meeting of the EARLI SIGs Instructional Design and Learning and Instruction with Computers (CD-ROM) (pp. 341-347). Tuebingen, Germany: Knowledge Media Research Center.

Sweller, J. (1994). Cognitive load theory, learning difficulty, and instructional design. Learning and Instruction, 4(4), 295-312.

Thalheimer, W. (2004). Bells, whistles, neon, and purple pose: When interesting words, sounds, and visuals hurt learning and performance--a review of the seductive-augmentation research. Retrieved January 21, 2006, from

www.work-learning.com/seductive_augmentations.htm

Trawick, L., & Corno, L. (1995). Expanding the volitional resources of urban community college students. New Directions for Teaching and Learning, 63, 57-70.

Valcke, M. (2002). Cognitive load: updating the theory? Learning and Instruction, 12(1), 147-154.

Visser, J., & Keller, J. M. (1990). The clinical use of motivational messages: An inquiry into the validity of the ARCS model of motivational design. Instructional Science, 19(6), 467-500.

Wolters, C. (2003). Regulation of motivation: Evaluating an underemphasized aspect of self-regulated learning. Educational Psychologist, 38(4), 189-205.

Zimmerman, B. J. (2001). Theories of self-regulated learning and academic achievement: An overview and analysis. In B. J. Zimmerman & D. H. Schunk (Eds.), Self-regulated learning and academic achievement. Theoretical perspectives (pp. 1-38). Mahwah, NJ: Lawrence Erlbaum.

MARKUS DEIMANN

Erfurt University

Germany

[e-mail]

JOHN M. KELLER

Florida State University 

USA

[e-mail]