Technology does not stand still. While the debate as to the efficacy of technology's impact on learning continues, microcomputers become more powerful and flexible. As opposed to the first microcomputers, today's classroom machines can easily have thousands of times the amount of internal memory available before. Audio and visual capabilities will soon exceed those of today's television, and auxiliary storage will soon be practically unlimited (Moore, Myers & Burton, 1994). Because of these (and related technological advances in software),' everyday users, and most particularly educators, have access to systems called multimedia and hypermedia (see 24.6.3). Yet the development of the interactive technologies that we now call multimedia has not been without controversy or unfulfilled promises (Gleason, 199 1).

Although the concept of multimedia has been present for a long time, educators and the technology industry cannot decide exactly what the concept of multimedia (see 21.1, 24.6. 1) includes (Strommen & Revelle, 1990). Until recently, the term has meant the use of several media devices, sometimes in a coordinated fashion (e.g., synchronized slides with audiotape). Advances in technology, however, have combined these media so that information previously delivered by several devices is now integrated into one device (Kozma, 1991, p. 199). The computer now plays a central organizing role in this environment. Questions remain: Does multimedia include, for example, interactive video, CDI, and DVI, as well as traditional slide shows supplemented by sound and many other media formats?

The most commonly accepted definition of multimedia appears to support the concept of computer-driven interactivity with the learner's ability to determine and control the sequence and content selection. Matchett and Elliott (199 1) argue that these "interactive multimedia" should include motion, voice plus data, text, graphic, and still images. This definition permits multimedia to "absorb" the historically older and somewhat broader notion of hypermedia, which will be discussed in more detail later. As such, interactive video is a "high-bandwidth" source in the sense that a great deal of information, in many modes, or channe * Is, are avail able at once (i.e., parallel fashion). DeBloois (1982) indicates that "it is important to realize that interactive video (multimedia) is not merely a merging of video and computer mediums; it is an entirely new media with characteristics quite unlike each of the composites" (p. 33). The attraction of interactive multimedia is that it includes two of the more powerful educational technologies: the computer and video. Unlike some of the earlier linear technologies that allowed the user to remain passive, the new interactive programs not only allow viewers to become involved but also demand it (Gleason, 1991). By doing so, these technologies have closed the gap between some of the earlier theories of learner control and learning styles. Interactive multimedia allows the user to see, hear, and do. Through this mix of presentation techniques, interactive multimedia can appeal to learners who prefer to receive information by reading, those who learn best through hearing, and those who prefer hands-on environments (Moore, Myers & Burton, 1994).

Research concerning the learning impact of this medium is still sketchy. Its potential is important because it can combine all the symbol systems discussed above. An important distinction in this medium, however, is that the computer controls the use of the various system states. Distinct potential advantages'accrue when using this media-rich envirorunent. The learner can develop pattern recognition skills from die video and access information (in all modes) in a random manner. The latter capability takes the learner out of the traditional sequential environment and into one in which he or she can explore the domain from multiple perspectives (Cognition and Technology Group, 1990; see 26.4.2.2). Using interactive videodisc, the learner can be placed into contexts that simulate the "real world." This type of learning has been referred to as "situated cognition" (Brown, Collins & Duguid, 1989) because the information learned is tied to retrieval cues in the environments in which it will be needed.

An excellent example of situated cognition is the use of the "anchored instruction" work (see 23.4. 1.) done by the Cognition and Technology Group at Vanderbilt (1990). They believe that young students learn better in meaningful, socially organized contexts. Their research indicates that problem-oriented approaches are more effective than fact-oriented approaches in overcoming inert knowledge (knowledge people know but often fail to use during problem-solving situations). The methodology is designed to help students develop rich mental models as the basis for future learning, create environments that permit sustained exploration by students and teachers, help students explore the domain from multiple perspectives, and develop integrated knowledge structures that help students transfer knowledge to more complex tasks. (It should be noted that the above comments are speculative and are not confirmed by direct research.)

This technology parallels mental models by forming associations or links between various ideas, then constructing meaning among these relationships (Kozma, 1991). Research suggests that a number of concepts can be explored by using hypermedia's cognitive flexibility (see 12.2.3.5.4, 2 1. 1). For example, users might be interested in pursuing information about land navigation. Searching on this area might turn up information about magnetic principles, topography, uses of the compass, terrain orientation, the coordinate system, and celestial navigation. The learner could follow one or all of these links-all of which would provide further links. There might also be an opportunity to watch a video of participants engaged in the sport of orienteering, or simulations using triangulation to determine location. While research on hypermedia is in its infancy, the learner will have access to a multitude of information. This information will allow the formation (and tracking) of mental models or schemata on unlimited types of domains.

Kozma (1991) suggests that "various aspects of the learning process are influenced by the cognitively relevant characteristics of media: their technologies, symbol systems, and processing capabilities" (p. 205). He also submits that learning is influenced by taking "... advantage of the medium's cognitively relevant capabilities to complement the learner's cognitive abilities and prior knowledge and cognitive skills" (p. 205). The discussion has considered basic cognitive teaming theory and the dual-code theory that links learning to the symbol systems inherent in multimedia. Also important is the strategy used by the instructional designer or teacher to take advantage of cognitive psychology in employing media. The discussion now turns to two approaches in which multimedia applications demonstrate the use of cognitive theory.

Does multimedia really work? To answer this question, it is necessary to note some of the earlier-mentioned learning theories (see 24.6.3) and also to note earlier media-related research. It may also be useful to differentiate between evaluation studies and research. Evaluation is practical and is concerned with how to improve a product or whether to buy/use a product Studies that compare one program/media against another (or a control for that matter) are primarily evaluations. Evaluation seeks to find programs that "work" more cheaply, efficiently, quickly, effectively, etc. Research, on the other hand, tends to be more concerned with testing theoretical concepts and constructs or attempting to isolate variables to observe their contributions to a process or outcome. Having said this, we should point out that the terms evaluation and research are often used interchangeably in the fields of education and media (Moore, Myers & Burton, 1994).

Multimedia is a combination of many technologies, most notably the computer, which allows for true interaction. Strommen and Revelle (1990) stress the importance of existing research literature on computer usage for understanding the pragmatic requirements of developing interactive tasks in the multimedia programs that were developed at the Children's Television Workshop. This literature helped "take children's special needs into account and ... (delineate) what the content of our interactive tasks should be and how those tasks should be structured" (pp. 77, 78). Smith (1987) indicated that there are three major sectors in our society that use and conduct research on the effects of interactive multimedia: the military, industry, and education. Educational use of multimedia programs is still limited and in most cases still experimental. Two multimedia formats (videodisc and videotape) are predominate in education. As you would expect, multimedia researchers are still debating their relative values and virtues (Smith, 1987). However, the marketplace may decide the winner, and DVI technologies such as CD-ROM and QuicktimeTm may well settle the debate in a practical sense. Despite the short duration of multimedia's availability, Smith (1987) reports evidence for both the effectiveness and efficiency of the interactive media on learning. Other researchers argue that there is little to support the contentions of the effectiveness of interactive media. They contend that little progress has been made since Clark (1983) argued that media in general have little substantial impact on learning (Hannafin, 1985; Slee, 1989). Hannafin (1985) asserts that while the interactive technology, as noted earlier, offers interesting potential, interactive video differs little from the allied technology from either "learning or cognitive perspectives."

Ragan, Boyce, Redwine, Savenye, and McMichael (1993) summarized the findings of seven major reviews of research on multimedia. The 139 reviews were from a variety of settings, but the majority concerned adults. Among (which are obviously not independent) their findings were:

1. Multimedia is at least as effective as conventional forms and has substantial cost benefits and efficiency.
2. Frequently, multimedia instruction is more effective than conventional instruction.
3. Multimedia is more efficient in terms of learning time than is conventional instruction (30% savings).

Ragan et al. (1993) stated that they were unable to determine why multimedia was appreciably more effective than conventional instruction, and cautioned that it would be inappropriate to say that multimedia is always the most effective delivery system. They suggested that certain instructional design features appear to enhance the quality of multimedia instruction. Among them are higher levels of interactivity, program or advised learner control, integration of multimedia with other delivery forms, and structured rather than totally exploratory learning.

Srnith, Hsu, Azzarello, and McMichael (1993) reviewed 28 group-based multimedia studies. They indicate that groupbased multimedia can be as effective as individualized multimedia, and it can be as effective or more so than traditional forms of instruction. They also found that learners prefer group-based multimedia to individualized multimedia and WAtional instruction. Again, Smith et al. (1993) stated that they were unable to predict which situations are appropriate for group-based multimedia, and that it would be erroneous to state that group-based multimedia is always superior to traditional instruction or individualized multimedia.

Though hypermedia is relatively new, there are hundreds of reports and studies about its implementation. However, most of them deal with the excitement of adopting this new technology or envision its potentials in education (Yang, 1993). Only a few of these reports are experimental studies. In these limited studies, some positive results of using hypermedia have been reported. Abrams and Streit (1986), as well as Jones and Smith (1989), reported significant gains in learning achievement. Janda (1992) found a positive attitude toward the use of hypermedia systems. Higgins and Boone (1992) reported a decreased demand on teaching time. Hardiman and Williams (1990) noted that the completion rate of courses was increased with the use of hypermedia. Liu (1992) found that hypermedia was very effective in the teaching of English as a second language. In a review, Smith (1987) summarized the findings: 'The effective evidence seems to indicate that the medium is both effective and efficient ..." (p. 2). Thompson, Simonson, and Hargrave (1992) also suggested that hypermedia was promising in a learning context (Yang, 1993). .

What does the research say about multimedia -and its interactive technologies? Unfortunately, not much. The term§' multimedia and interactivity are defined universally by neither the developers nor the researchers. Many of the current guidelines for the development of multimedia programs can be traced to just a few sources. One source is the behaviorist learning theory tradition of Thorndike and Skinner-, the second is existing research investigating computer-assisted instruction. The most prevalent sources, however, are assumption, intuition, and (apparently) common sense. After reflection on an extensive review of the literature, there appears to be little useful research on multimedia (Moore, Myers & Burton, 1994). Quite frankly, with few exceptions there is not a body of research on the design, use, and value of multimedia systems. TV few exceptions include the meta-analysis of some 60 studies of McNeil and Nelson (1991), the work at the Children's Television Workshop (Strommen & Revelle, 1990), and the reviews of Ragan et al. (1993). The lack of research concentrating on the interactive features that maximize learning effectiveness has been noted by both practitioners and researchers alike. Specific programs of research have been suggested to fill these gaps, e.g., Hannafin (1985) and Kozma (1991). Until these calls are taken seriously, multimedia development will have a less-than-adequate research base (Moore, Myers & Burton, 1994).