The concept of multi-image is closely akin to properties of cue summation research, which suggests increased learning from more cues within a single channel or using more cues across (multiple) channels. Multi-image research was very popular in the 1960s and 1970s. The multi-image format (see 16.4.2.4) in these earlier studies generally referred to the use of more than one image, with or without audio synchronization on single- or multiple-projection screens. Millard (1964) stated that simultaneous images can be used advantageously in instructional situations that require comparisons, the development of interrelated concepts, and illustrations of relationships, or in the presentation of dimensional and spatial characteristics of objects. Perrin's (1969) theory of using multi-images is based on the simultaneous presentation of images in which images interact; this may be of significance in making comparisons and establishing relationships. Film, slides, television, etc. (not current interactive multimedia forinats), presented content and images in a sequential, linear format. The meaning was based on the context (content that preceded) of the image. However, multi-image allows, as Perrin states:

... the viewer to process larger amounts of information in a very short time. Thus information density is effectively increased,. and certain kinds of information are more. effficiently learned (p. 369).

However, questions raised earlier by Hartman (1961), Hsia (1971), and others concerning the efficacy of simultaneously presenting information across (and within) channels also apply to the concept of multi-image presentations.

Burke and Leps (1989) indicated that there may have been a "failure" by multiple-image enthusiasts to prove its effectiveness. Multi-image, like other specific technologies, has always had to use traditional media comparison studies with their inherent problems (see 4.2.4). Fradkin (1974, 1976) noted that although there was wide use of multiimage in education, there was little empirical evidence in support of increased learning. Moreover, Burke and Leps (1989) note that little research on multi-image presentations has investigated the validity of aspects of Perrin's theory, and that many studies of multiple-image presentations have been limited to self-serving individuals involved in the hardware and production processes.

All of these instructional situations require association, which, according to Gagn6 (1965), is one of the basic mechanisms of learning. According to Perrin (1969), the number of instances available to the viewer to make associations by visual comparison are greater with simultaneously presented images than with sequentially presented images. Low (1968) pointed out that in single-image presentations one image follows another, thus determining the interrelationships between images. In multiple-image presentations, several images appear simultaneously and "interact upon each other at the same time, and this is of significant value in making comparisons and relationships" (Perrin, 1969, p. 90).

Perrin (1969) stressed that images are especially rich in information and in the range of associations they stimulate. Without careful control by the communicator, there is the possibility that some associations can conflict with the intended message, causing interference. Relevance, realism, and simplicity have been. found to be important in learning from book illustrations (Spaulding, 1956) and in learning from films (May & Lumsdaine, 1958). These factors are equally important in presentations utilizing multiple imagery (Perrin, 1969). A viewer's ability to determine relationships between images has an effect on memory and recall (Berger, 1973; Low, 1968). Low stated that no single image can establish certain memory combinations, but a group of images perceived simultaneously often recalls long-forgotten memories. Berger (1973) found that multi-image techniques are effective in expediting the recall of events and thought feeling associations in analytic psychology. The recall of memories and of events attributed to simultaneous images May be a function of the viewers' freedom to select their own sequence (Bruner, 1967; Gagn6 & Briggs, 1974 ). Therefore, as Perrin (1969) pointed out, presenting images simultaneously and allowing viewers to select their own sequential order may have an effect on the learning taking place. Roshka (1960), Malandin (cited in Perrin, 1969), and Allen and Cooney (1963) found simultaneous presentation of images effective in instruction with younger children. Roshka (1960) found that simultaneous images had less effect with older children, and Allen and Cooney (1963) stated that simultaneous images had a significant effect on learning of sixthgraders, but not eighth-graders. Malandin (cited in Perrin, 1969) found that primary classes had difficulty with recall from sequential images, but that grouping the images permitted an increase in the number of recollections and organization of the recollections. These studies support Perrin's (1969) view that image simultaneity is a significant factor in some learning situations. Beck (1983), in a study that supported Perrin's views, found that subjects exposed to simultaneous picture formats achieved significantly higher scores than subjects exposed to successive (linear ) formats. Goldstein (1975) stated that the simultaneous presentation of multiple images is in many respects "like the environment; it contains meaningful material, it surrounds us, and it is constantly changing" (p. 63).

A caution that emerges from the literature concerning the simultaneity of multiple images is that the theory of cue summation may not be valid in some contexts. Recall that cue summation, as noted earlier, is the general theory that posits that the more cues that are given through various communications channels, the more learning occurs (Whitley, 1977). Perrin (1969) notes that the use of simultaneous multiple images places a burden on the visual channel and that in the multiplication of visual stimuli, irrelevant as well as relevant detail is increased. Therefore, care must be taken to ensure that the visual stimuli are clear and simple and that detail included is relevant. Otherwise, the result is not cue summation but confusion. A study by Fradkin and Meyrowitz (1975) supports this hypothesis that cue summation and the avoidance of conflicting cues is important in the design of multiple-image presentations produced for cognitive learning situations.

The use of a large screen coupled with the simultaneous projection of two or more images has been cited as one of the major, inherent advantages of multiple imagery. A large screen provides better approximations of "real" environments by supplying the physical and psychological factors (see 16.11.1, 36.3.10) necessary for realism and involvement (Perrin, 1969).

Blackwell (1968) indicated that tasks requiring high visual acuity, such as detecting differences in texture or patterns, might benefit from the use of large-screen presentations. Two factors affecting usefulness of large screens were identified by Schlanger (1966): visual impact and visual task. Visual impact is the amount and forcefulness of information available to the sense of sight. The visual impact is proportional to the amount of the viewer's field of view that the screen occupies. According to Blackwell (1968), visual impact on the viewer is greater in large-screen presentations because more of the viewer's field of vision is occupied by the projected image-therefore limiting the chance of distraction from the surrounding environment. Schlanger (1966) stated that large screens can produce information rich in detail for the visual channel and simulate real environments, but Blackwell (1968) warned that any channel of communication loaded with information details may be distracting if the details are-irrelevant to the learning situation. Travers (1966), while attempting to deal with excess details, hypothesized that line drawings would be advantageous because they eliminated superfluous detail. His experiments with oversimplified drawings, however, indicate poor transfer of learning to real situations. Blackwell (1968) stated that the advantage of a large screen to reduce the visual task factor is conditional. Presented images, for example, must contain enough irrelevant detail to convey the proper message (which may not have been the situation in Travers's experiments), but not so much detail as to distract learners. Barr (1963) stated that a large screen opens up the frame and gives a greater sense of continuous space. The more open the frame, the greater the impression of depth; the image is more vivid. This suggests that simultaneous images produce an increase in information density during presentations.

A greater density of information is possible with multiple than with linear imagery. There are several dimensions to information density in multiple-image presentations (Whitley, 1977). Perrin (1969) believes that it is important to distinguish between the method of presentation and the mechanism of perception. He states that the theory of multiple images suggests that for making contrasts and comparisons, and for learning relationships, "simultaneous images reduce the task of memory (a dimension of visual task) and enable the viewer to make immediate comparisons" (p. 376).

Langer (1957) utilizes the terms linear and nonlinear to distinguish between verbal and iconic signs. She stresses the sequential ordering, the "strung-out" arrangement of linear (verbal) signs in time and contrasts this to the "all-atonce" (parallel) character inherent in pictorial signs (p. 83). Her position is that even single pictures shown in sequential order are essentially nonlinear (Whitley, 1977).

Nonlinearity and simultaneity go hand in hand. The use of visual images, inherently nonlinear, allows the presentation of a great deal of information simultaneously rather than sequentially, as with words arranged in sentences and thus bound to grammatical 'ordering and syntax. Perrin (1969) expands this line of analysis and hypothesizes that when visual images are combined in multi-image presentations, the result is an increase in the amount of information presented simultaneously, or in the infonnation density of the presentation.

Information density can be further increased if the information is organized properly (Whitley, 1977). McFee (1969) believes that visual organization is more important than the actual amount of information present. Much of our responding occurs so quickly that we are unaware of our own processing. Selecting and organizing visuals in advance makes the information for the user easier to assimilate (p. 85).

Investigative confirmation of the importance of organization is illustrated by the introduction of a carefully organized and automated televised instructional system called TeleMation at the University of Wisconsin. Hubbard found (1961) that information density could be significantly increased through proper organization without loss of material or loss of learning by students. A similar finding resulted when the Army Ordinance Guided Missile School conducted a series of evaluative studies in 1958 (U.S. Army, 1959). Instruction time was reduced 19.5% to 41% for a similar level of achievement, and an increase in learning was reported for the experimental groups 9 weeks later. Allen and Cooney (1963), however, suggested that time saved in instruction was as much a function of care in preparation as it was a function of the multi-imaged delivery of the subject matter.

Commercial producers claim that information density created through multiple imagery results in motivation and arousal. A serious question is whether or not this arousal is beneficial (Whitley, 1977). Research on motivation indicates that an increase in motivation improves performance (Smith, 1966) but that there is an optimum level. Eysenck (1963) found that for complex tasks, optimum performance is achieved when drive is relatively low; only for simple tasks is the optimum achieved with relatively high drive. Kleinsmith and Kaplan (1963, 1964) and Kleinsmith, Kaplan, and Tarte (1963) found that there is some confusion between learning and performance, with a person sometimes performing very poorly in highly arousing situations, yet tending to remember most vividly those incidents in his life that were most traumatic or arousing. These researchers measured skin conductivity, and their findings indicated that high-arousal associates showed stronger permanent memory and weaker immediate memory than low-arousal associates. Low arousal was accompanied by the normal forgetting curve. High-arousal responses showed poor immediate recall. This may explain some inconsistencies in research with regard to long-term retention. For example, VanderMeer (195 1) found that color films did not increase immediate learning but produced greater long-term retention. The findings of KJeinsmith suggest that the cause may have been the arousal produced by the color films.

Fleisher (1969) stated that the mind and eye have proved to be capable of tremendous speed and versatility in accepting multiple impressions, and that during a multi-image presentation the viewer's eyes explore the entire screen and keep the viewer very conscious of what is happening. In contrast, Goldstein (1975) indicated that multi-image presentation may cause information overload by presenting more information than the viewer can process and thus create arousal through frustration. This arousal may cause multi-mage presentations to be highly motivating but not very informative (Kreszock, 1981). Goldstein (1975) stated that when presenting specific concepts or highly technical information, multi-image presentations should be used with restraint. Perrin (1969) concluded that it is clear that great densities of information can be perceived during a multi-image presentation, but he went on to question whether great amounts of information were learned from these perceptions.

Several studies have compared different aspects of single-image and multi-image presentations. Lombard (1969) used both a single-image and multi-image format to teach synthesis skills in history to I I th-grade students. He found no significant differences in males between the singleimage and multi-image presentations at any achievement level, and the only female group to demonstrate any significant difference were the low achievers. These low-achieving females who received the multi-image presentations surpassed both the males and females in the average- and high-achiever groups who received the single-image format. Some of the procedures used in Lombard's study, however, make his findings dubious.

Conducting a study to explore the affective impact of multi-image presentations, Bollman (1970) experimented to see if there was any difference in the amount of shift in evaluative meaning of audiences viewing multi-image presentations and audiences viewing single-image presentations, and to ascertain if the persons' relationship to the screen had any effect on shifts in evaluative meaning. In his conclusions, Bollman (1970) stated that this experiment did not produce significant statistical evidence or conclusive answers.

Atherton (1971) conducted a study to determine if a multi-image slide presentation would result in greater affective and cognitive learning than similar content presented by a 16-min film. No significant differences were found between groups in the amount of attitudinal change elicited as a result of the presentation, or between treatment of groups relative to the cognitive learning resulting from viewing the presentations. These analyses indicated that one treatment was not significantly more effective (or even affective) than the other in producing positive increases in affective or cognitive learning (Atherton, 197 1). Didcoct (1958) conducted a study of the cognitive and affective responses of college students to single-image and multiimage presentations. He found no significant difference in attitude or cognitive retention between a group viewing a single-image presentation and a group viewing a multi-image presentation.

Westwater (1972), in conducting a descriptive study to gather information about the field use of a multi-image presentation, found that about 80% of the teachers who participated in the study would like to use such presentations to a greater degree. Westwater, however, pointed out two major limitations to the development of multi-image presentations. These are that few teachers were familiar with the characteristics and capabilities of large multi-image presentations, and they lack knowledge concerning their utility.

Jonassen (1979) states that it is generally believed that research on multi-image presentation revolves around linear vs. simultaneous presentation factors. Using Perrin's theory, most researchers predict that learning will increase (however it is measured) when "the viewer makes his own montage of different image elements, increasing the probability of learning comparative information" (Perrin, 1,969, p. 369). Jonassen (1979) indicates that the mere presentation of simultaneous images does not necessarily lead to simultaneous mental processing. The view still must provide a cognitive strategy for processing and make sense of the presentation order. Just as linear-sequenced material must be processed based on content and syntactic associations, multi-image presentations must also. Jonassen (1979) found that the literature on multi-image (simultaneous) presentations has yielded contradictory results. He feels that incomplete questions in the research hypothesis were asked instead of just questions about linearity vs. simultaneity. Researchers should consider "how simultaneous images can best be structured to facilitate specific types of learning behavior (p. 292). Jonassen (1979) continues by indicating that proponents have assumed that multi-image presentations are a unique form of communication. Multi-imagery is "not a medium," it is a presentation mode that can manipulate visual perception. Therefore, study on multi-image presentations should be based on established principles of concept learning. To date, little research in this area has been conducted with concept teaching in mind. An exception would be the study conducted by Whitley and Moore (1979) which found significant interactions between a student perceptual type (visual vs. haptics) and presentation mode (linear vs. simultaneous). Haptics scored higher with multi-image presentations. Another exception was completed by Ausburn (1975), which found that both haptics and visuals benefited from multi-image presentations.

Burke and Leps (1989), gleaning information from the limited (and possibly flawed) research on multi-image presentations (see 16.4.2.4), feel that multi-image as a concept offers little to learners to improve cognitive potential or "affective impact." This is due to conceptually weak studies. The limited number of reviews concerning multipleimage research (Allen & Cooney, 1964; Burke, 1987; Burke & Leps, 1989) have revealed few usable results. There is, of course, the seemingly ever-present problem of research design and implications. These basic problems included retention studies comparing single-image and multiple-image presentations that were flawed by the presence of unnecessary recall data in both sound tracks. In addition, "the comparisons were usually of single- and multiple-screen versions of the same material, thereby canceling out Perrin's theoretical call for multi-image to enhance a basic message" (Burke & Leps, 1989, p. 185). Burke and Leps, however, feel that multi-image presentations were given little opportunity to prove themselves due to cost and technical execution of the presentations.