29.3 Multiple-Channel Communication

Of major interest to communication theorists and instructional designers is whether humans can accommodate simultaneous audio and visual stimuli and, if so, the amount and types of information that could be so processed. Multiple-channel communication involves simultaneous presentations of stimuli ". . . through different sensory channels (i.e., sight, sound, touch, etc.) which will provide additional stimuli reinforcement" (Dwyer, 1978, p. 22).

Broadbent (1958) and later Feigenbaum and Simon (1963) espoused the single-channel theory, in which, if information arrives simultaneously in separate channels, information jamming will occur. Broadbent (1958, 1965) suggests that one reason for reduced learning in multiple-channel presentations is a result of the filtering process (bottleneck) occurring in an individual's information-processing system which reduces superfluous elements and permits only essential or basis information to be received; the nervous system acts as a single channel. Similarly, research conducted by Hemandez-Peon (196 1) has led to a hypothesis known as the Hemandez-Peon effect that contends that when information is being processed via one sense, this act may cause an impediment to the processing of a stimuli through other senses. Likewise, Jocobson (1950, 195 1) contended that the brain is able to process only small proportions of the large amounts of stimuli received. Thus, regardless of the amount of information presented in any sensory modality, learners are able to accept only limited amounts in the information-processing center (Atmeave, 1954; Brown, 1959; Dwyer, 1972; Livingstone, 1962). Broadbent (1958) asserts that information-processing of human beings can receive information from only one source at a time; the additional information is temporarily stored (in the sensory register). However, Hartman (1961b) also points out that Broadbent's thesis regarding the filtering of information in the central nervous system is based on data obtained from presenting unrelated information to learners through two or more modalities simultaneously. If after this momentary storage the information is not used, it is not retained. Thus, people viewing multiple-channel presentations are presented with the problem of switching from one channel to another (Broadbent, 1956, 1965). Other researchers, including Shannon and Weaver (1949), Spaulding (1956), and Cherry (1953) support this theory. Corballis and Reaburn (1970), Clark (1969), Herman (1965), and Welford (1968) have documented the reduction (impairment) of the processing of information in multiple-channels communication situations. Travers (1968) concurs in his review of multiple-channel communication. He suggests that there is no convincing evidence that multiple-channel communications were any more effective in producing learning than single-channel inputs. There appears to be major concerns, however, involved in determining the amount of information a human being can process at any one time. Travers (1968) indicates unequivocally that the human processing system is one of limited capacity (see also Miller, 1956). To recognize information simultaneously, the various receptors (eyes, ears) would have to analyze a great variety of different cues. At this initial stage, the system does function as a multiplechannel system. But once recognition has occurred and, hence, attention (see also Norman, 1969), the remainder of Operations on the incoming information is undertaken by a system with a limited capacity, the STM. The system from this Point on operates as a single-channel system, Travers (1968) states, ". . . unless the rate at which the incoming information being received is less than the capacity of the system for handling information. Only under the latter condition can two separate and distinct sequences of messages be received at the same time" (p. 10). Human beings are able to deal with the vast complexities of various types of data from the environment. These data are then simplified to be handled by the perceptual system. Much of the simplification of this huge amount and complex data involves the discarding Of redundant information. This process is referred to as "information compression" (Travers, 1968, p. 11).

It is also related to the information-processing system's strength: gist. Travers's perceptual model thus includes a high-capacity information system up to the point of recognition and a very limited system beyond. Lack of retention and understanding of many multiple-channel presentations are examples of this model in action. Travers's (1964a, 1964b, 1966) studies support this contention that human beings cannot receive more information if exposed to two or more sources simultaneously than if exposed to just one source, or if the information is transmitted by two different modalities. Van Mondfrans (1963), in a study using nonsense syllables and words, showed no advantage for an audiovisual presentation over presentations via audio and visual modalities alone. Cherry (1953) concluded that the utilization of information by the brain can be represented by a single-channel input. Travers (1968) continues and states that since the perceptual channel is very limited, we must assume that the receiver (learner) cannot process multiple-channel inputs as efficiently as "designers of audiovisual materials have commonly assume&'( p. 10).

Other researchers have supported the efficacy of singlechannel presentations. These include Fleming (1970) who reviewed research studies dealing with single- and multiple-channel presentations and noted the possibility that many instructional programs are already "perceptually overloaded." He suggests that additional "jamming" of the perceivers' senses through multiple-media (channel) may have negative results. Fleming suggests that the only possible instructional situation where "stepped-up sensory environments" are useful is when the desire is to "overwhelm, impress, or to exhilarate." Hartman (1961a) concludes that multiple-channel presentations do not produce increases in learning (however defined) over single-channel communication, unless the situation in which the learning takes place also contains the necessary additional cues. Hartman (1961b) has also expressed concern about the act of increasing the number of cues and/or the number of channels used with the expectation that more learning will occur. He states:

A common practice among multiple-channel communicators has been to fill the channels, especially the pictorial, with as much information as possible. The obvious expectation is for additional communication to result from the additional information. However, the probability of interference resulting from the additional cues is very high. The hoped-for enhanced communication resulting from a summation of cues occurs only under special conditions. Most of the added cues in the mass media possess a large number of extraneous cognitive associations. The possibility that these associations will interfere with one another is probably greater than that they will facilitate learning (p. 255).

Hsia (1971) drew several conclusions from an extensive review of literature comparing multiple- and single-channel presentations. These include: (1) Human information-processing functions as a multiple-channel system until the capacity of the system is overloaded; (2) when input becomes greater than the system's capacity, the system reverts to a single-channel system; and (3) an increase in the amount of information presented does not necessarily increase the rate of information transmission. Hsia (1971) asserts that, since all incoming information needs to be coded prior to being processed by the human processing system it would seem reasonable that all extraneous, irrelevant, and superfluous information be eliminated or reduced at that time. Hsia (197 1) contends that by reducing this "extra7' information, the learner is spared from having to discriminate the relevant from the irrelevant. In addition to filtering information, a large part of redundancy and noise are eliminated. Hsia (1971) and Carpenter (1953) feel that the processing capability of an individual is limited by the physiological aspects of the individual. A person can receive far more stimuli than he or she can effectively process. Clark (1969), Corballis and Reabum (1970), Herman (1965), and Welford (1968) indicate that there are a substantial number of research results that support the position that single-channel communication can be as effective as multiple-channel orientations. Dwyer (1978) cites approximately 50 studies in which the contention that additional cues----~'provided by the use of two or more information channels simultaneously--or excessive realistic cues within a single-channel may be distracting or even evoke responses in opposition to the desired types of learning" (pp. 29, 30).

There is also much criticism of the research which supports the single-channel view. For example, Norberg (1966) takes Travers to task for basing his assumption concerning single-channel communication on experiments using verbal material in both auditory and visual channels (i.e., no pictures presented). Norberg (1966) explains that Travers's studies:

... deal exclusively with verbal symbols, whereas most two-channel presentations actually used in instructional situations typically combine nonverbal signs in the visual channel with verbal auditory stimuli.... But it is still necessary to distinguish carefully between the actual experimental findings and theoretical statements regarding nonverbal "realistic " stimuli which have not entered into the experimental work cited.... It is one thing to say that the "density" of information in stimulus materials presented to the leaner may become a factor impeding efficient transmission; i.e., some presentations may be too realistic (p. 307).

Other criticisms of single-channel research are that much of the data collected were from studies where unrelated and/or contradictory stimuli was presented to the learners simultaneously. It would seem reasonable under these circumstances that a person would attend to one stimulus (message) and not the other. The following section looks at multiple-channel communication and the influences of the cue summation theories.