11.3 Message Design and Cognitive Processing

The vast majority of early instructional films and television programs were essentially documentary works that were developed by commercial, noneducational producers. At this early point in the evolution of instructional technology, little attention was given to the use of instructional techniques or design principles. Similarly, the technology of film or television was still in its infancy, and few, if any, editing or special visual effects were available to the producers of such materials. In light of this, it is not surprising that most of the earlier research focused on simple comparisons between the technology and some form of standard instruction. Since the two technologies did not incorporate many of the production elements that have become part of their unique symbol system as we understand them today, little attention was given to assessing the effects of specific media characteristics on student learning.

11.3.1 The Evolution of Message Design

During the period of the Pennsylvania State University film studies, however, some research was directed at determining how the intentional incorporation of instructional techniques and media characteristics interacted with learner achievement from the materials. The variables studied included: the use of inserted questions, color, subjective camera angle, sound track modifications, and the use of visual cueing devices (Greenhill, 1956; Hoban & van Ormer, 1951). Similar studies were further conducted on instructional television in the late 1950s and early 1960s, generally on adult audiences in controlled environments (Chu & Schramm, 1967; Greenhill, 1967). From this time on, a growing number of researchers have investigated, in increasingly greater levels of detail, the instructional effcctiveness of television productions incorporating specialized features that are intended to facilitate learning. The process of specifying and organizing these components has come to be called message design (see 26.1).

Fleming and Levie (1993) define an instructional message as "a pattern of signs (words, pictures, gestures) produced for the purpose of modifying the psychomotor, cognitive, or affective behavior of one or more persons" (p. x). Grabowski (1991) describes message design as "planning for the manipulation of the physical form of the message" (p. 206). The concept of message design was not used in the literature until the 1970s, although the general principles of message design were being synthesized from research on perception, psychology, and instruction. Early researchers focused primarily on visual perception (Norberg, 1962, 1966; Knowlton, 1966; Fleming, 1967); however, later researchers addressed auditory and print media as well. Fleming and Levie (1978, 1993) first defined the term message design and comprehensively articulated its general principles for instructional designers. Today, the concept of message design in television includes all of the scripting, production, and editing decisions that are made separate from the actual content of the program.

The design of the instructional television message (see 11.7.3.4) has become increasingly important as a greater understanding of instructional and cognitive principles has emerged from the study of learning and psychology, and with the growing sophistication of television production technology, particularly in broadcast television. The intentional use of various video effects such as zooms, cuts, dissolves, and the designer's manipulation of program pacing and use of various audio and graphic effects became a standard procedure among instructional designers wishing to maximize the effectiveness of television programming. For the most part, however, these production effects were not systematically investigated, and, consequently, the television producer had few reliable research guidelines on which to base production decisions.

During the nidd-1970s, Aletha Huston and John Wright used the term formal features to collectively describe the various production techniques employed in designing and producing the television message (Huston & Wright, 1983). They describe television as being distinguished by its unique forms, rather than simply by the content of the programming. These researchers and their associates at the University of Kansas began a systematic investigation of the formal attributes or features of television, particularly with respect to how these techniques interact with cognitive processes, such as attention and comprehension (Rice, Huston & Wright, 1982).

By the late 1970s, much of the television research focused on how children view television and those processes that relate to attention and comprehension of the televised information. This era of research can be best characterized as the conjunction of interest in both the developmental aspects of learning and in cognitive processing of information. TWo events in the area of children's television prompted this research: the initial success of Sesame Street and associated programming by the Children's Television Workshop (Mielke, 1990), and the increased criticism of television and its alleged negative effects by a number of popular writers (Mander, 1978; Postman, 1982; Winn, 1977).

With the advent of The Children's Television Workshop and Sesame Street, a number of researchers began to explore the value of using many of these production techniques. These studies were typically formative in nature, intended for in-house use to assess the adequacy of particular techniques, and consequently did not appear regularly in the research literature (Sammur, 1990). Thus, researchers began to focus on those unique features that promote children's attention and comprehension during television viewing. In this research, the cognitive effects of formal features such as pacing, audio cues, camera effects, anima-tion, and editing techniques were also explored with regard to the role that they played in attention and comprehension (Meyer, 1983).

During this time, public interest was also drawn to the possible negative effects of television programming on children. in addition to the continuing public concern for the effects of television violence on children, interest increased into the possibly debilitating effects on children's cognitive-processing abilities. Iii her book, 7he Plug-in Drug, Winn (1977) charged that television and the formal features inherent in the programming were causing excessive cognitive passivity and depressed processing capabilities.

Organized research, which was prompted by these events and criticisms, investigated the general effects on both attention and comprehension, as well as on the specific effects of television's formal production features in a fairly comprehensive manner. Such research has given us a remarkably thorough understanding of how television promotes cognitive activities (Anderson & Collins, 1988).

As interest in the cognitive aspects of children's television grew, hypotheses were developed to account for these effects in a broad manner, irrespective of particular types of programming. While a number of these theoretical perspectives are unconfirmed, they have provided the impetus and base for substantial, systematic research.

11.3.2 The Effects of Television on Cognitive-Processing Abilities

Television has been both lauded and criticized for the ways in which it presents information to the viewer, irrespective of the information itself (Anderson & Collins, 1988). It is this area, that of the relationship between the ways in which information is presented on television and the effect of that presentation on the cognitive-processing abilities of the viewer, which has continued to attract a great deal of theoretical as well as supporting research interest (Huston et al., 1992).

11-3.2.1. Theoretical Orientations. One critical view that has persisted over the years, despite contrary research findings, is that the television image and associated presentation effects are cognitively debilitating (Mander, 1978; Winn, 1977). The central assertion of this viewpoint is that the rapidly changing television image-enhanced by production features such as cuts, zooms, animation, and special effects-is cognitively mesmerizing. This is hypothesized to result in cognitive passivity, shortened attention spans, and, paradoxically enough, hyperactive behavior (Winn, 1977; Dumont, 1976, cited in Winn, 1977). Such a view is more conjecture than substantiated fact or articulated theory and has been drawn substantially from subjective observation rather than from extensive empirical research. The -notion, however, has appealed to many who associate these behavioral manifestations with general, adult entertainment forms of television and who are More critical of the content of television. programming rather than the presentation formats. It should be noted that most researchers in the area of cognitive science and educational technology have not supported these assertions, which remain, to a large degree, open to definitive and methodologicaUy rigorous research (Anderson & Collins, 1988).

11.3.2.2. Empirical Research. For the most part, research related to this aspect of television effects has been drawn from studies done in the area of advertising and marketing or in electroencephalography (EEG). Krugman (1970, 1971) compared the EEGs of subjects viewing rearprojected visual images and those of subjects reading, and concluded that television viewing resulted in different brain wave patterns than did reading. It is important to note that these studies were conducted on a single subject and only used the subject's EEG obtained while browsing a magazine as a baseline index. The length of time the EEG was recorded was also only 15 minutes, and readings were taken at only one location on the head. The two brain wave patterns of interest were the alpha rhythm, which is associated with an inactive or resting-brain state, and the beta rhythm, which is usually indicative of cognitive activity. These experiments were repeated by Krugman, using actual television images with similar results (Krugman, 1979). Similar findings were produced by several other researchers who indicated that television viewing produced more alpha activity than reading, which resulted in greater beta activity (Appel, Weinstein & Weinstein, 1979; Fetherman, Frieser, Greenspun, Harris, Schulman & Crown, 1979; Walker, 1980; Weinstein, Appel & Weinstein, 1980). In these cases, alpha activity was associated with periods of low cognitive activity, which was interpreted to be the mesmerizing effect described by critics.

Drawing from the work of Krugman (1979), Emery and Emery (1975, 1980) criticized television images as "habituating" because the continuously scanned image emitted an overload of light-based information, potentially resulting in an overload of the processing system. This claim was substantially refuted, however, in studies by Silberstein, Agardy, Ong, and Heath (1983), who, in methodologically rigorous experiments with 12-year-old children, found no differences in brain wave activity between projected text and text presented on the television screen. Furthermore, differences were found between text presented on the television screen and documentary or interview programming; whereas no differences were found between the two types of programming. A third interesting finding was that both the text and interview program produced right- and lefthemisphere effects, while the documentary alone resulted in greater right-hemisphere activity. A comprehensive and critical review of most of the EEG research was published by Fite (1994). In this report, Fite found virtually no substantiation of the detrimental effects of television evidenced by EEG-based studies.

Focusing specifically on viewer attention, Rothshild, Thorson, Reeves, Hirsch, and Goldstein (1986) found that alpha activity dropped immediately following the introduction of a scene change or formal. feature in the program material, which in these studies were commercial advertisements. Winn (1977) has further criticized children's television and Sesame Street, in particular, for contributing to shortened attention spans and hyperactive behavior. A study by Halpern (1975) has been frequently cited as providing evidence that programming such as Sesame Street contributed to hyperactive and compulsive behavior. This study has been seriously criticized on methodological grounds by Anderson and Collins (1988), and the findings have not been successfully replicated by Halpern. Other studies related to children's concentration and tolerance for delay reported moderate decreases in tolerance for delay associated with action programs (Friedrich & Stein, 1973) and actually increased concentration resulting from television viewing among children rated as low in imagination (Tower, Singer, Singer & Biggs, 1979). Anderson, Levin, and Lorch (1977) investigated the effect of program pacing on attention, activity, and impulsivity levels and found no differences in 5-year-old children's degree of activity, impulsivity, or perseverance levels. Salomon (1979), however, found that Sesame Street viewing, when compared to other general types of children's programming, produced a decrease in perseverance in a laboratory task. This effect may have been related to differences in the audience's age and the intended target age of the Sesame Street programming and the relative ease of the task.

11.3.3 The Television Symbol System or Code

For the most part, research into the cognitive effects of television has focused more specifically on how televised information is processed rather than on how television affects cognitive processing abilities (Anderson & Collins, 1988). This research is based on theory related to both the symbol system or formal features used in television and the ways that information is attended to and comprehended (see 26.4.3).

11.3.3.1. The Role of Filmic Codes in Processing. One of the most universal views of television as a medium was described by McLuhan (1964) when he suggested that the formal attributes of a medium, such as television, influence how we think and process information. Furthermore, McLuhan put forth the idea that different media present information in unique ways that are idiosyncratic to the individual medium. Goodman (1968) and Gardner, Howard, and Perkins (1974) furtlier elaborated on the function of such symbol systems, implying that similarities between the symbol system and mental representations of the content will facilitate comprehension of the instructional message. More recently, Kozma (1991) suggests that different media are defined by three characteristics: the technology, the symbol systems employed, and the methods of processing information. Of these, the symbol system is crucial to the mental processing of the person interacting with the medium. The individual symbol systems may be idiosyncratic to the particular medium and consequently may need to be learned by the user. This thesis has been elaborated on by Gavriel Salomon, who has attempted to test it empirically with regard to television (Salomon, 1972, 1974; Salomon, 1979; Salomon & Cohen, 1977). He suggested that different symbol systems or codes can represent information in different ways during encoding in memory, making it necessary to process the information in unique ways. Salomon contended that children learn to interpret these "filmic codes," which can be incorporated into cognitive activities in two ways (Salomon, 1979). The first function of symbolic or filmic codes is that they can call on or activate cognitive skills within the learner and can become internalized into the learner's repertoire of processing skills (Salomon & Cohen, 1977). In this way, such production features as montage or cuts can activate respective cognitive processes such as inferencing and sequencing. The second role of filmic codes lies in the assumption that these codes, which model cognitive processes, can actually "stand in" for or "supplant" the cognitive skills themselves, thereby facilitating learning (Salomon, 1974). In this manner, features such as zooms and dissolves can be used to model the cognitive skills they represent and consequently enhance the processing skills of the viewer.

Rice, Huston, and Wright (1983) further differentiated the types of representation within the television code into three levels. These include at the most basic level, literal visual or auditory portrayal of real-world information. At the second level are media forms and conventions that have no real-world counterpart, such as production effects and formal features. The third level consists of symbolic code that is not distinctive to the television medium. These thirdlevel codes consist of linguistic, nonlinguistic, and auditory code such as language, which may be used to "double encode" or describe the visual codes presented on the screen. Of the three, the media forms and conventions are of most interest to the researcher because they are idiosyncratic to the media of television and film and relate most specifically to the child's processing of the television message (Rice, Huston & Wright, 1983).

11.3.3.2. Research on Filmic Codes. There is not a great deal of empirical work related to the cognitive effects of television code. However, the work of Gavriel Salomon constitutes the most comprehensive series of empirical studies focused on the symbol system and code of television. Drawing on his theoretical position, he devised a series of experiments that explored the use of fihnic codes to both model or supplant cognitive skills and to call on or activate specific cognitive skills. He conducted the first group of studies with Israeli eighth-graders to determine if the camera effect of zooming could indeed model the relation of the part to the whole (Salomon, 1974). The results indicated that the children exposed to the experimental treatment performed significantly better than did those students either shown the individual close-up and overall pictures or those receiving no treatment. In this case, the use of explicit modeling of the cognitive skill improved the student's ability to focus attention on the detailed parts of the overall display. A second experiment, using fewer visual transformations, was not as effective as the first, possibly indicating that extensive modeling of these skills is necessary for this effect to occur. In a third experiment, Salomon confirmed that the internalization of the filmic codes could enhance the cognitive skills of the viewer by presenting scenes where the three-dimensional unfolding of an object was compared with the same representation in two dimensions. In this case, the three-dimensional animation effect modeled the cognitive analog of mentally unfolding the object from three dimensions to two dimensions more effectively than did simple presentation of the two-dimensional object. A study conducted by Rovet (1983) using a spatial rotation task with third-grade children further confinned Salomon's findings, although conclusive confirmation of this theory has not been provided through research.

The second assertion made by Salomon suggested that filmic, codes could also activate or "call upon" specific cognitive skills. In a series of studies, Salomon (1979) tested this hypothesis on groups of preschool, second-grade, and third-grade Israeli students using Sesame Street programming as the content. After 6 months, the groups of school-aged children demonstrated significantly higher comprehension scores. This was interpreted by Salomon to indicate that students were able to learn the meanings of the fihnic codes, and in so doing activated the respective cognitive skills. However, the effects were limited to the older children and have been qualified by Salomon to suggest that these mental skills can be activated by the appropriate filmic codes, but are not necessarily always activated in this manner.

11. 3.4 Children's Attention to Television

The effect of the television symbol system on learning has been addressed through two areas of cognitive processing: attention and comprehension. For each of these areas, we discuss theoretical approaches and empirical research.

11.3.4.1. Reactive/Active Theory. Two approaches to understanding the way in which children attend to television have emerged. These positions include the reactive theory, which generally views the child as passive and simply a receptor of information or stimuli delivered by the television, and the active theory,, which suggests that children cognitively interact with the information being presented as well as with the viewing environment (Anderson & Lorch, 1983). These two viewpoints generally parallel theoretical orientations to human information processing in that early concepts of the human information-processing system were reasonably linear and viewed attention as a relatively receptive process where the learner merely reacted to stimuli that were perceived (Atkinson & Shiffrin, 1968). Later conceptions of how we process information took the position that we are active participants in selecting and processing incoming stimuli (Anderson, 1980).

The first theoretical orientation, the reactive theory, is derived from Bandura's Social Learning Theory (Bandura, 1977). In this conceptualization, the salient formal features of the. television programming gain and maintain the viewer's attention. Continued attention and comprehension occur more or less automatically as the child's informationprocessing system functions reactively. Singer (1980) describes this process as one where the continually changing screen and auditory patterns create an ongoing series of orienting reflexes in the viewer. Key to this orientation is the role of the viewer as a passive, involuntary processor of information that is absorbed from the screen. The reactive theory of attention to television is supported by little direct research, with most of the foundation for the theory being based on the early human information-processing theories such as those described by Atkinson and Shiffrin (1968, 1971), Broadbent (1959), and Neisser (1967). The work of Singer (1980) included little direct research relative to this perspective, but rather drew on what was, at that time, a popular theory of memory that described the human information-processing system as one in which information was processed in the sensory store, received further processing in short term memory, and was then transferred to long-term memory, all without a great deal of active or purposeful selection, processing, or coding by the learner.

It is generally accepted today that the reactive theory requires much revision, particularly with regard to the learner's role in initiating and actively processing new information in relation to prior knowledge. For these reasons, little substantiation of the theory can be put forth, especially in light of the supporl that current research provides to the opposing theory, the active theory.

The alternative theory, the active theory, defines the child as an active processor who is guided by previous knowledge, expectations, and schemata (Anderson & Lorch, 1983). In this way, the child does not merely respond to the changing stimuli presented, but rather actively applies strategies based on previous experience with the content and formal features, personal knowledge structures, and available cognitive skills. Key to this view is the assumption that the child will apply existing schemas to the perception and processing of the televised information. Anderson and Lorch (1983) suggest that a number of premises underlie the functioning of the active theory. These include consideration of competing stimuli, the need to maintain a reasonable level of stimulus unfamiliarity, the role of auditory cues to refocus attention, and the effect of attentional inertia to maintain cognitive involvement (Anderson, Alwitt, Lorch & Levin, 1979). Additionally, a key component of the active theory is the role of viewing schemata, which Anderson and Lorch suggest develop through increased interaction with television forms, as well as with general cognitive growth.

'Me notion of representational codes or formal features and their role and effects in the processing of television information has become an area of particular interest and the central focus of much research regarding how children attend to and process the television message. Formal fea tures are defined by Anderson and Collins as characteristic attributes of the medium, which can be described without reference to specific content. In reality these include, but are not limited to, the visual features of zooms, camera movements, cuts and dissolves, montage techniques, animation, ellipses, program pace, and special visual effects, as well as the auditory features of music, sound effects, and unusual voices. A fairly comprehensive taxonomy of formal features has been developed by the research group at the Center for Research on the Influence of Television on Children (CRITC) (Huston & Wright, 1983; Rice, Huston & Wright, 1983).

Two constructs related to the visual message and the forms of television have emerged and become important to an understanding of how these forms function in the processing of the television message. These constructs, which include visual complexity or the amount and degree of change of information (Watt & Welch, 1983; Welch & Watt, 1982) and perceptual salience or those attributes of the stimulus that increase its intensity, contrast, change, or novelty (Berlyne, 1960; Rice, Huston & Wright, 1983), relate to both quantitative and qualitative characteristics of the message. Researchers associated with each of these positions have developed or adapted models that can be used to conceptualize the effects of these attributes on the message and how it is processed by the viewer. Watt and Welch employed an information theory model for entropy to explain the relationship between static and dynamic complexity and learning from television content (Watt & Welch, 1983; Welch & Watt, 1982). Rice, Huston, and Wright (1982) presented a model that described the relationship between attention and stimulus complexity. For the most part, however, the effects of the formal features of television have been considered with regard to the particular cognitive processes or skills with which they are associated, attention and comprehension, and consequently, they are best examined from that perspective.

11.3.4.2. Research on Attention. The variable of attention to the television program has received extensive research interest, of which the most comprehensive group of studies has been conducted by Daniel Anderson and his associates at the University of Massachusetts. This group of researchers was the first to propose that the process of attending to television programming was active rather than simply a reaction to the stimuli presented.

One of the first questions relative to attention to television is a qualification of exactly what attention is and how it can appropriately be measured. Anderson and Field (1983) describe five methodologies that may constitute an effective measure of this attention. These include: (a) visual orientation, the physical orientation of the viewer toward the television screen; (b) eye movements and fixations; (c) comprehension and recognition testing, which measures attention through inferences drawn from objective recognition and comprehension tests; (d) interference methods, which pinpoint attention as that time when a viewer responds to and removes some form of interfering information from the message; and (e) physiological measures that include cardiac, galvanic skin response, and electroencephalographic records of arousal. Of these, the most frequently employed have been visual orientation and the use of recognition and comprehension tests.

Anderson and Field (1983) identify a number of settings and contexts for viewing that impinge on the attentional process. They differentiate between the bome-viewing environment and laboratory settings in terms of the accuracy of data obtained. Home viewing generally results in overly inflated estimates of attentional time (Bechtel, Achelpobl & Akers, 1972). The use of monitoring cameras revealed that attention does not continue for long periods of time but rather consists of frequent interruptions, conversations, distractions, and the viewer's exits and returns to the room (Allen, 1965; Anderson, 1983). Allen used time-lapse movie cameras, and Bechtel, Achelpohl, and Akers videotaped in the home. The results of these studies appear consistent, indicating that children up to age 10 averaged about 52% of the time in the viewing room actually attending to the program, while - children aged I I to 19 years showed an average attention of about 69% (Bechtel et al., 1972). In all cases, attention to children's programs was substantially higher than to adult-level programming, although this may not remain true today because of changes in programming and the increased viewing sophistication of children. In laboratory settings, where more control over outside distractions could be maintained, it was found that children still were frequently distracted and demonstrated only sporadic attention to the program (Becker & Wolfe, 1960). In several studies, preschool children were observed to look at and away from the television 150 to 200 times per hour (Alwitt, Anderson, Lorch & Levin, 1980; Anderson & Levin, 1976; Field, 1083). The length of "looks" were also seen as important characteristics of attention. Anderson, Lorch, Smith, Bradford, and Levin (1981) found that looks of more than 30 seconds were infrequent and that the majority of look lengths were less than 5 seconds.

The viewing context was also identified as an influential factor in attention. Sproull (1973) suggested that toys and other activities were strong attention-diverting stimuli, in the absence of which attention rose to 80%. Studies by Lorch, Anderson, and Levin (1979) concluded that attention is strategic in children, because audio cues were used heavily to monitor program content and indicate instances when attention should be redirected to the television. The presence of other children with whom they could discuss the program and use as models of attention were also shown to be strong factors contributing to attentional control (Anderson et al., 198 1).

The factor of viewer age has frequently emerged as a variable of significance, particularly with regard to determining at what age children begin to attend to and comprehend the content of the television program. Very young children (6 to 12 months of age) appear to direct attention to the television screen about half the time in controlled situations (Hollenbeck & Slaby, 1979; Lemish & Rice, 1986), with a dramatic increase between 12 and 48 months (Anderson & Levin, 1976). In their study, Anderson and Levin observed an increase in look lengths by a factor of 4 at approximately 30 months of age. Other researchers have reported similar findings (Carew, 1980; Schramm, Lyle & Parker, 1961).

Attention appears to increase continuously beyond this age to about 12 years, at which point it plateaus (Alwitt et al., 1980; Anderson, 1983; Anderson, Lorch, Field, Collins & Nathan, 1986; Anderson, Lorch, Field & Sanders, 1981; Calvert, Huston, Watkins & Wright, 1982).

The unique role of the formal features of television has been the focus of much research on children's attention. Such features include both visual and auditory production effects that are integral to the television program composition and presentation. Formal features have significant implications for attention, comprehension, and, as has been discussed previously, modeling and activating cognitive skills. In terms of attention, the research has indicated that only some formal features, specifically special visual effects, changes in scene, character change, and high levels of action, are reasonably effective at eliciting attention, while conventional camera effects such as cuts, zooms, and pans have substantially less power to gain attention (Rice, Huston & Wright, 1983). The visual feature that most inhibited attention was the long zoom effect. Other prograin components, such as animation, puppets, and frequent changes of speaker, while not actually production features, were also found to promote attention. Those components that decreased attention were live animals, song and dance, and long speeches (Alwitt et al., 1980; Anderson & Levin, 1976; Calvert, Huston, Watkins & Wright, 1982).

Several researchers have observed that the sound track of the television program plays a major role in attention, particularly in gaining the attention of the nonviewing child (Anderson &. Field, 1983). With respect to the generalized use or effect of the audio track to direct attention, Lorch et al. (1979) found that auditory attention parallels visual attention and increases with age at a rate similar to that of visual attention. When the audio message was experimentally degraded so as to be unintelligible, either through technical reversal or substitution, children at ages 2, 3 112, and 5 years evidenced significant drops in attention to Sesame Street programs, with the most significant drop being observed with the older children (Anderson, Lorch, Field & Sanders, 1981). It has also been reported that children employ the audio message to monitor the program for critical or comprehensible content, which they can then attend to visually (Anderson & Lorch, 1983). Auditory attention to television is, to a large degree, mediated by the formal attributes of the auditory message, including type, age, and gender of voice, and the novelty of particular sound, sound effects, or music. Research conducted by Alwitt et al. (1980) revealed that certain audio effects were effective in gaining attention from nonviewing children. These included auditory changes, sound effects, laughter, instrumental music, and children's, women's, and "peculiar" voices; while men's voices, individual singing, and slow music inhibited attention (Anderson & Lorch, 1983). The researchers concluded that auditory devices such as those described cued the children that an important change was taking place in the program which niight be of interest, thereby prompting attention. They also reported that audio effects do not appear to have any significant effect before the age of 24 to 30 months, which parallels approximately the beginning of general attending behavior noted previously.

When all types of formal features, both visual and auditory, are considered in terms of their ability to facilitate attention, it becomes apparent that those which are most obvious are generally most effective (Wright, Huston, Ross, Calvert, Rolandelli, Weeks, Raeissi & Potts, 1984). These researchers contend that the more perceptually salient a feature is, such as fast action or pace, the more effectively it will gain attention. This was partially confirmed in research they describe in which those programs identified as high in feature saliency also had larger viewing audiences. Interestingly, Sesame Street, which has a high viewership and attention-gaining-power, has been found to be~ slower paced (in terms of shot length) than other entertainment programs (Bryant, 1992). Evidence was also found which suggests that violence per se is not necessarily attention gaining, but rather the high saliency of formal features in violent programs may be responsible for the higher viewer numbers (Huston & Wright, 1983; Wright et al., 1984; Wright & Huston, 1982).

The differential effects of both visual and auditory formal features have been cited by several researchers as significant evidence supporting the active theory of attention to television (Anderson & Field, 1983; Rice, Huston & Wright). They contend that for the reactive theory to be an apt descriptor of children's attentional behavior, all formal features should be effective at virtually all ages, because they should all automatically elicit an orienting reaction due to their movement, stimulus change, or salient visual patterns. Since the research consistently identifies only certain features at particular ages as attention gaining and conversely finds that other features are inhibiting to attention, this hypothesis is strongly rejected (Anderson & Field, 1983). With regard to the active theory, they describe the viewing child as actively and selectively in command of his or her own attentional strategies. For this reason, the child could be expected to respond differentially to the various stimuli and features, which is the case made by current research findings (Hawkins, Kin & Pingree, 1991). Alwitt et al. (1980) conclude:

An attribute (feature) comes to have a positive or negative relationship to attention, we hypothesize, based on the degree to which it predicts relevant and comprehensible content. A child can thus use an attribute to divide attention between TV viewing and other activities: Full attention is given when an attribute is predictive of understandable content and terminated when an attribute predicts irrelevant, boring, and incomprehensible content (p. 65).

11.3.5 Children's Comprehension of Television

Anderson and Field (1983) explain that formal features perform two significant functions: First, they mark the beginning of important content segments, and second, they communicate producer-intended concepts of time, space, action, and character (Anderson & Field, 1983). The notion that the formal features, which comprise such television effects as montage, are able to convey changes in time, place, or movement is integral to a viewer's ability to comprehend story content and plot as well as simply to gain or hold attention. It is in the area of comprehension that formal features appear to play the most important role.

11.3.5.1. Relationship of Comprehension to Attention. The basic theory related to children's comprehension of television relates to and derives from theoretical bases for attention (Anderson & Lorch, 1983). They cite the reactive theory for suggesting that once attention has been gained, comprehension will automatically follow as a natural consequence. Interestingly, Singer (1980) and Singer and Singer (1982), proponents of the reactive theory, suggest that the rapid pace or delivery of most television messages that gain or hold attention, may not permit the viewer to process adequately the information at a deep enough level to ensure high levels of comprehension. The active theory, on the other hand, maintains that attention itself is directed by children's monitoring of the program for comprehensible content, which serves as a signal to focus more direct attention to the message (Anderson & Lorch, 1983). To represent the relationship, Rice, Huston, and Wright (1982) offered the attentional. model presented in Figure 11-1. In this model, both high and low levels of comprehensibility inhibit attention. At the high end (incomprehensibility), the content is complex and not understood by the child and consequently elicits little interest or attention. At the low end (boredom), the content is familiar and lacking in information, making it less attention gaining. In this way, comprehension is interpreted to drive attention (Rice, Huston & Wright, 1983).

A good deal of the theory related to the formal features of television has relevance for~ the area of comprehension as well as attention. Of particular interest is the concept of montage, one of the formal features previously described. A montage is a series of scenes interrupted by special effects such as cuts, dissolves, changes in point of view, and overlays, the purpose of which is to show various shifts in time, place, or personal point of view. Such actions call on the viewer to maintain a sequence of events, infer changes of scene or time, and to relate or integrate individual scenes to one another (Anderson & Field, 1983). In this way, any two scenes can be joined together to generate a new idea or suggest a relationship that has not been explicitly shown.

Figure 11-1. A model of developmental changes in interest and attention. (From Rice, Huston & Wright, 1982.)

Piaget (1926) suggested that younger children (under 7 years) were limited in story comprehension because of weak seriation abilities and the inability to infer and comprehend transformations between events in a story which differ temporally. these limitations reduce the ability to develop complete schernas and consequently impair comprehension. Inconsistencies across theories such as these, however, have produced a dilemma among researchers concerning the ability of children to comprehend fully information presented in this manner via television (Wartella, 1979).

113.5.2. Research on Comprehension. Substantial research has addressed the interrelationship between comprehension and attention and the resultant support of the active theory suggested by Anderson and Lorch (1983). Lorch et al. (1979) compared different experimental attention situations in terms of recall of Sesame Street content by 5-year-olds. Their findings revealed that variations in the amount of attention a child demonstrated did not differentially affect comprehension scores. However, a significant positive correlation was found between the comprehension scores and the amount of attention exhibited during the specific program content that was related to the comprehension test items. These findings were further supported in research reported by Krull and Husson (1979) and Pezdek and Hartman (1981) who also identified the significance of audio cues in promoting comprehension as well as directing visual attention. A later study, however, by Anderson, Lorch, Field, and Sanders (1981, study 2), which controlled for extraneous confounding effects of formal features inserted in the programs, produced data that fully supported the earlier findings of Lorch et al. (1979). All in all, these studies provided strong support for the active theory over the reactive theory, in that attention appeared to be significantly directed by the comprehensibility of the program content.

Understandably, the role of formal features in comprehension is directly related to the active theory of television viewing. Anderson and Field (1983) suggest that the employment of formal features in a montage serve the purposes of the producers of the program to convey or infer changes in time, space, action, or point of view. They further contend that the active comprehension hypothesis is consequently supported, in that if children did not actively make the inferences, they would perceive the program as meaningless segments of video and would, therefore, not attend to it.

The earliest research on comprehension of film montage suggested that young children were incapable of comprehending the elements of montage (Piaget, 1926). Empirical research supported these contentions (Baron, 1980; Noble, 1975; Thda, 1969). In these cases, assessment of children's comprehension was made via verbal explanations of what had occurred, a process that has been criticized as being extremely difficult for younger children (Smith, Anderson & Fisher, 1985). In research that employed nonverbal testing methods such as reconstructing the story using dolls and the original television sets, these researchers found that children aged 3 and 5 years showed substantial comprehension of program content. It is interesting to note that no differences were found between treatments that employed the formal features of pans, zooms, fades, and dissolves and those treatments that relied solely on stillphotographic representation. Montage that incorporated formal features was apparently not necessary for comprehension of the story. Rather, children were able to comprehend the message presented via either montage or still pictures with equal ability. In a second experiment, Smith et al. (1985) examined the effects of specific montage elements in terms of the outcomes (ellipsis, spatial relationships, character point of view, and simultaneity of action) intended by the producer. In this case, both 4- and 7-year-olds demonstrated good comprehension via the nonverbal evaluation technique, with 7-year-old children showing greater comprehension. The researchers attribute this result to a greater amount of life experience on the part of the older children. A later study conducted by Huston and Wright (1989) indicated that formal features used in montage, such as those used to depict distorted perceptions, memory flashbacks, and instant replays were not comprehended well by school age children. Anderson and Collins (1988) have generally concluded that the features incorporated in montage are well comprehended by children, particularly those who are older and have greater prior experience and knowledge. Anderson and Field (1983) contend that the results of these studies indicate that young children make frequent, active inferences as they interpret montage effects in television programming. Furthermore, they suggest that this fact provides strong support for the active-comprehension hypothesis.

The comprehension of longer segments of programming that necessitated integration and inferencing skills was investigated by Lorch, Bellack, and Augsbach (1987). In two experiments, they determined that both 5-year-olds and 4- and 6-year-olds were capable of selectively recalling 92% of ideas that were central to the television stories. Much lower recall rates were found for incidental or noncentral information. In an earlier study, however, Calvert, Huston, Watkins, and Wright (1982) found that children recalled central content that was presented by means of highly salient formal features better than that which used low-salience features. In studies in which the programming content was of much longer duration, such as in commercially broadcast programs, older viewers were generally able to discriminate central content better than younger viewers (Collins, 1983). Collins further suggested that an inability to make inferences contributed to comprehension difficulties, although this research was conducted using entertainment programming that was intended primarily for adult audiences. Anderson and Collins (1988) concluded, however, that the poor comprehension of both central and implied content should be attributed primarily to lessdeveloped knowledge bases rather than to any cognitive disability. More recent research (Sell, Ray & Lovelace, 1995) suggests, however, that repeated viewing of the program results in improved comprehension by 4-year-old children. They attribute this effect to more complete processing of the formal features that enabled children to focus on essential inforination critical to understanding the plot.

11.3.6 Summary and Recommendations

Two theoretical orientations have emerged with regard to the cognitive processing of television program content and the effect of the formal production features on that processing. The earlier, reactive theory suggested that the child was a passive entity that could only react to the stimuli being presented. A number of writers accepted this theory and employed it to further describe the viewer as not only passive but also mesmerized by the flickering stimuli presented on the screen. Only modest data, however, reflect a negative effect of certain types of television programming on attention and cognitive processing, and virtually no reliable research confirms the strong, deleterious effects claimed by a number of popular writers and critics of television.

A second opposing position, the active theory (or the active comprehension theory), drew on more contemporary cognitive views of the learner and described the child as actively exploring and analyzing the program content being presented. This theory suggests that attention to the television program is not a reaction to stimuli but rather a monitoring and comprehension process to identify meaningful content requiring more directed attention.

Research has generally supported the active hypothesis, describing the attentional and comprehension processes as highly interrelated, with comprehension being a precondition to attention. Comprehension is further facilitated through the effects of formal features that function as elements of montage to infer meaningful changes in space, time, and point of view.

The television image has been shown to incorporate a unique symbol system that has certain specifiable capabilities it shares with no other medium. The modes of symbolic representation in television exist as a singular language that must be learned by the child. The specific effects of formal features have received substantial research attention with regard to both attention and comprehension processes, as well as to their ability to model and activate cognitive skills. The importance of formal features as they interact with content has also been underscored by many findings; however, their interaction with other variables has not been pursued sufficiently by researchers. Any research agenda should include continuing investigation of formal features, especially their complex interactions with other variables.

The simple act of a child viewing television has been demonstrated not as a response to stimuli but as a complex, purposeful cognitive activity that becomes progressively sophisticated as the child matures to adulthood. The cognitive effects of such activity have far-reaching consequences for both formal and informal educational activities.