The study of mass media effects has been characterized over time on the basis of systemic assumptions. Severin & Tankard (1979) depict the time between 1915 and 1939 as the era of the "Bullet Theory" (, also known as the "hypodermic needle" era) of strong media effects. Influential theories of this period, which included the development of propaganda, cinema, and broadcast radio industries, presumed that exposure alone to mass mediated messages resulted in cognitive, affective, and behavioral change. The years 1941 through 1959, the era of "Limited Effects Theory," were characterized by scholarship claiming that mass media actually had little or no strong effects on audiences, though they might contribute to changes through a nexus of mediating factors and influences. While mass media were seen as potentially serving an informative function, audiences were assumed to be selective in their perceptions. The present era, which is perceived having begun in the mid-1960s, is characterized as one of "Moderate Effects" or "Indirect Effects," in which strong effects are viewed as possible in certain situations. The study of interactive media as mass media falls neatly into this most recent category, especially in lieu of the capability for new interactive media formats to respond dynamically to contextual factors, including user input.
The uses and gratifications approach (Blumler & Katz, 1974; McQuail, 1984; McQuail, 1987) postulates that people actively select their media use habits on the basis of social or psychological disposition, assessments of the value of media use, and beliefs and expectations about possible benefits of such use. These expectations can result in attitudes regarding which media formats will be accessed and how they will be used to satisfy the information and communication needs of the user. This approach is mentioned by Rafaeli (1986) as being especially suitable for computer-mediated communication.
The need typology presented by McQuail, Blumler, & Brown (1972) divides such media-based needs into categories of diversion (including escape and emotional release), personal relationships (including substitute companionship and social utility), personal identity (including personal reference, reality exploration, and value reinforcement), and environmental surveillance. There are research findings which suggest that individuals recognize certain media, based on their attributes, as especially useful in satisfying these needs (Katz, Gurevitch, & Haas, 1973; Katz, Blumler, & Gurevitch, 1974). Books and cinema are perceived to fulfill needs of self-fulfillment and self-gratification by connecting individuals with themselves. Newspapers, radio, and television, on the other hand, serve to connect individuals with larger social groups. Expectations and attributions of a medium's function, whether defined socially or cognitively, can influence the value ascribed to information received through that medium.
It is important to note that the uses and gratifications framework has been the subject of frequent criticism (McQuail, 1984; Carey & Kreiling, 1974). A major criticism is that this perspective promotes tautological theories, moving from measured satisfaction back to an imputed need or forward from a need to a use and gratification, with no independent way of measuring need. Another criticism has been that uses and gratifications theory has a strong functionalist influence and that a conservative social model drives the perspective to ignore processes and effects which are not beneficial to the system. Lastly, the presumption of media audiences as being "active" in their viewing or listening decisions has been challenged.
There are two factors which should abate the criticism directed at functionalist influences. First, the uses and gratifications model usually discusses deviations from some system equilibrium which serve to generate need. Such deviations can certainly be modeled as processes which do not necessarily operate towards the benefit of the system. Second, a modeled gratification does not necessarily have to serve the ultimate benefit of the system, but rather a system's drive or motivation. The use of information for entertainment or escape, for example, might not always serve some higher good of an individual.
The presumption of user activity should not be a point of contention when considering interactive media which, by definition and design, cannot present messages to a participant who does not continuously make information-seeking choices. The accusation of a tautological approach can be refuted to some degree by methodologies which are made possible by the two-way interactivity, multimedia manipulation, and user tracking afforded by computer-based technologies. In an interactive multimedia environment, experimental or quasi-experimental contexts might be combined with the tracking of user choices to better distinguish needs from resultant usage. As Rafaeli (1986) points out with respect to CMC research, "the theoretical approach mandated by the uses and gratifications paradigm is toward a study of motivations of users mapped upon an empirical assessment of actual use" (p. 123).
While this dissertation's research does not deal directly with the assessment of use, it is important to consider that media expectations are important factors in the degree to which both knowledge and self-efficacy increase pursuant to a message's delivery.
Media Interactivity and Interactive
Media
Human beings, or any entity which acts to function within an environment
and influence reactions with other entities in that environment, can be
said to interact. In the broadest sense, the study of environmental
interactions might be considered the basis of scientific investigation.
Scientific studies of media effects have traditionally focused on how
exposure to communication media generates effects in individuals and in
social groups. Studies of interpersonal media, in contrast to mass media,
have generally focused on how communication processes affect the way
humans interact with one another through one-to- one discourse or within
small-group contexts.
In mass media studies, the descriptive label "interactive" has not generally been applied toward mass media mediated information and information formats which target large groups of people. Until recently, mass media channels, such as newspapers, radio, television, and recorded formats of videocassette and music CDs, have lacked characteristics of what is both popularly and academically considered "interactivity."
Over the past two decades, a merging of two technologies, information processing and communication, has produced collection of media formats which has been somewhat arbitrarily labeled "interactive media." Communication formats have undergone animism and anthropomorphization, i.e., producing a more "human-like" message through idiom usage and/or speech synthesis. Other yardsticks for interactivity have included bidirectionality of messages, user-control of computerized processing, and artificial intelligence. As Rafaeli (1988) points out, the unifying characteristic of interactive media exemplars, which include expert systems, hypermedia, video games, and interactive TV, is a relatively high level of responsiveness on the part of an information source, as solicited by queries or other input from the user. The interactive capabilities of these new communication technologies, along with other aspects of computer-mediated communication (Rafaeli, 1988), has served to break down the previously segregated concepts of mass media and interpersonal communication.
"Expert systems" are information databases which are structured to help human users solve problems within a "domain of expertise." They are often configured as variable sequences of questions which help pinpoint relevant information within a "problem space."
"Hypermedia" refers to computer-based texts that are accessed in a nonlinear fashion and are organized along multiple dimensions (Spiro & Jehng, 1990). Hypermedia uses various symbol sets to (a) convey information and (b) serve as indexes for further information retrieval. Many hypermedia information databases have been designed for and released to the consumer information market. Popular applications include - but are not limited to - family health, career development, academic preparation, personal finance, personal growth, historical/political retrospectives, and other types of self-help. "Hypertext" refers to the preliminary text-only precursor to hypermedia. Hypermedia documents are characterized by symbolic "links" to either specific sections of the same document or to other documents entirely. These links may be represented textually or graphically and provide information users with an ability to control the sequence of information as well as to "navigate" within a rich information environment.
"Interactive TV" refers to a developing technology intended to increase the selection and selectivity of programming available to television viewers. Rather than being content with what is scheduled by a program provider, viewers would be able to contact the provider directly and choose what to view from within a library of content as well as when to view the content.
For Rafaeli, interactivity is "an expression of the extent that in a given series of communication exchanges, any third (or later) transmission (or message) is related to the degree to which previous exchanges referred to even earlier transmissions." In other words, interactivity refers to a communicative process which is recursive in nature. The level of recursiveness is a measure of the extent of interactivity. Full interactivity, according to Rafaeli (1988) can only occur when communication roles are interchangeable among all real or surrogate participants.
A fully interactive communication system, therefore, is one which displays information based on the progress of discourse with the user. Taking this model further, a factor which makes interactivity possible is the extent to which a communicator can construct a model the other communicator's relevant goals and characteristics based upon prior messages. Such a system is assumed to coopt the information goals of the user to the extent that its responses reflect certain characteristics of the sum of the user's previous responses. Media channels which facilitate full interactivity include face-to-face, telephone, and written correspondence, including computer-mediated communication such as bulletin-board conferencing and electronic mail.
At the other end of Rafaeli's interactivity scale, non-interactive communication refers either to jolting, incoherent conversation or to one-way communication exemplified by broadcast and print media. The middle of the scale is occupied by "quasi-interactive" communication, that which facilitates a reaction to a user's input, but without consideration to the depth, content, or context of discourse.
For the purposes of this study, the term "interactive media", as well as "interactive database," will refer to computer-based information systems which allow users some measure of control over the content and/or sequence of presentation. This definition fits the use of the term when applied to hypermedia and interactive multimedia. While this would certainly not qualify as sustaining "full interactivity" (Rafaeli, 1988), it satisfies the descriptive criteria of a quasi-interactive communication system. Furthermore, research of quasi-interactive systems is helpful and perhaps necessary as scientists attempt to isolate the causal factors of more complex, more fully-interactive systems.
The interactive nature of interpersonal communication has been widely studied. In fact, the concept of classifying media interactivity on the basis of recursiveness has its roots in interpersonal discourse analysis (Dance, 1967, p. 297; van Dijk, 1985; Winograd, 1977, p. 64). Because of the traditionally one-to-many modal configuration of mass media, though, models of interactivity have generally not been applied to these communication channels. Discourse with a printed newspaper is impossible and, aside from selecting channels, human beings do not generally "interact" with a television set. It is conceded that viewers have responded to mass media entities, often in the form of letters to the editor, however because of the low rate of media institution responsiveness, such exchanges can hardly be termed "interactive. The independent development of two new forms of mass media, interactive computer-mediated communication (CMC) and CD-ROM, has made communication scholars rethink the issue of mass media interactivity.
Computer-mediated communication, or CMC, is the label which describes the multidirectional discourse which takes place when PCs are used as receivers/transmitters of digital information. CMC users and information hosts are interconnected within the existing telecommunication infrastructure, through the global umbrella network of the internet, enabling computers to compose and transfer messages. Within different CMC contexts, users can engage in one-to-one, one-to-many, and many-to-many communication. Because CMC messages can be transmitted to an almost unlimited number of receivers, it qualifies as a mass medium, albeit the first, and presently only, many-to- many mass medium (Rafaeli, 1986).
CD-ROM is technically not a mass medium; it is a digital data storage technology. However, certain characteristics of this technology, including environmenal robustness, convenient size, high storage capacity, and universal diffusion as a music recording medium, have associated "interactive CD-ROM" closely with the packaging and distribution of interactive multimedia entertainment and information applications. These applications generally include textual information as well as graphic and sound information which occupy a great deal of storage space. Most interactive CD-ROM applications, which might contain the digital equivalents of multi-volume encyclopediae, fit onto a single CD- ROM. CD-ROM-based information is digitally encoded, easily and quickly processed by a PC in a manner that facilitates user control of information presentation and, in some cases, a type of discourse which resembles full interactivity. CD-ROMs have the capacity to be truly "mass" media because of the possibility for such applications to be mass distributed. However, while the information on every interactive CD-ROM may be identical, the individual ways in which users choose to interact with the application can potentially make each experience, and each message sequence, unique.
Original inspiration for interactive multimedia might be attributed to Vannevar Bush, former Director of the Office of Scientific Research and Development, who presented a vision to post World War II scientists of machines to help humans manage a growing, bewildering storage of knowledge (Bush, 1945). His concept of the "memex" system introduced the possibility for documents to include hypertext-like links for annotation and access to related text and graphic documents. Twenty years later, Ted Nelson (1987; cited in Berk & Devlin, 1991) coined the term "hypertext," referring to documents, constructed in part by the author and in part by the reader, which encourage latter to move from topic to topic rapidly and nonsequentially. While CD-ROM-based applications generally do not provide users with authoring capabilities, they do afford information navigation functions.
It is worth noting that recent CMC developments have produced a communication format which combines key aspects of interactive CD-ROM and CMC. The World Wide Web (WWW) is an internet-based communication protocol which facilitates interactive hypermedia links to documents including text, graphics, video animation, and sound information. The WWW presently enjoys mushrooming usage and popularity. As of February 1995 there were 27,000 WWW sites and the population is estimated to be doubling every 53 days. There are also estimated to be 5 million documents accessible through WWW servers, a figure which reflects a doubling in volume every six months to a year (Business Week, 2/27/95, p. 78). The author does not consider it incidental that shortly after its acceptance as fulfillment of an academic requirement, this dissertation will likely be available for reading by any and all interested parties who have internet access to the WWW.
Media and Learning
There is presently a large and growing effort to develop educational
applications which employ multimedia and interactive techniques to convey
educational information. Financial commitments and investments on the
parts of computer manufacturers, including Apple and IBM, have resulted in
the growth of new industries intended to capitalize on the acceptance and
proliferation of interactive, multimedia education. However,
justification of the multimedia learning industry rests on the premise
that media (, or multimedia,) choice has a significant relation with the
learning process. This presumption has not gone unchallenged.
According to Clark (1983), "... summaries and meta-analyses of media comparison studies clearly suggest that media do not influence learning under any conditions." This assertion was a response to researchers who proposed "media selection" models (e.g., Reiser & Gagne, 1982) based on observations of the relative achievements of groups who received similar subject matter from different media. On an historical note, Clark further claims that whenever a new medium has been introduced, including radio, TV, and presently computers) advocates materialize to claims for improved learning through new media. It is not media choice which influences learning, states Clark, but rather instructional method as reflected by media choice. Different media may be grouped into categories according to media attributes which facilitate certain instructional methodologies. For instance, video and film might be considered equally appropriate or effective at presenting animated visual information.
Kozma (1991), in rebuttal, advocates a theoretical framework towards explaining the relationship between media and learning. He presents an image of the learner as an active collaborator with the medium in constructing knowledge. The learner "strategically manages the available cognitive resources to create new knowledge by extracting information from the environment and integrating it with information already stored in memory" (Kozma, 1991, p. 179). Clark's schism between medium and method, according to Kozma, is unnecessary. Medium and method have an integral relationship in that both are part of instructional design. "Within a particular design, the medium enables and constrains the method" (Kozma, 1991, p. 205). As Salomon (1979b) puts it, a medium is conceptualized as the interaction between a technology and a symbol set. New media become distinct when a new symbol system, which is a actually a new blend of existing symbol systems, evolves and creates new modes of expression .
Kozma (1991) also describes the ways different media influence learning. Books provide for stable information presentation as readers can slow down their information intake as they construct mental models and be more selective about the information they attend to. Television is presents information geared for low level engagement. This low level of involvement results in the construction of shallow, simplistic representations of information presented. The television medium is especially suited for carrying information about situations, though those with limited prior contextual knowledge might not have the background to process the context-related visual and audio cues on a transient, moment-to- moment basis. Time-based media are also very well suited for clarifying "what happens next," i.e., transitional processes and dynamic processes of mental models. Computer applications, which might include simulations and text-to-graphic transformation, have this capability as well, perhaps even to a higher degree as they can be used to create dynamic, symbolic representations of abstract, formal constructs frequently missing in the mental models of novices. Furthermore, computers allow learners to manipulate these representations within computer "microworlds" (Kozma, 1991, p.199).
Salomon (1979a) points out that a prime consideration in theoretical treatments of media and cognition is the way a medium's essential modes of representation, its symbol systems, relate to modes of representation in human beings. This observation is essential to the way media is integrated into both information processing and distributed cognition explanations of learning. Computer-based interactive multimedia, including hypermedia, present the potential of combining all known media visual and audio communication formats. Interactive media properties can enable users to view alternate symbol sets in the sequence of their choice or simultaneously. However, as this hybridized medium is still driven by relatively new technology, there are still many outstanding questions of both theory and practice which ask how it's potential can be maximized. Preliminary findings indicate that hypertext both calls on and develops cognitive skills beyond those associated with processing static text (Gay, Trumbull, & Mazur, cited in Kozma, 1991; Marchionini, 1989; Egan, Remde, Landauer, Lochbaum, & Gomez, 1989). On the down side, there are preliminary indications (e.g., Charney, 1987) that the nonlinear nature of hypertext might require novices to make information-seeking decisions for which they are ill-equipped due to a lack of sufficient domain expertise.
One fundamental difference between the perspectives of Kozma and Clark, pointed out by the former (Kozma, 1993), is one of presumed activity of the learner. Clark's "delivery truck" metaphor reflects a behavioral stimulus-response approach while Kozma (1993) and others (Greeno, 1988; Pea, 1993; Perkins, 1993; Salomon, 1993b) conceive of knowledge and learning as being a reciprocal interaction between the learner's cognitive resources and aspects of the external environment.
When considering how a communication medium might influence distributed and solo cognitive processes, it is important to consider the medium's technology, the symbol systems it supports, and its processing capabilities (Kozma, 1991, 1993). The technology describes the surface or physical features of the medium which constrain the its symbol systems and processing capabilities. Symbol systems are sets of symbolic expressions, such as spoken languages, printed text, graphics, and music, by which information is communicated about a field of reference (Goodman, 1976, cited by Kozma, 1993). Processing capabilities refer to the ways a medium operates on available system sets in specified ways. A television receiver can operate on television signals encoded in radio waves to produce animated graphic and audio content. Its processing is limited to tuning a particular channel frequency and translating the signals in "real-time" only. Combined with a video-cassette player, however, the television can both process the electronic signal and manipulate the information by replaying the sequence, altering the speed and temporal direction of play, and freeze the frame to a still image. Although text can be displayed on television as graphic information, it cannot be operated on as a symbol set.
While not all CD-ROM-based interactive multimedia applications afford the same capabilities, the base technology allows users to select information on the basis of text searches as well as graphic and textual hyperlinks. Animated graphic and audio sequences can be processed much the same way as on videocassette. Perhaps the most distinguishing characteristic of interactive media is that of dynamic feedback, whereby a medium provides direct responses to user inputs. Internet-based multimedia protocols such as the World Wide Web add the processing references to documents beyond the closed system of the CD-ROM and into the internet's communication network.
It is the user's ability of manipulating the searching and sequencing parameters of information which constitutes the "interactivity" of different media in this study. Furthermore, it is the distributed cognitive processing afforded by this interactivity which theoretically results in the cognitive and affective residue, namely gains in knowledge and self-efficacy.
Cognitive Processes, Involvement, and
Interactive Media
To understand the role of media in learning and other cognitive and
affective processes, it is important to ground a theory of media in
cognitive processes, rather than strictly behavioral models. Schema
theory (Quillian, M. R.) explains that memory and thought function through
cognitive structures of schemata. Schemata consist of interconnected sets
of ideas which are further linked to other schemata. Meaningful learning
takes place when new ideas becomes linked and assimilated with existing
schemata. Similarly, the philosophy of constructive, or generative,
learning asserts that learning itself is a constructive process in which
new meanings are generated by activating and altering existing knowledge
structures (Wittrock, 1974; Brown, Collins, & Duguid, 1989). This
activation and alteration is facilitated through a learner's experiences
in using the new knowledge. Historically, the promotion of involved,
active learning may be viewed as a response what Whitehead (1929) referred
to as "inert knowledge." Inert describes knowledge, exemplified by rote
memorization, which is not retrieved under authentic conditions of use
and, therefore, may not be useful to an individual in actual circumstances
requiring its retrieval. Schematic linkages are defined in terms of
context and, as a result, cause us to remember and process information
differently in different situations (Borsook & Higginbotham-Wheat,
1992). The role of the environmental involvement in generative or
constructive learning is explained differently in the cognitive models of
information processing and distributed cognition.
Information processing theory considers cognitive functioning within the closed system of the learner. The stages of information processing include recognition of patterns in the information, short-term memory storage (STM), and long-term memory storage (LTM) (Borsook & Higginbotham-Wheat, 1992). Increasing the probability that information is acquired, sorted and retrieved involves promoting the processes of attention, selective perception, rehearsal, encoding, and retrieval, all of which are conceptualized as entirely internal cognitive activities. As processing continues, information and skills are recoded to form stronger links between internal memory nodes of stored information. Instructional aids, essentially based in communication media, rely on learner retention. Therefore, they should encourage extra processing of information in order to increase the speed and strength of such memory pathways.
"Distributed cognition" and "person-plus" are labels of the perspective which examines the role of organism-environment interaction in explaining cognitive processes. In this framework it is generally acknowledged that descriptions and explanations of cognitive processes cannot be context-free. "People appear to think in conjunction or partnership with others and with the help of culturally provided tools and implements" (Salomon 1993a). A distributed cognition view of interactive media would explain that cognition does not occur only inside the head of the interactive media user but, rather, in the interaction between user and implement. In hypermedia applications, for instance, conceptual associations are incorporated in the hyperlinks, not only in the head (Salomon, 1993a) of the user. Thus, different situations engender different modes of cognitive functioning.
A distributed cognition perspective is especially appropriate for the study of interactive media because of its emphasis on the organism-environment interaction. It is also important, if lasting cognitive effects on human beings are presumed, to distinguish between what Salomon (1990) calls effects with computer technology, i.e., the cognitive load that is distributed, and effects of computer technology, i.e., the cognitive residue which remains after the environmental interaction. Although this dissertation examines mainly the effects of the interaction between user and interactive media, explanations of these residual effects must consider distributed cognitive processes as well. It is the effect with the computer of being involved, i.e., interacting while being mindful that the interaction brings about goal-seeking consequences, which helps bring about the cognitive and affective effects of this involvement.
A distributed-cognition view of interactive media presumes an involved partnership between the user and the interactive information application. The manner of involvement is an important factor in whether effects will be lasting. As Salomon (1990) points out, some effects with computers allow relatively short-lived, superficial residual effects. The presence of visually fascinating or interpersonally involving cues within an interactive application can lead users to concentrate more on the superficial "glitz" accompanying information than the information itself. When interacting with a relatively new information medium, a user may concentrate more on exploring its characteristics than on the intended messages.
According to Salomon (1990), there are two possible mechanisms by which a computer program might leave a lasting cognitive residue: skill activation and skill internalization. Skill activation is a method by which a computer program calls upon the user to repeatedly exercise particular thought processes and the strategies needed to solve a specific type of problem, thereby letting the learner develop an abstraction of underlying principles. In contrast, skill internalization postulates that knowledge and skills become learned, or internalized, through interpersonal communication. This is based on the assumption that humans use culturally defined symbolic devices to "affect, regulate, and guide" their social environments. In short, through a process of internal symbolic transformations, external conversations become internal speech. Such external discourse need not be limited to contexts among humans. Salomon (1979a) has shown that children are capable of internalizing symbolic elements of television and use them for cognitive- representational functions.
The present study focuses on human cognitive partnership with an interactive medium which presents information in an encyclopedic fashion in which there are no simulations presented to subjects. Therefore, as skills are not actually being exercised or activated, but rather observed, it is the skill internalization process which is presumed. Salomon (1990) presents four criteria necessary skill internalization to develop. First, skills to be internalized must already be in the process of maturing in the individual, i.e., the individual must be developmentally ready to acquire the skills. Second, these skills must be of a generalizable nature such that they can be retrieved from memory in appropriate situations outside of the immediate instructional context. Third, the skills must be considered useful and novel enough to be worth internalizing. Finally, the skills to be learned must be encountered in a straightforward manner, rather than alluded to in some non-explicit or assumed way.
Assuming that a target skill meets these criteria, an instructional model of an active, rather than passive, learner should also discusses what types of learner activities promote cognitive residue. Salomon (1981) addresses this in his construct of amount of invested mental effort (AIME). AIME is defined by Salomon as the number of mental elaborations executed in the processing of information, discounting the proportion of automatic or habitual elaborations in the learner's cognitive repertoire.
An interactive medium might be instrumental in maximizing the presence of these criteria. Based upon evaluation of learner inputs, an interactive information application might consult internally programmed models to determine the learner's level of maturity regarding the target skills, and present information appropriate to that level. An interactive application also has the potential to respond to user queries regarding the general situational factors which guide skill usage, which might help learners to see the generalizable nature of these skills. Regardless of the interactive processing functions of interactive media, any possible cognitive influence is directly dependent on whether a user actually uses the interactive capabilities afforded by the media application. Regarding AIME, there are important questions to address regarding interactive media. First, does media interactivity contribute to conscious mental elaboration? Second, are there ways of designing such media to promote such cognitive activity?
Csikszentmihalyi's (1975) concept of "flow" describes a state in which an individual's cognitive situation is balanced between his or her capabilities to perform and the degree of difficulty or challenge posed by a particular task. Flow is closely associated with "intrinsic reward-seeking," performing a task whose very activity generates positive feelings because it fully utilizes one's physical and sensory potential (Evans & Clarke, 1991). Two affective states summarize the situation when the balance between ability and challenge is upset. On the one hand, boredom is characterized by being "under-challenged" whereas, on the other hand, bewilderment characterizes a situation in which a task's challenge is too overwhelming for the individual's capabilities. Both of these latter states are relatively unpleasant and unmotivating.
Reaching a state of flow requires a significant degree of invested mental effort. The expectation of flow, or of some related functional competence, might also be an important predictive factor of successful learning. Salomon (1981) theorizes that self-efficacy bears a curvilinear causal relationship with AIME. Learners who are very confident about their grasp of a domain of knowledge or skill may not feel the need to invest effort into learning about this domain. On the other hand, those who have very little confidence in their relevant capabilities might feel that any effort they expend will be wasted. The challenge for an interactive information application in promoting cognitive involvement for a wide range of users may be addressed by presenting an information landscape which is both (a) rich enough to allow exploration at different levels of expertise, and (b) navigable through an intuitively understandable interface.
The phenomenon of being overwhelmed in an information environment is closely related to the broader concept of "infoglut." This refers to the ever increasing volume of information available, and indeed constantly being transmitted, to individuals and the public. As documented by Pool (1983), with respect to the output of text alone, public media information grew at an annual rate of more than seven times the increase of per capita consumption of words between 1960 and 1977. The widespread diffusion of cable television and, for many, of the internet has almost certainly accelerated the momentum of information. The mass communication concepts of "selective exposure" and "selective perception" speak of the manner in which people, who all have limited information processing capacities, filter through the volumes and reams of environmental data to obtain the information they seek. An interactive medium which responds effectively to information search queries can provides an information-seeker with the means to be more selective in the search process, thereby serving as its own filter with the user's guidance.
While the issue of involvement is certainly an important factor in the evaluation of interactive media for learning purposes, there is a variant perspective which explains why the manner of involvement afforded by interactive media is especially useful in the construction of new knowledge. According to Spiro and Jehng (1990), traditional methods of instruction rely on linear media. This approach is not problematic when the subject matter to be taught is well-structured and relatively simple. Ill-structured information domains, however, demonstrate more varied relationships across the cases in which the knowledge has to be applied. As information becomes more complex and ill- structured, greater amounts of important information are lost with linear approaches and the unidimensionality of organization that typically accompanies them.
Cognitive Flexibility Theory (CFT), postulates that one learns by "criss-crossing conceptual landscapes" (Spiro & Jehng ,1990, p 169). This navigation in, and manipulation of, the information environment exercises cognitive flexibility, the human ability to spontaneously restructure one's knowledge in adaptive response to changing situational demands (Spiro & Jehng, 1990, p. 165). Instruction spurs this process by providing materials that enable the user to actively explore a multidimensional information landscape. The process of exploration, which occurs during learning, forms knowledge representations. Gaining expertise involves the development of flexible representations of knowledge which promote deep conceptual understanding and the ability to adaptively use knowledge in new situations (Jacobson, 1994). The ability to use the same information in a variety of situations distinguishes active, useful knowledge from Whitehead's (1929) conception of "inert knowledge," the rote memorization of details.
Cognitive flexibility is a function of both the way knowledge is represented (e.g., along multiple rather than single conceptual dimensions) and the processes that operate on those mental representations (e.g., processes of schema assembly rather than intact schema retrieval). Personal computer technology development has facilitated random access information retrieval, the display format which allows user-controlled nonlinear sequencing of multimedia information presentation. Random access instruction refers to nonlinear learning with random access media. Learning with random access instruction might entail making conceptual "jumps" from one text to another or revisiting the related content material in a variety of contexts or sequences, each bringing out additional aspects of the content's complexity. The mastery of complex bodies of information is greatly facilitated by random access instruction, including hypermedia instruction (Spiro & Jehng, 1990; Spiro, Feltovich, Coulson, & Anderson, 1989; Feltovich, Spiro, & Coulson, 1989).
As cited in Kozma (1991, p. 200), the Cognition and Technology Group at Vanderbilt University has developed a series of interactive video-based information environments. These researchers contend that this presentation helps learners generate more complex, richer models of the problem situation. More recently, Swan (1994) has found that hypermedia presentation of information improves a learnerŐs ability to draw conceptual links in the study of history.
Jacobson (1994) outlines seven theory elements which prescribe how hypermedia applications might be designed to maximize the learning potential presented by cognitive flexibility at advanced stages of learning. Based on CFT, such learning systems should (a) employ numerous cases and examples using a variety of symbolic representations, (b) use multiple forms of knowledge representation, including different opinions and perspectives, (c) link abstract concepts to case examples, (d) demonstrate conceptual complexities and irregularities across cases, (e) stress web-like conceptual interrelations through navigational mechanisms such as hot-words, (f) encourage knowledge assembly from different conceptual and thematic case sources, and (g) promote active learning.
An empirically supported distinction between hypermedia instruction and that instruction form more traditional, linear media is the type of learning which is enhanced from the experience. Several studies (e.g., Spiro et al.,1987; Hartman & Spiro, 1989; Jacobson, 1990) have found that while control group learners outperformed cognitive flexibility learners on reproductive memory tests, the cognitive flexibility learners were better able to apply the new knowledge in different contexts. Hence, we see the potential for hypermedia learning, an archetypal cognitive flexibility communication format, to zero in on the challenge of knowledge transfer, i.e., the imparting of efficacious knowledge.
Linking Cognition and Affect in Interactive
Media
One foundation of this dissertation is that cognitive activity can have
affective influences upon an individual. This follows from a
meta-theorietical presumption that cognition and affect, while
distinguishable in their ramifications towards behavior, are closely
related in the process of knowledge construction. Whereas Salomon (1981)
and others have examined the roles of cognitive effort and involvement in
knowledge retention and comprehension, there are also theories which speak
of effects cognitive effort have upon attitude formation, including
attitudes towards oneself. Whether implicitly or explicitly addressed,
the role of communication is crucial as it deals with the transmission of
the raw materials, namely information, from which attitude and knowledge
are formed.
In studies examining the role of cognitive effort on message retention and affective influences of television, Hawkins and Pingree (1986) point out that children who score high on inferring television content not explicitly presented in a program are the most affected by implicit social reality messages on television. The authors assert that the potential for media messages to influence affective processes is closely related to a greater tendency to draw inferences and may reflect the results of prior cognitive effort.
Cognitive consistency approaches (e.g., Festinger, 1957; Heider, 1946) assert that new information disrupts the cognitive organization developed by an individual. The disruption is intolerable and leads to a form of unpleasant mental tension until the new information becomes a complementary part of the individual's knowledge structure (Reardon, 1991). Humans therefore try to avoid or, if need be, reconcile gross inconsistencies between thoughts, beliefs, attitudes, and behaviors (Zajonc, 1960, cited by Reardon, 1991, p. 59). An illustrative example presented by Cialdini (1988, p. 83) is the phenomenon of college fraternity initiation ceremonies. The decisive behavior to undergo different degrees of hazing increases the attitudinal value individuals attach toward fraternity membership. Limitations of consistency approaches lay in a reliance on the tenuous assumption of human rationality and in the difficulty in predicting which of a number of inconsistent relationships will hold sway in behavioral decisions. However, if one can isolate the relative strengths of certain beliefs and attitudes, it is reasonable to assume that cognitive consistency can be useful in predicting cognitive, attitudinal, and behavioral change.
Consistency approaches clearly support the notion that cognitive effort can result in attitudinal or behavioral change. Part of this dissertation's research asks whether the cognitive effort involved in navigating within an interactive database can influence such change on the basis of a learner's tendency to consistently reconcile behavior with attitude. In other words, if a learner works to obtain information through constant decision- making, will he or she be more inclined to believe that the material was learned than if there were no such decision-making demands imposed by the medium.
The Elaboration Likelihood Model (Petty & Cacioppo, 1986), or ELM, specifies a number of ways in which the processing of a message, its source, and other variables have an impact on attitude change. Attitude formation is theorized to proceed by either central route processing, i.e., mindful, elaborate consideration of relevant information contained in a message, or by peripheral route pressing, i.e., resorting to simpler heuristics and inference which might be related or peripheral to the message itself. Attitudes formed or changed via peripheral route processing are less persistent, predictive of behavior, and resistant to change than those formed through central route processing. Some variables which influence the likelihood of elaboration include topic relevance, an individual's tendency to attend argumentation, and environmental distraction. A theoretical question relevant to this dissertation is whether or not media interactivity can be used to promote central-route processing of attitude-forming information.
The potential for interactive media to influence affective processes might also be explained on the basis of the evolutionary development of human environmental preference. Cognitive map theory is a variant of schema theory based on the theorized course of human cognitive evolution, described by Kaplan (1973). Essentially, the model presents humans as animals who have survived in a visually rich environment by developing cognitive abilities to organize and retrieve information quickly and efficiently within an associative, highly interconnected structure of concepts. This model points out the prominence of visual navigation in mankind's development and, citing supporting neurophysiological evidence (Hebb, 1968), the spatial structuring of both spatially and non-spatially related information in the brain.
According to cognitive map theory, four major types of knowledge have helped humans to survive through natural selection: (a) identification of where one is, situationally, as a starting point for adaptive behavior, which requires both perception of the present stimulus array and memory of immediately preceding events, (b) identification of what is likely to happen next, (c) evaluation of the goodness or badness of each predicted situation, and (d) a ready mental cache of possible courses of action. Therefore, the feeling of being informationally lost had to be extremely unpleasant and to be avoided, if possible. Thus, the model explains human preferences for (a) informed participation in, and some control over, one's environment, (b) being aware of one's possible actions and the likely outcomes of those actions, (c) information environments made up of familiar objects and situations, and (d) environments novel enough to be worth exploring which also exhibit signs of being navigable . Navigability is the degree to which an environment presents information in coherent, intelligible patterns which may be used toward the solution of individual human needs (Kaplan, 1973).
Interactive media provide more control to the learner than non-interactive, or passive, media. From a cognitive map perspective, as well as that of CFT, the increased control capability would explain why users might prefer the experience of using an interactive, multimedia database to that of using information channels and formats which offer no capability to navigate. Becker & Dwyer (1994) conducted research to determine whether hypermedia users perceived the greater degree of instructional control afforded by that information format. The authors found that not only did the learners perceive that control, they also exhibited greater indications of intrinsic motivation for the material. This latter finding supported Cognitive Evaluation theory (Deci & Ryan, 1985, cited in Becker & Dwyer, 1994), which postulates that a major component of intrinsic motivation is the amount of control the learner feels over the learning situation.
On the other hand, presenting information within a context allowing virtual navigation, as in hypermedia, also presents the possibility of getting lost in the information space (Nelson, 1989). Getting lost refers to the unpleasant feeling of bewilderment experienced when one either (a) becomes disoriented within the available options, or (b) cannot recall or understand the conceptual path which led to the information currently displayed. In both situations, the information-seeker cannot return to a recognizable, comprehensible information location.
Self-Efficacy
Perceived self-efficacy is defined as "people's judgments of their
capabilities to organize and execute courses of action required to attain
designated types of performances. It is concerned not with the skills one
has but with the judgments of what one can do with whatever skills one
possesses" (Bandura, 1986, p.391).
Self-Efficacy and Human Agency
Knowledge and information processing skills, including conceptual
transformation skills, are necessary but insufficient for the performance
of tasks. Sometimes, people do not perform tasks that they know how to
do, even if they are highly motivated. Perceived self- efficacy, also
referred to as "self-efficacy," is an important self-referent factor
governing the interrelationship between knowledge and action.
Self-efficacy is distinguished from "response-outcome expectation," the predicted outcome of successful performance of a skill which also plays an important motivational role in purposeful behavior. For example, a surgeon might feel very confident in her abilities to perform a delicate operation on a very sick person. However, the patient's pre- operative condition might be so critical that even a successful operation might be too taxing for the patient to survive. In this case, the surgeon has a high level of self-efficacy but a low response-outcome expectation, which might likely result in a decision to avoid performing the surgery. Conversely, a basketball player might believe that slam-dunking, i.e., jumping high enough to literally push a basketball through the hoop, would yield the positive outcomes of winning a game as well as the respect and gratitude of teammates. However, the player feels that he is either too short or lacks the strength to make such a jump. In such a case, there is a low-level of self-efficacy even though the response- outcome expectation is high.
Although self-efficacy and response-outcome expectation represent different cognitive constructs, both of these concepts can influence one another. People see outcomes as contingent on the adequacy of their performances, and care about those outcomes, that they rely on self-judged efficacy in deciding courses of action (Bandura, 1986, p.392). A driver who has doubts about his ability to safely steer through a winding road will imagine a tragic accident. On the other hand, a response-outcome expectation can influence self- efficacy. For example, the aforementioned surgeon who foresees a lethal outcome to an operation might further rationalize the decision against surgery by re-evaluating her abilities.
Self-efficacy plays an important role in influencing the outcomes of learning. Efficacy, or competent functioning, is characterized as a generative process, in which cognitive, social, and behavioral subskills become part of integrated strategies for use in different situations. Success in learning a new skill is often attained after the repeated trial and error of different strategies and often demands perseverance. Self-doubters, those with low perceived trait self-efficacy, low self-efficacy in particular contexts, and/or low self-efficacy regarding specific skills and subskills, will quit a learning process if initial efforts don't yield some level of perceived success (Brown & Inouye, 1978; Schunk, 1984). The relationship between low self-efficacy can potentially develop into a negative helical process, in which low self-efficacy leads to less cognitive learning effort, leading in turn to reduced learning success, resulting in even lower self-efficacy. Conversely, a learner with a high level of self-efficacy, who feels that he or she has "little to learn" about a topic or skill, might see little relative advantage in exerting effort in the process of learning. This fits very will with the aforementioned concepts of AIME and flow mentioned above. However, outside of learning contexts, high levels of self-efficacy serve to intensify and sustain the perseverance required for difficult tasks (Bandura, 1986, p.394).
Bandura (1986, p. 399) maintains that self-efficacy, whether accurate or faulty, is based on four principal sources of information: enactive attainment, vicarious experience, verbal persuasion, and physiological state. Enactive attainment refers to authentic mastery experience and is cited as the most influential source of efficacy information (Bandura, Adams, & Beyer, 1977; Biran & Wilson, 1981; Feltz, Landers, & Raeder, 1979; cited in Bandura, 1986). Success raises efficacy self-evaluations and failure lowers them, especially if one is a novice or at some early point in the learning sequence. The influence of new experiences upon self-efficacy depends on the nature and strength of the pre-existing self-perception. Once established, enhanced self-efficacy tends to generalize to other situations, though generalization effects occur most predictably in activities that are most similar to those in which self-efficacy has been improved.
Vicarious experience, the observation of other people performing successfully, can raise self-efficacy within an individual by internal proposing to the beholder that he or she possesses the capabilities to master comparable activities (Bandura, Adams, Hardy, & Howells, 1980; Kazdon, 1979; cited in Bandura, 1986, p.399). On the other hand, observing others fail despite high levels of effort can lower an observer's own self- efficacy judgment (Brown & Inouye, 1978; cited in Bandura, 1986). Vicarious experience can be especially influential when the observer has little experience with the subject matter or has very limited feedback related to past performances. Although this source of information is less influential than attainment experience, it can produce "significant, enduring changes through ... effects on performance" (Bandura, 1986, p.400).
Verbal persuasion has a greater impact on undermining self efficacy (when refuting the persuadee's efficacy) than on promoting self-efficacy (when supporting said efficacy). Illusory boosts in self-efficacy are generally disconfirmed by one's own actions whereas those who have been persuaded of their inefficacy tend to avoid attempts to perform the activity in question. When supporting a person's efficacy, verbal persuasion is more effective when the message receiver has some reason to believe that he or she can produce desired results (Bandura, 1986, p.400).
A person's physiological state can influence self-efficacy to the extent that an individual interprets somatic arousal, e.g., headaches and muscular tension, as an ominous indication of vulnerability (Bandura, 1986, p.401).
Research concerning the effects of self-efficacy on goal attainment and behavior have examined the construct in various contexts, including academic performance (e.g., Schunk, 1984), career choice and attainment (e.g., Betz & Hackett, 1985), teaching performance (e.g., Ashton, 1985), phobic dysfunction (e.g., Bandura, Adams, & Beyer, 1977), pain management (e.g., Reese, 1983), computer literacy (e.g., Miura, 1984), reaction to physical threat (Ozer & Bandura, 1990), and health behavior (e.g., Ewart, Taylor, Reese, & DeBusk, 1983). This research is generalizeable to the degree that it addresses how perceived self-efficacy can reduce the incidence of intrusive negative thinking and anxiety arousal. Convergent evidence from these and other divergent domains lend broad support to the notion that perceived self-efficacy operates as an influential mechanism in human functioning in the environment (Bandura, 1986, p.430).
Effectance motivation, a theoretical construct closely related to self-efficacy (Bandura, 1986, p.410), is conceptualized as an intrinsic drive for interaction with the environment as this arouses feelings of both pleasure and efficacy (White, 1959). Successful actions build effectance motivation yet further. A more recent version of the theory postulates that the motivation for interaction is strongly influenced by a drive for mastery motivation, i.e., effective action in dealing with the environment (Yarrow et al., 1983). Critics of the theory point out that while experience and practice are linked to both efficacy and self- efficacy, there is no treatment distinguishing successful from unsuccessful performance (Bandura, 1986, p.410).
Self-Efficacy and Communication
Media
Self-Efficacy and Communication Media
Within the movement of health promotion, health communication and health
education are synonymous (Freimuth, Edgar, & Fitzpatrick, 1993). The
terms reflect a process in which messages are communicated to influence
individual, social, cultural, and political agents to make community
environments more supportive of health behaviors. The centrality of
individual involvement within this contemporary philosophy underscores the
importance of individual self-efficacy, the judgment of each individual
regarding his/her ability to enact recommended behaviors and otherwise
apply what he/she knows.
Maibach & Flora (1993) produced experimental findings supporting the hypothesis that symbolically modeling risk-reduction techniques and cognitive rehearsal of this information can enhance self-efficacy beyond the presentation of risk-reduction information alone. Symbolic modeling helps an observer divide complex behavior into its component parts to facilitate processing, organization, and retention (Sheffield & Maccoby, 1961; cited in Maibach & Flora, 1993) For theoretical purposes related to the research, symbolic modeling is synonymous with the aforementioned self-efficacy source of vicarious experience. Cognitive rehearsal adds the dimension of active processing which enhances the retention necessary for observational learning (Bandura & Adams, 1977; cited in Maibach & Flora, 1993).
The design was based on three experimental groups treatments. The control group saw a televised video explaining the risk-reduction technique of discussing AIDS prevention with a sexual partner. The second group saw the same video as well as a symbolic modeling video showing actors illustrating the technique. The third group saw the same content as the second and completed cognitive rehearsal exercises. The study concluded that (a) mediated health information combined with symbolic modeling enhanced self-efficacy more than mediated information alone, and (b) mediated health information combined with both symbolic modeling and cognitive rehearsal enhanced self-efficacy more than mediated health information combined with symbolic modeling content.
The communication medium employed by Maibach and Flora (1993) to present the information and promote the symbolic modeling was non-interactive. This dissertation research serves, in part, to extend this research by considering potential effects of communication media interactivity upon self-efficacy.
Self-Efficacy and Media Interactivity
Interactive multimedia formats, including hypermedia, are predicated on
the active involvement of the information recipient in the course of
acquiring said information. The affordance of user involvement combined
with other symbolic processing capabilities of interactive multimedia has
the potential of magnifying three of the four information sources of
self-efficacy.
Enactive attainment may be facilitated by interactive media in at least two ways. Firstly, computer-based media has the technological capability of presenting interactive simulations in which a learner receives feedback on his or her decisions or actions as communicated to the computer application. Flight and automobile simulators, for instance, provide for personalized virtual experiences which might have a positive influence on self-efficacy. Second, it is conceivable that by making navigational choices through the information space of a hypermedia application, a learner can associate the success of acquiring pertinent, appropriate information with successful use of this information.
As with passive media, interactive multimedia can afford learners with the possibility of vicarious experience by illustrating examples of other similar people performing tasks using skills to be learned. One important differential advantage of interactive media is the potential for learners to have some control over the some of the symbolic information presented. A learner could potentially choose to zoom in on particular parts of the body which are more relevant to the task at hand, or one might choose to add or remove captions describing the activity. Learners can more easily repeat all or part of a sequence at will in order to enhance comprehension and retention. Learners might make useful associations by choosing to view different techniques which share some relevant conceptual factor, a practiced explained by cognitive flexibility theory above.
Interactive multimedia formats have the capability of providing not only passive verbal persuasion, but the type of persuasive feedback which directly addresses a learner's choices and responses while he or she navigates the information space or interacts with simulations. This more personalized feedback, while not necessarily exhibiting "human- like" qualities might be perceived, more importantly, as more relevant and realistic than messages presented with no connection to the individual responses of the learner.
Hypertext emphasizes knowledge construction rather than transmission, which requires learners to make meaningful representations and, hence, practice cognitive rehearsal (Jonassen, Ambruso, & Olesen, 1992). This reasoning may be extended to hypermedia, which affords the same relevant knowledge-construction operations, including associative linking, semantic network modeling, and dynamic user control. Therefore, hypermedia should enhance observational learning beyond that promoted by passive media. This assertion is made, on the basis that interactive multimedia users are constantly called upon to make information-seeking choices and, perhaps indirectly, are compelled to consider the nature of the information presented to them.
There are characteristics of interactive media design which might cause attempts to enhance self-efficacy to backfire. One of these issues is that of navigability. Navigability of an interactive media application depends upon attributes of the application as well as the user. On the one hand, the method by which a learner enters responses or queries might be difficult to use or understand. On the other hand, the information sequencing routine might force more advanced learners to sit through information presentation which is boring in either content or delivery. This would likely result in an aversion and avoidance of the use of the information application, thereby effectively precluding both the learning of the skill and attainment of self-efficacy. Also, to a novice the task of making information choices about a domain not well understood might be cognitively overwhelming. As mentioned above, this feeling of being informationally lost would likely bring on an unpleasant feeling of helplessness which the learner might associate with his or her ability to use the information itself. For the reasons stated here, it is crucial that instructional interactive application designers take care to consider the range of expertise and symbol manipulating capabilities of the user/learner.
The information contained herein is made available on a strictly non-commercial basis and for the express purpose of furthering academic scholarship and research. Printed, photocopied, or electronic distribution (outside the domain of this particular web-site) of all or parts of this document is expressly forbidden without the tacit consent of the authors.
Click here to return to Associated Theory and Research Table of Contents