Towards a Theoretical Framework for the Design of Interactive Online Distance Learning Systems
Introduction
Designing systems informed by current research
With the advent of ever more sophisticated communication technologies, there is an increasing desire to incorporate these new technologies into the distance education classroom. Although new communication tools have already been adopted by many distance learning programs, few of them were developed based on a theoretical framework that could have informed their design. All too often new technologies are incorporated into the distance education program without including a careful evaluation of the relevant theories and research in the planning process. If technology is supposed to improve students' learning however, much more attention needs to be devoted to this planning process (Elmore & Jafari, 1995). Indeed, as Elmore and Jafari (1995) note, "projects are too often initiated without adequately addressing the conceptual and system design issues" (p. 30). This reluctance to let research inform system design can be observed throughout the distance education literature. In their guide to the use of conferencing technologies in distance education classrooms, Burge and Roberts (1993) outline six steps, which the planning and design process should follow. None of the steps advocated by these authors, however, suggests consulting the research on the use of conferencing tools in an effort to inform the system design. While Burge and Roberts (1993) make some valid points about the necessity of determining learning needs and designing learning objectives, they fail to incorporate theory as an integral part of the planning and design process. Eastmond (1995) identifies this atheoretical position as a major limitation in the practice of online distance education.
What is needed then, is a distance online learning system that is based on a theoretical framework able to inform its design. This paper will demonstrate how the theory of hyperpersonal online communication (Walther, 1996, 1997, 2001) and its theoretical predecessor, social information processing theory (Walther, 1992a, 1992b), can be used as a framework to guide the development of an online learning system designed to enhance interpersonal communication, information exchange, and student learning. Before discussing the implications of basing the system design on this theory, it is first necessary to provide a brief synopsis of both theories and explain the main assumptions underlying them.
The Theory of Electronic Hyperpersonal Communication
Findings from early research on computer-mediated communication (CMC) often concluded that because of its reduced cue nature, the online medium encourages highly impersonal communication (Hiltz, Johnson, & Turoff, 1986; Rice, 1984). Since the online medium cannot easily transmit nonverbal or relational cues, this limitation of the medium was usually invoked to explain the impersonal communication that researchers observed in computer-mediated interactions. Culnan and Markus (1987) referred to this explanation as the "cues-filtered-out" perspective. As Walther (1992, 1996) has pointed out however, the conclusion that CMC is an impersonal medium because it can only provide few, if any, nonverbal and relational cues, might be misleading. Rejecting the cues-filtered-out perspective, Walther (1992) instead argued for a social information processing approach to the study of online communication.
Social information processing theory (SIP) is based on the following three assumptions: (1) "communicators are motivated to develop impressions and relations despite hindrances that alternative media may impose", (2) "users adapt their efforts to present and acquire social information to whatever cue systems a medium provides," and (3) " relational processes take time, and CMC is relatively slower than face-to-face (FTFT) communication" (Walther & Tidwell, 2001, p. 7). The key variable in social information processing theory is time. According to Walther (1992), the findings of impersonal communication reported in early CMC studies, may be more accurately explained by time constraints facing online communicators than a lack of cues available to them. Since social information cannot be easily transmitted through the online medium and needs to be communicated in writing, CMC groups take longer to communicate and can therefore be expected to require more time to develop relationships and reach an interpersonal communication stage. Social information processing theory (Walther, 1992) assumes that CMC users, just like face-to-face communicators, are motivated to develop social relationships, but that they simply need more time to do so. One-shot studies that examine limited-time interactions between CMC users are therefore likely to conclude that CMC is inherently impersonal.
If given enough time, CMC users should be able to exchange sufficient social and task-related information to develop interpersonal relationships and avoid impersonal communication. Due to the limited number of cues capable of being transmitted, "the medium cannot convey all the task-related as well as social information in as little time as multichannel FtF communication" (Walther, Anderson, & Park, 1994, p. 465) and therefore slows down the process of relational development. While social information processing theory explains under what circumstances computer-mediated communication may result in impersonal, or interpersonal interactions between CMC users, Walther's (1996) theory of hyperpersonal online communication extends this framework by identifying the factors necessary to the development of hyperpersonal communication. Walther (1996) refers to hyperpersonal communication as online communication that is more socially desirable and more intimate than FtF communication.
According to Walther (2001), the hyperpersonal communication framework "offers an approach to the ways that CMC users in some conditions exceed the intimacy, affections, and interpersonal assessments of their partners, relative to parallel FtF activities or alternative CMC contexts" (p. 8). In order to achieve such hyperpersonal levels of communication, psychological processes related to four elements of the traditional communication model need to be activated (Walther & Boyd, unpublished manuscript). These four elements are: (a) the receiver, (b) the sender, (c) the channel, and (d) feedback. The hyperpersonal communication model argues that CMC affects these four variables in ways that are not always possible in FtF interaction and thereby allows computer-mediated interactions to surpass FtF interactions in terms of intimacy and affection.
Message receivers in online contexts are said to contribute to the hyperpersonal phenomenon by idealizing their online partners and engaging in overattribution. Given the reduced cue nature of the medium, CMC users build their impressions of one another on any bit of information they receive about their communicative partner, and in doing so often overestimate the similarities between themselves and their partner. Walther (1996) invokes SIDE (social identity/deinidividuation) theory (Spears & Lea, 1992) to explain this overattribution phenomenon. According to SIDE theory (Spears & Lea, 1992), communicators who interact without seeing one another, experience an increased perception of group identity and therefore look for cues about their partner that would highlight their similarities rather than set them apart as unique individuals. When interlocutors cannot see one another, as in the case of most computer-mediated communication, message receivers are thus likely to construct idealized pictures of their communicative partner, because they are looking for similarities in the social cues emitted by the message senders. This idealization of the communicative partner may in turn affect attraction and intimacy levels.
Message senders furthermore contribute to the development of hyperpersonal communication by engaging in selective self-presentation strategies (Walther, 1996). Selective self-presentation refers to the act of presenting a more socially desirable, idealized version of one's self to one's online interlocutors (Walther, 1996). Several factors facilitate the use of selective self-presentations strategies in online contexts. First, affordances for nonverbal leakage, that is "the accidental transmission of unintended nonverbal behavior or physical appearance" (Walther & Tidwell, 2001, p. 10), are reduced or even eliminated in online contexts. Not only is it difficult for unintentional nonverbal messages to be transmitted through the online medium, but verbal communication also rarely contains unintended messages. Because CMC, and especially asynchronous CMC, offers its users more time when formulating messages, CMC users are more likely to edit and mindfully construct their messages before sending them (Walther, 1996; Walther & Tidwell, 2001). Even more so, CMC users only have to focus on their writing and do not have to attend to a number of factors people engaged in face-to-face interactions have to attend to. Indeed, in computer-mediated interactions, users do not need to monitor their own nonverbal behaviors, or worry about providing adequate feedback to their interlocutor (Walther & Tidwell, 2001). Freed of these normal FtF tasks, CMC users have more cognitive resources at their disposal and are thus able to reallocate these resources to the message construction task. Cognitive reallocation processes, combined with affordances for editing and mindful construction of messages and prevention of nonverbal leakage, thus all contribute to the construction of selective self-presentation messages and to the resulting impressions that are formed by message receivers.
The communication channel is another element of the communication model likely to affect the development of hyperpersonal communication. Asynchronous computer-mediated communication, by definition, allows users to read and compose messages at times that are convenient for them without having to be online at the same time as their interlocutor (Walther, 1994). In non-mediated synchronous work groups, such demands on synchronizing time between group members often pose a big problem. McGrath (1991) refers to this problem as "entrainment". Parks and Walther (1994) further explain that entrainment refers to the synchrony among group members with competing time demands. When group members are pressed for time, the focus of the interaction shifts to task rather than social concerns. Having to decide between getting the task done and maintaining social relations, group members often sacrifice a socio-emotional orientation for the sake of a task orientation. In asynchronous computer-mediated interaction, where users interact at times that are most convenient to them, such problems are unlikely to be observed. As a matter of fact, the hyperpersonal communication model argues that asynchronous CMC actually allows for dis-entrainment to occur. As a result of dis-entrainment, CMC users can attend to both task and social dimensions in their interactions, which in turn should result in "relatively more relaxed and enhanced message processing" (Walther & Tidwell, 2001, p. 11).
Although the sender and receiver behaviors discussed above, may encourage interpersonal attraction and intimacy to develop, it is through psychological processes related to feedback that these behaviors are intensified to the extreme level characteristic of hyperpersonal communication. Through a process of behavioral confirmation, CMC users influence their perceptions of one another and end up creating a powerful self-fulfilling prophecy. Because of the selective way in which CMC users present themselves to their communicative partner, the message recipient is likely to form a favorable impression of his or her interlocutor and to act upon that impression. By doing so, the message recipient invites the message sender to actually enact the behaviors he or she is believed to possess, thereby creating an intensification loop capable of producing hyperpersonal outcomes. As Walther and Tidwell (2001) explain, "the hyperpersonal perspective depicts how senders select, receivers idealize, channels promote, and feedback increases enhanced and selective communication behaviors" (p. 11). It is important to note that the four variables that lie at the chore of this theory do not act in isolation, but are all interconnected in numerous ways.
Besides describing the key variables and examining their relation to one another, Walther's (1996) hyperpersonal communication model also identifies the conditions under which hyperpersonal communication is likely to occur. According to Walther (1996), such communication is more likely to be observed when users (a) experience a social identity and are self-aware, (b) when they are physically separated, (c) when they communicate through a limited-cues medium that enables them to use selective self-presentation strategies without risking to be perceived as deceptive, and (d) when they anticipate future interactions with their communicative partners. Because of its increased affordances for offline editing and selective message construction, hyperpersonal communication is also more likely to occur when users communicate through an asynchronous computer medium rather than a synchronous one.