Chapter 6 - Communication Media Selection for Remote Interaction of Ad Hoc Groups
Introduction
Nowadays, no one works completely independently. Almost everyone is part of at least one group, typically several groups at any point in time. Figure 1 shows a typical cooperative work framework [1]. Groups of two or more participants (P) communicate together, share information, generate and organize ideas, build consensus, make decisions, and so on. Being engaged in some common work, participants interact with tools and products (i.e., artifacts of work, A). The main purpose of communication is to establish a common understanding of the work shared between participants. The development of the understanding happens both indirectly and directly. The arrows that link participants to the artifacts denote indirect communication. It happens through the manipulation of shared tools and work objects (e.g., a document, a piece of code). Feedback represents the information gained by the participant who directly controls a shared artifact. Further, the changes applied to an artifact convey information also to the other participants (feed through). Direct communication is denoted by the arrow between the participants and happens by speech or over communication media, such as telephone, fax, and e-mail.
While communicating face-to-face (F2F) by speech is easy for individuals, conducting a long-running, productive conversation through the digital medium is difficult, especially as the group size increases. The difficulty of computer-mediated communication (CMC) and collaboration stands in stark contrast to our natural ability to easily communicate and collaborate with one another in the physical world [2]. As such, there is a need to further our understanding of the effectiveness of the many available synchronous and asynchronous communication media (e.g., e-mail, videoconferencing, or specialized collaboration tools) to support activities in distributed teams. However, not only media properties, such as synchronicity, affect the performance of groups collaborating from a distance. In fact, also who does what matters, that is, the characteristics of groups (e.g., size, history) and tasks (e.g., idea generation, decision making) play an important role.
Mainly because of economic factors, nowadays work is becoming predominantly distributed, bringing significant challenges to effective communication of geographically dispersed groups. In fact, multisite work presents considerable loss of opportunities for rich interaction and a very substantial reduction in frequency of both formal and informal communication between coworkers [3]. Following the trend to business globalization, also software development has increasingly become distributed, with little or no possibility for developers to meet. Among the software development activities, requirements engineering is one of the most communication-intensive and then, its effectiveness is greatly constrained by the geographical distance between stakeholders [[4], [5]].
The definition of requirements is a highly collaborative, interactive, and interdisciplinary process, involving heterogeneous teams of stakeholders [[6], [7]]. It provides another example of a dynamic collaboration that can be accomplished by a virtual, ad hoc group, where some members (e.g., representatives from the customer organization) join the developer group when they can add a value (e.g., to take part in the elicitation of the requirements, in a prototype demo session), and disengage at the end of the task. These groups create temporary networks of independent companies and collaborate as virtual organizations, using information technology to share skills and costs. Thus, such teams are ad hoc in the sense that they tend to be highly dynamic in creation, participation, and release, other than typically being geographically dispersed and cross-organizational. Other common scenarios of ad hoc group collaborations are provided by the partner consortium formed by representatives from different organizations in various sectors (e.g., academic institutions, industry), who have to coauthor a funding proposal for applying to the Framework Programme of the European Commission. Also in the field of software development, several processes, such as document inspections and reviews in general, can be carried out by ad hoc groups. The first contribution of this chapter is the proposal of a new definition of ad hoc group, which builds on the previous ones given in the existing literature on group research and is compliant with the emerging scenario of short-term distributed collaborations.
Due to their temporary nature, ad hoc teams need tools with infrastructure and administration costs kept at minimum. Instead, multipoint audio–video communication poses significant practical barriers to deployment (e.g., expense, infrastructure, support). As such, short-term groups like ad hoc teams often fall back to textual communication only. However, rich media theories on CMC, namely Social Presence [8], Media Richness [[9], [10], [11]], and Common Ground [12], have hypothesized group effectiveness to decrease when media other than F2F are used to accomplish equivocal tasks that require relational cues to be exchanged. They have reported about the inadequacy of text-based communication, as compared to rich media, like F2F and video. Lean media, such as e-mail and instant messaging, lack the ability of conveying nonverbal cues that contributes to the level of social presence (e.g., gaze, tone of voice, facial expressions), which in turns fosters individuals’ motivation and mutual understanding. Nevertheless, these theories have also been criticized for considering the task to execute as an atomic activity. In addition, both Social Presence and Media Richness theories have generally been supported when tested on traditional media, such as F2F communication and telephone, whereas inconsistent empirical findings have resulted when tested on e-mail and video.
These inconsistencies have encouraged a reconsideration of the descriptive and predictive general validity of such theories. Thus, more recent theories have asserted that the effectiveness of CMC depends also on factors other than media richness, such as the degree of synchronicity, task typology, and group temporal scope. Media Synchronicity theory [[13], [14], [15]] and Cognitive-Based View [16] have started to investigate on media effects, looking at the underlying communication processes that happen in group tasks. On the one hand, Cognitive-Based View represents a sort of “Copernican revolution,” which capsizes the existing perspective of CMC theories, looking at communication as a cognitive process: Not only must the sender's comfort with the communication medium be taken into account, but also the motivation of receivers and, above all, their ability to process the message properly. Furthermore, Cognitive-Based View argued that the use of rich media high in social presence should be used to assure attention for small amounts of information, whereas the use of lean media low in social presence causes a decreased motivation, but increases the ability to process large amounts of information during longer periods of time. On the other hand, Media Synchronicity theory distinguishes between the interplay of two different communication processes (the conveyance of additional information, and the convergence to shared views), which vary with the degree of synchronicity of the medium. Furthermore, since a task is not actually atomic, but rather constituted of several subactivities, Media Synchronicity theory suggests that the synchronicity level of media should be aligned with the degree of conveyance or convergence of each subactivity.
The concept of alignment between task and media characteristics is the very basis of the theories of Time–Interaction–Performance [17] and Task/Technology Fit [[18], [19]]. The frameworks proposed by these theories evaluate the appropriateness of task-medium matches, considering tasks no more as somewhat atomic activities, like in Media Richness and Social Presence theories, but rather as complex sets of subactivities and subprocesses, each having different characteristics. Likewise, also group and media characteristics have to be aligned for opportune collaborations to take place. The theories of Common Ground and Channel Expansion [[20], [21], [22]] argue that groups without a history of previous collaborations, like ad hoc groups, do not share any experience and thus, have not established a level of common ground (i.e., shared understanding) sufficient for communicating effectively over lean media. Conversely, members of long-term groups are expected to communicate more effectively over impoverished media, using their shared experiences to compensate for the media leanness.
Drawing upon these theories, we argue that, by understanding the paradoxical effects of rich media high in social presence, groups may be better able to select and use the most appropriate media to accomplish their goals. Hence, the second contribution of this chapter is presenting a critical review of the very many existing, and often conflicting, theories on CMC, which have been combined in a comprehensive theoretical framework for predicting, evaluating, and comparing the goodness of Task-Technology Fits. The proposed framework also builds on McGrath's Task Circumplex [23], which is the most widely used reference model in group research for task analysis, comparison, and categorization.
Finally, the third and last contribution of this chapter is the definition of a high-level research model, adapted from Ref. [24] (see Fig. 2), which can be used to support empirical studies on distributed group research. The theoretical background outlined later in this chapter will show that providing evidence of group task effectiveness can be overly challenging: The effects of technologies are contingent on many factors that differ from situation to situation, according to the context of a group process—that is, group composition, task typology, and communication medium. Thus, also the outcome of a group task (e.g., efficiency, effectiveness, product quality) depends upon the interaction between the group process and these varying contextual factors. Therefore, results from empirical study with communication media must be qualified by the context—the group, the task, and the medium—to which they apply. Through the rest of this chapter, we will update such model to include the variables that define the contextual group-, task-, and media-related factors.
To summarize, the three main contributions of this chapter are:
- (A)
the definition of ad hoc groups, which builds on the previous definitions given in the existing literature on group research;
- (B)
the design of a high-level research model for remote group performance evaluation;
- (C)
the design of a comprehensive theoretical framework, built upon the Task Circumplex model and the very many existing theories on CMC, which can be used to predict, evaluate, and compare the goodness of Task-Technology Fits.
The remainder of this chapter is structured as follows. 2 Task-Classification Frameworks, 3 Group Research deal with task-classification frameworks and ad hoc groups research, respectively. Section 4, instead, frames the complex background of CMC by reviewing the most prominent theories on media effect. In Section 5, we merge the contribution of the previous sections, thus creating two general frameworks relevant to group research on distributed collaboration. Finally, we conclude in Section 6.
Section snippets
Task-Classification Frameworks
When differences in group performance are studied, differences in group tasks must be taken into account with the due regard as well. A widely accepted, general definition of group task is the one given by Campbell, who defined it as “the behavior requirements for accomplishing stated goals, via some process, using given information” [25]. Such definition acknowledges that task characteristics define not only what is to be accomplished, but also how it is to be done. In fact, because required
Teams with No Past and Future: Ad Hoc Groups
Besides task type, another contextual factor that influences group studies is temporal scope, that is, “the extent to which groups have pasts together, and expect to have a future” [17] (p. 149).
Work groups are today increasingly nimble and subject to frequent changes [39]. This underlying idea in ad hoc groups is that of a small entity, highly dynamic in creation, participation, and release, formed to accomplish the goal at hand (e.g., solve a specific problem), and then, disband as soon as
CMC Theories
As geographically dispersed individuals more and more communicate via computer, understanding the effectiveness of the very many available media has become vital. Media are usually classified in the time/space matrix (see Fig. 6), according to both the spatial dimension (collocated/distributed, i.e., where interaction occurs) and the temporal dimension (synchronous/asynchronous, i.e., when the interaction occurs). For instance, F2F communication allows synchronous interaction and requires
Managing the Context: The Effects of Task, Media, and Group Factors
The theories discussed in previous sections have framed a complex theoretical background for the selection of communication media for opportune remote group collaboration. Messages communicated to a group on channels that are inappropriate to the context may be misinterpreted by recipients or may be otherwise ineffective with regard to their intended purpose [[71], [72]]. In group research, context is defined by the group, task, and media factors. In 2 Task-Classification Frameworks, 4 CMC
Conclusions
In this chapter, we have reviewed a large body of theories related to group and group tasks, as well as CMC theories. In particular, we reviewed McGrath's Task Circumplex framework, the most widely used model to categorize tasks, and objectively evaluate and compare their complexity in group research. This chapter also contributed to the study of a particular kind of short-term, dynamic groups, namely ad hoc groups, for which we have reviewed the existing literature and proposed a new
Acknowledgments
We are grateful to Daniela Damian for reviewing and helping us to improve the two frameworks proposed here.
References (74)
Toward understanding the role of tasks in behavioral research
Acta Psychol.
(1969)Task complexity: Definition of the construct
Organ. Behav. Hum. Decis. Process
(1986)- et al.
When does the medium matter? Knowledge-building experiences and opportunities in decision-making teams
Organ. Behav. Hum. Decis. Process
(2003) - et al.
Human–Computer Interaction, third ed.
(2003) - et al.
Distance matters
Hum. Comput. Interaction
(2000) - et al.
- et al.
Requirements engineering challenges in multi-site software development organizations
Requirements Eng. J.
(2003) - et al.
An exploratory study of facilitation in distributed requirements engineering
Requirements Eng.
(2003) - et al.
- et al.
Requirements elicitation: A survey of techniques, approaches, and tools