Department of Computer Science
University of Saskatchewan
Department of Computer Science
University of Calgary
Real-time distributed groupware is coming of age, and it is no longer unusual to see people separated by distance working together concurrently through their computers. These systems typically provide two basic services: a communication channel (text chat, audio, or desktop video, or a combination) and a shared visual workspace (akin to whiteboards or tabletops) where people create and manipulate task artifacts. Unfortunately, collaboration within a groupware visual workspace is often awkward, stilted, and frustrating compared with collaboration in face-to-face settings, particularly when the workspace is larger than the screen and people navigate independently through it. As a result, the majority of users are either early adapters who enjoy experimenting with technology or people with a great need to work together over distance.
To understand why groupware workspaces are awkward, we can contrast them to how people work together in a face-to-face setting over a common surface such as a tabletop. There, people notice and act on a wide variety of perceptual cues to stay aware of what others are doing. People see motions made by others, they hear sounds as people move about and manipulate artifacts, and they track what others are saying. People actively collect information; they quickly shift their gaze from their own work to where others are working and they glance around to get a big picture of what is going on. In essence, people easily and continually track who is around, where others are working, and what others are doing. This awareness of others in the workspace is what we call workspace awareness, the up-to-the-moment understanding of another person's interaction with the shared space. Workspace awareness is used in collaboration to coordinate activity, to simplify verbal communication, to provide appropriate assistance, and to manage movement between individual and shared work.
We believe that workspace awareness is necessary for natural and smooth collaboration in a shared workspace. Current groupware systems, however, provide only a fraction of the information people need to maintain workspace awareness, and this is one reason why they are still awkward. Consequently, we hypothesize that increased support for workspace awareness will improve the usability of real-time distributed groupware. Our goal in this article is to evaluate this hypothesis and to describe an experiment to assess the effects of workspace awareness support on a realistic groupware system.
We compared people's performance on two versions of a groupware interface. Both versions provided users with a scrollable viewport into a portion of the workspace. This allowed individuals to look at different areas of the workspace. Both versions also supplemented this display with a workspace miniaturea small window showing an overview of the entire workspace. This con
figuration is now common in some systems, particularly video games. However, our two versions used different types of miniatures (see figure 1). A basic overview provided only rudimentary awareness information and a radar view provided a higher level of support for workspace awareness. Particular differences were as follows.
- Update granularity. The radar showed workspace objects as they moved; the basic overview was updated only after the move was complete.
- Viewport visibility. The radar showed both people's viewports (the area of the workspace visible in each person's main view), whereas the basic overview showed only the local user's viewport.
- Telepointer visibility. The radar showed miniature telepointers for both users and the overview did not show any telepointers.
We measured three general aspects of groupware usability: how well groups perform with each interface, the efficiency of their collaboration, and the group's satisfaction with the system. We also looked at the strategies that the two different groups used. The study shows that groups using the radar view had significant improvements in speed, communication efficiency, and satisfaction for some of the tasks and that people used more effective strategies for completing tasks. Observations of the sessions and participant feedback provide some explanation for these results and indicate additional design directions.
The paper provides considerable further detail, including how the study was done, the particular metrics used, and the detailed results. The message we want to convey, however, is simple: For researchers, we now have empirical evidence that support for workspace awareness improves the usability of groupware. For practitioners, we show that a miniature overview of the entire workspace is effective for communicating workspace awareness, particularly when it includes the location of others' viewports in the workspace, the location and motion of people's cursors, and the motion of workspace objects as they are moved.
School of Computing Science
Bounds Green Road
London, N11 2NQ, United Kingdom
Phone: +44 (0) 20 8362 5000
F. Paternö and C. Santoro
Via S. Maria, 36
56126 Pisa, Italy
Phone: +39 050 593 289
Avenue Édouard Belin, 7
31055 Toulouse, France
Our work aims to form bridges between techniques such as task analysis, a social view of collaborative activity, and the work of systems designers who require systematic methods to evaluate design choices. An important rationale for this approach is that evaluation of interactive systems is more economically carried out early in the development lifecycle, when redesign in response to problems that have been identified is more feasible. We have applied the method to the design of new communication technology in an air traffic control environment, focusing on the access to information and communication facilities that is given to workers.
The Method for Evaluation of Cooperation, Hazards, and Allocation (MECHA) is a method used to evaluate and compare design options for allocating communication media, especially support for cooperation in an interactive safety-critical system.
Briefly, the method involves the following activities (see Figure 1):
- Task analysis and modeling of the current system and identification of the possible main design options. This activity allows designers to model cooperative applications and capture a wide variety of possible temporal relationships among activities. The design options differ by the media available, how the tasks are performed and allocated, and the choice of artifacts and representations that are appropriate to support such tasks.
- Selection of specific sequences of tasks and development of corresponding scenarios. This activity involves creating a contextwith actions, technology, responsibilities, and environmentto perform the sequence of tasks identified for a particular design option.
- Analysis of potential hazards and problems belonging to a predefined set of categories. For each design option we form hypotheses about potential problems in allocating tasks and performing them. The result of this analysis is tabulated, indicating for each problem the associated causes, consequences, protections, and design recommendations.
- Analysis of the coordination required among participants and how the design supports such coordination work. Many types of communication can occur in a work environment, such as a control room, and they have to be carefully captured in order to effectively support them in an environment enriched with new technologies.
In short, using this method the designer compares the different options according to the implications that design choices have for individuals' performance of tasks, for how they prevent or mitigate possible deviations in user behavior, and for the coordination of the activities of several collaborating individuals. To capture the commonalities of and differences in design options, each is characterized by the specific media and representations they provide for controllers, the specific ways that tasks are performed and allocated, and the scenario of use that is considered. Comparison of the design possibilities will be guided by a collection of criteria that address usability (for example, task efficiency, effective coordination, and mutual awareness) as well as safety (including analysis of user errors and their impact).
This analysis shows how it is possible to improve the design of new communication technology in a safety-critical context in order to support users even when their behavior differs from expected behavior and still ensure flexible co-operation mechanisms. Our work has been carried out in the MEFISTO European Project (giove.cnuce.cnr.it/mefisto.html). We envision further work on developing tool support for the proposed method.
James "Bo" Begole, Mary Beth Rosson, and Clifford A.
Blacksburg, VA 24061
firstname.lastname@example.org, email@example.com, firstname.lastname@example.org
Application-sharing systems such as Microsoft NetMeeting and SunForum allow people to work together simultaneously using existing single-user applications. This form of data conferencing is referred to as collaboration transparency because the shared use is transparent to the application and its developers; no application source is modified or inspected to achieve collaborative use. Several collaboration-transparency systems are available free on widely used platforms, allowing even casual computer users to try them out. As a result of their wide availability, we became interested in how well current systems support collaboration.
We found that, primarily because of the implementation architecture used by conventional collaboration-transparency systems, several factors limit users' ability to work together effectively. First, all participants see the same view of shared data at nearly the same time. Second, to avoid conflicting inputs, only one user at a time is allowed to control the application. These two factors limit collaborators' ability to work independently in parallel. Additionally, only rudimentary activity information is provided, limiting group members' awareness of one another's activity. As a result, the usefulness of conventional collaboration transparency is restricted to tightly coupled styles of collaboration.
To support loosely coupled styles of collaboration, we designed an alternative approach for object-oriented application platforms called flexible collaboration transparency. Our approach is based on a replicated architecture in which selected single-user interface objects are dynamically replaced by multiuser equivalents. We incorporated our approach into a Java-based collaboration-transparency system called Flexible JAMM (Java Applets Made Multiuser), which is available at simon.cs.vt.edu/jamm/. Figures 1a and 1b show a text editor before and after being shared using Flexible JAMM.
To validate that flexible collaboration transparency is truly an improvement over conventional systems, we conducted an empirical study of collaborators performing both tightly and loosely coupled tasks using Flexible JAMM versus a representative conventional system, Microsoft NetMeeting. Participants were able to complete the loosely coupled task significantly more quickly with the same accuracy when using Flexible JAMMwhich allows simultaneous text entrythan when using NetMeeting, which requires turn taking. For the tightly coupled task, we were concerned that having to use new multiuser interface elements would hinder performance when using Flexible JAMM, but we found no significant difference in completion time or accuracy between the two systems. Overall, participants preferred Flexible JAMM and commented that they felt involved in the collaboration, whereas with NetMeeting, interface difficulties and explicit turn taking made them feel like they were not working together. The results support our aim of providing flexible support of multiple styles of collaboration in a collaboration-transparency system.
By supporting the range of tightly to loosely coupled collaboration styles, flexible collaboration transparency aims to allow collaboration to flow naturally between the two styles. This capability enhances the collaborative use of the ever-growing body of legacy, single-user applications.
Figure 3. (a) A single-user editor called Stylepad. A
scrollable panel contains the document. (b) Stylepad shared via
Flexible JAMM. The single-user scrollable panel has been replaced
by a multi-user version, which provides the original scrollbar
interface plus a miniature view of the document in the window on
the right, called a radar view. A uniquely colored and shaded
rectangle indicates each participant's scroll position. Three
participant positions are shown here: PC User and Solaris User
overlap near the top of the document, and Unix User is near the
center. Although there is no visible indication, the text editing
component has also been replaced by an extension that allows
simultaneous editing by multiple users.
©2000 ACM 1072-5220/00/0700 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2000 ACM, Inc.