< CSCW'98 Home
 < Program Top Level

 - Workshops
 - Tutorials
 - Doctoral Colloquium
 - Monday Program
 - Tuesday Program
 - Wednesday Program
 - Plenaries
 - Papers
 - Panels
 - Demonstrations
 - Videos
 - Interest Groups
 - Title Index
 - Author/Presenter Index

Saturday, November 14, 1:30-5:30, and Sunday, November 15, 9:00-4:30, in Suite 3953

Ten participants were accepted from the disciplines that form CSCW, including computer science, cognitive science, sociology, etc. The abstracts of their work appear below. During the Colloquium, participants give a short informal presentation of their work, followed by an open discussion of their work. At the end of the session is a general discussion of career issues and research in CSCW.

Students welcome interest in their work. To contact Colloqium participants, please use the email addresses provided below. For more information about the Doctoral Colloquium, please contact Colloquium Co-Chairs Gary M. Olson (gmo@umich.edu) and Judith S. Olson (jsolson@umich.edu).

Panelists

• Judith S. Olson, University of Michigan, USA
• Gary M. Olson, University of Michigan, USA
• Carl Gutwin, University of Saskatoon, Canada
• Atul Prakash, University of Michigan, USA
• Joerge M. Haake, German National Research Center for Information Technology, Germany
• Yvonne Waern, Linköping University, Sweden

Student Presentations

• Grayson: The Effect of Proxemic Information in Video Mediated Communication
• Fry: Computer Mediated Communication Across Divergent Research Networks
• France: A Comparison of Video-Mediated, Face-to-Face and Audio-Only Group Communications
• Garbis: Communication and Co-ordination through Public Representations
• Nam: An Investigation of Multi-user Design Tools for Collaborative 3-D Modeling
• Hettinga: EVOLVE: EVOLutionary Aspects of Vidoeconferencing Explored
• Campbell: Concurrency Control for Real-Time Diagramming
• McDonald: Supporting Dynamic Recommendations in Organizational Information Systems
• Raybourn: Designing the DomeCityMOO Collaboratory
• Bergman: High-Level Requirements Analysis for Systems in Complex Work Settings

Abstracts

The Effect of Proxemic Information in Video Mediated Communication

David Grayson (david@mcg.gla.ac.uk), Multimedia Communications Group, University of Glasgow, U.K.

Proximity is arguably the most basic form of non-verbal communication and is known to affect the way that an interaction occurs, influencing factors such as dialogue, persuasion, trust and length of an interaction. So far it is unknown however whether proximity can have similar effects in Video Mediated Communication (VMC).

To investigate the possible effects of a stranger appearing close or far away, an experiment was devised simulating a financial advice transaction using a multimedia banking kiosk, where the financial advisor appeared either very close or far away. The experiment showed that when the financial advisor appeared very close, interactions were longer, with the customer saying more, making more interruptions, and having more turns than if the advisor appeared far away.

While this research indicates that perceived proximity may indeed have behavioral implications for interacting across a video link, future research aims to investigate further the precise nature both of the consequences and the reasons behind them. As part of this, Conversational Games Analysis (CGA) is used to examine the functional differences in the dialogue as well as the structural. Other issues raised by the research include the nature of familiarity and social presence.

Computer Mediated Communication Across Divergent Research Networks

Jenny Fry (j.fry@bton.ac.uk), School of Information Management, University of Brighton, U.K.

Developments in electronic networks, such as the Internet, provide the potential to alter scholarly communication patterns and work organisation radically. The focus of this study is the mutual interaction between electronic networks and disciplinary culture and the consequences of cultural differences for the uptake and use of such networks. Knowledge domains within academia are not homogenous, each discipline has a distinctive social and epistemological structure which leads to variations in the communication system which underpins academic research. These domains can be categorised into four general types: Pure science; applied science; arts and humanities; and social science. A number of authors have devised typologies that outline the social and epistemological processes which span the disciplines within each group. The relationship between these differential cultures and electronic networks will be explored using in-depth interviews with networks of researchers from several divergent specialisms. Analysis of pilot interviews has revealed domain differences in the purpose, frequency, and perception of electronic network use.

A Comparison of Video-Mediated, Face-to-Face and Audio-Only Group Communications

Emma France (emma@mcg.gla.ac.uk), Multimedia Communications Group, University of Glasgow, U.K.

Few studies of technology mediated group communication exist. This paper describes a laboratory-based information exchange task comparing the communication and task performance of 36 three-person groups in face-to-face, audio-only, and video-mediated communication (VMC). Analyses revealed no statistically significant differences in dialogue length or performance between the three conditions. However, VMC conversations tended to have most words and speaking turns and those in face-to-face communication the least. This trend was explored using Conversational Games Analysis (Kowtko, Isard & Doherty-Sneddon, 1991), an exhaustive form of coding of the functional use of utterances. The content of 12 face-to-face and 12 video-mediated dialogues was coded. This showed that significantly more interactive work tended to be required in VMC to complete the task. It is proposed that impoverished visual feedback cues, novelty and remoteness in VMC make it more difficult for the participants to reach mutual understanding, and hence more difficult to complete the task.

Communication and Co-ordination through Public Representations

Christer Garbis (christer.garbis@tema.liu.se), Department of Communication Studies, Linköping University, Sweden

In my thesis work I am investigating the way in which teams of operators engage in co-operative process management, such as nuclear power plant control, use 'public representations', i.e., the artifacts representing information in such a way that it is commonly accessible and available to all team members at the same time (for example, a wall mounted electronic display). I am particularly interested in the role that these representations play for the operators' collective assessment and awareness of the state of the system that they are responsible for operating. In addition, I am researching the differences between accessing information through a 'public representation,' such as a fixed line diagram in the underground line control, and through 'private representations,' such as a single-user computer screen. It is my belief that the role and function of 'public' and 'private' representations in the above settings should be carefully studied so they can be designed in a more tightly coupled and integrated way in order to provide the operators with a sufficient and flexible mode of information.

An Investigation of Multi-user Design Tools for Collaborative 3-D Modeling

Tek-Jin Nam (Tek-Jin.Nam@brunel.ac.uk), Department of Design, Brunel University, U.K.

The objective of this research is to help designers working in teams by providing an improved collaborative design environment. The focus is on the investigation into specific issues and requirements for the development of multi-user CAD systems for collaborative 3-D modeling. By examining means for incorporating shared design workspace into conventional design workspace, we propose new mechanisms to transform existing CAD tools into collaboration-aware systems. From an initial experimental study of the team design process and a series of prototype development of collaborative CAD systems, a theoretical framework has been proposed and applied to the new collaboration-aware design systems. The result of the research will lead to the new generation of design tools to support team design tasks improving efficiency and effectiveness of team working.

EVOLVE: EVOLutionary Aspects of Vidoeconferencing Explored

Marike Hettinga (hettinga@telin.nl), Telematica Instituut, The Netherlands

The EVOLVE project focuses on evolutionary processes that take place after the introduction of videoconferencing in medical teleconsultation sessions. Evolution refers to what happens with patterns of work, including patterns of using a technology for particular purposes. As these patterns often divert from the patterns initially expected by designers, we believe that evolutionary processes are an important factor for the successfulness of the introduction of new technology. EVOLVE aims at yielding design guidelines for a better support of these evolutionary processes. These guidelines concern the technology (the "technical system"), as well as the organization of the use of the technology (the "social system"), and the relation between the technical and social system.

Concurrency Control for Real-Time Diagramming

Jeffrey D. Campbell (jeffc@sis.pitt.edu), Department of Information Science and Telecommunications, University of Pittsburgh, USA

Diagrams represent a design, concept or object. Multiple users working together simultaneously to create a diagram can interfere with each other's work. At a minimum this results in lost productivity. In undetected, the interference can cause inconsistencies or errors in the diagram greatly reducing its value. A concurrency control mechanism is needed to maintain integrity for collaborative diagramming. The method described here focuses on identifying logical units of work for collaborative diagramming. These units are analogous to transactions in a database system. This emphasis on transaction identification is a key distinction between this technique and prior CSCW concurrency control approaches. The improvement in transaction identification along with an implementation of split transactions reduces resource blocking, a problem generally found in applying locking techniques from database to CSCW applications.

Supporting Dynamic Recommendations in Organizational Information Systems

David McDonald (dmcdonal@ics.uci.edu), Department of Information and Computer Science, University of California, Irvine, USA

This work explores how information systems can be augmented to assist users in finding other individuals who are likely to have specialized, expert information that they need. In particular, this work considers the social and cognitive mechanisms that people use to find candidate sources of expertise. I recently completed a field study of information finding and sharing in a software development organization. The social and cognitive mechanisms identified during the study will be used to design and implement a system that can assist users with finding potential experts. The design and implementation of a system concomitant with an analysis of subsequent data are work in progress.

Designing the DomeCityMOO Collaboratory: A Multi-User Simulation in a Text-Based Networked Virtual Environment That Supports Non-Scripted Interactions Toward Intercultural Understanding

Elaine M. Raybourn (emraybo@sandia.gov), Sandia National Laboratories, USA

This proposal argues that designing a multi-user social-process simulation in a shared virtual environment offers unique opportunities to explore intercultural issues such as identity, power, and prejudice because its collaborative environment is much less threatening than face-to-face. In the proposed Text-Based Networked Virtual Environment (TNVE) also known as a Multi-User Dimension Object Oriented (MOO), players' narratives and experiences provide the basis for discovery and exploration in a virtual "collaboratory." The DomeCityMOO environment is unique in that it advances our state of knowledge of the effects of designing a non-scripted collaborative social process simulation which supports both group and individual intercultural learning in a shared virtual space.

High-Level Requirements Analysis for Systems in Complex Work Settings

Mark Bergman (mbergman@ics.uci.edu), Department of Information and Computer Science, University of California, Irvine, USA

Large scale intra- and interorganizational information systems development has failed to yield useful systems ~50% of the time. It appears that one cause of these failures is the misunderstanding or overlooking of organizational and institutional requirements in the design of these systems. High-level requirements analysis is being created as one way to start to address this problem. It is a new methodology that can be used to gather organizational, institutional as well as technical information system requirements. Research is being performed to determine how to build a high-level requirements analysis framework. Then, research will continue in applying the framework to identify critical individual or combinations of technical, organizational, institutional requirements for a "real life" project that have been misunderstood or overlooked. Beyond this, high-level requirements analysis should allow for increased insights in (1) requirements analysis, (2) complex system design which contains technological, organizational, and institutional factors, (3) how technology, organizations, and institutions co-evolve over time, (4) policy creation in the development and governance of these types of systems, and (5) how to build, maintain, and improve a high-level requirements extraction and analysis system. Altogether, this sets up a research framework that may eventually yield solutions to become much more successful at either implementing large scale intra- and interorganizational information systems or knowing when not to build them.