Johannes Gärtner, Edeltraud Hanappi-Egger
Despite the importance of temporally and spatially distributed work, face-to-face meetings remain important (compare also ). For more than 10 years, group decision support systems (GDSS) was the major approach used to design computer support for such meetings. These approaches reflected the dominant understanding of managing-as-decision-making . In the last years, the "D" was dropped from GDSS, and now the approach is known as group support systems (GSS).
GSS express the idea of supporting not only the decision process, but also nearly all types of cooperative work . About 1990, electronic meeting rooms became commercial products. Many people critiqued the earlier approach of computer-mediated group interaction (see, e.g., [8, 15]). The new idea was to support face-to-face meetings by allowing the participants to have:
- Anonymous input and evaluation,
- Simultaneous data entry,
- Electronic recording and display, and
- Structured interaction.
As a result of these improvements, most of the work concentrates on the design of rooms and the design of supportive software (e.g., [8, 14]). With a few exceptions, these approaches rely on a shared whiteboard and a computer for each participant.
GSS has three crucial drawbacks.
1. Electronic meeting rooms are not widely available and not easily transferred from one place to another, thereby restricting practical usage. Participants need to travel, which is expensive and time consuming.
2. Electronic meeting rooms are expensive facilities (in both building and maintaining them). Given their costs, they are hardly used for ordinary employees for regular meetings.
3. Because all participants have to use a computer, a relatively high and homogeneous level of qualification is needed.
One solution is a simple but little explored alternative: Liquid crystal display (LCD) projection with facilitation. With this approach, a facilitator works with the computer. The content of the monitor is projected onto the walls to make it visible for all participants. Besides driving the computer, the facilitator also structures the meeting and visualizes the contributions of participants. Facilitators thereby use techniques described in Crane  and Klebert et al. . (See the sidebar for technical issues.) LCD projection with facilitation is much cheaper than GSS, and only one personthe facilitator or a member of the grouphas to be able to use the software. Therefore, it can be used in more areas and with more people. A typical setting of LCD projection is illustrated in Figure 1. A facilitator uses a laptop and an LCD panel together with traditional facilitation tools such as a flipchart and a blackboard.
We used LCD in many projects as external consultants, ranging from systems design to consulting, in a variety of contexts. The projects are described in the next section. Drawing on these experiences, we try to tackle two questions. First, we discuss the question of how the use of LCD projection technology changes the way participants actually participate in the meeting. The analysis takes into account types of meetings (opening, agenda setting, etc.). We compare facilitation with LCD projection with two other approaches: facilitation without LCD projection and working with GSS. We then consider how the technology changed the content of meetings. Additional practical details are discussed in the sidebar.
The following reflections are based on the experiences of both authors with facilitated GSS and on the experiences of Gärtner using LCD as a consultant for 3 years. The latter projects were done partly as a researcher at the Vienna University of Technology, and to a growing part as consultant in a small company, Arbeitszeitlabor. For 2 years LCD panels on overhead projectors were used (see Figure 1). Since spring 1997, LCD projectors instead of LCD-panels have been used because they display brighter pictures.
The projects undertaken cover a variety of assignments. About 50 projects dealt with computer-supported shift scheduling. The Arbeitszeitlabor group developed the system Shift-Plan-Assistant (SPA) to the design of shift-schedules for employees doing shiftwork (examples of such design problems are given in ). Typically, the group spent 1 to 4 days on each project. Systems design was addressed in five projects. Development ranged from small applications (a few days of development) to medium-sized systems (up to 10 months).
We typically worked in groups of 5 to 10 people (at most 25). We found that working with larger groups presented more difficulties because of our facilitation technique. Cost was another factor in the decision to keep groups small.
Our guidelines for selecting members of the working group involved two questions: Who is critical for the success of the development or later use? and Who is affected strongly? In nearly all cases the customers already had similar guidelines, accepted our guidelines, or asked us how to select participants. The actual selection involved a compromise between involving people and keeping the group size small. A typical working group in shift scheduling consisted of the plant manager, the personnel manager, two foremen, and two shop stewards. A typical systems design group had several members of the development team, two or three customer representatives, and two to five potential users. Projects took place mainly in rooms at the customer's sites.
The consultant had a dual role as a domain expert and as a facilitator (Fach- und Prozeßberatung). Such a mixture of specific domain knowledge and skills as facilitator is not unusual in our fieldat least in German-speaking countries.
The facilitation techniques used were practically the same in all cases. Our group used typical facilitation techniques (compare [10, 11]). Such facilitation strongly shapes the course of action within the meetings but has to be considered standard to run such meetings. Crucial to this approach of facilitation is the neutral position of the facilitator concerning content. Serving as both facilitator and expert creates the risk of role conflicts. In our approach we restricted our expert role to input concerning technical alternatives and information about consequences of decisions but tried not to prescribe the decisions of the workgroups.
Dickson et al. compared different modes of facilitative support for GDSS. First, they analyzed the "chauffeur-driven" mode; that is, an individual drives the system at the direction of the group members. He or she implements features of the GDSS for the group but does not assist the group with the process. Second, they analyze a facilitator-driven mode, whereby the facilitator directs the group members to what GDSS features to use and when to use them. Dickson et al. found the chauffeur-driven facilitation mode to have an advantage but suspect that facilitation, to be effective, must be much richer and more flexible than it was employed in their study. These assumptions agree with our approach of flexible facilitation.
At the beginning of each meeting, we spent a lot of time refreshing the general ideas and results of earlier meetings. In systems development, we also devoted meetings to clarify notions and to gain a common understanding of the problem area. If feasible, we spent time at the workplace. We worked with many scenarios and used mockups and prototyping.
Designing and running meetings tackles a huge number of issues. The following section concentrates on four issues:
- Planning general aspects of the meeting (determining optimal working group size and setting up the stage),
- Using the laptop and writing the draft report,
- Addressing specific issues and techniques of facilitation, and
- Deciding what software to use.
Planning general aspects of the meeting
Determining optimal working group size
Each working group typically had 5 to 10 members. This is comparable to the size that small group facilitators work with and to the number of people many GSS are designed for. In our experience, smaller groups are easier to work with. In large groups, of 20 or more, we started to split facilitation between a facilitator who worked with the laptop and a regular facilitator. This is also about the size at which facilitators (at least in our experience) started to work in teams of two.
Members of the working group
GSS are only feasible with members who have minimal computer experience, because they have to drive the system at least partly by themselves. This implies severe restrictions: People who do not qualify, and people who fear that they have inadequate skills cannot participate. No such restrictions apply when using LCD.
Setting up the stage and introducing the LCD
Using LCD during workshops was not itself a big issue, but it nearly always made customers a little curious. However, their attention soon was occupied by issues of the workshop itself or by solving technical problems, (such as where the power supply is or on which wall to project best).
The reactions of different groups of participants were similar. After a few words by the participants, the projection system was hardly an issue. If there were problems with readability (e.g., direct sunlight, or the facilitator's forgetting to switch on the LCD while working on the laptop), the theme of LCD popped up again. After the meeting or during breaks the issue often resurfaced. Typical remarks were "We should use something like this! It's good to have the draft report growing while working!"
Using the laptop and writing the draft report
Setting the agenda and displaying the draft report
As in many facilitated meetings, facilitators proposed an agenda at the beginning and the working group then refined it. The preparation typically was a matter of efficiency. If for any reason no agenda had been prepared in advance, regular facilitation techniques were applied to develop the agenda (see the discussion of brainstorming and voting in the next section, "Addressing specific facilitation issues and techniques").
The agenda typically was written on a flipchart and in the text processing system. A flipchart is necessary so that the agenda is visible all the time, The agenda is also needed on the laptop to make sure that contributions are written in the same order. If agendas changed strongly during the meeting, it was sometimes difficult to update two lists at the same time. If we could not keep up both, the agenda on the flipchart was the one that was maintained.
Working with the laptop
The literature says that people compete to access the keyboard and the public screen [1, 16]. In our experience, this never happened in a single project with external clients. We assume that in our setting, group structures were already highly defined and the facilitators' role were unchallenged. The situation described happened only when there were two facilitators.
Typing skills of facilitators are highly relevant. It is less a question of making no mistakes (participants often laugh and it seems to reduce pressure if even the consultant makes mistakes) so much as speed of writing. Slow typing (or taking too much time to refine layout and eliminate bugs) leads to humbling in the group. When we began using this work approach, it was difficult to facilitate and use the computer at the same time only in the first workshops.
Writing the draft report
Draft reports developed interactively tend to become more extensive than normally. Typing is quicker than writing on a flipchart and there is no space limit, so more things are written down and things are written down in more detail.
It is difficult to structure the draft report; for example, formatting and layout may take a lot of time and interfere with the discussion. If discussion gets heated, people stop waiting for things to be written down. Then the facilitator has to summarize or ask for a short break. Such breaks or even repetitions were always accepted by the participants without complaint.
Writing draft reports this way has two main advantages. First, it is immediately evident whether something was included in the draft report. If people realize that their contributions are written down, they hardly repeat them (as known from regular facilitation). Second, there is no fear of changes by the editor. ("It is good to know that the draft report will not state something different.") People often referred to earlier draft reports as being very good, complete, and correct.
Reference to earlier notes
Taking notes changes strongly when using LCD. Typically, space for posting (for example, flipcharts) is highly restricted in everyday facilitation settings. Compared to such settings LCD projection is an improvement because there is no space limit.
Conversely, referring to earlier notes becomes more difficult. Whereas in regular workshops each participant can check the contributions on the wall directly, with LCD projection it is necessary to ask the facilitators to switch back in the document. This is difficult in two ways: first, participants must talk with the facilitator, and second, the facilitator has to find the note within a large amount of text.
Bringing a small printer into the meeting did not help much. Single printouts are of little use. They can not be posted on the wall, as letters are too small. Printouts for each participant are feasible only during breaks. Instead of bringing our own printer, we ask customers to print if printouts are needed. This typically was possible as meetings were at customer sites.
Addressing specific facilitation issues and techniques
Brainstorming and other idea-generation techniques; anonymous input
In regular facilitation, brainstorming is conducted using two typical methods. Either the facilitator writes contributions from participants on a flipchart, or the participants write their ideas on small slips of paper. The slips are stuck on a wall either by the participants themselves or by the facilitator. Problems typically arise from lack of space (only a few words can be written on a slip of paper to maintain readability) and clarity (it is often difficult to understand the idea). In theory, inputs can be anonymous in larger groups if the facilitator collects the slips. In very small groups, this would not work anyhow. Nevertheless, even in large groups this only partly works. Often contributions have to be explained, and contributors have to choose between two alternatives: either to stay anonymous (then the slip of paper typically is dropped) or to explain (and hence lose anonymity). In our experience, participants nearly always prefer to explain.
Using LCD allows for both techniques. Flipcharts are substituted by an LCD screen; slips of paper are used in the traditional way. When reviewing the slips the facilitator types the contributions. Reordering, adding headlines, and duplicating slips to attach them to several headlines become much easier.
The problem of anonymity and assessment of contributors occurred too when we used GSS. Also in GSS, people had to explain some of their ideas. GSS were not much quicker either. Participants needed time to read the contributions. The only difference was who had to type (participants with GSS, the facilitator with LCD). There are some advantages of typing into a GSS instead of writing on slips of paper (e.g., more space and better readability); however, there are also drawbacks (symbols and drawings are more difficult to produce when typing).
The support for anonymous voting is considered to be one of the main advantages of group support systems. We think that this advantage is exaggerated.
In regular facilitation voting is done in several ways, for example, anonymous voting and direct voting, which uses so-called "1-point questions," where participants attach a colored point to a flipchart (see Figure 2).
In our case, the most important aspect of voting was to get comments to questions like the one in Figure 2: What should we do to improve the content of the workshop? What is the reason for the one very negative vote? Such voting takes only a few minutes; the discussion takes longer.
The focus of such a voting session is feedback on reasons for a specific judgement and ideas for improvement. Similar considerations arise in most voting; for example, assume a voting of 5 to 3 for the question "Should we apply strategy A?" In company settings, such a voting typically should not lead to going to the next issue. On the contrary, a discussion is needed to understand the reasons for differing judgements and find a common ground. Anonymity is not important in such discussions; unanimity is.
Voting on quantitative issues and selecting from a large number of options are difficult. In these cases, again, anonymity has never been a big issue, but portraying the voting in a way that allowed for clear and simple decision making was difficult, especially for the facilitator.
In all the workshops we conducted, the number of anonymous votes was less than five. The number of informal votes was very high. Typically, there were two to five informal votes in each workshop.
Summing up, facilitation supplies a high number of easy and quick voting techniques. The lack of anonymous votes (compared with GSS) was no real issue in our projects.
Active participation and discussions between participants
The degree of active participation strongly depends on the facilitator (e.g., how and how often he or she asks for input). These limitations are comparable with limitations in regular facilitation, and in facilitation using GSS.
Working with subgroups
It turned out to be very difficult to work with subgroups if they used a laptop to write their draft report. Different styles of editing, different levels of preciseness, and a number of technical problems made it difficult to combine the draft reports. After negative experiences on two projects and given the amount of technical hassle, we stopped asking subgroups to write their draft reports with laptops. In further meetings, subgroups worked with slides or flipcharts. Their notes were incorporated into the general draft report either after the workshop or while they presented their results.
More advanced facilitation techniques
The technical requirements for using LCD are not very high. Therefore, the room may be used in many different ways, allowing for more advanced facilitation techniques such as dramatization, drawing, giving talks, or other physical activities to mobilize participants. GSS typically restrict such activities by the furniture of the room. Additional rooms or very big rooms were necessary.
Deciding what software to use
Evidence exists that using sophisticated software might raise problems in computer-supported groups. Mark et al.  questioned whether users without much experience and situated in face-to-face meetings can portray their ideas in a hypermedia format within reasonable training time.
Similar considerations apply for LCD projection. Keeping an overview for a single document (e.g., the draft report) is already difficult. Sometimes this alone confuses participants; moving between programs only adds to the complexity.
We tested documents that reflected the expected course of the meeting (specific input boxes, layout, etc.). We only used it in a few workshops. The actual course of the meeting was different from the expected way and the prepared layout, input boxes, etc. hindered a lot.
Therefore, we typically work with simply structured software (e.g., word processing and spreadsheets and simple or unstructured documents). Only under specific circumstances did we introduce specialized simulation and design tools.
The use of LCD projection works well with group processes. Facilitation becomes more difficult but in our experience remains feasible. Using LCD projection also eases the writing of draft reports. The lack of features of GSS (anonymity, parallel input, etc.) does not substantially hinder group work.
We do not discuss the general advantages of computer support for accomplishing tasks, but rather we outline only the main differences of LCD projection versus regular facilitation and GSS.
Differences between LCD projection and regular facilitation
There are three main advantages of using LCD projection in workgroups versus regular facilitation: (1) issues that are otherwise too technically complex can be tackled, (2) more people can participate simultaneously, and (3) documentation of the process becomes easier.
Tackling technically complex issues
Technical complexity may hinder working groups a lot. E.g. if the working group has to wait for several minutes until a calculation is finished, then the discussion is interrupted. If this happens often the participants tend to get very upset and the work becomes inefficient. Computer support may resolve these problems by the speed of calculation.
Problems may technically be too complex to be handled otherwise in the workshops. For example, the design of a new shift schedule for shiftworkers induces heavy calculations. At the same time, a high number of variations have to be assessed to be able to get a feeling for the influence and combination of factors.
Another example for technical complexity is that problems might get too complex as a result of the number of single issues involved. Space on walls is limited, and it is burdensome to rearrange contributions. Performing this task on laptopusing tables and outlines, footnotes, or commentsmakes it possible to tackle a higher number of single issues.
Involving more people in systems design
With LCD, more participants can be involved in the evaluation of prototypes of computer systems, making tight interaction between a higher number of designers and users feasible. The role of the facilitator remains crucial: he or she has to make sure that participants can make their contributions and understand which issues are discussed
Making documentation easier
Screenshots using LCD are extremely useful. Prototypes and simulations can be more easily captured than with other media. A further advantage is the reduction of work in writing the draft report later.
Differences between LCD projection and GSS
The differences in both approaches regarding anonymous input, parallel editing, and voting were not found to be crucial in the settings where we used the system (see earlier discussion).
McLeod  discusses differences in efficiency between field and lab studies of GSS. One possible reason noted is that the efficiency benefits of GSS will not be realized until the task schedule is spread over several days. Jarvenpaa  et al. stress the need for appropriate training. Our analysis leads us to an even more restricted assumption. GSS has an advantage if all participants are highly qualified and the tasks at hand exploit working in parallel and over longer periods of time. This indicates that highly qualified teams that work regularly or over a longer period together and that have a high number of tasks will work more efficiently with GSS than with normal facilitation or even with LCD.
Another issue for which the advantages of GSS remain unaffected is simulation, whereby participants have to act together.
About 1990, so-called "electronic meeting rooms" became popular, the focus of which was the design of screens and desks that would not interfere in face-to-face discussions. However, in recent years, face-to-face meetings have remained important in decision-making processes, even though technical tools allow for geographically distributed group discussions. The importance of face-to-face meetings has been recognized in the development of according technical support. Lately the popularity of electronic meeting rooms has decreased, partly because of the problems of usability and cost.
Because decision-making processes are based on intensive information processing (compare also ), the computer support in facilitated meetings with LCD-projection provides participants with several applications such as spreadsheets, simulation programs, graphical presentation of changes, and the like. In addition, LCD projection can be used to work with prototypes during system design. This means that LCD projection has the already mentioned benefits of electronic recording and display.
LCD projection works well with facilitation. In some cases, it is necessary to switch to regular facilitation. In contrast to GSS, LCD projection with facilitation does not support anonymous input and evaluation or simultaneous data entry. But, the loss of anonymous input is a disadvantage only under specific circumstances.
From a practical viewpoint, LCD projection is an affordable alternative, because it combines the features of an electronic meeting room and the traditional blackboard (see Table 1). LCD projection provides participants with many technical features without making the face-to-face meeting overly technical.
8. Hiltz, S.R., Johnson, K., and Turoff, M. "Experiments in Group Decision Making: Communication Process and Outcome in Face-to-Face versus Computerized Conferences." Human Communication Research 13, 2 (1986): 225-252.
13. Mark, G., Haake, J. M., and Streitz, N.A. "The Use of Hypermedia in Group Problem Solving: An Evaluation of the DOLPHIN Electronic Meeting Room Environment." In Fourth European Conference on Computer-Supported Cooperative Work (Stockholm, September 1014, 1995), Marmolin, H., Sundblad, Y. and Schmidt, K., eds., Kluwer Academic Publisher, Dordecht, 197213.
16. Streitz, N.A., Geißler, J., et al. "DOLPHIN: Integrated Meeting Support Across Local and Remote Desktop Environments and LiveBoards." In Computer-Supported Cooperative Work (CSCW-94), Chapel Hill, NC, 1994.
Department of Computer Science
Vienna University of Technology
Phone: +43 1 58801-18714
Department of Computer Science
Vienna University of Technology
Phone: +43 1 58801-18713
Methods & Tools Column Editors
Lotus Development Corp.
55 Cambridge Parkway
Cambridge, MA 02142
P.O. Box 260
- Strength of light is most critical. It is awful to work in dark rooms and a nuisance to always have to turn lights on and off when switching between the flipchart and the computer. Additionally, to project the picture you often have to cope with walls that are not white. Therefore, we stopped using LCD displays (LCD panels) and switched to LCD projectors as soon as they were affordable.
- If LCD projection is used at a customer's site, the weight of the system becomes critical. Good systems are available for less than 5 kg (10 pounds).
- Typically, we worked with a resolution of 800 x 600 pixels. A lower resolution made it difficult to project enough information. Higher resolution often resulted in overly small letters and symbols because the size of the projected picture typically is restricted by the height of the room. Font size had to be increased. Generally we could not project more information than with 800 x 600 pixels.
Computer and LCD
- Sometimes the two systems do not work well together. Some computers do not support LCD and the monitor of the laptop at the same time.
- Problems with synchronization sometimes cause flickering
- Computer crashes were not a severe problem. In several meetings we had to restart the computer and lost a little information. Typically, the group considered it an additional opportunity to discuss things or to take a short break.
- Rooms large enough to work in with such working groups are not always available or sometimes do not even exist at a customer's site. It pays to check room size in advance.
- Lighting in projectors has improved strongly in the last few years; however, protection from sunlight is still necessary.
- Bring an extension cord with you.
©1999 ACM 1072-5220/99/0300 $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 © 1999 ACM, Inc.