Exploring the Boundaries of Successful GSS Application: Supporting Inter-Organizational Policy Networks. Hans de Bruijn, Gert-Jan de Vreede. Faculty of ...
Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999
Exploring the Boundaries of Successful GSS Application: Supporting Inter-Organizational Policy Networks Hans de Bruijn, Gert-Jan de Vreede Faculty of Technology, Policy and Management Delft University of Technology, The Netherlands email: {hansb, gertjanv}@sepa.tudelft.nl Abstract Technologies used to support group work are based on and contain underlying assumptions regarding how people work together. The appropriateness of such assumptions is an important factor in determining the successful employment of the technology. This paper uses an action research approach to explore the boundaries of effective GSS application by challenging the basic assumptions built into GSS. This exploration is carried out in the context of a particular arena in which groups have to interact to reach a certain goal: inter-organizational policy networks. From 9 cases it appears that GSS are most effective in the orientation phase of inter-organizational policy making, while GSS should be avoided during the separation phase where winners and losers can be identified. During the package deal phase of an inter-organizational policy making process, GSS may have added value to offer, but it should be employed with caution. To this end, a number of meeting design guidelines is presented.
1. Introduction Group Support Systems (GSS) are one of the few examples of an information technology that was conceptualized by university researchers and subsequently with reasonable success commercialized and applied in industry [13]. Since their commercial inception in the 1980s, GSS have become an important and popular area for research. GSS are believed to offer added value for collaborative problem solving by providing support for parallel communication, anonymous interaction, and automatic recording of meeting minutes, see e.g. [21]. A substantial body of knowledge has emerged concerning the effects of GSS on group processes and outcomes, but this knowledge is somewhat inconclusive. As much as field studies paint a predominantly positive picture, lab experiments present mixed results [10, 11]. This is confirmed in a number of meta-analyses of lab studies. Although GSS seem in general to have a positive effect on group
productivity and decision quality [4, 9, 19], the extent to which this positive effect occurs seems to depend on a number of other factors, such as task-fit and group size [9]. Also, meta-studies show varied results with respect to meeting participant satisfaction, see also [24]. Often, the inconsistencies in results are explained by pointing out the incomparability of tasks and GSS setup studied and the differences in experimental design and method of analysis. Yet, the positive findings from the field indicate that GSS do have a distinct potential to provide effective support for collaborative problem solving, if they are skillfully employed see e.g. [22]. Therefore, the conclusion must be that GSS do not have an added value under all circumstances. In other words, when studying GSS, it is especially important to get a clear sense of the context in which the GSS is applied, see also [21]. For the purposes of this study, we believe it is valuable to look at contacts in which GSS are applied in unsuccessful as well as successful manners. Accounts of GSS application failures (or “almost failures”) may be more insightful than yet another account of a successful GSS application. It can be argued that one can learn more from unsuccessful experiences by comparing them to successful experiences than from successful experiences alone [26]. Also, lessons learnt in critical situations can have serious implications for GSS meeting design and for the design of GSS meeting facilities. In other words, we have to explore the practical boundaries of effective GSS application. The application area that is addressed in this paper concerns interorganizational policy networks. Policy making is known to be an area in which there is a need for group support but effective GSS application is anything but straightforward [15]. To study the application of GSS and its added value in the context of inter-organizational policy networks we take a number of assumptions regarding group interaction and behavior underlying GSS as a point of reference. Hence, we examine the extent to which these assumptions hold. An example of such an assumption is that people in an (electronic) meeting wish to share information. We take this
0-7695-0001-3/99 $10.00 (c) 1999 IEEE
1
Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999
perspective because experiences show that the application of group support technologies or the technologies themselves sometimes fail because of incorrect or incomplete assumptions of how people work and collaborate. Examples of such technologies are Xerox CoLab and the early version of the Coordinator [3]. In summary, the research objective of our study is to get insight into the use of GSS in a “difficult” application area. Specifically, we attempt to identify situations in which the application of GSS should be stimulated or avoided. To this end, our study aims to derive a set of GSS meeting design guidelines. These guidelines may go beyond the chosen application area. We pursued our research objective by applying an action research strategy. Action research, ‘an inquiry into how human beings design and implement action in relation to one another' [2, p.4], is intended to have the dual outcomes of action (or change) and research (understanding). Action researchers participate or intervene in the phenomenon studied in order to develop a theory or to apply a theory to evaluate its worth [1]. Action research studies are carried out in four phases: Plan, Act, Observe, Reflect [28]. The remainder of the paper will describe the background of our study and the results of the four action research phases. The next section will introduce the basic assumptions on group interaction and behavior underlying GSS. Section 3 will describe the context of the study, inter-organizational policy networks, and the way in which we choose and carried out the cases that were part of this study (the Plan and Act phases). Section 4 will illustrate in what way and to what extent the GSS assumptions hold (the Observe phase). Section 5 will discuss our findings and propose a number of meeting design guidelines (the Reflect phase). The paper concludes with a short summary and a description of the limitations and applicability of our findings.
2. Assumptions underlying GSS application There are various assumptions on group behavior in meetings that can be identified to have guided the design and application of GSS. Below, a number of these assumptions are presented, organized in six categories. The sources for identifying these assumptions are threefold. First, we looked at GSS product information as presented by vendors of GSS and in popular computer magazines. Second, we looked at general reports on electronic meetings in popular computer and business magazines. Finally, we investigated the functionality of two GSS, GroupSystems and MeetingWorks. We selected these systems because (1) they can be considered the most widely applied commercial GSS, (2) they are comparable in terms of functionality, (3) most academic research has been done with GroupSystems, and (4) we used GroupSystems in our cases, see section 3.3. We recognize that focusing only on two GSS developed from a
predominantly technologically point of view limits the applicability of our findings to situations in which systems are employed that are developed from a social or negotiation point of view such as Resolver and Group Explorer. Assumptions on group behavior are seldom made explicit in the sources studied. We identified them by looking at the implicit implications of certain described features and advantages of GSS. For example, in success stories it is often pointed out that groups generated hundreds lines of output in so few minutes. This implies that it is desirable to generate a massive amount of output. After identifying the list of assumptions, we presented it to a number of experienced GSS facilitators working in university and commercial environments. Their feedback suggested (1) that the list of assumptions was complete, and (2) that they recognized the assumptions from popular business publications and promotional materials. However, recognizing an assumption did not mean that they believed that it always held. Obviously, the fact that an assumption on group behavior can be identified does not imply that all groups behave like that or will behave like that. It just means that certain group behavior is considered to be desired or expected behavior by some of the GSS designers and practitioners. It is important to bear in mind that a number of the assumptions presented below have been subject of various experimental studies by university researchers. It is not the purpose of this paper to compare and contrast these studies with the assumptions, nor to evaluate the assumptions as such. Rather, we use these assumptions as a reference point to allow for a guided and focused inquiry into the application of GSS in the chosen context.
2.1. Meeting processes should be "fair" One of the most frequently stated advantages of using GSS is that they supposedly contribute to the fairness of the meeting process. GSS are promoted by claiming that because of the possibility to communicate in parallel, participants have equal time and equal opportunity to contribute to the meeting process. Also, because of the combined use of parallel and anonymous communication, single individuals will not be able to easily dominate an electronic discussion. Finally, fairness of the meeting process is also implicitly promoted in most of the voting tools. During a vote, the opinion of each group member counts the same. The investigated GSS do not offer dedicated functionality to determine whether someone’s vote should count, for example, three times as much as someone else’s. (In GroupSystems this can be done manually, by writing the vote of a person down and having him or her re-enter the vote the number of times required.)
0-7695-0001-3/99 $10.00 (c) 1999 IEEE
2
Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999
2.2. Meeting processes should be "open" A second category of assumptions relates to the openness of meeting processes. Openness can have different meanings. First, participants should have an open mind toward each other. Second, the main stakeholders should not be excluded from the session. We only refer to the first meaning. The assumption on openness is most prominently illustrated by the often-quoted characteristic of GSS that an accurate, literal record of what went on in the meeting is immediately available. In other words, the GSS facilitates a permanent record of the electronic part of the meeting process. Another illustration of the desired open nature of meeting process concerns the fact that normally contributions are shared among the participants immediately. In other words, all participants can freely see all contributions during the meeting, except for individual votes. An exception is the “use private list” option in GroupSystems that allows participants to generate ideas individually first before contributing them for everyone to see. However, this feature is primarily meant to support Nominal Group Technique [25].
2.3. Meeting processes should be "rational" Often, meeting processes are stimulated to have a certain degree of rationality. This is most noticeable in the anonymity feature of GSS. This is advertised as having the advantage that ideas are only judged and commented upon based on their merit, not on who contributed it. In other words, the only important attribute of an idea is "content". It is not important, even undesirable, to know who formulated it or how it was communicated. Also, body language is often stated not to influence the decision making process. Another indication of the desire to have a rational meeting process is the stress on structuring meetings according to a predefined agenda and setting a number of behavioral and communication rules for the meeting participants to adhere to. Finally, it is often advertised that consensus on issues can be explored and achieved by using the GSS voting tools. The results of a vote indicate to what extent a group has consensus. Subsequent discussions and new voting rounds are proposed to lead a group to consensus.
2.4. Groups should be guided by a process facilitator The fourth category of assumptions underlying the design and application of GSS concern the role of a process facilitator (not a ‘technographer’) in an electronic meeting. In commercial writings the facilitator is often pointed out as being critical to the success of a meeting. The meeting process is assumed to be mastered by a skillful facilitator. The facilitator should be the captain of the meeting process. This is embedded in GSS such as GroupSystems and MeetingWorks. All electronic communication opportunities
are to be controlled by a facilitator, possibly using an assistant (technographer). Moreover, any change in the meeting agenda or process that has an effect on the GSS setup should be made by the facilitator, possibly following the consent of the group as a whole. It cannot be done by individual participants. (A change of the agenda can mean that new issues are put on the agenda, or that existing problem definitions are amended, or that new solutions are taken into account.)
2.5. Groups should exchange as much information as possible One aspect that is continuously advertised with respect to GSS concerns the participants’ productivity. Statements like “The group was able to generate 700 lines of output in under an hour of meeting time” are common. The assumption behind this is that groups should be focused on generating a lot of ideas; more is better. Participation should be as intense, i.e. actively concentrating, as possible to improve meeting efficacy. Also, "outside the box" contributions should be encouraged because this will encourage people to generate more ideas. Another illustration of the productivity focus is the argument that groups in electronic meetings should be large. It is assumed that larger groups reach better decisions and achieve better results because they have more interaction channels and as a result can share more information.
2.6. People are cooperative by nature, with respect to each other and to the meeting process The final category of assumptions addresses the consent meeting participants demonstrate during electronic meeting with respect to each other and the process they participate in. Claims that GSS enable organizations to draw all kinds of people to a meeting to work together in solving a complex problem are grounded in the assumption that people with individual backgrounds, expertise, and reasons for attending a meeting, will work together to reduce unfamiliarity and uncertainty with a problem domain. Also, it is assumed that people will synthesize their ideas with those of others, whenever they are confronted with the ideas of others, e.g. on their own screen or on the public display. Furthermore, anonymity is frequently advertised as a GSS characteristic that lowers the threshold for people to contribute to a meeting. In other words, if participants are not identified during the meeting process, they are expected not to be afraid contributing ideas and to be contributing these honestly and openly. A final assumption relates to the notion of electronic communication using keyboards. GSS designers assume that people are willing to express their ideas in writing.
0-7695-0001-3/99 $10.00 (c) 1999 IEEE
3
Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999
3. The analysis of GSS in inter-organizational policy networks
be aware actions to get the support of B, C, D, and E can interfere with E’s strategies.
In the literature in recent years a great deal of attention has been devoted to so-called inter-organizational policy networks (hereinafter called networks). A network is a collection of organizations that are dependent on one another for the realization of their goals. In this section we first indicate what the characteristics of a network are and we describe the nature of decision making in networks. Research has shown that decision making in networks has certain specific characteristics that differ from decision making in so-called rational models. Next, we indicate that the parties in networks often have a need for electronic group support, such as GSS. Finally, we describe the method of analysis for our study.
Table 1. Different decision making processes Analytical-Rational Problem Problem Solving in a Solving Network Focus on Problems Focus on Interest Focus on Information Focus on Relations Search for the Right Solution Search for Support for a Solution
3.1. Inter-organizational policy networks The literature contains many definitions of networks, see e.g. [7, 18]. In this paper a network is defined as a structure of interdependence between organizations or parts of organizations [23]. The focus is on inter-organizational networks: organizations need each other to realize their own goals and therefore they will have to cooperate. However, cooperation does not come automatically: • The organizations in a network may have different interests. Organization A needs the support of organization B, but B is not interested in cooperation. • The organizations in a network may have different perceptions of their own dependencies. A does not realize that C has strong ties with B, so being provocative against C could harm the relation with B. • There may be a temporal misfit between the dependencies. A depends on B on Task1, but B depends on A on Task2. • Organizations behave strategically (see below) Decision making processes in a network differ from analytical-rational decision making (see table 1). In an analytical-rational approach, an organization A will define a problem X, gather information and try to find the right solution. When A depends on B, C, D, and E his problem is nothing more than a problem perception. B, C, D, and E may have other problem perceptions or may even be convinced that there is no problem at all. A will have to find out which interests the other organizations have and under which conditions they will be willing to cooperate with A. In order to get the support of B, C, D, and E, A may have to broaden the problem definition or will have to wait for a so called policy window. In addition to that, A has to be aware that E has problem Y and will simultaneously try to get the support of B and C. This requires a strategic behavior of A: he has to
For a more detailed description of problem solving in a network, we refer to [6, 8, 17]. From this body of knowledge, it appears that decision making and problem solving in a network has three important characteristics: • Decision making is, at least for a part, a matter of negotiation. Organizations will try to solve their problems in a process of wheeling and dealing with other organizations. • Organizations will have to behave strategically. They have to be aware that other organizations are trying to maximize their interests and will have to develop a strategy to cope with this. • The result of a decision making process may be favorable for organizations B, D, and E and may harm the interests of A and C. So a decision making process may result in winners and losers (which does not necessarily mean that the winners have the right solution for the right problem and that losers have wrong solutions or poor problem definitions).
3.2 The need for Group Support In a network there may be a great need for electronic support for meeting activities. The reasons for this, we propose, are threefold. First, as explained earlier, normally many parties are involved in the decision making process. Each of them has its own interests, goals, problems, and desired solutions. As a result, the combination of goals, problems, and solutions in a network is complex. An organization A that desires to make a package deal with a number of organizations in collection B – G must form a picture of the goals, problems, and solutions of these organizations. GSS appear to enable the efficient exchange of information to build such a picture. Second, in line with the first reason, during the course of a session the participants may learn about the flexibility and adaptability of the goals, problems, and solutions of the other parties. In this way, they obtain a better sense of the room available for a package deal. Finally, the parties involved in the process may have a need for structuring. A final package deal may contain a wide variety of solutions. As a result, participants in a network normally have a strong incentive to contribute to a very great
0-7695-0001-3/99 $10.00 (c) 1999 IEEE
4
Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999
extent their own goals, problems, and solutions. Hence, there is a need to structure the variety of contributions made by the participants, for example, into likely and less likely solutions. For this, GSS can be helpful.
3.3 Method of analysis In order to study to what extent the assumptions presented in Section 2 are appropriate within the context of interorganizational policy networks, we investigated a number of cases that met the following criteria: (1) participants represented a number (more than 2) organizations, (2) that have different interests, and (3) that are inter-dependent. Using these criteria, sessions were selected from a set of all sessions that were organized at the GSS facilities of the authors’ institute in the period from 1996 through the first half of 1998. In total, 9 cases met the criteria described above, together representing 12 GSS sessions. Some characteristics of these cases are summarized in table 2. GroupSystems was used in all cases. The largest number of parties involved was 15 during case 4. Note that the number of parties is not necessarily equal to the number of participants. In some cases, a number of participants represented the same party. However, there were no situations in which participants had to share a workstation during a GSS meeting. A more elaborate description of each case can be found in appendix A. Table2. Summary of selected cases. CaseTitle # of Length # of Sessions (days) Parties 1 Ministerial strategy 4 4 6 development 2 Optimization of 1 0.5 10 waste/recycling chains 3 Inter-municipal cooperation 1 0.5 6 4 Oil and Gas Supply Industry 1 1 15 5 Strategy development 1 0.5 6 university board 6 Environmental Law 1 0.5 6 Enforcement 7 National physical planning 1 0.5 4 8 Provincial working process 1 0.5 4 9 Prison privatization 1 0.5 4 Data was collected from a number of sources. During our study, we used both open and semi-structured interviews (questions are listed in appendix B) to converse with participants and meeting owners. Interviews were done immediately after the session and in some cases up to 10 weeks after the session. To be interviewed participants were in principal selected at random. However, we made it a point to interview participants that made critical remarks about the GSS process and technology during the sessions. Also,
participants filled in satisfaction questionnaires after each meeting, consisting of open questions and questions on a 5 point Likert scale [5]. These sources were augmented by the facilitators’ observations. The facilitators wrote their observations down during and immediately after each session. In some sessions the technographer added a few observations. We used a selective coding technique for analyzing the collected data. This means that we closely examined all collected data, broke them into discrete parts, and labeled these parts. The labels we used were directly derived from the list of assumptions presented in Section 2. If a data element did not seem to relate to any of the listed assumptions, it was not included in the analysis. After labeling the data for each case, we checked whether the data suggested that an assumption did not hold. If this was the case, the case scored a “-“ for that particular assumption (see tables 3-8). If not data was found that conflicted with the assumption, the case scored a “+” for that assumption. It is important to note that the role of the researchers during the cases was only that of a facilitator. Interventions during the session were aimed at supporting the participants in achieving their meeting goals. In other words, during the “Plan” and Act” phases of our action research study, no interventions were planned nor made from a research perspective.
4. Results In this section we describe the results of the “Observe”-phase of our study. For each set of assumptions, the characterization of the 9 cases we investigated is summarized in table 3. The “-“ and “+” symbols can be read as “a little” and “well” respectively.
4.1. Meeting processes should be fair Table 3 indicates to what extent the assumptions applied to the above-mentioned sessions. It appears that assumption 1 is accepted in almost all the sessions, except when there are very large conflicts of interest, such as in cases 3 and 7). Feedback from interviews suggests that participants accept this assumption as a rule of the game that they do not want to challenge. However, they state that equal time contributing to a meeting is not identical to equality in position of power! What is striking from the information in table 3 is that assumptions 2, 3, and 4 score very mixed. Observations and feedback from the participants suggests that there are several explanations for this. First, it appears that position of power is relevant. In a number of sessions the participants were not interested in the content of the opinions generated in the brainstorm. The assumption of equality appeared to clash with the existing power relations in the network. The participants knew which organizations they needed to realize
0-7695-0001-3/99 $10.00 (c) 1999 IEEE
5
Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999
a package deal and were especially interested in the opinions of these participants. For example, in case 7 the central government was more interested in the opinions of large municipalities than of small municipalities. Second, sometimes an idea is only interesting when the sender is known. Occasionally the inclination arose to break through the anonymity and to ask who had which opinion. Some participants mentioned in the brainstorm session their name and the ideas they generated. A number of case 1 sessions were attended by civil servants who restated the standpoints of political administrators. Every now and then they mentioned this, thus indicating that everybody should take these views into account. This clearly was an example of dominant behavior, which was, however, often appreciated by the participants. They knew where they stood as a result. Third, organizations are not equal. In some cases, the equality standard in voting produced lists of solutions, for which however there was insufficient support. The one-manone-vote system clashed with the power relations in the network. One participant clarified this as follows: “I always have a big mouth! That helps [to get my way, GJdV].” It appears that the ‘best’ solution in a network is not democratically chosen. During all cases in which the parties desired to reach a decision, the idea of one-man-one-vote was rejected. This was, for example, very clear in case 8, where the participants had to conclude with a number of mutual agreements. The one-man-one-vote system produced several agreements that did not do full justice to the real power relations and hence were not accepted by the participants. This resulted in a negative perception of voting activities. Said one participant: “I find voting about these issues unclear. Which question are you asking? How do you vote when you don’t agree with an issue?” Finally, the positive scores with respect to assumptions 2, 3, and 4 concerned cases that focused more on the starting phase of the decision making process. This suggests that in this situation, equality gives room to learn about possible package deals. In a number of cases the participants clearly accepted the assumption of equality opportunity and avoidance of dominant behavior. Said a participant in case 6: “Especially in meetings that traditionally last a long time, you have a tendency to keep your mouth shut.” The assumptions appeared to hold especially when “orientation” was a primary goal of the meeting. One participant in case 4: “GSS is good for climbing to cruising level. I now have a feeling of the standpoints and possible solutions of everybody. However, the real work (i.e. decision making, authors) must still begin.”
4.2 Meeting process should be open The assumptions regarding an electronic meeting record score overwhelmingly positive (see table 4). Feedback from the interviews and questionnaires suggest that the
participants often accepted the openness of the process in all sessions simply because the facilitators introduced it as a rule of the game. At the same time, some participants indicated that the process of wheeling and dealing during the creation of a package deal does not agree with the assumption of openness during the meeting (assumption 3). They stated that openness harmed the negotiation process. In one case, some participants were clearly unwilling to input their ideas and suggestions to the system. The views of these parties -which we knew as a result of preliminary conversations- were not contributed because (1) they would undermine too much the negotiating position of the party concerned or (2) would have a negative impact on the desired partnership. In another case, certain opinions were specified very cautiously or not entered at all because they were expected not to go down well with the European Commission. It also occurred that certain ideas and suggestions were not intended for all participants present but only for those with whom the sender wanted to make a deal. Finally, observations in all cases suggested that “openness” was an important factor in preventing full package deals to be realized. Although parts of package deals were achieved in some of the cases, more traditional meeting processes were required by the participants to reach a full deal.
4.3. Meeting process should be rational Overall, the assumptions with respect to the rational nature of the GSS meeting process score very poorly (table 5). The idea of a rational process with a focus on content appeared to clash in almost all cases with the power relations and the nature of the decision making process in a network. Power relations appeared to be of great importance for the participants. Being proved right (sufficient support from the other participants) was more important than being right (a good idea in terms of content). The cases and the feedback from the participants showed that decision making as negotiation clashes in a number of ways with the assumptions. First, it occurred often that more and more issues were introduced in order to make a package deal. During a session the parties seek a package deal, implying that unexpected issues are occasionally added to the agenda. For this reason, it appeared to be difficult in a number of cases (especially cases 5, 7, and 8) to structure the agenda in advance. During such sessions, the participants introduced a number of new issues which appeared to be conducive to the creation of a package deal but which were not on the original agenda. Moreover, these issues were introduced in the final part of the session and at that moment there was no longer a possibility to brainstorm on them. For example, in case 8 the participants were, at the end of their session, in danger of not
0-7695-0001-3/99 $10.00 (c) 1999 IEEE
6
Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999
st
Table 3. Results of the case characterization regarding the 1 set of assumptions. Assumption Case Meeting processes should be "fair" 1 2 3 4 5 6 1. Participants should have equal time contributing to a meeting. + + - + + + 2. Participants should have equal opportunity contributing to a meeting. - + - + - + 3. Dominant behavior should be avoided. - + - + - + 4. Participants' votes should weigh equally. - + - + - + Table 4. Results of the case characterization regarding the 2 Assumption Meeting processes should be "open" 1. It is desirable to have a permanent meeting record. 2. It is desirable to have an immediate and literal meeting record. 3. Participants should freely see all contributions during the meeting.
nd
set of assumptions. Case 1 2 3 4 5 6 + + + + + + + + + + + + + + - + - +
7 8 9 - + + - + + - - + - - +
7 + + -
8 + + -
9 + + +
rd
Table 5. Results of the case characterization regarding the 3 set of assumptions. Assumption Case Meeting processes should be "rational" 1 2 3 4 5 6 1. The only important attribute of an idea is content. - - - - - + 2. Body language should not influence the decision making process. - - - + - + 3. Meeting processes should be structured (agenda, behavioral rules, communication rules). - + - + - + 4. Consensus on issues can be achieved by voting. - - - - - +
7 8 9 - - + - - + - - + - - +
th
Table 6. Results of the case characterization regarding the 4 set of assumptions. Assumption Case Groups should be guided by a process facilitator 1 2 3 4 5 6 1. The facilitator should master the meeting process + + + + + + 2. The facilitator should control electronic communication opportunities. + + + + + + 3. Only the facilitator or the whole group makes changes in the meeting agenda or process. - + - + + +
7 8 9 - - + + + + - - +
th
Table 7. Results of the case characterization regarding the 5 set of assumptions. Assumption Case Groups should exchange as much information as possible 1 2 3 4 5 6 1. Groups should generate lots of ideas. + + - + - + 2. "Outside the box" contributions should be encouraged. + + - + - + 3. Participation should be as intense as possible. + + - + - + 4. Groups should be large to reach better results. + + - + - +
7 8 9 - - + - - + - - + - - +
th
Table 8. Results of the case characterization regarding the 6 set of assumptions. Assumption Case People are cooperative by nature with respect to each other and the meeting process 1 2 3 4 5 6 1. People with individual backgrounds, expertise, and reasons for attending a meeting, will - + - + - + work together to reduce unfamiliarity and uncertainty with a problem domain. 2. People will synthesize their ideas with those of others, whenever they are confronted with - + - + - + the ideas of others. 3. If not identified, people will contribute honestly and openly. - - - - - + 4. If not identified, people will not be afraid to contribute ideas. - - - - - + 5. People are willing to express their ideas in writing. - - - - - +
0-7695-0001-3/99 $10.00 (c) 1999 IEEE
7 8 9 - - + -
- +
-
- + - + - +
7
Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999
reaching consensus on a number of agreements concerning the way they would cooperate and the most important issues of their partnership. In order to reach consensus, new agreements had to be added. These new agreements had not been part of any of the preceding meeting activities. Second, we experienced that content structuring can be very difficult because sometimes the outcome of a session cannot be determined in advance. Negotiation may lead to a result that by nature is never entirely predictable. The parties can make a “content” deal, but if this fails, “process” deals can also be made or it can merely be agreed that the parties will meet again. For example, in case 5 the parties were going to develop an integral strategy (consisting of the topics ‘research’, ‘education’, and ‘management’) but in the end they could not reach consensus on the strategy with respect to management and no statements about this were made. Said one participant afterwards: “A number of additional sessions are needed to reach consensus”. Third, winners and losers start to behave strategically. In some cases, a pattern of winners and losers became visible. Both parties assumed their stereotypical behavior patterns. The winners were inclined to trivialize all arguments that were contrary to the package deal. The losers were inclined to magnify these arguments. Sometimes the loser was even inclined to frustrate the whole session: “Better a disrupted session than a session where I end up as loser”, is what seems to be the tenet. In case 1, environmental issues were threatened to lose from the other issues on the table. From that moment on, the defenders of the environmental issues were exceedingly critical of every suggestion for a strategy. They attempted to block the creation of a strategy because they were on the verge of losing. Only when the other participants gave in a little and made several environmental issues a part of the strategy, the blockade was lifted. In these situations, participants displayed a tendency to go back to oral interaction. Feedback from participants suggests that in this way (1) they felt more in control of the process and (2) they thought to stand a better chance of having their interests accommodated. This was illustrated by frequent remarks during interviews such as “I would like more oral discussion” and “There was too little personal contact”. Also the anonymity features appeared to get in the way. “For me, anonymity is not an issue,” explained one participant who also stated that he would have felt more at ease with identified communication during the session. Notable exceptions to the above considerations are cases 6 and 9. Here the assumptions did apply. The most likely explanation is that the parties were at the start of the negotiation process. The participants felt and expressed that there still was no need for decision making.
4.4. The facilitator In many cases the assumptions with respect to the facilitator appeared to apply (table 6). Assumption 2 applied in all cases. Assumption 1 did not hold in cases 7 and 8, in which there were large conflicts of interest and the parties had to reach a decision. The explanation for this is that a facilitator may be regarded as a party. When there are winners and losers, consensus is no longer possible. The winners will support a package and the potential losers will block this package. It was clear from participants’ feedback that a facilitator who attempts to force a decision in such a situation (e.g. because the agenda prescribes this) is considered to be taking sides with the winners or losers. A facilitator may then get reproached; one party may need a scapegoat and choose the facilitator to fulfill that role. An important assumption is that a change of the agenda is a decision taken by the facilitators or the group (assumption 3). However, this assumption did not apply in a number of cases in which (1) there were conflicts of interest, (2) the parties had to reach a decision in the form of a package deal, and (3) the process of wheeling and dealing produced a different result than initially foreseen. However, only winners or only losers desired changes in the agenda; the winners because, for instance, a change is conducive to the creation of a package deal and the losers because they want to frustrate a package deal. If this situation occurs and the facilitator makes a choice, again he risks being reproached for taking sides.
4.5. Groups should exchange as much information as possible The picture with respect to the assumptions on information exchange is varied (table 7). On the one hand, a number cases scored positively for all assumptions. These were cases in which final decision making was not yet on the agenda (cases 2, 6, and 9) or in which the final decision making activities represented a very small part of the whole process (cases 1 and 4). In these cases it was obvious that much information makes the possibilities for a package deal clear. Based on the participants’ feedback, extensive information exchange appeared to be useful and conducive to the creation of a package deal. The shared information contained several solutions which the parties had never considered and which simplified the making of a package deal. For example, in case 2 the parties from the waste processing chain appeared to have all sorts of ideas which the initiator of the process had never considered. This simplified the making of a package deal. On the other hand, cases 3, 5, 7, and 8 scored negatively for the listed assumptions. The explanations based on our observations and participants’ feedback are twofold. First, it appeared that in the final decision-making
0-7695-0001-3/99 $10.00 (c) 1999 IEEE
8
Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999
phase too much information can be disruptive. These sessions were attended by the parties who had already gone through a negotiation process and hoped to make a package deal in the near future. However, during the meeting process all kind of ideas came up for discussion that had been ruled out previously by the negotiating parties. From a “content” perspective this may be useful. However, the parties involved felt it frustrated the process of wheeling and dealing. The impression arose that the entire process could start all over again and that earlier decisions could be re-negotiated. But what these parties needed was a limitation on information. Said one disappointed participant: “Few new ideas were generated.” Other participants got frustrated because for them the amount of generated information blurred the decision making process instead of helping it ahead: “When there are 100 to 140 statements, I loose overview. In my opinion it should be limited to 60 to 80 statements” and “We had little time. Hence we could not read everything which prevented us from getting an overview”. Second, it appeared that information was not divulged because of strategic behavior. Although information exchange seems to be a prerequisite for the creation of package deals, the strategic behavior of the parties appeared to be such that they were not always willing to divulge information. For example, the municipalities in case 8 held certain views about the role of central planning. However, they did not want to broadcast these views because this would solve the problems of the central government and they wanted to make a package deal without the central government.
4.6. People are cooperative It may not be surprising that the assumptions held with respect to the cooperative nature of meeting participants apply to the context of networks only to a limited extent (table 8). A number of remarks have been made in the preceding sections that are also valid here. In addition, the nature of networks, involving parties with conflicting interests, contradicts the nature of the assumptions listed in table 8. One participant hinted at this, saying “The impact of the session will be limited as long as the relationship between Party X and Party Y remain underdeveloped”. (Names of parties taken out by authors). The exception to this were cases 6 and 9. As mentioned earlier in section 4.3, in these cases the parties were at the very beginning of the policy process and were behaving cooperatively in order to support their own orientation. An additional explanation why most of the assumptions did not hold in most of the cases is the following: We observed that the participants over time learnt to behave strategically. It was striking that participants who attended a number of sessions began to behave strategically over
time. For example, they generated ideas which they did not advocate but which could strengthen their negotiating position. Some participants generated ideas in the style of one of the other parties in the hope that this would mislead other parties. Certain organizations at one time also sent representatives with insufficient commitment power: they were not allowed to make agreements on behalf of their organizations but were fully informed at the session. In other words, participants consciously behaved uncooperative during various occasions.
5. Discussion and reflection This section is devoted to the “Reflect” phase of our study. We look back at the results of the “Observe” phase from our study as presented in the previous section. From the data that we collected it appeared that the extent to which certain assumptions with respect to group behavior and GSS application held depended on the phase of the decision making process in which the network found itself in each case. We propose that a distinction can be made between three moments in the decision making process: 1. An orientation phase. Here the parties come together and find out about each other’s positions, goals, requirements, etc. 2. A separation phase. In this phase it becomes clearer that the process is going to result in winners and losers. Based on the argumentation and preferences expressed by the various parties in the network, each party itself can assess where it is likely to end up at the conclusion of the decision making process: the winning or the losing end. 3. A package deal phase. Once the winners in the process are obvious, the losers are left aside. In the last phase, the winners together try to make a package deal. The 9 cases that we investigated can be characterized in terms of these three phases as follows: Cases 2, 6, and 9 were in the orientation phase. Cases 3, 5, 7, and 8 were in the separation phase. Cases 1 and 4 were in the package deal phase. If we now take another look at the extent to which the various assumptions held in each case, organized according to the phase of each case, we get the results presented in table 9 (assumption “x.y” refers assumption y in category x). The results in table 9 show an interesting picture. It appears that there is a distinct difference in the extent to which the assumptions hold for each of the three types of cases. In general, the assumptions hold very well in the orientation phase of decision making in networks, and very poorly during the separation phase. In the package deal phase, there appears to be more balance between assumptions that hold and those that do not. Given the positive results in table 9, network sessions in the orientation phase appear to be good candidates for
0-7695-0001-3/99 $10.00 (c) 1999 IEEE
9
Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999
effective GSS application. It is recommended to qualify a GSS session in this phase as a ‘power free’ session in which only the generation of ideas is put on the agenda. When the parties in a network are at the start of a decision making process, openness, content focus, and equality can be attractive meeting characteristics. The observations in the cases studied show that because the final decision will be a package deal, the parties are in this phase interested in as many problem definitions and solutions as possible. In this way, they obtain an insight into the room available for concluding a package deal later on in the process. On the contrary, network sessions in the separation phase appear to be very poor candidates for GSS support. Here GSS seem to have a very limited role to play. As was apparent in the cases that were included in our study, the strategic behavior of the parties involved will be very strong at this point in the decision making process. It represents the stage in which processes of power formation occur. If GSS are introduced into this arena, it can lead to a situation in which none of the assumptions apply. Assumptions such as openness, rationality, and cooperation simply do not hold during this phase. Finally, in the package deal phase there appears to be potential for the effective application of GSS. During this phase the parties involved have indicated that they want to conclude a package deal. GSS may be functional at several moments in the process of achieving this deal: • When the discussions have reached a stalemate, interaction can take place about possible solutions. • The parties can brainstorm and vote on all kinds of process aspects of the meetings, for instance, the sequence in which they bring up issues for discussion (e.g. first the ‘heavy’ issues and then the ‘light’ issues). The facilitator can make several draft packages and ask the parties to indicate which changes must be introduced to each of the packages before they become acceptable to the parties concerned. This can give each party a feeling of the available room for negotiations. • The parties can use voting methods that offer the possibility of impassioned voting in order to give a feeling of what the ‘heavy’ points are for specific parties. However, as became apparent from the observations in the various cases and the overview in table 9, GSS have to be applied with caution in this phase. Successful GSS application does not seem to be straightforward. Below, we offer a number of GSS meeting design guidelines for network meetings in the package deal phase. We stress that these guidelines are meant as a proposal, based on the results and insights from the cases that we studied. It remains to be seen in future cases whether the guidelines are appropriate and effective. In other words, these guidelines can be seen as the “Plan” phase of a new action research cycle:
1.
2.
3.
4.
5.
6.
Gather commitment power at the meeting table. The facilitator should become informed of the commitment power of the participants. They must have the ability to bind their own organization, because during the package deal phase, the meeting process involves negotiations and decision making, not noncommittal brainstorms. Assist with facilitation AND negotiation skills. The facilitator should not only possess GSS facilitation skills but also negotiation skills. It is conceivable that two process facilitators are involved in sessions in a package deal phase: one for the GSS process and one for negotiations. Place GSS in a secondary role. Negotiations play a primary role. A GSS is only used when this is functional for negotiations. In contrast to sessions in the orientation phase, GSS must play a secondary role. Use an open agenda. The agenda used in these GSS sessions should be open, as the meeting process is only partly predictable. For example, a brainstorm may result in a deadlock situation, which can be followed by another GSS session, oral discussions, or an adjournment of the meeting. Hence, the meeting agenda should be prepared in such a way that unexpected activities can be accommodated. Make process agreements concerning GSS use. In order to achieve some structuring of GSS sessions, the parties make a number of process agreements prior to the package deal phase concerning the use of GSS. For example: ‘When 4 of the 6 parties want to use GSS during the negotiations, a GSS session is organized in which all the parties participate’. Make no deadline and process agreements concerning the end of a session. GSS sessions in this context should not have a predetermined deadline. For the negotiations are difficult to plan and so are the GSS activities. The date and length of GSS meetings should not irrevocably be determined beforehand. Instead, process agreements must once more be made. For example: ‘At the start of a GSS session the parties indicate at which moments during the session they may decide to end it’. Offer GSS support for sub-groups. There should be room for GSS activities involving a limited number of participants, e.g. the members of one of the delegations. Process agreements must also be made about this, for example: ‘When a number of parties make it known that they want to hold a GSS session in which several parties do not participate, this session is organized if the non-participating parties do not object to this. The results are only available to the participants’.
0-7695-0001-3/99 $10.00 (c) 1999 IEEE
10
Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999
Table 9. The extent to which assumptions held in each type of case. "Orientation" "Separation" "Package Deal" 2 6 9 %+ %- 3 5 7 8 %+ %- 1 4 %+ %Assumption + + 100 0 + + 50 50 + + 100 0 1.1 + + + 100 0 + 25 75 + 50 50 1.2 + + + 100 0 0 100 + 50 50 1.3 + + + 100 0 0 100 + 50 50 1.4 + + + 100 0 + + + + 100 0 + + 100 0 2.1 + + + 100 0 + + + + 100 0 + + 100 0 2.2 + + + 100 0 0 100 + + 100 0 2.3 + + + 67 33 0 100 0 100 3.1 + + 67 33 0 100 + 50 50 3.2 + + 100 0 0 100 + 50 50 3.3 + + + 67 33 0 100 0 100 3.4 + + 100 0 + + 50 50 + + 100 0 4.1 + + + 100 0 + + + + 100 0 + + 100 0 4.2 + + + 100 0 + 25 75 + 50 50 4.3 + + + 100 0 0 100 + + 100 0 5.1 + + + 100 0 0 100 + + 100 0 5.2 + + + 100 0 0 100 + + 100 0 5.3 + + + 100 0 0 100 + + 100 0 5.4 + + + 100 0 0 100 + 50 50 6.1 + + + 100 0 0 100 + 50 50 6.2 + + + 67 33 0 100 0 100 6.3 + + 67 33 0 100 0 100 6.4 + + 67 33 0 100 0 100 6.5 91.4 8.6 19.6 80.4 60.9 39.1 Average%
6. Conclusion During our study, we applied an action research approach to investigate the applicability of GSS in the context of inter-organizational policy networks. The interventions we applied in the cases we studied were limited to facilitation support from the groups involved. Our study reveals some areas in which the application of the technology has to be carefully planned and evaluated before entering the electronic meeting arena. Often unsuccessful applications of GSS originate in the process with which GSS are applied and not in the GSS itself, see e.g. [15, 26]. Our study shows that it is paramount to structure the use of GSS to meet the group’s needs. This is an important message for most (prospective) facilitators as they are often presented with showcases of GSS use and commercial information about the advantages of the technology. From the results of our study, it appears to be especially important that the facilitator and the network parties involved ascertain in which phase of the decision making process they are. This may provide the basis upon which a decision to use GSS and, if so, in what way, can be made.
Our findings suggest that GSS seem to provide sufficient support for the orientation phase of a decision making process to be effective. Meetings in the separation phase are unlikely candidates for effective GSS application. With respect to meetings in the package deal phase, we offered a number of meeting design guidelines to stimulate the effective application of GSS. The assumptions on GSS group behavior that we used as a reference point were derived from commercial publications and product information. Apart from acting as a reference point for our inquiry into the application of GSS in the context of inter-organizational policy networks, the assumptions also provide a focus for comparison of our findings with other GSS research. For example, some studies [12, 14] found that larger GSS groups generated more ideas per person than smaller GSS groups and were more satisfied. Our study illustrates that the additional productivity may not always lead to increased perceptions of satisfaction. Some GSS studies found that the technology appears to be more useful for simple tasks than for complex tasks, see e.g. [20]. Our results are consistent with this view. During the orientation phase, in which relative simple generative and evaluative tasks are
0-7695-0001-3/99 $10.00 (c) 1999 IEEE
11
Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999
performed, GSS appear most useful. During the separation and package deal phase, in which more complex negotiation and evaluative tasks are performed, GSS appear to be much less useful. Although our study differs from most published GSS studies in context (interorganizational policy networks) and method (action research), further attention is required to compare and contrast our findings with other GSS studies. The limitations of our study are twofold. First, our observations relate to Dutch groups using GSS. Most GSS research involved North American groups [27]. Although culturally, North America and the Netherlands bear many similarities, there are significant differences as well, for example in the way people seek compromises [16]. Hence, it is unclear to what extent our findings are comparable to findings in North American GSS research. Second, following an action research method, we were both researchers and facilitators during the study. This interplay between roles may present a problem. During the sessions we stayed mostly focused on the session itself, because the topics were often sensitive, while the discussions, both electronic and oral, were very rich. We found it too hard to facilitate the session process and write up research observations simultaneously. Hence, most of our observations were written up during breaks in the process or immediately after a session. As a result we may have missed a few important observations. The applicability of our findings may extend beyond the network context in which our research was carried out. The meeting phases we distinguish as well as the design guidelines that we propose may be helpful in other meeting arenas as well. The characteristics of networks (multiple independent parties with different but interrelated interests) are not uncommon. Often meeting participants, even within a single organization, bring their own interests to a meeting, pursuing their accommodation. In other words, meetings often can be considered as a process in which participants assess the feasibility of certain package deals while maximizing the accommodation of their own interests and, at the same time, trying to conform as much as possible to the overall goal of their group or organization.
Acknowledgments The authors gratefully acknowledge the helpful comments from three anonymous reviewers. References [1] Argyris, C., D.A. Schön, Participatory Action Research and Action Science Compared - A Commentary, American Behavioral Scientist, 32, No. 5, pp. 612-623, 1989. [2] Argyris, C., R. Putnam, D. McLain Smith, Action Science Concepts, Methods and Skills for Research and Intervention, Jossey-Bass, San Francisco, 1982.
[3] Bannon, L.J., Group Decision Support Systems: An Analysis and Critique, Proceedings of ECIS’97, Cork, Ireland, 19-21 June 1997, pp. 526-539. [4] Benbasat, I., Lim, L.H., The Effects of Group, Task, Context and Technology Variables on the Usefulness of Group Support Systems: A Meta-Analysis of Experimental Studies, Small Group Research, 24, 4, pp.430-462, 1993. [5] Briggs, R.O., G.J. de Vreede, Measuring Satisfaction in GSS meetings, in: Kumar, K., DeGross, J.I. (eds) Proceedings of the 18th International Conference on Information Systems, pp. 483-484, Atlanta, December 1997. Briggs, R.O., Nunamaker, J.F. Jr., Sprague, R.H. Jr., 1001 Unanswered Research Questions in GSS, Journal of MIS, 14, 3, 1998. [6] Bruijn, J.A. de, Heuvelhof, E.F. ten, Policy Networks and Governance, in D. Weimer (ed.), Institutional design, Boston 1995. [7] Bruijn, J.A. de, Heuvelhof, E.F. ten, Netwerkmanagement, Lemma, Utrecht, The Netherlands, 1995. (In Dutch) [8] Cohen, M.D., J.G. March, J.P. Olsen, A Garbage Can Model of Organizational Choice, in Administrative Science Quarterly (1972) vol. 17, nr 1, 1-25. [9] Dennis, A.R., Haley, B.J., and Vandenberg, R.J., A MetaAnalysis of Effectiveness, Efficiency, and Participant Satisfaction in Group Support Systems Research, Proceedings of the Seventeenth International Conference on Information Systems, Cleveland, 1996. [10]Dennis, A.R., J.F. Nunamaker Jr., D.R. Vogel, A comparison of laboratory and field research in the study of electronic meetings systems, Journal of Management Information Systems, Vol. 7, No. 2, pp. 107-135, 1991. [11]Dennis, A.R., R.B. Gallupe, A History of Group Support Systems Empirical Research: Lessons Learned and Future Directions, in: Jessup, L.M., J.S. Valacich (eds), Group Support Systems - New Perspectives, Macmillan Publishing Company, New York, 1993. [12]Dennis, A.R., Valacich, J.S., Computer Brainstorms: More Heads Are Better than One, Journal of Applied Psychology, 1993, 78:4, 531-53. [13]Dickson, G.W., Personal Communication, 1995. [14]Gallupe, R.B., Dennis, A.R., Cooper, W.H., Valacich, J.S., Bastianutti, L., Nunamaker, J.F. Jr., Electronic Brainstorming and Group Size, Academy of Management Journal, 1992, 35:2, 350-369. [15]Herik, C.W., Group Support for Policy Making, published doctoral dissertation, Delft University of Technology, The Netherlands, 1998. [16]Hofstede, G., Cultures and Organisations: Software of the Mind, McGraw-Hill, London, 1991. [17]Kickert, W.J.M. (ed.), Managing Complex Networks, Sage London 1997. [18]Marin, B., Mayntz, R. (eds), Policy Networks: Empirical Evidence and Theoretical Considerations, Frankfurt am Main, 1991. [19]McLeod, P.L., An Assessment of the Empirical Literature on Electronic Support of Group Work: Results of a MetaAnalysis, Human-Computer Interaction,7, pp.257-280, 1992. [20]McLeod, P.L., Liker, H.K., Electronic Meeting Systems: Evidence from a Low Structure Environment, Information Systems Research, September, 3(3), 195-223, 1992.
0-7695-0001-3/99 $10.00 (c) 1999 IEEE
12
Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999
[21]Nunamaker, J.F. Jr., A.R. Dennis, J.S. Valacich, D.R. Vogel, J.F. George, Electronic Meeting Systems to Support Group Work, Communications of the ACM, Vol. 34, No. 7, pp. 4061, 1991. [22]Nunamaker, J.F. Jr., Briggs, R.O., Mittleman, D.D., Vogel, D.R., Balthazard, P.A., “Lessons from a Dozen Years of Group Support Systems Research: A Discussion of Lab and Field Findings”, Journal of MIS, 13(3), 163-207, 1997. [23]O’Toole, L..J., Treating Networks Seriously: Practical and Research-Based Agendas in Public Administration, Public Administration Review (1997), Vol. 57, no 1, 45-52. [24]Shaw, G.J., User Satisfaction in Group Support Systems Research: A Meta-Analysis of Experimental Results, Proceedings of the 31st HICSS, IEEE Press, Los Alamitos, CA, 1998. [25]Van de Ven, A.H., Delbecq, A., The effectiveness of nominal, Delphi, and interacting group decision making processes, Academy of Management Journal, Vol. 17, pp. 605-621, 1974. [26]Vreede, G.J. de, P.C. Muller, Why Some GSS Meetings Just Don’t Work: Exploring Success Factors of Electronic Meetings, in: Galliers, R., Carlsson, S, Loebbecke, C., Murphy, C., Hansen, H.R., O’Callaghan, R. (eds) Proceedings of the 7th European Conference on Information Systems (ECIS), Cork, Ireland, Vol. III, pp. 1266-1285, June 1997. [27]Watson, R., Ho, T., Raman, K., Culture: A fourth dimension of group support systems, Communications of the ACM, October 1994, 45-55, 1994. [28]Zuber-Skerrit, O. (1991). Action Research for Change and Development, Aldershot: Gower Publishing.
•
•
•
•
•
Appendix A. Background information on selected cases In this appendix we give a brief description for each case that was included in the analysis described in section 4: • Case 1: Ministerial Strategy Development. The Ministry of Housing, Physical Planning and the Environment had to develop a middle long term strategy. There were conflicts of interest among the departments of the ministry (environment vs. physical planning vs. housing). In addition, strong conflicts of interest exists with respect to important external stakeholders, such as industry and other governmental authorities. • Case 2: Optimization of waste/recycling chains. A number of waste processing organizations wanted to optimize the waste/recycling chain. There was strong opposition from waste disposal companies, waste transport companies, and certain municipalities. • Case 3: Inter-municipal cooperation. A municipality wanted to take a stand with respect to regional cooperation with other municipalities. Conflicts of interest existed among these municipalities, and within the organizing municipality, about the most desired form of cooperation.
•
Case 4: Oil and Gas Supply Industry. The Oil and Gas Supply Industry had to bring forward a unified standpoint for the European Commission about the impact of this industry on the European economy and environment. The supply industry consists of dozens of different branches, that have diverging interests. Case 5: Strategy development university board. The board of a university had to develop a middle long term strategy. Among the key players in this process there were many different opinions about the desired and potential topics of this strategy. Case 6: Environmental Law Enforcement. A national province had to enforce environmental laws. A new strategy had to be developed in order to improve the efficacy of the law enforcement efforts. Important stakeholders included the conservation movement, the companies in the region, and competing enforcement bodies. Case 7: National physical planning. The national government wanted to restructure the physical planning of a country. This lead to new roles for municipalities and provinces. However, the latter two did not agree with the national government. In addition, conflicts of interest among municipalities and provinces were evident. Case 8: Provincial working procedures. A number of departments of a province wanted to change their working procedures. The intention was to arrive at a more participative approach to work instead of a hierarchical top-down approach. The departments involved had extensive autonomy, but were forced to start working together. They had to reach agreement on the changes that were going to be made in the organization. Case 9: Prison privatization. Representatives of various ministries had to take a stand on the privatization of the prison system. In the country concerned, this is a very sensitive topic with passionate advocates and opponents.
Appendix B. Interview questions The questions in the semi-structured ex-post interviews were aimed at obtaining overall insight into the respondent’s perception of the process and technology. Also, they were focused on the assumption categories in particular: 1. Is this the first time you’ve participated in an electronic meeting? 2. What did you hope to produce in this meeting? 3. Did you meet the objectives you had for this meeting? 4. Did you feel comfortable with “open” nature of this meeting?
0-7695-0001-3/99 $10.00 (c) 1999 IEEE
13
Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999
5.
Did you feel comfortable with the “fair” nature of this meeting? 6. Would GSS adequately support your usual meetings? 7. If given the choice, would you prefer other tools and methods over GSS tools and methods? 8. What is your perception on the role of the facilitator? Did he/she support the process well? 9. Are you satisfied with the work process used today? 10. Are you satisfied with the outcome of this meeting? 11. What do you think will be the impact of this meeting? 12. Do you have any other remarks or observations?
Appendix C. Satisfaction questionnaire The satisfaction questionnaire used in this study consists of 18 questions that relate to the four constructs of a causal model of individual satisfaction with team effort [5]. Analyzing 173 questionnaires from US meeting participants (non-students) showed Cronbach alphas between 0.90 and 0.91. Interest accommodation 1. Today, my interests were (not accommodated accommodated) 2. Thinking about what I needed from this meeting (I did not get it - I got it) 3. In this meeting I personally (lost - gained) 4. The outcome of today’s activities (does not meet my personal needs - meets my personal needs)
Product value 5. The work we accomplished today was (not worth - worth) the effort. 6. The results of this meeting are worth the resources it cost to produce them. (disagree - agree) 7. The value of the meeting’s outcomes justifies our efforts. (disagree - agree)
Process satisfaction 8. I was satisfied with the way we did things today to achieve our goals. (disagree - agree) 9. The meeting methods we used today (did not meet - met) my expectations. 10. Today’s meeting process was (inadequate - adequate) to meet our goals. 11. How satisfied were you with the work process we used today? (dissatisfied - satisfied) 12. The group used its time well, devoting enough attention to the important issues. (disagree - agree) 13. How satisfied are you with the design of the session in the sense that the agenda items followed one another logically? (dissatisfied - satisfied)
Product satisfaction 14. The product of this meeting (did not meet - met) my expectations. 15. The outcome of today’s activities (does not meet the meeting’s objectives - meets the meeting’s objectives) 16. With respect to the outcome of today’s meeting, I have (many complaints - no complaints at all) 17. The outcome of today’s meeting is (unsatisfactory satisfactory) 18. The results of today’s meeting are (inadequate - adequate)
0-7695-0001-3/99 $10.00 (c) 1999 IEEE
14