GroupWare and Group Decision Support Systems for Wood Procurement Organisation. A Review

Many kinds of decision support systems (DSSs) have been suggested for use of wood procurement organisations, but few meet the real needs of team managers in group decision-making process. Therefore, it has been concluded that the important features of group decision support systems (GDSSs) should be developed for teamwork-based organisations. Electronic meeting systems (EMSs), Computer-aided Visualisations (CAVs) and heuristics as well as other numerical approaches as combined with optimisation seem to be some of the most promising elements of GroupWare, because decisions are made in distributed groups and they deal with human behaviour. Relations between GDSSs and spontaneous decision conferencing (SDC) for modern organisations are also discussed, and suggestions for future research of management approaches are also given.


Introduction
Applications of telematical or telemechanical technology have provided organisations advanced information and communication systems. They have been shown to be very helpful electronic systems (GroupWare) in all kind of group work.
Recently, attention has focused on the relations between group decision-making and communication in organisations' dispersed structures. When organisations are being streamlined and more focused on advancing logistics, the value of group decision-making will constantly increase. In this respect, domestic research has also reported team managers' needs for education of helpful Group-Ware (Leppänen et al. 1999, Toivonen andPalander 2001). Furthermore, the reinforcing effects on individuals of observing other managers succeeding well in co-ordination of responsibilities in teamwork-based organisation using Group-Ware are well established in international literature.
The theory of GroupWare approach supporting small group communication and collaboration has been discussed for about ten years, while the technology for conferencing systems has been available for management of organisations for twenty years. Recently, these systems have been developed to support everyday decision-making, which is referred to as Spontaneous Decision Conferencing (SDC) (Hämäläinen andLeikola 1995, Palander 1998c). SDC can be used to manage whole decision-making processes of a group, in which Group Decision Support Systems (GDSSs) can be used to avoid groupthink and Electronic Meeting Systems (EMSs) in turn can especially help to support group dynamics. All of these tools are understood as subsets of Group-Ware. Characteristics of the decision-making situation of Finnish wood procurement organisations have been described as presented in Fig. 1 (Palander 1998b). At all organisational levels team managers' real needs for GroupWare were assumed according to this description list. Actually, it has been found out that group decision-making in teamwork-based organisations requires more communication than traditional decisions made without group in organisations based on functional divisions (Leppänen et al. 1999). This is because the transformation of the organisation to a team structure requires changes to several organisational features, focusing attention on the functionality of the team instead of on the actions of the foreman. These changes are especially extensive for an organisation with its personnel geographically dispersed (Mohrman 1999). Thus, if geographically dispersed team-based organisations are used, in addition to effective communication, the need for possibilities for collaboration is especially high (Cohen and Mankin 1999). However, extensive time is still spent at different meetings and in the travel to these meetings. Since also distances between managers are further extended, more GroupWare for collaboration should be used in decision-making (Toivonen and Palander 2001).
Decision Support Systems (DSSs) are largely used and discussed in literature. Generally, they are interactive computer-based systems, which present decision alternatives (Harstela 1997). With DSS a decision-maker can better understand and learn a decision-making process, and thus make better and perhaps also faster decisions ( Scott-Morton 1971, Keen 1981, Turban 1988, Silver 1991. For some time, the defi nitions of DSSs have included properties like interaction, the ability to solve ad hoc problems, and the use of models (Jelassi 1986, Turban 1993. However, theory behind DSS is quite narrow, because GroupWare technology has developed. Therefore, more comprehensive systems have been developed for groups, which need to use many ways of communication. These systems are called GDSSs, which are in that sense subsets of Group-Ware, because GroupWare may provide electronic collaboration facilities needed in GDSS. Originally, Rao and Jarvenpaa (1991) laid out the foundation for future research in the GDSS topic area.
Theoretical research into group decision-making is recent within the fi eld of wood procurement. Consequently, studying GDSSs and the features adhered to them are also just beginning. However, EMSs, Computer Aided Visualisation (CAV), optimisation and some numerical approaches of GroupWare have been found useful in the formulation of models in the context of GDSSs (Palander 1998b(Palander , 1999a. Although most of GDSSs systems are undeveloped, there are expectation values to study them, since big forest industry companies are interested in developing them. Reviewers of DSS suggest that diffi culties in the decision-making processes of wood procurement include the remoteness of planning models from the actual decision making and a lack of heuristic and cognitive components (Robak 1991, Harstela 1997. In the sense of electronic systems, they obviously mean numerical approaches, when they mention heuristic and cognitive components. Following the approach of GDSS research, Toivonen and Palander (2001) suggest that these diffi culties should also include so called group aspects. They showed that these diffi culties are partly due to the lack of human and social interaction caused by the geographical separation of managers involved in the process, as well as the attitudes of managers towards the new tools. Partly, of course, the diffi culties are caused by the general unpredictability of the natural environment, in which wood procurement functions are performed (Palander 1995). Anson et al. (1995) have reviewed various group researches. They found that structured communication and decision-making procedures have been effective in enhancing social interaction, although the procedures have been used without computer support. In order to fi nd more support for group decision-making, potential computer applications related to GroupWare have also had to be found out. They have been directly reviewed under several well-known subject areas, such as Decision Analysis (Corner andKirkwood 1990, Spector 1993), Group Decision Processes and Mathematical Programming (Lewis and Butler 1993), Group Meetings (Anson et al. 1995), EMSs , and Conferencing (Hiltz et al. 1991). These reviews provide knowledge about the potentials of communication and collaboration tools, which could also effect positively on managers' attitudes.
The emphasis of this review will be on the group decision-making related to the team-based structure of the wood procurement organisation, and its relationships with customers. The contribution of the article is to produce a theoreti-cal understanding for the support of it in the wood procurement process. GroupWare, SDC, GDSSs, EMSs, CAVs, and some other numerical approaches are to be described and defi ned to accomplish this objective. The necessity of GDSSs is going to show by the over twenty years of development of group decision-making. The aim is to analyse literature related to group decision-making within the context of management science, decision science, psychology, education science, sociology, and forestry in order to synthesise the features, which GDSSs need to be suitable for wood procurement applications.

Group Decision-making and GroupWare
GroupWare is organisation's software for computer-supported co-operative decision-making (Anson et al. 1995). For GroupWare decisionmaking models and methods have been adapted to group dynamics so that decision-making groups can actively reach consensus through equal participation. Thus, the decision-making groups' broad needs to aggregate information and to choose among decision alternatives are satisfi ed in group processes (Hackman and Kaplan 1974). When group decision-making is used, Group-Ware applications of SDC can be used to provide possibilities for the use of EMSs. In studies of this area, conferencing is defi ned as a subset of the Computer Mediated Conferencing Systems, but actually both of them are taking on more of the features of conferencing methods (Hiltz and Turoff 1985). Therefore, many distinctions between simple electronic messaging systems and group-communication-oriented conferencing systems will become negligible. Then GroupWare would facilitate brainstorming, decision-analytic problem structuring, prioritisation of criteria and analysis of the alternative decisions (Hiltz et al. 1991, Hämäläinen and Leikola 1995, Palander 1998c. In wood procurement, managers work in geographically separated offi ces. To support decision making, they could collaborate using GroupWarelike telecommunication in their remote deci-sion-making sessions (Toivonen and Palander 1999). Anson et al. (1995) have suggested that in this kind of session group members could have electronic videoconferencing support, or simply teleconference. In practice, using Group-Ware technology, e.g. shared written or visual information, model bases, audio-and videoconferencing, decision-making groups could share ideas, aspirations and preferences, and consequently enhance collaboration.

Group Management and Spontaneous Decision Conferencing
In the 1990s, organisations have been under pressure to manage information fl ows effectively and effi ciently in order to be able to respond to the changing aspects of decision-making. Therefore, it is not surprising that use of information continues to be an important element in the management of decision-making processes of teams (Kärhä 1998). For example, for decision making, a supervisor must be concerned about how each of his busy team managers interprets information and interacts in communication. However, at present there is an ideological controversy between the fl exibility of teamwork, and the control inherent in the traditional leadership (Hayes andWalsham 2000, Wiesenfeld et al. 1999).
In order to improve the use of information, group collaborations even without GroupWare can be adapted to multiple decision maker situations. For this purpose managers' preferences must be aggregated into a single group or consensus preference (Lewis and Butler 1993). According to Sen (1970), this aggregation requires three distinct and interrelated activities: individual preference measurement, interpersonal preference comparison, and group preference determination. Sen (1970) constructed a utility model providing the aggregation principle to structure human interaction.
However, group meetings are often not as effective as they could be (Shaw 1981). Meetings may lack a clear focus, because making a clear difference between group decision-making and negotiations is very diffi cult. Often managers may even hold back from participating because they are apprehensive about how their ideas will be received. Particularly, team managers are experts in the fi eld of the decision task, but they are rarely experts in the theories involved with decision making. Therefore, meetings may end without a clear understanding or record of what was discussed.
Despite the diffi culties, little computer support is available for group meetings, which is surprising, given the ubiquitous nature of computer support in modern organisations (Meriläinen et al. 1995). According to the group studies, a group in an organisation may use many approaches to adapt a participatory phase of collaboration to a management policy. Often, a method of full-scale conferencing is used. Generally, it is understood as a two-day meeting in which a decision-making group tries to solve a strategic decision problem with the help of a facilitator and a decision analyst. Unfortunately, two-day meetings last too long and strategic decisions are rarely made. Therefore, a method has been developed to support everyday group decision-making, which is referred to as SDC (Hämäläinen andLeikola 1995, Palander 1998c).
Often DSSs as kind of Management Information Systems (MISs) have been recommended without the needs for SDC or GDSSs, because time for meetings is characterised as a critical factor at the top hierarchical level of wood procurement organisation. Similarly, setting a date for the lower level managers' meetings is also diffi cult. However, decision-making groups of them are characterised as ongoing groups. Thus, spontaneous and ad hoc group meetings are characteristics of their decision-making process. The above mentioned special characteristics encountered in the meetings would suggest that computerised conferencing and the related decision-making approaches should be join with them and especially from the managers point of view (Huber 1991, Palander 1998b. Therefore, they suggest GroupWare-like SDC to manage group collaboration in meetings. According to Hiltz and Turoff (1985), an organisational solution to constrict the fl ow of information and communication is not without its costs. On the other hand, meeting outcomes are contingent on the balance of an electronic meet-ing's gains and losses (Connolly et al. 1990). Accordingly, group-collaboration characteristics, e.g. group features, organisation's infrastructure and the Decentralised Information Processing Technique, could establish an initial balance of decision consciousness, which the decision-making group may alter by using EMSs and SDC , Palander 1998a. Therefore, Ancona's (1987) research on boundary management (i.e. the management of a decision-making group's performance with situations and individuals external to the decision-making group) raises an interesting theoretical point with respect to management. Ancona (1987) found that teams equally matching the characteristics could be differentiated based on boundary management.

Group Dynamics and Electronic Meeting Systems
Generally, the advantages of EMSs compared to an ordinary face-to-face meeting include: reducing the impact of social obstacles (passivity, domination, etc.), more careful preparation for a meeting, and a better structured meeting process (Fish et al. 1993, Tan et al. 1998. Research into the processes of group collaboration has also shown that there is no difference between the outcomes of technically sound videoconferences and face-to-face meetings, but without video, electronic meetings produce signifi cantly worse results. This conclusion bases on the fact that the difference between the impact of video conferencing compared with that without video mainly relates to communication, which is apparent in the long run.
In group collaborations, managers can use EMSs for advancing group dynamics. It is possible due to a built-in provision to allow anonymous suggestions and to applications of heuristic decision practices (Beck and Lin 1983). These features act to eliminate some managers' dominating behaviour and produce a more innovative atmosphere (Ellis et al. 1991, Stefi k et al. 1987. Using advancing features of EMSs a group may be effi cient in decision making. As a result, a signifi cant portion of research on GDSSs is related to EMSs that primarily support the communication process between decision-makers, e.g. teleconferencing, electronic mail, and different networks.
A large amount of research into group dynamics comes from the fi elds of social psychology and human behaviour in organisation (Williams 1978, Shaw 1981, Finholt 1997. In order for a group to be effi cient and to meet its goals, the support of communication between the group members is needed. In general it is known that communication is effective within an effi cient group. If it is not, it can be improved, e.g., with non-verbal communication by means of visual cues like gestures and facial expressions. Furthermore, in an effi cient group, consensus must exist on the division of participation and management among the members; appropriate decisionmaking methods for specifi c situations should be applied in a fl exible way; the leadership should be equal or acceptable to the group as a whole. For these features developing of EMSs is ongoing.

Groupthink and Group Decision Support Systems
In the recent DSS research, criticism has been directed toward the lack of interaction between the cognitive and psychological aspects of human problem solving and decision support (e.g. Carlsson 1991, Vanharanta et al. 1997). Therefore, a special form of DSSs, active decision support system, was studied, in which the DSS played an active role in dealing with ambiguous and complex problems (Manheim 1989, Carlsson andWalden 1995). The active decision support system does not follow specifi c orders, but has the ability to respond to non-standard requests and commands. However, real applications dealing with immaterial investments are still waiting for future innovations (Keskiäijö et al. 1996). Among other social issues, DSSs have failed to provide features for addressing problems of groupthink (Palander 1988b). In particular, it may be associated with the following: incomplete generation of decision alternatives, incomplete understanding of goals, failure to examine the risk of preferred choices, poor quality of search for information, bias in the interpretation of informa-tion, failure to appraise alternatives. To avoid these problems, the features of decision support need to have not only the features of DSSs, but also hardware, software, and models necessary to reveal the negative aspects of groupthink. GDSSs represents this technology (Anson et al. 1995, DeSanctis and Gallupe 1987, Sauter 1997. They suggest structured decision-making processes with numerical approaches, which would support the aim of reaching satisfactory decisions instead of rational decisions following the principles fi rst described by Simon (1955).
Research into groupthink has revealed an interesting phenomenon. Confl icts caused by controversial ideas and different opinions should be encouraged, because they increase the quality and creativity, as well as the commitment to decisions. The drawbacks to confl ict within group decision-making include the excessive impact of dominant personalities on outcomes and erroneous consensus. In the latter situation, decisions are not criticised in order to avoid confl ict, thus maintaining a good atmosphere. These disadvantages should be controlled during the decision-making process (Johnson and Johnson 1987, Hogg and Abrams 1988, Baron 1992, Couger et al. 1993, Hogg and Vaughan 1995, Sosik and Avolio 1998.
According to the literature, several comprehensive theories for group decision-making have already been applied in DSS research. Most of them, e.g. multiple attribute utility theory, social judgement theory, and social choice theory, are usable in GDSS research. In fact, GDSS as the concept adopted for research, when the development of elements of GroupWare started for decision making. These elements resembled computer supported collaborative work systems. They included work patterns, as well as tools for the specifi c organisation's co-ordination, control, and norms. When the elements are deeply embedded into the social practices of an organisation, they strengthen the organisational social constructs.
GDSSs are especially convenient for situations with several nearly equal alternatives (Piippo et al. 1999). Recently, it has been stated that at the local level of wood procurement organisations, decision making may be a tactical choice between alternatives that have been calculated to be equal (Palander and Toivonen 1999). Accord-ing to them, further knowledge of different situational conditions infl uencing decision making is therefore needed. In addition, procurement plans and decisions may need to be modifi ed, sometimes rapidly, during their implementation due to changes in timber requirements or the natural environment. Under these circumstances, GDSSs supporting interaction and communication of managers are better able to manage the groupthink in the logistic management problems ( Palander 1996Palander , 1998bPalander , 2000b. Kärhä (1998) proposed use of DSSs for timber buyers' work. The aspect of timber buying is based on customer satisfaction (satisfaction of a seller), the monitoring of which is usually included as a part in a whole quality management system. Palander (1998cPalander ( , 1999a has proposed using GDSSs for all kind of managers' work. Particularly, he developed an adaptable system model for the determination of balanced stocks in the logistics fl ows. The model was applied according to the theory of Simon (1955). Thus a decision-making group can determine bounded rational sizes for the buffer stock. The model could also be used as a business process model, because for supply chain management both material and monetary fl ows of the model were depending on satisfaction of mill customers. Testing broader GDSS applications that could support actual decision-making and a whole environmental quality management is just beginning.

Co-operative Groups and Computer-aided Visualisation
Current, major challenges for managers of wood procurement include: changes in the forest owner population (Ripatti and Reunala 1989, Sikanen 1999, Leinonen 1998), a hectic pace of working (Klen 1998), forest owner satisfaction (Kärhä 1998), mill satisfaction (Palander 1998c), allinclusive service (Kärhä 1999), and socio-economic logistics (Palander 1999a(Palander , 2000b. To respond to these challenges, CAVs have been understood as a potential feature of GDSSs to provide more effi cient support for managers to cooperate with forest owners and public. A search for new more appropriate CAV models has therefore started. It also seems that new innovations and development in communication and information technology can, at least partly, meet the needs for efforts that CAV models require in wood procurement. In decision making of wood procurement, CAVs mean any visual way to give information to decision-makers, in which e.g. fi gures, images and photos have been used to compress information before the knowledge has been formed. In the fi rst place, CAVs have been suggested for forest industry companies, because decision making about logistic decision alternatives could be made easier, if group managers could also be motivated with participatory and interactive planning ( Palander 1996Palander , 1997Palander , 1998b. Particularly for mill satisfaction, which is the main responsibility of wood procurement organisation, CAVs could support managers dealing with abundance of information. Palander (1997) has also stated that Geographical Information Systems (GISs) with its internal CAV will probably be a part of these kind of decision-making and planning systems.
In addition to managers' viewpoint, the need for CAVs may arise from the following reasons: increased interest in social environmental issues, increased public calls for environmentally friendly forest operations, decreased knowledge about forestry among forest owners, and education of communication, computer, and information technology. These are mainly the viewpoints of forest owners and public. Karppinen (2000) has classifi ed forest owners onto four types: recreationists, self-employed, multiobjective, and investors. According to him, the share of recreationists and investors will grow in the future. Recreationists differ from the other groups due to their greater concern for the protection of the landscape values, which relate to outdoor activities. Ripatti and Reunala (1989) have suggested that forest owners are mainly concerned about the visual effects of logging operations planned for their holdings. By the same way public evaluate operations by the visual impacts they cause (Schauman 1988). According to Kilvert and Griffi th (1996), perceptions of environmental quality are also primarily visual.
When considering forestry knowledge, a forest owner, public, and managers are unequal in their meetings. With CAV of forestry novices (forest owners or public) and professionals (managers) could cooperate in the proposed forest operations interactively in a way all involved could understand (Orland 1988, Cox 1990). It seems that decision making could be easier when impacts of different operations on the landscape were presented as pictures beforehand. Also Johnson et al. (1994), Pukkala et al. (1995) and Nalli et al. (1996) have stated that CAV of forestry could be useful for solving forest management problems. According to Pykäläinen and Kangas (1996) and Pykäläinen (1999) participatory and interactive planning could then be used in decision-making.
During timber and pulpwood trading, there are two major problems: fi rst to fi nd enough stands suitable for sale and then to complete a purchasing transaction. To support the fi rst problem, GISs have already been developed, in which satellite pictures, maps, boundaries of forest stands, and information on ownership are connected. When these features of CAVs have been combined with modern marketing skills of managers, they have proven to be useful ways to locate and buy forest stands (Bergström 1998, Sikanen andOikarinen 1998). In these systems features of GDSSs for purchasing transactions is still undeveloped. Companies currently use only registers of forest owners with some kind of satisfaction measurements made for specifi c situations.
In Leinonen's (1998) analysis of future wood procurement, 69% of the respondents estimated that by the year 2030 trading will be accomplished through telecommunications, with no face-to-face meetings between managers and forest owners. Therefore, the innovation of forest owner satisfaction by Kärhä (1998) would pre-sume more applicable systems than registers and GISs can provide for monitoring it in practice. Fortunately, Toivonen and Palander (1999) have developed managers' EMSs using GroupWarelike telecommunications. According to them, the features of CAVs could be integrated into this kind of GroupWare.

Group Consensus and Numerical Approaches
When a satisfactory compromise is not forthcoming using only EMSs or face-to-face decision making without GDSSs, numerical problem restructuring, i.e. redefi ning the group decision and negotiation task presentation, is a key approach (Chatterjee et al. 1991). This can also be conceived from alternative theories of the literature of management science. Accordingly, numerical approaches are fundamental in structuring group interaction and communication processes with computers. Chatterjee et al. (1991) focus indirectly on such approaches that use i) cognitive process theory and game-theory models, ii) group decision theory and negotiation support systems, and iii) management theory and artifi cial intelligence. These categories have been studied over the last ten years. Although they all have been represented via DSSs, or more specifically, via GDSSs (Anson et al. 1995), here the approach of management science are omitted. Game theory provides various approaches for obtaining compromise solutions in group decision-making (Cramton 1991, Tolvanen-Sikanen et al. 1995. Recently, these non-co-operative game-theory models have come in for criticism for two reasons: i) they are not robust enough to withstand even small changes in the assumptions, and ii) co-operative models have been found to support negotiations of group decision-making better. Furthermore, combining social choice and game theory to multi-objective mathematical programming situations, a consensus can be obtained only under powerfully controlled conditions (McKelvey andWendell 1978, Wendell 1980). However, in some situations, non-cooperative models can convey some rich strategic fl avour of real-life (Chatterjee et al. 1991).
The decision-making approaches of group studies often focus on mathematical programming. It is used as an integrated part of GDSSs or alone via optimisation. In optimisation made alone, three standard multiple-objective mathematical programming methods for compromising the decision problem are i) generating effi cient points (e.g. Yu and Zeleny 1975, Wendell and Lee 1978, Dauer and Liu 1990, Dauer and Saleh 1990, ii) goal programming (e.g. Dyer 1977, Harrald et al. 1978, and iii) parametric righthand side optimisation (e.g. Hadley 1962, Cohon andMarks 1975). These methods can also include an interface for iv) interactive optimisation (e.g. Benayoun et al. 1971, Dyer 1972, Geoffrion and Hogan 1972, Geoffrion et al. 1972, Benson 1975, Zionts and Wallenius 1976, 1983, Steuer and Choo 1983, Nakayama and Sawaragi 1984, Reeves and Franz 1985, Wierzbicki 1986).
The standard methods are also used as integrated GDSSs. According to Jelassi (1986) and Lewis and Butler (1993), interaction in optimisation methods can be implemented as standalone models with no support for communication and information exchange, i.e. as DSSs. Furthermore, most of the previous methods either used direct exploitation of a decision maker's utility function (e.g. Zionts and Wallenius 1976) or setting of aspiration levels (e.g. Benson 1975). Therefore, Siskos and Despotis (1989) developed a DSS for compromise programming of multiple-objective linear programming problems. Actually, the idea of embedding multiple-criteria decision-making models in DSSs was suggested earlier by Zeleny (1982), Korhonen and Laakso (1986) and Keen (1987). Recently, other integrated systems have been developed to improve the participation of a decision-maker in the methods' interactive procedures (Lewandowski and Wierzbicki 1989, Steuer et al. 1993, Vetschera 1994. The main drawback of compromise programming remains the same as is in the preceding non-co-operative approaches, namely that the compromise obtained in a group meeting is not a case of a group consensus. It should be agreed through a structured group decision-making task (e.g. Kersten 1987, Iz andJelassi 1990). Then, due to decision-makers' criteria, the consensus can be the result of two-way communication and participatory democracy supported by using an interactive decision system (Blahna and Yonts-Shephard 1989, Knopp and Caldbeck 1990, Creighton 1993. To facilitate this, according to Watabe et al. (1992) and Iz (1992), GDSSs uses a coordinator or some interactive model to aggregate participants' preferences. Furthermore, the participation of managers during decision making could be properly addressed by both approaches. Thus, the interaction and communication, common to most decision-making tasks of decentralised (distributed) organisations ( Hoffman andMaier 1959, Iz 1992), could be most easily included in decision-making using these approaches.

Discussion
In Fig. 2, results of analysis of literature are presented in cross-sectional surface to frame the synthesis of the potential of GDSS to wood procurement. The following discussion about Group-Ware and its subset GDDS is concentrated to needs of managers (Fig. 1.) which is diffi cult to facilitate using DSS. The discussion is also related to future research trends. It can be stated that the conversion to teamwork-based organisations has succeeded surprisingly well in Finnish wood procurement. However, it seems, as Leppänen et al. (1999) suggest, that the share of the management based on approaches of group decision-making is still increasing, so as a conclusion, also teamwork is still looking for its niche. For example, there are needs for shifting from small procurement teams to very large fi eld teams, which would be responsible for large geographical areas. These teams could also be vertically integrated with the upper levels of the organisation on the basis of function. In this kind of arrangement, face-to-face communication may become more diffi cult, so the reliance on SDC and EMSs will probably increase. Therefore, research should be focused on the use of the both for internal relationships of groups and between different hierarchical levels of organisations.
Although SDC is not yet structured for computer use and in that sense useful in wood procurement, it seems to be helpful and effective meeting strategy for management. Related to GDSSs, a very important research problem of SDC structuring would be to examine how two new leadership styles, transactional leadership and transformational leadership, effect on the effi ciency of a group during SDC. It is obvious that one of these leadership styles should be selected, but it is not clear on what grounds. The selection could be based on research into GDSSs; it has already been determined that the different styles of leadership infl uence the creativity of GDSS groups (Couger et al. 1993, Sosik and Avolio 1998. It has also been noted that the creativity and performance of a group have a positive impact on each other. Then, the effect of the behavioural components of the leadership styles as grounds on the performance of a group in a task requiring creativity could be found out. Despite new structures of organisations and management, expectations for at least as fast wood fl ow remain, which may demand faster decision making. In this kind of situation, Linked GroupWare and EMS could provide communication facilities for increasing use of GDSSs. Furthermore, the increasing needs of decisionmaking groups for collaboration may lead to integration of EMSs and GDSSs. In theory, desktop GroupWare of videoconferencing could support communication processes between teams in their remote locations. Therefore, for future improvement of GDSS, the most important technical research problem related to EMSs would be to fi nd out how and to what extension GDSS can be combined with desktop videoconferencing. So far, the data conferencing properties of desktop videoconferencing systems have turned out to be useful for decision process (Koskinen 2000).
In that case GDSS included not only the use of software and document sharing, but also collaboration. This is in accordance with the suggestions of Cohen and Mankin (1999), that technical tools are needed during the decision process to solve confl icts, to support the process, and to fi nd optimal solutions. There are several collaborative phases needed for decision-making processes of wood procurement (Palander 1998c, Toivonen and Palander 1999, Toivonen et al. 2001. Therefore, besides the features related to EMSs, numerical technology is also needed for GDSSs to process information fi les to be used before a meeting as well as after it. Information is needed to aid in producing alternatives (e.g. brainstorming, optimisation) and in analysing them (e.g. visualisation). It seems that the elements of GroupWare applied to optimisation, visualisation, information storing and handling could be programmed for GDSSs based on the available literature. However, it is better to keep in mind that the optimisation models should include special heuristic support and heuristic guidelines, so that group decision-making could be effective and also innovative.
Aside the technological features for GDSSs, knowledge about the applicability of GDSSs for the social communication is needed for group processes and group dynamics. Furthermore, research fi ndings from psychology, education science, sociology, and forestry should be utilised to determine how GDSSs infl uence human behaviour in organisation. According to literature, if new GroupWare is established, it may have two kinds of effects: either it creates new obligations for the workers, in which case it will be considered an extra charge and will be rejected, or it replaces traditional work, then its use will be easily adopted (Finholt 1997). So far, only the attitudes of wood procurement managers towards EMSs have been surveyed (Palander et al. 2001).
According to the results, there are plenty of prospects for application, if technical function of EMSs is reliable. However, there are also a lot of prejudices and resistance to change in working competence, which conservative managers should overcome either alone or using supportive training and demonstration.
Use of CAV as an element of GroupWare is just its beginning. To support timber trading and customer meetings in future, CAV of forests could be used for presentations to participants of meetings. Then, it could be an effort to promote understanding of wood procurement processes. To facilitate both, it would be useful if remote sensing methods, and through them the stand selection system, could be joined to a graphically simulated virtual forest. More comprehensive CAVs could also be used for decision making about customer satisfaction. In that respect, CAV of forest alone could help participants by giving them information about forest area and by illustrating changes in the landscape caused by the implementation of various forest management and logging operations on the area and on different stands. If CAVs were a part of GDSSs, more advanced support would be available for customer satisfaction, but managers would have to be educated to show the effects of logging on other forest values. Probably then, GDSS with CAV could show the effects of wood procurement on biodiversity, landscape, logistics, multiple-use values, as well as on the future production profi tability of a forest stand.
An important research task would be to study whether customers need or desire to use CAVs for customer satisfaction, when wood procurement functioning is discussed with customers. The initial research would discern whether CAVs could encourage customers to be more receptive to co-operation. There is no doubt that the theory of group decision-making may be applicable to relationships with customers and interest groups. Furthermore, wood procurement seems to consist of interactive functions between forest owners, public and managers, between whom broad use of GroupWare-like telecommunication is possible. In order for this vision to become a reality, further research must be done before conclusive GroupWare for this purpose can be developed.
Based on this selective review of literature, the decision making of wood procurement could be supported by active elements of GroupWare. However, development of it supposes experiments and empirical studies in the entire group decision-making process. In addition to the already mentioned study areas, testing the practicality of user interfaces for specifi c GDSSs is a necessity for different organisations and interest groups. By analysing interviews and written reports of experiences, conclusions could be made about the innovations needed for development. Later study could then evaluate what and how the elements of GroupWare have been adopted and accepted, as well as how GDSSs are realistically used. Moreover, modern knowledge about the process could produce a general theory describing GroupWare and GDSSs.
It can be concluded that research should be carried out to investigate experiences of managers, customers and test groups to create 'core' knowledge for educating group members. First of all, the behavioural sides of organisation and interest groups should be studied, as well as management competence, to observe how personnel development for the utilisation of potential computer applications has been undertaken. Perhaps main contribution of this paper will be to stimulate others to treat the topic or subtopics in more depth. The fi nal aim could be to develop a virtual reality model to be used for research and education. This kind of virtual environment could even be used as part of a distance learning/ teaching technology directed at the personnel of organisations (Palander 1999b(Palander , 2000a. According to the vision, repetitive conclusions could be made on how to improve the competence of the decision-making group.