Decision-support and Complexity in Decision Making ACADEMIC TRACK

Organisations increasingly have to deal with complex problems. Multidisciplinary teams are needed to cope with such problems. In such teams, different people have different perspectives, knowledge and approaches. For decision-making on complex problems, this knowledge has to be shared, and new knowledge has to be constructed, in order to develop solutions for the problem. Theory suggests that ICT-tools can support the quality of decisionmaking on complex problems. ICT-tools may facilitate knowledge elicitation, knowledge sharing, reaching common ground, and, ultimately, knowledge construction. Furthermore, this facilitation may positively affect the quality of the proposed solutions. We propose research on external representations used in multidisciplinary teams to test these expectations. Decision-support and Complexity in Decision Making 3 Business organisations are increasingly confronted with complexity in decision-making situations (e.g., Courtney, 2001; Rotmans, 1998; Sterman, 1994). Changes come at everincreasing pace, and developments in communication and transportation allow for worldwide competition. In such a competitive environment, novel approaches are needed to attain and maintain business advantage. However, traditional approaches to decision making (i.e. adaptive, incremental approaches) fail to generate those innovative solutions necessary for keeping such advantage (Lomi, Larsen, & Ginsberg, 1997). ICT-tools are used to enhance the quality of the decision-making process. These generally aim at facilitating formal and informal communication, harvesting knowledge, and building knowledge repositories (Courtney, 2001). Such ICT-tools can be regarded as responses to complexity. However, it is not clear by which mechanisms ICT-tools are able to support decision-making for complex problems. In this paper, we present a conceptual framework for the analysis of ICT-tools. Such a framework can guide in distinguishing between a variety of ICT-tools, and act as a basis for development of new ICT-tools for support of decision-making. We explore how ICT-tools can be used to facilitate decision-making on complex problems in multidisciplinary teams. The leading thread in this paper is the question how ICT-tools can effectively support the quality of decision making for complex policy problems. First we describe various aspects of complexity and decision-making, from which we gather requirements for decision making on complex problems. We then discuss group processes in decision-making teams, to identify key processes for decision support. After that, tools for supporting these key processes are dealt with. From an integration of these topics we gather research hypotheses, for which we propose some first designs for empirical study. Complexity and decision-making The higher the number of factors and relations within a system, the more complex this system is (cf. Evans & Marciniak, 1987). Small changes in parts of a complex system may result in considerable changes in the system as a whole. This is caused by the intricacy of the various Decision-support and Complexity in Decision Making 4 relationships. Feedback mechanisms and delays can cause non-linear behaviour (Rotmans & Van Asselt, 1999). Complex problems in decision-making often exist across disciplinary boundaries, and thus require a multidisciplinary approach (Rotmans, 1998). To enable such an approach, decision-making for complex problems is often done in multidisciplinary teams. In such teams, every team member has his or her own perspective, which can be seen as a coherent and consistent description of the perceptual screen through which (teams of) people represent the world (cf. Van Asselt, 2000). The way people represent problems is subject to their individual perspective, hence multiple representations of a decision-making problem exist in multidisciplinary teams. Decision-makers in multidisciplinary teams confronted with complex problems have to take the existence of multiple problem representations into account. Multiple representations lead decision-makers to regard problem solutions they normally (i.e. when considering only one problem representation or perspective, or only a limited synthesis of multiple perspectives) would not regard. Or, as Vennix puts it, “the more different perspectives are taken into account, the smaller the chances of premature problem definition and ‘solving the wrong problem’” (Vennix, 1996, p. 1). The risk of only considering business-as-usual strategies (i.e., adaptive, incremental) is averted, and innovative designs can be developed if multiple representations are taken into account. The nature of complex problems requires identifying and articulating multiple problem representations, and taking both innovative and adaptive solutions into consideration. We hypothesise that effective decision-support will help articulation of multiple problem representations, and developing both innovative and adaptive solutions. Group processes Decision-making teams engage in several social and knowledge processes, which result in shared and newly constructed knowledge, and in the development of problem solutions. During the process, social relationships between team members evolve, influencing the decisionDecision-support and Complexity in Decision Making 5 making process. We see the developed solutions as the result of knowledge processes, which in turn are affected by the evolving social processes. Both need to be regulated for effective decisions to ensue (cf. Mulder & Swaak, 2001). One way of decision-support aims at the regulation of group reasoning processes. For example, Suthers uses a tool called Belvedere to enhance scientific argumentation in groups (Suthers, 2001). Belvedere asks evidence for every statement made by the group. It also asks for evidence opposing a statement. Thirdly, Belvedere prompts checking whether stated evidence supports statements other than the one it was given for. Such measures may support decisionmaking by helping teams distinguish between strong and weak statements. With respect to knowledge processes, regulation refers to structuring the processes of knowledge elicitation, knowledge sharing, and knowledge construction, and balancing their respective importance. The same sort of government is needed with respect to the social processes, because of the relationship between social and knowledge processes. A positive social environment in a decision-making group is a prerequisite for effective knowledge processes. In other words, regulative processes can be seen as the guidance of group attention to specific knowledge and social processes, and decision-support as an example of such guidance. Regulation also refers to monitoring the decision-making process. Dealing with complex problems means performing a number of decision-making steps, like problem definition, articulation of perspectives, development of solutions and alternatives, testing those solutions, and implementing them (cf. Van Asselt, 2001, and Courtney, 2001). Regulation then refers to monitoring the performance of these steps by the decision-making group. Knowledge processes To be able to develop solutions for complex problems, decision-making teams have to construct a shared representation of the problem. Individual team members, with individual problem representations and mental models, embark on discussions about the problem. A number of knowledge processes are important to these discussions. These involve knowledge elicitation, knowledge sharing, and, through reflection and elaboration on, and synthesis of Decision-support and Complexity in Decision Making 6 shared knowledge, knowledge construction. All these processes are aimed at dealing with the present problem. Knowledge elicitation, knowledge sharing, and knowledge construction each play a role in a specific knowledge transformation. Knowledge elicitation refers to the transformation of implicit knowledge into explicit knowledge. Knowledge sharing means that internal knowledge becomes external, or sharable knowledge. If various group members add to a pool of external knowledge, they can all internalise each other’s knowledge, enabling reflection upon the pooled knowledge. Finally, knowledge construction can occur (see Figure 1). Figure 1: Knowledge elicitation, sharing, internalisation and construction change knowledge from implicit state to shared state, and, through group reflection, enable knowledge construction. Knowledge construction is portrayed here using the circular arrow. Common ground and knowledge construction Decision-making for complex problems requires decision-makers to share their individual knowledge with the rest of the group they are working in. Sharing knowledge is not simply a transmission of information from one person to another. Knowledge from one member Knowledge processes and transformations