1990 - Environments for Supporting Decision Processes

Budapest, Hungary, 18-21 June

Proceedings edited by H.G. Sol & J. Vecsenyi.

Title Pages
Mumford, E.

This paper discusses the difficulties of effective decision making in complex and volatile organizational environments. It evaluates the contribution of information technology and of participative design.

View Full Text (PDF)
1-12
Williams, K. and Zmud, R. W.

While it has long been recognized that the strategic success of a business depends not on just one person but on the entire top management team, serious research examining the senior executive team has only recently surfaced. It is thus not surprising that the DSS literatures have tended to overlook both the promise and nature of a technological environment appropriate for this critically important organizational context. Still, the rich and growing design literatures regarding executive information systems (EIS), which support individual executives, and group decision support systems (GDSS), which support a group's decision behaviors, provide potentially valuable insights which can be applied to the design of a technical environment directed toward executive team support. Once a sufficiently rich understanding of the unique requirements of this particular organizational environment has been developed, the EIS and GDSS literatures are examined in order to identify design strategies and guidelines for an appropriate technological environment. These strategies and guidelines culminate in a technological architecture for executive team support. The paper concludes with suggestions for DSS research which should be undertaken if success is to be achieved in developing and implementing a technical support environment for the executive team.

View Full Text (PDF)
13-28
El Sawy, O. A.

This paper seeks to develop a perspective of decision environments that is structured favorably for the effective support of the strategic decision making process in turbulent environments. The paper adopts a view that issues are more appropriate units of analysis than decisions for considering decision support at the strategic level, and proposes a decision support environment that is structured around managing the lifecycle of strategic issues.

View Full Text (PDF)
29-36
Vamos, T.

Information society introduces a relevant change in the citizen's life similarly to all other activities. Its major features are: the availability of on-line, direct information which enables to surpass most previous geographical and time limitations; direct access of earlier detached knowledge; and by all that a possibility and need for new cooperative structures with all social, psychological, cultural consequences. The environment for this change is created by the computer and communication hardware, by the software services for different user requirements but many other supporting efforts are needed; These include procedural support as facilitation of cooperation, negotiation procedures which are going to be the kernel of every cooperative activity and a high-level cultural support for enabling the citizen to have an equal chance in the changing world. This new literacy which is much more a cultural educational background than the presently understood computer literacy will be the most important device for preserving durable values and entering a new epoch.

View Full Text (PDF)
37-46
Rajkovič, V. and Bohanec, M.

In the paper, the explanation of decision making knowledge is discussed in the light of understanding decisions. An expert system shell is taken as a tool for intelligent decision support. In this sense, decision knowledge is modeled as a knowledge base. The corresponding inference engine suggests the best decisions and explains the way by which the decision has been achieved. The main intention of the paper is to propose some techniques for explanation of the whole knowledge base, i.e., not only its part which is used in a particular problem-solving procedure. The so-called limited transparency of the traditional explanation is extended. The proposed knowledge explanation is illustrated by a practical example of evaluating programmers' performance using DECMAK, an expert system shell for multi-attribute decision making. Knowledge explanation is combined with knowledge acquisition and utilization in order to achieve a better fit between the decision maker and the system as a contribution to the quality of decision making.

View Full Text (PDF)
47-58
Berkeley, D., Humphreys, P., Larichev, O. I. and Moshkovich, H.

This paper describes the background to the development of ASTRIDA, and its forefathers MAUD and ZAPROS, each of which is a system capable of supporting individual, group and organisational decision making at a strategic level. We start by rejecting the philosophy of providing strategic decision support by “bootstrapping the decision maker” in favour of focusing on problem structuring, and we review how this idea was developed in MAUD within the context of multiattribute utility theory (MAUT). The advantages of procedures based on MAUT are discussed, as are their drawbacks, particularly their tendency to land the decision maker in “black holes” in the preference space. We examine how this problem was avoided in ZAPROS in constructing partial orderings of multiattributed alternatives within a verbal decision model. Remaining difficulties in providing support are traced to MAUD's and ZAPROS's inherent process modelling limitations. The basis for improving process modelling is explored in the context of dominance search theory and its practical application in ASTRIDA is detailed. We show how ASTRIDA's process model and thoroughly developed interactive dialogue facilitates problem structuring and choice between alternatives on the basis of the decision maker's preferences expressed in his own natural language. We conclude by examining how ASTRIDA supports the generation and reality-testing of new potential alternatives on the basis of an analysis of the difficulties experienced while considering the current ones.

View Full Text (PDF)
59-82
Bots, P. W. G.

This paper describes an approach to the design and implementation of information systems to support decision processes. The main characteristics of this approach are the fact that it focusses on supporting information workers, and the way it addresses the gradual shift from problem understanding to systems design. Understanding is reached through a model cycle of conceptualization and specification, resulting in a dynamic model of the problem situation. Using a technique called task analysis, decision processes of information workers are described in terms of tasks and decisions that in turn make reference to objects in the dynamic model. This combination of a specification of decision processes with a dynamic model of the problem situation makes it possible not only to analyze these processes, but also to test modified or newly designed ones. Furthermore, the task descriptions prove to be a suitable means for where and in which way information technology can be applied to improve task performance. Thus, the transition from problem understanding to systems design is made. As an example, task analysis is applied to the problem of managing a factory. It is shown how a task structure can serve as a blueprint for a supporting information system. The flexibility of task structures is illustrated by simulating a general strike and its impact on the day-to-day decision processes of the firm's management and its need for decision support. Finally, the outline is given of a problem solving support environment that provides the tools that are needed for dynamic modeling, task analysis, and the construction and management of support objects.

View Full Text (PDF)
83-104
Hill, T. R. and Jones, B. H.

The purpose of this paper is to introduce the long-term research plan at the University of Hawaii designed to explore the use of information systems tools in the support of multiattribute negotiations. This paper presents a conceptualization of a Theory of Negotiation Support in terms of an objective, characteristics and an approach to development. It is argued that the objective of such a theory would be to decompose the negotiation process into the underlying cognitive processes and the corresponding constraints which limit negotiators' rational behavior. The decomposition approach suggested by Todd (1990) for Decision Support System Design may be applicable to the “softer” area of negotiation, given the potential of Neural Network Technology. Negotiation involves many “higher order” Elemental Information Processes, such as noise filtering and pattern recognition, which are more abstract than those appropriate for much of traditional decision support. However, Neural Networks have great potential for modeling such “higher order” processes and thus for capitalizing on the decomposition approach for negotiation. A Negotiation Support System prototype is being developed to address the basic EIPs involved in negotiation, through a color graphics, mouse-driven windows environment. Preliminary findings suggest the potential of this approach and future extensions.

View Full Text (PDF)
105-114
Sol, H. G.

The main object of Information Systems is to increase the effectivity of knowledge workers by making appropriate use of information technology. Heading for a global, networked society, substantial improvement is likely in the area of primary organizational processes by effectively tuning decisions to supporting activities. The field of information systems should therefore be provided with adequate design philosophies and approaches, aimed at a continuous structuring of organizations at micro, meso and macro levels.

View Full Text (PDF)
115-128
Savolainen, V.

It is of great importance for the decision maker to know the preconditions and the overall favourable atmosphere which should prevail for effective and successful Information Technology (IT) decisions in an organization. The overall favourable atmosphere includes a favourable state of contingency factors of the information systems development activities in the organization. Other preconditions of successful IT decisions include a proper analysis of decision alternatives and design measures followed by specification of alternative implementation plans. The IT decision classes as well as definitions of favourable circumstances and preconditions are constructed starting from long range strategic level decisions in organizations and ending up with the main project control decisions in the information systems development life cycle.

View Full Text (PDF)
129-140
Dur, R. C. J. and Sol, H. G.

Technological tools currently applied within offices are usually designed to support office activities at a low level. How these tools are to be used within the wider scope of office tasks, is left to the information workers themselves. As the complexity of information processing tasks within organizations increases, there is a growing demand for adequate support of information workers in performing their tasks. To address this problem, a modelling technique based on information processing tasks and on the notion of coordination of tasks, is presented.

View Full Text (PDF)
141-156
Docherty, P. and Ivanov, K.
Keywords:Decision Support Knowledge-Based Systems Expert Systems Cooperative Work Professional Skill Constructive Learning Action Control Theory Contrastive Task Analysis Social Systems Science Social Welfare Assistance

This paper identifies some challenges which arise in the context of developing an administrative decision support system for handling requests for social assistance. The focus of the inquiry will be on the social environment of a particular decision support technology as it is applied in an unusual environment which requires special considerations of organizational support. A main goal of the analyzed project has been to create a tool or support system, using both conventional DP (data processing) and KBS (knowledge-based systems) techniques, which strengthens the professional skill of the user rather than attempting to automate part of it. This has required an analysis and understanding of the administrative work in order that the system should mirror its structure and the user's actual working procedures. The emphasis of this paper, however, is to gradually shift the attention from the technical-factual description of the system to the social environment and to the need of organizational support. With this purpose in mind a survey is made of the explicitly stated objectives of the system; mainly improved service through speedier and more nuanced case handling for clients, improved case handling skills for social secretaries, and provide a basis for future evaluation and redesign. The objectives are then analyzed and related to the handling of the applications and to the possibilities of using the system in the future for other purposes, and with other consequences that may not have been anticipated by the designers. Some conclusions are then drawn regarding efforts to create and use methods and theories for development of support systems, including not only the computer field but also the context of social and economic systems such as planning-programming-budgeting. Methods for systems development tend to be too much conditioned by the appearance of new support technologies, and the real challenge is to see how such new technologies relate to older established knowledge and to unusual social environments.

View Full Text (PDF)
157-176
Kobashi, Y.

In an effort to help members of a volunteer organization engage in more effective conversation concerning the qualities of the social service programs they provide, a MAUT-based decision aid was modified and incorporated in a method called the “Championship Technique”. This technique exploited an event which was conceptually analogous to an athletic championship meeting where the programs were pitted against each other. Organization members with various standpoints participated in preparing and refereeing the competition, which was conducted using a set of questionnaires. Evaluation structures elicited in this process were used to build MAUT decision models which in turn rank-ordered the programs. The technique turned out to be not only useful in its immediate effect on this particular evaluation process but also satisfied general quality criteria of support.

View Full Text (PDF)
177-186
Taylor, A. D. and Weaver, J. C.

Some of the most important, scarce and expensive human resources employed by organisations are not involved in either production or management decision making. They are providers of intellectual analytical services and are used in an advisory and ‘enabling’ capacity by decision makers. Prime examples include lawyers, policy advisers and consultants. To date, these areas of work have not generally been considered as potential applications for DSS – except with regard to their practice management. We suggest, however, that DSS tools and techniques could have a far greater impact on organisations if their application was extended to these ‘conceptual’ support activities. This paper describes, contrasts and explains the thinking behind two major case studies of decision support systems for such 'professional' or 'knowledge' workers - one for solicitors in a major London law firm, the other for policy advisers in UK Government Departments. On the basis of the lessons derived from these two projects, we believe that a new generic type of DSS is emerging.

View Full Text (PDF)
187-202
Kottemann, J. E. and Remus, W. E.

One common capability of Decision Support Systems (DSS) is to allow decision makers to develop formal models and to “what if” scenarios of the problem. This capability is purported to improve decision making. The experiment reported here investigates the effect of DSS formal models and “what if” analysis on performance both when learning is occurring and when learning is completed. The experimental task used was production scheduling. We found that neither formal models or “what if” analysis helped improve decisions.

View Full Text (PDF)
203-214
Holt, C. C.

The time is rapidly approaching when it will be economically efficient for all organization workers at every level to be supported by relevant computer, data and communications systems in tasks involving information. Then full effectiveness will require joint design of organization structure and its supporting information technology. If the decisions and communications in organizations are to be supported comprehensively and successfully by modern information technology, a common conceptual framework must be formulated and agreed to by the managers, users and information system designers. This paper explores processes for making decisions that are good in overall performance including the costs of information and analysis incurred in making the decisions. Six models are presented in the order of their applicability to problems of increasing complexity on the continuum from isolated decisions to systems design problems. Both cognitive decision processes and organization culture are addressed.

View Full Text (PDF)
215-238
Persson, I.

This paper is based on a research project which started with the goal of understanding the need for information and knowledge in the investment process. The method was case studies where the needs in the process were defined as those which were found in the chosen projects. Based on the collected data, the possibilities which could be created by information technology are explored. The analysis in the paper is limited to initiation phase and calculus. The initiation phase and which managers initiated the project was found to be very important. In the analysis of this phase, information is separated into three groups - general information, information about the situation and perspective/solution -areas. The possibilities of decision support is then described. The technical possibilities in calculation, offered by DSS, are well known. In the project, however, it was probably more important which calculations were done and when they were done. The hypothesis is, that a GDSS can be a tool to increase commitment and to prepare for management of the ready system.

View Full Text (PDF)
239-256
McCosh, A. M.

The special group decision support system design problems which arise when the group's commonality involves applying several expertises to a single decision class are considered. The decision domain is the international bond market, and the internal decision environment possesses a clear and unitary channel of communication to the client market. This feature simplifies the prototype bond switching model discussed.

View Full Text (PDF)
257-276
de Jong, R. and Sol, H. G.

In solving vehicle routing and scheduling problems little attention is paid to punctuality of departures and arrivals. Customers often have strict time windows, but these are violated frequently. An environment with a TRIP DATA ANALYZER and a TRIP EXECUTION SIMULATOR is described. Both tools form part of a feedback connection between trip execution and trip planning. Feedback information improves the creation of trip plans that actually do meet time constraints. Also, a short survey is given of our experiences with respect to the process of designing an environment to support operational trip planning.

View Full Text (PDF)
277-294
Leclerc, A. and Paris, J.

The PIMS project (Project Integrated Management System) was a 3.5 year ESPRIT project that ended in July 1989. Its goal was to support software project managers in their activities of management. In this paper, we will present the PIMS approach for methodological support of software project managers. The method support allows a method engineer to define a management method and provides guidance to a project manager in managing his project according to a method. The method definition consists, on the one hand, of the definition of a network of activities of management (the activities, their relationships and their triggering events), and, on the other hand, of the connexion of the terminal activities to the PIMS tools which support their realisation. The method guidance proposes a control board showing both the progress of the management activities and a set of possible activities to be performed, and, activates automatically the tools supporting a chosen activity. In this way, the project manager is aided in the process of project management, i.e. in the choice of the activity to be performed (What to do, when and why?) by a method guide and in the achievement of an activity (How to do it?) by management tools.

View Full Text (PDF)
295-308
Fekete-Szucs, L.

Over the past few years the management style of top manager has changed in many organisations. As an “exercise of discretion within prescribed limits” discussing strategic issues has became rather a cooperative game of mission oriented people of an organisation. This game helps groups of managers to take into account uncertainty, preferences, make judgements and take decisions. Decision Conferencing (Figure 1.) is one approach to group decision support systems. Its main pillars, as disciplines are the decision theory, group processes and information technology. Normally it is an intensive two-day session of those organisational people who realising a live decision problem, can be interested and involved to formulate a corporate strategy and fulfill them. The management team work is aided by an experienced decision analyst team, who helps the managers to structure their thinking, think creatively and imaginatively about the problem and interpret the results. The output of such a two-day meeting, developed by the management team: a shared understanding of the problem and an agreement about who, what, when to do in order to implement the accepted solution. The model, created in a Decision Conference reflects the scheme of thinking of the management group, therefore it demonstrates what the organisation can do tomorrow. Decision Conferencing work, based on the scenario of Dr. L. D. Phillips (LSE/London) has been successfully tested in banks, software houses, industrial companies and other organisations over the past two years in Hungary, as well. Most of the cases were about business policy rather than resource allocation. A brief description of Decision Conferencing, a few words about SZÅMALK-ICL Decision Conference Center project, two case studies, highlighting the Hungarian peculiarities of preparing, running and post analysis activities in connection with the two days, and several conclusions by suggesting a number of outstanding problems that deserve attention in the future, are described by the paper.

View Full Text (PDF)
309-318
Leatherwood, M. L., Dilla, W. N. and Boland, R. J.

This study investigates an important feature of information processing environments, communication structure, and its effects on decision makers’ perceptions of both the task and the social environment under varying incentive schemes. Using a classic retailing decision task where purchasing and merchandising managers are faced with quantity and price decisions (Ackoff, 1967), three types of networked communication environments were manipulated: no communication, unidirectional communication, and bidirectional communication. The incentives provided participants were also varied with each of the communication conditions and were either conflicting, based on individual department performance, or nonconflicting, based on division-wide performance. As predicted, managerial uncertainty was reduced by allowing managers communication access even when other aspects of the task environment (database information) were held constant. Communication access facilitated task understanding and increased managers’ perceptions that they had better information about their own and the other manager's payoffs. Finally, although unidirectional and bidirectional communication had similar effects on task understanding and perceived information adequacy, unidirectional information affected the extent to which managers a) were forced to consider the actions of others and b) felt the need to cooperate in order to increase their payoffs. These perceptions of the social environment were significantly different when bidirectional communication was allowed.

View Full Text (PDF)
319-334
Ching, C., Holsapple, C. W. and Whinston, A. B.
335-356
Meier, J. and Sprague, R. H.

This paper explores the relationships among the adaptability of organizations and the systems which support them, as they respond to changes in their environment. It develops the rationale, the key concepts, and the reality of adaptive organizations and systems. A major section develops a conceptual model based on the traditional views and assumptions of how organizations and systems adapt to each other, and to changes in their environment. It reveals some popular misconceptions about these relationships and suggests that they result in a structural level of organizational ineffectiveness that is intolerable. It also proposes a major research plan to develop new knowledge on this subject, and summarizes what we have accomplished with this paper.

View Full Text (PDF)
357-370
Carlsson, S. A.

Learning and unlearning are critical processes in organizations. The relationship between Decision Support Systems (DSS) and learning have been pointed out in several conceptual and empirical DSS-studies. DSS benefits like better understanding of a problem and improved insight reflect learning. Based on what has been missing in previous DSS and learning studies this paper focus on organizational learning, unlearning, and organizational learning systems. A case study of spreadsheet program users is used to exemplify the concepts.

View Full Text (PDF)
371-385