DELPHI, CMC, INQUIRY SYSTEMS, etc.
Supplementary lectures notes for cis 679.
Gives material which augments current lecture tapes
© Copyright Murray Turoff 1998

Delphi: Delphi is a method that can be used to gather requirements from much larger samples of users than would other wise me possible. Delphi can be done through the computer if you have the right software setup. It allows systematic input from users and a true sense of participation in the early phases of the design.

DELPHI METHOD

Prepare an initial survey which summarizes what is already known about the problem. Send this to the participants. Each question is open ended so they can add items that were not considered. They answer anonymously. They know they will be given a chance to change their view. Get the inputs back and summarize them along with providing an chance now evaluate all the items from the first round.

For example, each requirement that was suggested might be evaluate on scales of "desirability" and "feasibility." On the third round arrange all requirements from most desirable on down. Ask people to comment on reasons for low feasibility of ratings of some desirable requirements. Point out disagreements on desirability or feasibility and ask those with extreme views to justify their view.

REASONS FOR DELPHI

DELPHI CHARACTERISTICS

DELPHI OBJECTIVES

DELPHI JUSTIFICATION

TREND ANALYSIS & PLANNING I

TREND ANALYSIS & PLANNING II  

TREND ANALYSIS & PLANNING III

Structural modeling is the concept of allowing users to build models without need to be able to program. There are a wide range of approaches in the literature. Collaborative modeling is a new application area of group communications.

Note that an accurate model of the future would require 10 million pieces of subjective data from a decision maker to represent the interaction among 10 possible future events. Therefore we do structural models that are approximations to the real world and which have emerged form a wide range of disciplines.

Memory means that the occurrence of an event in the future is dependent upon the sequence of events that occurred first. So every possible sequence must be considered. Non memory methods (e.g. Markov models) assume no memory. It still leaves a many body problem and 5000 pieces of data for a model of 10 events are still very large. Most methods are matrix oriented to examine only principle relationships among each pair of objects and ignore three body and higher interactions.

STRUCTURAL MODELS

ELECTRONIC DELPHI

Unrealized opportunities
Prior lack of graphics
Shared memories
Integration of Structural Modeling tools
Should be tailorable in general Computer Mediated Communications (CMC) Systems
On going (continuous, asynchronous) Delphi?s and Collaborative Expert Systems

The literature on CMC (Computer Mediated Communications occurs under a long list of different names that have been invented for it since 1971 when the first system was created. The use of the term mail is like the use of the term "horseless carriages" when automobiles first came out. That metaphor of mail limits the imagination of people in realizing what is possible once you put human communications through a computer system.

CMC: A rose is a rose

CMC OBJECTIVES  

We can design anything we want from free and open communications to a dictatorship.

DICTATORSHIP DESIGN

NEW COMMUNICATION MEDIUM

NJIT TAILORED SYSTEM EXPERIENCE

Proven Applications in Organizations

Project Management
Crisis Management
Collaborative Composition of Reports
Task Scheduling and Tracking
Collaborative Budget Estimation
Planning and Forecasting
Product and Customer Support
Exploration of complex situations
Resource Allocation Analysis
Administrative Tracking
Strategic Analysis

ORGANIZATIONAL CRISIS SITUATIONS

Levels of GDSS Support

1. Providing alternate communication channels for the group. This is the process of allowing a group to work more efficiently and/or more effectively with text, structured data, and graphics without regard to the limitations of synchronous meetings. It includes the ideas of such facilities as "group memory" and "group graphics."
2. Providing process structuring for communication protocols and human roles. This includes the improvement of the group process by addressing such concerns as: (a) Software support for a leadership or facilitation role (appointed or elected); (b) Participation equality (e.g., requiring inputs from each participant, as in nominal group technique); (c) Anonymity to encourage free exchange of ideas and opinions as in Delphi techniques; and (d) Voting protocols (e.g., one vote for preferred option, rating of each alternate, rank ordering of alternative).
3. Supporting the collection, organization, filtering, formatting, feedback, and retrieval of any material (text, data, graphics) generated or required by the group to support its deliberations.
4. Integrating and using sophisticated decision aids in support of the group process (e.g., structural modeling, games and simulations, and statistical analysis and forecasting).
5. Integrating problem structures into the communication process (e.g., group calendars, task tracking structures, pert structures, etc.).

KEY ISSUES IN CMC

THE TRADEOFF

ISSUES OF STRUCTURE

TYPICAL HUMAN FACILITATION FUNCTIONS

GROUP COMMUNICATIONS STRUCTURES FOR INFORMATION WORK

ORGANIZATIONAL HUMAN ROLES

PROBLEM SOLVING PARTS

The following dimensions of human communications results from a multidimensional scaling study as to how people think about a communication situation. Most human considerations are handled by fewer dimensions. But communications are particular important to the individual and there for quite complex mentally.

DIMENSIONS OF HUMAN COMMUNICATION

FACTORS INFLUENCING STRUCTURE

CMC META PROCESSES

OBJECTS OF DISCOURSE

GROUP PROCESS GAINS:

GROUP PROCESS LOSSES:

CMC DESIGN CONCEPTS

ORGANIZATIONAL IMPACTS OF INFORMATION SYSTEMS VS. CMC

SUPERCONNECTIVITY

WHY A SLOW PROCESS

  A FACTORS ORIENTED CONTROL MODEL OF GDSS

Contextual Factors:

Outcome Factors:

Intervening Factors:

Adaptation Factors:

In the above the degree of complexity of a problem depends on which of the above relationships (links) are explicit or implicit in the problem solving process and which of the data elements (nodes) deal with subjective or objective data and information.

DIMENSIONS FOR DESIGNING CMC SYSTEMS

STRUCTURED: There is an existing structure that is accepted as the framework for addressing this type of problem. This structure serves as a basis into which all the relevant factors or considerations can be placed. The interrelationships of the components of the structure are also well understood.

SEMI-STRUCTURED: There is structure that can be applied to the problem that is generally accepted. However, the interrelationships among components of the structure are not fixed and are somewhat of an arbitrary function, depending on the nature of the problem and the determinations made by the individuals involved.  

UNSTRUCTURED: There is no accepted structure for examination of the problem; however, there is acceptance that examination of the problem by the group can produce sufficient insight to evolve a structure for consideration of the problem by the group.

WICKED: Not only is there no accepted structure, but there is also general lack of confidence that the group can gain complete insight into all the relevant considerations. Sometimes this is expressed as the existence of "Unk Unks" (Unknown Unknowns) that could invalidate any group outcome. Furthermore, there is no assurance that the group can reach agreement on the structure or tools for dealing with the problem.

DEDUCTIVE: A logical process or deductive procedure can be arrived at by the group to make the connection between the observations available about the problem and the findings of the group. While there may still be a need for a consensus on the final result, the relationships exposed in the problem solving process will largely be accepted as obvious by the "rational" group members.

INDUCTIVE: Many connections to be made between the observations and the findings of the group are largely subjective in nature. There must be a consensus not only on the findings, but also on the relationships needed to reach the findings. Once again, rationality is assumed.

RELATIVE: It is assumed that there is no optimum or single truth to emerge. Truth must be found by comparing alternative findings in relationship to agreed objectives or goals. Either inductive or deductive methods may be employed to connect observations to findings and findings to objectives. However, it is the relationships between findings and objectives that dominates the validation process. The results are not inherently "true" outside of the context of the problem specifics.

NEGOTIATED: A group "negotiates" what is to be considered true. Truth does not have to be tied to any external realities or observations. Reaching an agreement on values as expressed by policies and goals is a very typical negotiating situation. Reality is socially defined, based upon the group judgment process.

CONFLICTUAL: Truth can only be derived from the strongest possible conflict for each finding. Truth emerges only from the examination of the strongest possible opposing views being considered for the acceptance or rejection of each finding. Out of this conflict it is hoped a "synthesis" of truth will emerge; however, there is no guarantee of such.

No need to use a more complex inquiry process than is dictated by the problem. It is the feedback loops in organizations that provide for changes in the interpretation of past history and future objectives that allow for a Singarian or negotiated reality inquiry process.

PARALLEL: Each individual approaches the problem entirely independently of the other members of the group. There may be information provided from each member to the group on the status of his or her progress, but there is no imposition of any group process upon the performance of the individual members. There may be aids and structure to provide for the organization and filtering of the information generated by the individuals involved, but individual judgments and problem solving behavior are disjunctive in nature. There is no imposition of any collective group result or standardization for producing such. The resulting group process is largely one of information exchange.

POOLED: The group imposes a structure or standard to capture and represent individual contributions into a collective group representation of the efforts of each member. This may include voting on and assessment ratings of the material generated in the problem solving process. There is, however, no constraint on sequence of activities that each member may undertake to arrive at individual results to be incorporated into the collective group results. The process is highly conjunctive in nature.

SEQUENTIAL: The group imposes phases on the problem solving process that must be undertaken in a sequential manner by all the members of the group. Members of the group must deal with and resolve a phase of the problem before moving on to another phase. There is usually a well-defined strategy or plan for the nature of the phase and their sequence of occurrence. In this sense, one may view this as a "static" group process.

RECIPROCAL: In this mode, there is some form of consistency imposed upon the various aspects of the problem, so that any change in one part of the problem solving process can impose changes in other parts. The sequence in which phases can be dealt with depends upon the results that occur in any given phase. In this sense, the nature of the group process can be considered "adaptive" as opposed to "planned."

References:

I have added references for the above and some others for other concepts discussed in the lectures.

Bandyopadhyay, R., Information for Organizational Decision Making: A literature Review, IEEE Transactions on Systems, Man and Cybernetics, Volume SMC-7, number 1, 1977. Reviews the schools of information theory.

Churchman, C.W., The Design of Inquiry Systems, Academic Press, New York, 1971. Original source for inquiry system material.

Hiltz, S. R. & Turoff, Murray, The Network Nation, Revised Edition, MIT Press, 1993. Chapter 14 discusses the dimensions for characterizing CMC

Kerr, E., and Hiltz, S. R., Computer-Mediated Communication Systems, Academic Press, New York, 1982. Summary of many early findings for group communications.

Kurke, L.B., and Aldrich, H.W., "Mintzburg Was Right: A Replication and extension of the Nature of Managerial Work," Management Science, (29:8), August 1983, pp. 975-984. How do managers spend their time.

Lendaris, G., "Structural Modeling: A Tutorial Guide," IEEE Transactions on Systems, Man & Cybernetics, (SMC-10:12), December 1980, 807-840. Tutorial on Structural Modeling

Lindblom, C.E., "The Science of Muddling Through, Public Administration Review, (19:2), Spring 1959. A must for anyone interested in how to manage in bureaucratic organizations.

Linstone, Harold & Turoff, Murray editors: The Delphi Method: Techniques and Applications, Addison Wesley Advanced Book Program, 1975. This is considered the standard reference text for those interested in doing Delphi studies. There is also a good philosophical section which covers inquiry systems and negotiated reality.

Mowshowitz, A., 1994, Virtual Organizations: A vision of management in the information age. Information Society, Volume 10, 267-288.

Mowshowitz, A., 1992, The market value of information commodities, Journal of the American Society of Information Science 43, 225-232.

Mowshowitz, A., 1992, Virtual feudalism: A vision of political organization in the information age., Information and the public Sector 2, 213-231.

Pyhrr, P.A., "Zero-Base Budgeting," Harvard Business Review, November-December, 1970. The original concept explained.

Pyhrr, P.A., Zero-Base Budgeting: A Practical Management Tool for Evaluating Expenses, John Wiley and Sons, 1973.

Rana, Ajaz R., Turoff, Murray, & Hiltz, S. R., Task and Technology Interaction (TTI): A Theory of Technological support for Group Tasks, Proceedings of the 29th Annual Hawaii International Conference on System Sciences, Vol. IV, Los Alamitos, CA: IEEE Computer Society Press, 1996. Follow on paper in the 30^th meeting for more detail on how to understand CMC systems

Rohrbaugh, J., "Improving the quality of group judgment: Social judgment analysis and the Nominal Group Technique," Organizational Behavior and Human Performance, (28), 1981, 272 -288. Group facilitation methods

Turoff, Murray, Virtuality, invited paper for special section of CACM, Volume 40, Number 9, September 1997, pp. 38-43. Related to work on Virtual Organizations, this issue has a set of papers on virtual organizaions.

Turoff, Murray and S. R. Hiltz, (1995), Computer Based Delphi Processes, in Michael Adler and Erio Ziglio, editors., Gazing Into the Oracle: The Delphi Method and Its Application to Social Policy and Public Health, London, Kingsley Publishers, pp. 56-88. A version of this paper is at: http://eies.njit.edu.~turoff/

Turoff, Murray, S. R. Hiltz, A. N. F. Bahgat, and Ajaz Rana. (1993), Distributed Group Support Systems, MIS Quarterly; December 1993, 399-417. History of early work in CMC.

Turoff, Murray, Rao, Usha, & Hiltz, S. R., (1991) Collaborative Hypertext in Computer Mediated Communications, Proceeding of the Hawaii International Conference on Systems Science, Volume 4, January, IEEE Computer Society, 357-366.

Turoff, Murray, (1990), Computer Mediated Communication Requirements for Group Support, Journal of Organizational Computing, Volume 1, Number 1. General paper on the design of CMC systems.

Turoff, M., "The Anatomy of a Compute Application Innovation: Computer Mediated Communications (CMC)," Journal of Technological Forecasting and Social Change, (36), 1989, 107-122. Historical overview.

Turoff, Murray, (1985), Information & Value: The Internal Information Marketplace, Journal of Technological Forecasting and Social Change, Volume 27, Number 4, July, 357-373. Concept of marketplace organizations.

Turoff, Murray, & Chinai, Sanjit, (1985), An Electronic Information Marketplace, Computer Networks and ISDN Systems, Volume 9, Number 2, February.

Van deVen, A., and Delbecq, A.L., "Nominal and Interacting Group Processes for Committee Decision Making Effectiveness," Academy of Management Journal, (14:2), 1971. Group facilitation methods.

Van deVen, A. H., and Delbecq, A.L., "The Effectiveness of Nominal, Delphi and Interacting Group Decision Making Process", Acad., Management Journal, 17, 1974.

Vennix, Jac A.M., Jan W. Gubbels, Doeke Post, and Henk J. Poppen, "A Structured approach to Knowledge Elicitation in Conceptual Model Building," System Dynamics Review, (6:2), Summer 1990, 31-45. Structural modeling using system dynamics.

Warfield, J.N., "Toward interpretation of complex structural models," IEEE Transactions on Systems, Man and Cybernetics, (SMC-4), 1974, 405-417.