Modeling System Evolution:
A Means of Cutting Through Obstacles to Desirable Change to Large Business (Information) Systems, Their Many Computer-Human Interfaces, and User/Operator Responsibilities
(abstract of poster presented at CHI '92)
Too often, attempted change and change planning suffers from inadequate insight, coordination, and involvement. Decision makers frequently get caught up in power/political contests, utilize and are affected by cognitive (Anderson, 1991) or organizational (Grudin, 1991) structures and procedures that hinder beneficial cooperation/participation, and focus excessively on visible technology end-states and inadequately on business and human processes, underlying infrastructure, and enabling intermediate-states. Hence, resulting change, if achieved at all, is often change without readiness, buy-in, or gain.
To counter or avoid these problems during attempts at improving and planning improvements to the systems of large corporations, an approach to modeling system change was developed. Ingredients of this approach that are believed to be essential include: an understanding of (the need for) evolving systems and an evolving orchestration of system evolution (Sacerdoti, 1991); a means of depicting/constructing a "picture plan" of system change -- one that enables seeing the entire picture (at a particular level) in one view (Nelson, 1991); a means of identifying system "character" and intermediate character states through which a system should evolve; continual model development with involvement/input from everyone (to be) affected by the change.
Pieces of a change model developed for a large corporation's procurement system are presented to reveal the possible products/tools of such an approach. Among them are textual descriptions of planned system changes arranged on a large, two-dimensional surface in such a way as to reveal category of business process each affects, phase of system evolution during which each is to occur, and temporal order of planned occurrence. Windows presenting rationale and elaboration are superimposed on the surface when requested. Windows presenting change rationale are accompanied, as appropriate, by lines connecting the change to prerequisite changes and/or to changes it enables. Change elaboration often appears as conceptual snapshots consisting of computer-human interface mockups/prototypes and/or graphical depictions of work processes.
Such a model facilitates consideration of the full picture of what is needed to get from point "a" to point "b," and it provides a vehicle for interaction and a framework to structure thinking about system change. However, tools need to be developed specifically for change modeling to reduce the cumbersomeness and time of model construction, to ensure maintenance of necessary relationships among system changes, to guide the construction of good models, and to facilitate multi-user model development and the tracking of model changes.
References
Anderson, R. I. (1991) "Models we live by:" Two models of computer-human interaction that govern the degree and nature of user involvement in system development. Proceedings of the Bellcore/BCC Symposium on User-Centered Design: Making It Happen, 23-24. Piscataway, NJ: Bellcore.
Grudin, J. (1991) Interactive systems: Bridging the gaps between developers and users. IEEE Computer, 24, 59-69.
Nelson, T. (1991, July) The design of interactive software is a branch of movie-making. Presentation made via BayCHI, Palo Alto, CA.
Sacerdoti, E. D. (1991, April) Case studies of AI successes and failures. Presentation made at Pacific Bell, San Ramon, CA.
Change Modeling's Strengths
(handout disseminated at poster site)
Change modeling is an innovative approach to designing and enabling desirable change to large business (information) systems, their many computer-human interfaces, and personnel/client responsibilities (Anderson, 1992a). I developed change modeling to counter or avoid many of the "human problems" that typically arise during and following the implementation of computer-based information systems (Eason, 1987) as well as during their design and because organizations are increasingly deciding to use the power of computers to change business processes rather than to automate old ways of doing business (Gulden & Reck, 1991).
Much of the strength of change modeling comes from the fact that it doesn't divorce computer-human interface (re)design from business process (re)design or from employee process (re)design or from information system (re)design, nor does it divorce business process (re)design from employee process (re)design or from ... In short, change modeling attempts to embed the design of any aspect of a system within the context of the design of everything else that it will impact or that impacts it.
Additional strength comes from change modeling being a combination of:
a means of coordinating change across all relevant business processes
Organizational boundaries too often define the limits of design/development efforts. As a result, such efforts are often "piece meal" - angering or frustrating people in other organizations and ultimately frustrating everyone. Change modeling addresses needs and impacts and opportunities across all relevant business processes regardless of organizational boundaries.
a means of accomplishing change in accordance with organizational goals
Design/development efforts often plunge ahead without organizational goals that can guide design (Harker, Olphert, & Eason, 1990). Change modeling requires an understanding of an organization's desired nature to reveal advisable (and the advisability of) design/development efforts and characteristics.
a means of "seeing" the basis and implications of (the timing of) proposed system changes
Much should be attended to when designing system change and when evaluating change proposals. Because there is too much to "keep in mind," humans need "external" support, support that is too often inhibited by conventions of the paper-based world even when that support is provided by computer (Nelson, 1991). Change modeling helps design participants/evaluators "see the entire picture" by removing the constraints of standard paper size and by providing means of recording and "seeing" design rationale - an important product of the design process too often unrecorded or otherwise inaccessible (MacLean, Young, & Moran, 1989).
a means of changing systems in accordance with principles of reengineering
Employee effectiveness and fulfillment and successful system change are often undercut by antiquated approaches to management and by antiquated views of the proper employee-system relationship (Adair-Heeley, 1991). Change modeling facilitates transition to characteristics of highly effective jobs (e.g., Hammer, 1990) and to effective computer support of those jobs (e.g., Fischer, 1989).
a means of structuring change as evolution
System change is often attempted in big, unmanageable chunks - unmanageable by those responsible for system design/development and unmanageable by those who must make too great a change in how they go about their work in order to use the new or modified system. Change modeling requires identification of evolutionary stages through which change should progress and during which system change is incremental to enable change that is both welcome and lasting (Anderson, 1992b).
a means of initiating task analysis, user interface prototyping, user testing, ...
The needs of and impact on the people who (are to) use information systems continue to receive inadequate attention during system design in spite of heavy promotion of user-centered design techniques (e.g., Bellcore, 1991) and published evidence of the techniques' benefits (e.g., Karat, 1990). Change modeling utilizes such techniques and calls for their continued use throughout the system design process.
a means of involving and obtaining input from everyone (to be) impacted
When those (to be) impacted by system change are not adequately involved in system design, the design of and the support for change are often weak. Change modeling provides a needed vehicle for cooperative design (Kyng, 1991) and for considering and balancing the concerns, values, and perceptions of all stakeholders (Rouse, 1991).
a means of managing the entire change process
Keeping track of large business (information) system change is a big task. Change modeling provides a means of orchestrating system change via links of the big change picture to project plans and progress reports.
a means of innovating to counter problems as they are discovered
Change modeling is NOT a rigid set of techniques with a step-by-step manual. Such common but unnecessary rigidity and linearity stifle innovation (Grudin, 1991; Poltrock, 1989). Change modeling is an innovative approach that is itself open to innovation; it is a means of change design and change management that is ongoing and modifiable as influences on/within organizations fluctuate and as more is discovered about the systems to which it is applied.
References
Adair-Heeley, C. B. (1991) The Human Side of Just-In-Time. New York: American Management Association.
Anderson, R. I. (1992a) Modeling system evolution: A means of cutting through obstacles to desirable change to large business (information) systems, their many computer-human interfaces, and user/operator responsibilities. Interactive poster presented at the CHI '92 Conference, Monterey, CA.
Anderson, R. I. (1992b) Understanding and facilitating cultural change. Position paper for the CHI '92 "Cross-Cultural Perspectives on Human-Computer Interaction" Workshop, Monterey, CA.
Bellcore (1991, November) Proceedings of the Bellcore/BCC Symposium on User-Centered Design: Making It Happen. Piscataway, NJ: Bellcore.
Eason, K. D. (1987) Methods of planning the electronic workplace. Behaviour and Information Technology, 6, 229-238.
Fischer, G. (1989) Intelligent support systems. CHI '89 Tutorial, Austin, TX.
Grudin, J. (1991) Interactive systems: Bridging the gaps between developers and users. IEEE Computer, 24, 59-69.
Gulden, G. K. & Reck, R. H. (1991) Combining quality and reengineering for operational superiority. Perspectives on the Management of Information Technology, 8(1).
Hammer, M. (1990) Reengineering work: Don't automate, obliterate. Harvard Business Review, July-August, 104-112.
Harker, S. D. P., Olphert, C. W., & Eason, K. D. (1990) The development of tools to assist in organisational requirements definition for information technology systems. Proceedings of the IFIP TC 13 Third International Conference on Human-Computer Interaction, Cambridge, U.K., 295-300.
Karat, C. (1990) Cost-benefit analysis of iterative usability testing. Proceedings of the IFIP TC 13 Third International Conference on Human-Computer Interaction, Cambridge, U.K., 351-356.
Kyng, M. (1991) Designing for cooperation: Cooperating in design. Communications of the ACM, 34(12), 65-73.
MacLean, A., Young, R. M., & Moran, T. P. (1989) Design rationale: The argument behind the artifact. CHI '89 Conference Proceedings, 247-252.
Nelson, T. (1991, July) The design of interactive software is a branch of movie-making. BayCHI Meeting, Palo Alto, CA.
Poltrock, S. E. (1989) Innovation in user interface development: Obstacles and opportunities. CHI '89 Conference Proceedings, 191-195.
Rouse, W. B. (1991) President's forum: Human factors in the 1990s and beyond: Sociotechnical problems and challenges. Human Factors Society 35th Annual Meeting, San Francisco, CA.
