Knowledge that Transforms

Data dictionary systems traditionally have been analyzed and compared against idealized standards that are independent of both the special requirements of the organization and the environment in which they will be used. This has resulted in the selection of data dictionary systems that appear to be the best product in terms of the features offered, but which may not be the most effective alternative for meeting the needs of the organization or supporting its data administration effort. This article presents a new framework for analyzing data dictionary systems that encompasses environmental dependencies and organizational requirements. Also described is a structured methodology for selecting a data dictionary system based on this framework. This methodology is then applied to a bank holding company example.

Planning is a basic organizational decision-making activity often requiring the judgement and expertise of a group of organization members. Computer and communications support for the planning process can help individuals and groups who may have different perspectives and priorities, to communicate and coordinate their activities. This article examines how groups organize their planning and decision-making activities in a number of application areas, and identifies their requirements in terms of communication structures and online information systems capabilities. Examples are given from engineering projects, computer projects, scientific communities, company planning, and crisis management. The applicability of a distributed decision-making approach to organizing information systems support is assessed as a way of meeting these requirements.

The trend toward distribution of information resources throughout an organization requires an increased awareness of, and sensitivity to the organization’s structure, strategy and constraints. However, organizational requirements may be in conflict with technical concerns involving distributed systems. This article presents a mechanism for analyzing the degree of consistency regarding MIS activities across three levels of management: corporate, senior MIS, and operational MIS. The procedures are developed in the context of data management in distributed environments. The reconciliation mechanism presented can be used to provide information of either a qualitative or quantitative nature. This information helps to ensure that computer-based systems are consistent with organizational requirements.

The field of data administration originated as an adjunct to the expanding use of database management systems. With the increasing amount of data that data processing personnel are faced with managing, and the realization that data is an important corporate resource, the concept of data administration has continued to evolve and change. At several points in time over the last few years field surveys have been conducted to determine the state of practice of the data administration function. Using data derived from those surveys (and emphasizing the two which were conducted by this author) as well as empirical evidence, this article traces the development of the concept of data administration, concentrating on the period from 1981 to 1985.

An empirical experiment was conducted in order to compare a third generation, procedural language (COBOL) with a fourth generation, nonprocedural language (Focus) in terms of programmer productivity and program efficiency. Six applications were developed in two languages by different programmers who had been matched according to their level of expertise. The results of the experiment showed that the applications were programmed significantly faster, particularly by beginning programmers, using the nonprocedural language. On the other hand, it was found that the procedural language was significantly faster than the nonprocedural language in terms of CPU execution time.

A prototyping development method is presented here which has the potential to reduce the growing application development backlog. Prior research and our findings indicate that a prototyping process can assist in the efficient development of application systems by breaking a complex problem into several comprehensive parts. A state-transition model of the IS development process is presented and discussed. A two-prototype method is explained in the context of this model. Two projects are described which are typical of development efforts made by end users in a microcomputer environment and IS specialists in a mainframe environment.

As the business environment becomes more complex and dynamic, it becomes increasingly vital for top executives to scan the information environment to identify strategic threats and opportunities. This article seeks to understand the strategic scanning behaviors of top executives in order to provide some guidelines for the possibility of designing computer-based systems to support and enhance these scanning processes, The characteristics of the strategic scanning process are investigated through an empirical study of 37 chief executive officers of small to medium-sized high technology companies. Results show that they are very systematic scanners when it comes to strategic information, and that their information sources are limited, mostly personal, and external to the organization. This suggests that a computer-based system for strategic scanning would probably have to be a customized personal system, not tightly coupled to the organizational information system.