P. S. B. Harsh, D. J. W. Lloyd, Larry R. Borton
This paper explores the difference between models for decision making and those for other purposes such as designing production systems and modeling the growth process of plants. Reviewed are the evolution of models for decision making and the relationship of these models to the overall structure of a modern information system. This paper has three emphases: a) the importance of the decision maker in building models, b) problems related to modeling for decision purposes and c) the problems related to user acceptance. It also addresses problems of integrating decision models with databases and the need for support resources. The last section of this paper relates some of our experiences in building decision making models are conveyed and how these experiences have influenced our thoughts regarding future directions.

The title suggests that models designed and built for decision making have unique characteristics that differentiate them from other types of models. What are these characteristics? One of the major factors is that the models' main focus is decision making. This means the model was designed and built to address the specific problem of a decision maker. A concept that has withstood the test of time is a basic one proposed by Davis, who makes a distinction between data and information (see Figure 1). He indicates information is used to make decisions, whereas data are just raw facts, figures, objects and opinions. Data must be processed into information before they are used for decision making. The processing is often done with an analytical model and with the assistance of computer technology.



Since information is used to make decisions, it becomes the starting point in building models. For the information to be useful, one must know what type of information is desired by the decision maker and how it is used in the decision making process. As soon as these facts are discovered, then to build and design a processing component (e.g., model) to supply the desired information is possible, given the constraints of the available data. The alternative approach of trying to adapt a model built for other purposes (e.g., a research model) to supply information to decision makers is often plagued with problems. For example, information supplied is not in the desired format, the processing of the data is too complex, or the data demands are excessive.

Harsh, P. S. B., Lloyd, D. J. W. and Borton, Larry R. (1989). MODELS AS AN AID TO DECISION MAKING. Acta Hortic. 248, 27-48
DOI: 10.17660/ActaHortic.1989.248.2

Acta Horticulturae