Selecting an End User Programming Language for DSS Development

A bstract Today managers and policy makers are confronted with an overwhelming range of choices of computer software to develop decision support systems (DSS). The authors argue that DSS language evaluation and selection should be a multi-step process involving most, if not all, of the following: The objectives of each of these activities are described , as well as specific procedures for accomplishing them. In addition, the authors discuss the usefulness of a multidisciplinary task force to accomplish the DSS language evaluation and selection process. Decision support systems (DSS) represent relatively new way of thinking about managerial use of computers. A decision support system is a computer-based information system that is designed to help managers in private corporations and policymakers in public sector organizations solve problems in relatively unstructured decision-making environments. Long-range and strategic planning, merger and acquisition analysis, policy formulation, policy evaluation, new product development, marketing mix planning , research and development, and portfolio management are a few areas where the DSS concept has been successfully applied [see 6, 9, 14 and 19]. Unstructured decision-making environments are those where the global problem is not well enough understood for a complete analytical description. A DSS is a system which provides computational and analytical support in situations where it is necessary to integrate judgement, experience , and insight of managerial or policy decision makers along with computer-supported modeling and presentation facilities. DSS focuses on achieving productivity improvements from managers and policymakers, rather than from the reduction of clerical and administrative costs. As used in this article, a DSS language is one very important example of the generic set of computer application development tools generally known as end user programming languages (or in some cases, fourth generation languages). Other examples include relational database facilities with powerful report generation and ad hoc inquiry facilities; general purpose statistical data analysis languages; and broad based graphics generation languages. While each of these four types of languages usually address different user needs, they often share some subset of similar capabilities and characteristics such as:-Integrated database management (sometimes relational)-User friendliness to nontechnicians-Both procedural and nonprocedural command structures-Interactive on-line utilization-Support of prototyping and adaptive development-Modest training requirements for end users-Easy debugging and intelligent default assumptions MIS Quarterly~December 1984 267 Selecting an End User Language Quantity of code required only a fraction of Cobol, Fortran, etc. Internal documentation generation support Understandable code for non-developers [5]. The methodology presented here …