THE ABILITY TO ACCURATELY ESTIMATE SOFTWARE DEVELOPMENT COSTS AS EARLY AS POSSIBLE IS CLEARLY OF GREAT IMPORTANCE IN CHOOSING BETWEEN DIFFERENT SYSTEM ALTERNATIVES.

THE ABILITY TO ACCURATELY ESTIMATE SOFTWARE DEVELOPMENT COSTS AS EARLY AS POSSIBLE IS CLEARLY OF GREAT IMPORTANCE IN CHOOSING BETWEEN DIFFERENT SYSTEM ALTERNATIVES. UNFORTUNATELY, THE PRESENT STATE OF THE ART LEAVES MUCH TO BE DESIRED, FOR AT LEAST TWO BASIC REASONS: THE ABSENCE OF GOOD SIZE/COMPLEXITY ESTIMATORS WHICH CAN BE USED AT THE FUNCTIONAL SPECIFICATION STAGE; THE INABILITY OF CURRENT COST ESTIMATE MODELS TO ADEQUATELY CAPTURE THE EFFECTS OF DIVERSE SOFTWARE DEVELOPMENT ENVIRONMENTS. THE PROPOSED RESEARCH WILL: A) FORMULATE A MODEL FOR ESTIMATING BOTH THE SIZE AND THE COMPLEXITY OF THE SYSTEM TO BE DEVELOPED, TAKING AS ITS INPUT A FUNCTIONAL SPECIFICATION OF THE SYSTEM DEVELOPED USING A MODERN STRUCTURED ANALYSIS METHODOLOGY; B) ANALYZE THE EFFECTS ON SOFTWARE DEVELOPMENT COSTS OF LIKELY MAJOR CHANGES IN DEVELOPMENT ENVIRONMENTS, NAMELY THE USE OF ADA/APSE AND THE USE OF RAPID DEVELOPMENT METHODOLOGIES.