Measuring functionality and productivity in Web-based applications: a case study

Abstract

We explore the variation of the cost of writing code when object oriented framework based development of Web applications is encountered for the first time. Managers need such information to justify their investments in innovative development strategies. Size measurements are essential in this task and a number of metrics, namely: lines of code, classical function points, and object oriented function points, are employed. It is argued that lines of code and object oriented function points are more suitable in this case. Data analysis reveals that learning influences mainly the cost of writing new code, consisting of continuous calls to components provided by the framework. We also explore the applicability of an already proposed effort prediction model that is based on different reuse types. A cost estimation model is the by-product of this study, providing a helpful tool for managing the first projects in which the framework is employed.