An Observational Study on the Maintainability Characteristics of the Procedural and Object-Oriented Programming Paradigms

Abstract

This paper presents an observational study that takes an initial step towards answering the question of how the selection of a programming paradigm affects the maintainability of a software system. Answers to this question are important because they can provide insight on how the maintainability characteristics of different programming paradigms could be leveraged to improve software-engineering education. The observational study compares developers' effectiveness and speed performing predetermined maintenance tasks on two equivalent variations of a web-based application: one based on Procedural Programming and one based on Object-Oriented Programming. Presented within this paper are the study's design, metrics measuring maintenance effectiveness and speed, and a statistical analysis of the results, which provides evidence that the maintainability characteristics of the two paradigms are different within the context of this research. In addition, as an observational study, some of its important contributions are lessons learned about the experiment design process and ideas for new hypotheses. These are presented in preparation for a future, broader research, and larger-scale experiment. Finally, we discuss some forwardthinking ideas about how differences in the maintenance characteristics of programming paradigms may influence on software-engineering education.