Date of Award

1998

Degree Type

Thesis

Degree Name

Master of Science

School

School of Computer and Information Science

Faculty

Faculty of Communications, Health and Science

First Advisor

J.E. Terry

Abstract

As society and the world economy moves into the second millennium. the service industries involving knowledge workers will continue to increase. Software is the enabling technology that is driving the knowledge industry. As the development of software is mostly a design process, where new artefacts are conceived and built, the prediction of outcomes in the process is fraught with difficulties. Software project estimating is one of the essential Software Engineering techniques that will enable the rationalisation of decision-making regarding software development. Estimates that are more accurate will increase the probability of success and lower the risk. This thesis analyses the current software project estimating techniques available to practitioners and examines current practice in the estimating of software projects within the Western Australian industry. The principal techniques examined are Function Point Analysis and COCOMO and these are shown to be flawed in their construction. The practices adopted by expert and experienced practitioners are analysed and it is shown that the formal algorithmic models are not widely used. It is also shown that estimates are required in a project's lifecycle before the full requirements are known. The Western Australian practices are also compared to similar analyses conducted in other countries.

Share

 
COinS