Skip to main content

Software Productivity Evaluation Model

Objective

The main objective of the SPEM project was to build an evaluation model of software development productivity for use across the spectrum of the software development industry.
The main goals of the project were to:
-contribute to the measurement, comparison management and improvement of software development productivity in an industrial context
-contribute to the awareness and understanding of software productivity models and modelling, throughout the software development industry
-construct, formalise, and offer as a standard, data definitions and data collection procedures proved as usable on software projects
-develop, use and validate a prototype of a new model of software development productivity
-cooperate with other projects in the metrics and models area, aiming at creating a mutually beneficial synergy with them.
The main objective of the project was to build an evaluation model of software development productivity for use across the spectrum of the software developement industry. The main goals of the project were: to contribute to the measurement, comparison management and improvement of software development productivity in an industrial context; to contribute to the awareness and understanding of software productivity models and modelling, throughout the software development industry; to construct, formalize, and offer as a standard, data definitions and data collection procedures proved as usable on software projects; to develop, use and validate a prototype of a new model of software development productivity; and to cooperate with other projects in the metrics and models area, aiming to create a mutually beneficial synergy with them. Development cost models were reviewed and classified, salient cost and quality determinates identified, and the measurement techniques and data definitions for the drivers reviewed and collated. A preliminary phase, to test data definitions, collection procedures, etc, was specified and performed, and the results studied. A second phase, incorporating the lessons of the first, was defined. The second phase included new date definitions, collection procedures, and new data verification and analysis techniques. Data was collected, verified and analysed from 30 industrial projects. Cost and quality determinants in the collected data were analysed for their importance and a prototype productivity evaluation model built.
Development cost models were reviewed and classified, salient cost and quality determinants identified, and the measurement techniques and data definitions for the drivers reviewed and collated.
A preliminary phase, to test data definitions, collection procedures, etc., was specified and performed, and the results studied.
A second phase, incorporating the lessons of the first, was defined. The second phase included new date definitions, collection procedures, and new data verification and analysis techniques.
Data was collected, verified and analysed from 30 industrial projects.
Cost and quality determinants in the collected data were analysed for their importance and a prototype productivity evaluation model built.
Exploitation
The model should allow:
-the measurement of software development productivity across a spectrum of software application areas
-comparison of the software development factors of any specific project with a general database of such factors, to support more accurate planning and prediction on further projects
-identification of detrimental development factors in a project, and indicating the need of intervention to achieve favourable development factors, supported by evidence from a database of existing projects.
The data collection should allow:
-data definitions and a collection manual to be offered as a standard to international bodies, offered as a standard for other metrics projects, and exploited commercially as a data collection tool for the software development industry
-the database to be offered as a starting point to other metrics projects, and exploited commercially as a database of "anonymous" projects from the software development industry.
At the time of writing, there have already been approaches from other projects with respect to the data definitions, collection manual and database.

Coordinator

FUIGI ITALIANA
Address
Via Benedetto Croce 19
00142 Roma
Italy

Participants (5)

CERCI
France
Address
56 Rue Roger Salengro
94126 Fontenay-sous-bois
O DATI ESPANOLA S.L.
Spain
Address
Paeso De Gracia 2,4
08007 Barcelona
SOFEMASA
Spain
Address
C/princesa 1
28008 Madrid
United Kingdom Atomic Energy Authority (UKAEA)
United Kingdom
Address
Culcheth Laboratory Wighsaw Lane
WA3 4NE Warrington
VERILOG SA
France