Kaizenlog

Lecture 2
Software Metrics
Book: Software Engineering by Roger E.Brasman
Why Software Metrics?
1 Effective management of any process requires quantification, measurement and modelling.
2 Software metrics do provide a quantitative basis for the development and validation of models of the software development process.
3 Metrics can be used to improve productivity and quality.
4 Results with metrics improve the more the metric is used and fine tuned.

Software Crisis Problems
1 Inability to produce correct and reliable software, within given time frames and a given budget.
2 Software development for large projects is inherently a complex process.
3 There is no proper analytical description for software development.
4 Use of software metrics will help to improve the development process.

Case of Software Metrics
1 Many metrics have been created.
2 Some have not been properly tested.
3 It is difficult to identify which properties of software will be measured
2 Even though many theoretical metrics exist, few are really acceptable in

practice. 3 When using metrics, it is important to understand any underlying assumptions and the environment of application.
Prerequisites for using software metrics
1 One should have an idea of basic statistics and experimental design.
2 One should understand the common software development life cycle model.

3. Ideally one should have practical experience as a project team member. Why? Because one has to know which methods are available and how to select the best metric. Possible measures of correlation, best fit etc are all possible.
Definition of a Software Metric
“ A software metric deals with the measurement of the software product and the process by which it is developed “.
The software product needs to be seen as an abstract object that evolves from an initial requirement of the need to finish a software system. The software product includes the source and object code, and documentation produced during the development.
Need for Software Metrics
1 Software development is a long and tedious process
2 Accurate costing and project scheduling are needed
3 Software management needs tangible measures
4 Metrics used properly will help prevent software crisis.

Qualities of good metrics
1 Simple, precise and definable
2 Objective to the most possible extent
3 Easily obtainable at a cheap cost
4 Valid: i.e. the metric should measure exactly it’s intended purpose
5 Robust: i.e. insignificant changes in the product should not affect it!!
6 Have data values that belong to appropriate measurement scales e.g. months, weeks, effort in man days etc.

Classification of Software Metrics
1 Product Metrics: These are measures of the software product at any stage if it’s development, from requirements to installed system. Product metrics measure the complexity of the software design, final program size (source/object code), number of documentation pages produced.
E.g. COCOMO, LOC, KLOC, defects/KLOC etc.
2 Process Metrics: These are measures of the software development process e.g. development time, average level of programming experience.
Further Classification
3 Primitive metrics
4 Computer metrics
5 Subjective
6 Objective

E.g. Primitive subjective product metrics
E.g. of compound metric defects/KLOC

COCOMO
Basic COCOMO Intermediate Advanced
Function Points Raleigh Model

Welcome to Kaizenlog.com If you're new here, you may want to subscribe to my RSS feed , Twitter You can contact us by using the contact form or submitting a comment. You can also share this post with your friends by clicking on the 'ShareThis' button above. Thanks for visiting!

 Print This Post

This entry was posted on Tuesday, October 31st, 2006 at and is filed under Computer.You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.