Customer Reviews

Software Measurement and Estimation: A Practical Approach (Quantitative Software Engineering Series)

5 star:		(3)
4 star:		(2)
3 star:		(0)
2 star:		(0)
1 star:		(1)

 

 

It is amazing how much information could be packed into a book of only 250 pages (approximately). This is probably the best book one can currently buy for a good introduction to software measurement and estimation. It is both a guide for students as well as a reference for software measurement practitioners. The book contains small imperfections that can be fixed and by no means reduce its value. They will certainly be corrected in a future edition. Comments below:

- Chapter 1 / Section 1.4 about the CMMI mentions the Project Planning and Project Monitoring and Control process areas (PAs) but fails to mention the Measurement and Analysis process area - that has everything to do with the book's theme.

- Chapter 2 provides a short (and good) description of the GQM (Goal/Question/Metric) approach but fails to mention the PSM (Practical Software & Systems Measurement) methodology, even though the PSM book is listed as reference [3] at the end of the chapter. Also, Section 2.3 is titled "Method 3: Standards Driven Metrics" but does not mention the ISO/IEC 15939 standard (Software Engineering - Software Measurement Process), which is related to the CMMI Measurement and Analysis PA and also to the Practical Software & Systems Measurement initiative. Any book on Software Measurement should cover the ISO/IEC 15939 standard.

- Chapter 3 contains a very good explanation of measurement scales, measures of central tendency, measures of variability, measurement errors, accuracy, and precision.

- Chapter 4 is about software sizing. Even though the coverage is very good, the chapter fails to inform the reader that 3 of the FSM methods - IFPUG, NESMA (Netherlands Software Metrics Users Association, not even mentioned), and COSMIC have become ISO standards. It should also be said that IFPUG considers the Value Adjustment Factor optional since 2003. The main IFPUG Function Point reference, the Counting Practices Manual, is not listed as a reference at the end of the chapter.

- Chapter 5 features a good introduction to complexity measures, including McCabe's cyclomatic complexity, Halstead's measure, and the Agresti-Card-Glass system complexity metric. Object-oriented design metrics are also covered in this chapter. The author explains those measures very clearly including examples.

- Chapter 6 is about effort estimation. Again the author provides excellent and comprehensive explanations of several estimation models, such as expert opinion (including Delphi), benchmark data, analogy, proxy points, custom models, and algorithmic models. On page 96 the author directs the reader to reference [10] for the IFPUG website. As it turns out, reference [10] on page 116 points to the website of a consulting company. It would be nice to have this fixed in the next edition of the book. On page 104 COCOMO II is mentioned but the example on page 105 uses COCOMO 81. The reader should be aware that COCOMO 81 has been replaced by COCOMO II. On page 106 some estimation tools are listed, namely COCOMO, SPR, and Checkpoint. COCOMO itself is not a tool so it would have been better to use the USC COCOMO II free tool as an example, as well as commercial tools such as COSTAR and Cost Xpert. SPR is listed as a tool when it is actually a company that used to sell the Checkpoint tool that has been replaced by KnowledgePlan.

- Chapter 7 provides the reader with a very good introductory treatment of defects. Figure 7.6 on page 122 depicts a size-defect relationship that has been "linearized" through the use of a log-log scale. That should be clearly explained (the text refers to the relationship as linear). Overall this is an excellent chapter.

- Chapter 8, 9, and 10 are, respectively, about reliability, response time and availability, and measuring progress. They are all very good and I can only recommend them. Excellent explanations and references.

- Chapters 11 and 12 - Outsourcing and Financial Measures - seem a little displaced in this book, even though they can still be useful for students.

- Chapter 13 on Benchmarking is perhaps too short to give the theme a fair treatment.

- Chapter 14 on Presenting Metrics to Management is very interesting and potentially useful to measurement analysts. The author teaches the reader how to create a dashboard for all metrics to be viewed as a whole.

What I like about this book is that it deals with the most important topics of software measurement and estimation without being bulky or tedious. The language is simple and keeps the reader interested throughout the 250 pages. Recommended.

I hope this review helps potential readers as well as the author of this very good book.

I bought this book after going through the Table of Contents. After reading some chapters, I am convinced that it is a good buy. The book is written in a simple language and is well structured with good examples thrown in. I found the chapters 'Measuring Size' and 'Estimating Effort' especially well presented. The References listed at the end of the chapters is extremely useful and allows one to explore certain topics in greater detail.

One drawback is that the problems listed at the end of the chapters do not have solution key and the FTP site is not available despite the claim on the front cover. The good news is that one of the authors (Linda) emailed me to state that creating the FTP site was her next project !!!

Overall, an excellent book on a yet to be fully understood software engineering topic.

I've been working on a web application project. I needed to analyze and estimate the reliability of my project. Using the techniques presented in this book (especially chapter 7 - Defects and Defect Metrics), I was able to predict the reliability of my project. Starting at the beginning of the project, I used my history data from other projects for calculations. As I progressed to the test phase (I used incremental model), I was able to estimate the reliability based on the defects collected during the testing phase. I highly recommend this book especially if you are software engineering major.

It is a good book which provides enough theoretical foundation based on which correct metrics / measurements can be defined and used.

This is the probably the first book to contain most of the information I needed know about Software Measurement and Estimation in simple and easy to understand language. It contains excersises to practice what you have learned. Overall an Excellent book. Well written Linda. Thanks

There are better chances for this book to sell, if Amazon dropped the kindle price to 9.99. Why would I pay 65 USD for an electronic copy?