Customer Reviews

Software Testing Fundamentals: Methods and Metrics

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

 

 

Traditional software testing focuses two key metrics - defect removal efficiency and, in mature environments, defect density. This book takes a different approach that, on first read, may seem like anarchy to experienced test professionals who work on internal projects. However, when you stop and consider the context in which Ms. Hutcheson is using, the methods and metrics she sets forth make perfect sense.

The context is delivery of commercial products or getting company-critical applications rolled out for purposes of competitive advantage. Within this context she defines the following goals: (1) first to market with the product, (2) optimum pricing model, (3) products with the right features, (4) keeping unacceptable bugs to an absolute minimum. For the last she has a corollary, "Make sure your bugs are less expensive and less irritating than your competitor's", which is not the 'party line' in testing, especially in organizations that strive for zero defects, but is realistic, and especially so when time-to-market is critical. In some ways this book reflects software testing in marketing-driven product development. This is further reinforced by the following definitions and objectives:
- definition of quality is customer satisfaction.
- system for achieving quality is constant refinement.
- measure of quality is the profit.
- target goal of the quality process is a hit every time.

Achieving the above within the software testing domain does require a departure from conventional thinking and methods, and how to go about achieving them is thoroughly covered in the body of the book. The key approach is to develop a test strategy that is based on 'Most Important Tests'. Supporting activities covered include building a test inventory, managing risk, and a feedback loop of data analysis.

In addition to being aligned to product line development, this book's approach can also be easily tailored to rapid, iterative development approaches such as agile methods. If you are working in an internal development environment that uses 'heavier' development lifecycles this book is not going to fit; however, if you work in a product-oriented environment this book will not only change your thinking, but will provide the basis for an integrated development-marketing approach that could make a real difference in competitive advantage.

This text describes MITs (Most Important Tests), a risk based test methodology. MITs makes use of prioritized test cases, which collectively are referred to as a test inventory. The book is organized in 3 sections. Chapters 1 through 5 focus on background concepts. Chapters 6 through 8 focus on the test inventory and how to create it. Chapters 9 through 14 discuss risk analysis, test techniques , and test planning and estimation.

Although the author describes how her methods fit with either traditional or agile software development, I feel it is geared towards tradtional software development with its heavy investment in upfront planning. Agreed upon test inventories she asserts are important to communicate both the value and cost of testing, as well as to establish a contract of what will and what won't be tested for the project.

Although the subtitle of the text is "methods and metrics", there was just one chapter devoted to test metrics.

Overall, I appreciated the "how - to" orientation of the text; this is a book for the test practioner. There are several examples which can be used for sample test templates and checklists. There is also a companion web site, a glossary of terms, and comprehensive end notes.

Recommended book for software test engineers. Since the author also discusses political aspects of testing in today's real business world, where the CFO rather than the CEO rules, this is an interesting book for other test engineers, too.

One of the main topics of the book are what the author calls MITs, the "most important tests". The MITs are determined using various methods, including path analysis, boundary value analysis, expert interviews, and test ranking.

While there are nice explanations of the other methods, my first impression was that the author's explanation of her ranking method was lacking technical details. Actually, that ranking method is rather simple. Use of consistent terminology and some detracting errors in examples just make it more difficult than necessary for the reader to understand.

Good test for those wanting to learn the basics on testing software. As someone who has been thrust into software testing, this book has served me well in getting my feet off the ground.

I am a graduate of Computer Science and found a job in software testing. I bought this book based on the Recommendations on Amazon. However, after reading a few chapters, I am disappointed:

The topic of the book is about Software Testing. However, many times the author goes off the topic, discussing something like:

"A software bug has properties much like a real insect: height, length, weight, type or class( family, genus, spider, beetle, ant, etc ... ), color, and so on. It also has attributes like poisonous or non poisonous, flying or nonflying, vegetarian or carnivorous ..." (page 109)

I wondered: Am I reading a Biology book?

In many other places, the author introduced the content of other books. Several times, he quoted the definitions from the Webster dictionary. Of course, references or quoting is good, but the author has overused them.

In order to save time, I have to skip many pages of the book when the author goes off the topic. Overall, the book has around 400 pages, but in my opinion, the author can reduce the book's size to one third and the book can still be able to carry the same information.

My suggestion for the new edition: keep the content as precise as possible by going straight into the details. It will save the time of the readers.