Software inspections are commonly used to help identify and eliminate defects early in the product development life cycle. This chapter looks at past publications on software inspections to reveal patterns that can result in more effective inspections. The main questions are initially formulated, and the data from the various experiments are used to arrive at the answers. The book chapter reveals insightful results on the effectiveness of an inspection, like the impact of team size, types of documents, and different reading techniques.

Many product development companies gather data from inspections, but using this data to optimize development processes is not straightforward. This chapter provides an approach that software development organizations can use to collate and categorize information across a wide range of data collected from different projects. Another significant contribution is that the chapter attempts to look at results over three different levels: the organization, project, and individual. The process described in this chapter can be effectively used to derive results that would help make the development processes more effective at all levels in an organization.

Apart from the conclusions drawn in the chapter, the description of the process of categorizations, classifications, benchmarking, and analysis is excellent. The authors provide counter viewpoints about why the methodologies may not be considered conclusive by some. This leads to a very straightforward narrative style that is easy to read and understand. I would highly recommend this chapter to organizations that want to create knowledge-based software inspection data.