Bad Testing Metrics—and What To Do About Them
Many organizations use software testing metrics extensively to determine the status of their projects and whether or not their products are ready to ship. Unfortunately most, if not all, of the metrics in use are so flawed that they are not only useless but possibly dangerous—misleading decision makers, inadvertently encouraging unwanted behavior, or providing overly simplistic summaries out of context. Paul Holland reviews Goodhart’s Law and its applicability to software testing metrics. Paul identifies four characteristics that will enable you to recognize the bad metrics in your organization. Despite showing how the majority of metrics used today are bad, all is not lost as Paul shares the collection of information he has developed that is more effective. Learn how to create status reports that provide details sought after by upper management—and avoid the problems that bad metrics cause.