Test design is a science, not an art…

My brother was obsessed with cars when he was a kid. He built a huge collection of model cars of various colors, shapes and categories over many years of his childhood. At some point of time, he even started designing newer models and thought of selling them to the toy companies so that they can actually make those cars!

I must admit, he did an awesome job as a designer when he was a third grader!

However, when we both look back now, we realize that there was one fundamental problem with those designs. They were merely sketches, not designs. They were more arts than science. They were about how he sees the cars, instead of how the cars work. There were absolutely no measurements about anything that were specified in the design.

Based on my experience of interacting with various software testing teams, I feel the same way about the test cases creation process. The creation and review of test cases are still very subjective and thus qualitative. They are more art than science. No one cares to quantitatively measure the quality of test cases, as part of the review process, before they are actually executed. The only time the quality of test cases are assessed, unfortunately, is when there is a bug found in production and management has doubt about how much was the test coverage when the application was tested in lower environments.

Very recently I co-founded Testing Algorithms, LLC. with a friend to help organizations with an automated test case design algorithm that makes the test case creation a scientific, logical, quantitative and automated process, as opposed to artistic, subjective, qualitative and manual one.