Empirical software engineering research produces lots of results. But what do all these results mean? Do we know?

Theories are important because they can help us understand and predict empirical observations. But where are the theories that we need for our next experiments? Who creates them? Many SE researchers implicitly believe they can’t “do” theory, and I posit this common belief lacks basis and does not withstand scrutiny.

In this talk, I will describe how we used empirical research as a springboard to theory-building, explain how theory positively impacted our empirical research in software testing, and share some surprising concrete ideas for how you, as an empirical SE researcher, might also add theoretical methods to your own research repertoire.

(Throughout, I will refer to our TestLoop theoretical process model as well as our TerzoN and NaNofuzz composite software testing tools as concrete examples that illustrate and provide context for this journey.)