Programming Languages as a discipline has traditionally been approached using the research methods of mathematics: formal definition and proof. There are some who challenge us to approach it as a science, through experiment and measurement (e.g. user studies).

What would it look like to apply *design* research methods to PL? (I do not have a Design background, but I have become familiar with some of these methods, and I'm excited about the idea.) What do you think is most appropriate? A combination of these? Something else entirely?

What counts as a "rigorous evaluation" is often contentious, coming up e.g. when reviewing papers and deciding on the curriculum for the PhD program. The expectation is usually that your advisor teaches you the research methods they consider appropriate...

...but that tends to mean that researchers are never exposed to other methods in their training and so, by default, regard them all as un-rigorous or inappropriate to the discipline.

Academics of any stripe, I invite you to share in the replies: what are the research methods or evaluation methods you use? What is their disciplinary origin? What makes them rigorous?