My concern with most research is that it often attempts to generalize/control for "type of programming language" with "a specific implementation of that type of

My concern with most research is that it often attempts to generalize/control for "type of programming language" with "a specific implementation of that type of programming language". At least in my experience, the implementation matters. And then the number of papers that generalize results about "an object oriented language" when they mean "Java" is sickening. See also research on productivity w/ "static typing" vs. "dynamic typing". One of the worst things about learning to program at school in the oughts was that Java was the language of choice. I have no problems (more than any other language) with Java as a professional tool, but Java as a first language is a disaster with its conceptual overload.