18

from Wikipedia:

Software engineering (SE) is the application of a systematic, disciplined, quantifiable approach to the development, operation, and maintenance of software, and the study of these approaches.

I was wondering if there is something similar for research, something that we could call Research engineering. I imagine it to be a research field on its own, with students "researching on how to do research". I believe software development has benefited a lot from research in SE. Maybe research could also benefit from Research engineering.

The questions are:

  • is there some institute or some university department in the world where they work on Research Engineering?
  • in which faculty you would position such department/institute?

Edited: After getting a couple of good answers, I am still not completely satisfied, so I would like to clarify my question. What I am really interested in is indeed a "software engineering" approach. I am not interested in philosophical or sociological research. In fact, the question I had originally in mind was whether it's possible or not to apply actual software development methodologies to research. In more concrete terms, I am wondering whether anyone has studied the application in research of models similar to the waterfall, or the spiral model, or things like extreme programming, Scrum, etc... (Note: these are just examples, please don't comment to each of them one by one).

  • 1
    Interesting question, and interesting concept. – eykanal Apr 23 '12 at 12:10
  • 7
    The name "Research Engineering" biases things a little bit by suggesting an engineering approach. A more neutral name might be "Research Research", i.e., research about research. Your question would itself fall under Research Research Research. Research Research sounds like the academic equivalent of management consulting: outsiders come in with less domain knowledge, but they are supposedly smarter and with broader perspective, so they can offer valuable advice. I'm skeptical that it would be useful, and I don't know of anyone making progress on this, but I can't rule it out in theory. – Anonymous Mathematician Apr 23 '12 at 15:28
  • 1
    @AnonymousMathematician: I was indeed suggesting an engineering approach. "Research Research" sounds too meta. And so does "Research Research Research" :P – Alessandro Cosentino Apr 23 '12 at 15:39
  • 1
    Perhaps we need a Department of Departmental Name Engineering. On a more serious note, see ncbi.nlm.nih.gov/pmc/articles/PMC2539276 and theatlantic.com/magazine/archive/2010/11/…. They are specific to medical research, but they amount to studying how and why research does or doesn't work. – Anonymous Mathematician Apr 23 '12 at 16:00
  • 1
    Yo dawg I herd you like research... – JeffE Apr 24 '12 at 8:51
14

There is plenty of "research on research" (or "science of science").

There are dividid into different fields, e.g.:

  • Scientometrics - measuring citations, networks of collaborators, relations between topics and other quantities characterizing the scientific output.
  • Sociology of science - treating science as a activity of groups of people, with its history etc.
    For example there is a great book Ludwik Fleck, Genesis and Development of a Scientific Fact.
  • Research on collective intelligence, innovation (when, where and how does it happen), etc.
  • General fields related to education, didactics, teamwork and management.

As you asking about engineering (i.e. how things work in practice), I don't even mention things like philosophy of science. Also, as you see, the answer depends on scale - from an individual, through a group or an institution to a country, the world nowadays or our civilization.

Typically it is done under umbrella of complex systems, complexity, network science, econophysics or data-mining and modeling in sociology. There are institutes doing it, see e.g. the front page of the Santa Fe Institute. Also, there are some projects on it, e.g. QLectives.

ADDED:

As you are interested in the optimization (not only the observation) and on the micro scale: some findings may implicitly give hints, e.g.:

Other things may be harder to find, as in science it works a lot in apprentice-master mode, with approaches differing from a group to a group. So it may be not as easy to be serialized (as in different fields, countries, etc. one may need to have different approach); and when you don't a large enough sample, you cannot use quantitive methods in a meaningful way.

Moreover, now we are in the phase preceding formalized studies, as only recently people started to share with the world their soft and subjective findings on that matter, e.g.:

and on things like academia.SE, for a bit of self-reference.

  • 2
    Great answer. In the comments above, I was thinking about research aimed just at improving how research is done, but as you point out, it can be much broader than that. – Anonymous Mathematician Apr 23 '12 at 18:26
  • hi Piotr, thanks for the answer. Interesting, but not quite what I mean. I guess it's my fault, I should have been more detailed in explaining the question. I'll edit it soon. – Alessandro Cosentino Apr 24 '12 at 15:46
  • 1
    @AlessandroCosentino Before editing (it is good to clarify a question, but not to change it), what exactly do you have in your mind? If I change 'software' to research in you question, then the answer is on it. – Piotr Migdal Apr 24 '12 at 16:00
  • 1
    What about research on research on research? – Jase Jan 25 '13 at 1:27
  • @Jase only if the peer reviews there are peer reviews of peer reviews. Btw peer reviews of peer reviews actually exist. – dgraziotin May 31 '14 at 9:46
4

I've never heard of a field dedicated to the study of research methods. There are journals dedicated to advances in methodology (e.g., [1], [2]), but the closest concept I've encountered to a field dedicated to researching research is Thomas Kuhn's classic work The Structure of Scientific Revolutions and other similar philosophical works, which discuss how science, as a field, progresses and evolves.

I would guess that the reason for the lack of such a field is such research is part and parcel of the actual work done in the field. In order to study biochemistry, one must understand how to study biochemistry; in order to study mathematics, one must know the types of questions and the methods used to find answers in mathematics. Each field is unique, and each field will find specific methods that will optimally serve the needs of that specific field. While there may be broadly-applicable research techniques, each field will solve the problem of "how to do research" differently, in the way that best suits that field.

  • Hi eykanal, thanks for the answer and for the comment. Well, you could say the same in software engineering, you could say that methods applicable to, say, Web programming, are different from methods for desktop applications programming. Or that some patterns can only be used when programming in some languages, rather than others. I still believe that in research there are common patterns applicable among all the fields. If we don't know them yet, it might be just because no much research "on how to do research" has been done yet. What do you think? – Alessandro Cosentino Apr 23 '12 at 13:14
  • @Alessandro - You're correct that there are some "universal truths" in research, so to speak. However, those are fairly well understood, without the need for further research. The details, as I described, are left to the individual research fields. – eykanal Apr 23 '12 at 13:22
  • I would be happy to know those "universal truths" :), or even better, to see them somewhere written down. I must disagree with you, though, I think there is too much improvisation in the field and those few known methodology truths are revealed only at some introductory graduate skill seminars for first year grad students. – Alessandro Cosentino Apr 23 '12 at 13:39
  • Check out virtually any course on this list. Many disciplines have their own "Research Methods" course, and that's where they share both research fundamentals and field-specific techniques. – eykanal Apr 23 '12 at 13:44

protected by Community Feb 20 '17 at 15:38

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.