One of my PhD objectives involves an evaluation of the suitability of an emerging technology for deployment in geographical area (where it has not yet been deployed before -- mainly because it's new). I have been asked "Why evaluate this tech? Why not the others?" I have found this tech to be significantly superior to the other alternatives in several parameters of importance (in the lab), which is probably already a good justification itself to proceed. My question though is: The other technologies have been around for years and been fairly evaluated already. This tech is new and untested in this area. Is the fact that it is new a justification in itself for me to bias myself to it? Do I need empirical results to justify? Isn't it enough that the results, good or bad, are a significant contribution since there was no knowledge before?
-
I guess the answer depends in part on the domain, but if it's a scientific domain, then I'd be dubious of your plan which doesn't seem to contribute to scientific knowledge. Sure, demonstrating the utility of a technique to some ends is an interesting methodological contribution, but at least in the fields with which I am familiar, that is insufficient– TeuszMay 10, 2017 at 7:13
-
Looks to me like, in itself, this is a task more suited to the private sector– TeuszMay 10, 2017 at 7:14
-
2A new technique can be groundbreaking. Or lost time and effort. The problem with this is, if there are enough well-established ways to do the same thing then you at least need to point out some flaws of these technologies that yours might not have. Otherwise you are re-inventing the wheel, maybe even a very complicated to use wheel. You need at least one valid answer to "But why would we need this instead of all of the other methods?"– skymningenMay 10, 2017 at 8:19
-
1Yes, also a testable hypothesis!– TeuszMay 10, 2017 at 10:10
-
Thank you very much for your insightful input. I will proceed with providing empirical justification, since it exists and is strong anyway.– DrMaxBMay 16, 2017 at 11:08
2 Answers
Usually, in life there are multiple options and methods to do things. Some are more common, others are rather uncommon. Why are the uncommon ones uncommon? Sometimes they are just more inconvenient than the common ones for most people, so the uncommon option is not "catered for" and just falls behind, although it is perfectly fine. Other times, they have been tried and discarded a long time ago. In very rare cases, nobody really thought about it, although it is great.
It is similar in science.
A new technique can be groundbreaking. Or lost time and effort. The problem with this is, if there are enough well-established ways to do the same thing then you at least need to point out some flaws of these technologies that yours might not have (to make it similarly convenient). Otherwise, you are re-inventing the wheel, maybe even a very complicated to use wheel (that might have been discarded previously, as failures are unfortunately still rarely published you wouldn't necessarily know). You need at least one valid answer to "But why would we need this instead of all of the other methods?"
Yes, you should have a solid scientific argument if you choose to only evaluate a single method. If I read a paper where only a single method is discussed, my first question is "but how does it compare to other methods?" If I can't find any reference in the article that gives information on that, I won't consider the paper to be of much quality. The fact that it is new is not enough justification for not comparing it to established methods. (Obviously this doesn't include a technical paper describing the method, but even in those papers you almost always find at least a simulation study or so comparing it to other methods.)
Which brings me to my second point: in clinical research it is even obligatory to compare a new medicine to the readily available alternatives. Because without controlled comparison in an experimental setting, you have no way of drawing solid conclusions about the relative performance of the different techniques.
And honestly, "is it really better or am I fooling myself here?" is probably the most interesting question to be asked about any new method. So why not looking at it?