6

Consider a scenario where one (say Mr. X) has implemented a method for solving a problem, experimented it on a dataset D and obtained significant results. The obtained results seem to be better than a set of results published in a paper P in a reputable journal J which solves the same problem experimented on D by some another author.

Mr. X decides to cite P and directly compare the claimed results published in it with his own.

The question is how much necessary (in the perspective of a reviewer) is it for Mr. X to reproduce the results published in P by himself if he is to publish in

  • a) the same publisher as journal J?
  • b) a totally different journal publisher?

Most importantly, what would be the case if J is an IEEE journal?

8

In theory, if a result is published you can use it as is.

In practice, it is better to replicate it yourself. That way you can really make certain you are comparing like with like. Research often involves a very long chain of decisions that cannot all be published in an article. Good research documents that somewhere else, but it is easy for honest mistakes or omissions to seep in. So that way it could easily happen that Mr. X's design is subtly different from the design used in paper P, and the difference in result are due to that difference in design rather than the difference in method. You can rule that out by replicating the results in paper P yourself.

  • And (to make this relevant for the last three questions), while in practice there could be one, in theory there should not be a difference based on the fact if the journal the person cited and the person citing published in match. – skymningen Feb 8 '17 at 12:41
  • Thanks for answering. I understand the practical implications and the need to replicate. I was concerned about how will it affect the publishability per se if the results were not self-replicated (in an IEEE journal). – Ébe Isaac Feb 9 '17 at 3:47
  • In general, good research is more publishable. In your particular case, if you are improving on a given method, then chances are that someone involved in the original method will be a reviewer, and know in detail what they did, and find what you did is different and thus uncomparable. – Maarten Buis Feb 9 '17 at 9:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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