-2

Basically, in this IEEE paper, the writer states an equation that he has developed and consequently plots a graph to validate such an equation. The equation is a model of a wireless system.

At first, I didn't give it much thought, but then I notice something wrong. I started plotting some of his equations and they don't match what he plots. It seems as if he has used the plots to trick people, cause there are many inaccuracies in the paper.

My question is, how can it have been published? Should I contact the author?

3
  • 11
    And how do you know they are lying instead of just sloppy? – Captain Emacs Mar 22 at 11:44
  • 6
    What steps have you taken to check your analysis of the plots and equations to ensure that it is not you who is making a mistake? – user2705196 Mar 22 at 12:00
  • 1
    Well, I always tell all my students, do not believe everything authors said in their papers. I've read many papers that are just rubbish with unrealistic assumptions or parameter values that ensure their system works. In most cases, I would say that it is due to incompetent authors. Many authors are students, and hence, it is quite likely they the theory wrong. Also, do not expect their supervisors to check their work carefully. – Prof. Santa Claus Mar 23 at 5:55
12

In my experience, when there is a situation like this, in which a low level student thinks they have found a fatal flaw in the analysis of a peer-reviewed technical paper, 95+% of the time the student has just made a mistake in their own analysis. If the student instantly believes the author of the paper is lying, without making inquiries about the substance of the analysis, I would bump that probability up to 98+%. If the student makes a clear spelling/grammatical error in their description of the issue, bump that probability up a bit more.

Now, perhaps you're one of the small number of cases where the student is correct and the paper is actually fatally flawed. Perhaps this case is even in the extraordinarily rare class of cases where there is a deliberate misrepresentation in the paper by the author, as opposed to an unintentional error. I'm going to play the odds and suggest that you should proceed with caution, making absolutely sure you are correct before you make any contact with the author or the journal. Start from a position where you do not assume malice, and show your work to an academic at your school to see if they understand the analysis any differently to you. If you find that some other skilled person agrees with you then you can write to the author (politely) to inquire into the difference between your own results and the material in the paper.

It is possible to get to a point where it is reasonable to infer an erroneous result in the paper due to deliberate malice by the author. However, that is something that should happen at the end of a long line of sober inquiry and consideration of all possibilities.

11

"Lying" is a very, very strong accusation. Maybe there are some details that you missed? Maybe the description just isn't complete? There are many possibilities here, and you cannot prematurely assume malicious intent. Of course you can contact the author, but you should state how you tried to reproduce the result and politely ask for clarification instead of jumping to accusations. If it turns out you're wrong, you will otherwise look very bad.

3

Contacting the author seems like a polite thing to do first. It may give them a chance to clarify that perhaps there's an implementation difference, or there are some steps that were omitted from the paper write-up for clarity, etc.

1
  • 3
    Well maybe --- it is not the polite thing to do if he contacts him to immediately accuse him of lying. – Ben Mar 22 at 13:23
0

Use PubPeer, find the paper (e.g. by using its DOI), and add your claims there.

PubPeer is specifically designed for such kinds of 'post-publication reviews'.

To be on the safe side, you could keep your comments confined to factual observations, without inferring motives (e.g. "lying").

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.