34

I looked into an article published rather widely in the media, but on a fairly marginal topic (perceptions of grammar use), so noone will die or ever bother to replicate or correct it.

The contents of the paper looked dodgy in analysis and setup, but more than anything else, the statistics look very suspicious and possibly erronous or at least probably mistaken in assumptions. I contacted the corresponding author for their data, and their editor for an overview of their data sharing policies. No answer from either for more than a month.

What can I do? Is there maybe an online web-environment, where I could flag potentially dodgy studies for further discussion? If the contents are as dodgy, what leverage is usually in process for a retraction to happen? Thanks!

The question is about this particular paper, but this kind of thing can come up a lot. If this was a paper in my own specific area of interest, I would maybe write a public answer, but as this is not, I'd just point it out to other interested parties.

Improve this question
7
  • 11
    Unless it directly impacts your own research, I would ignore it.
    – David Ketcheson
    Commented Jan 25, 2016 at 17:39
  • 4
    Practically, very much on target! Philosophically, it leaves rather grim outlooks to how seriously anyone should take stuff that's "published".
    – puslet88
    Commented Jan 25, 2016 at 18:46
  • 8
    @puslet88 "published" simply means that 2-3 people related to the field thought it is reasonable. The review process can sometimes help improve the work and filter out work with glaring holes. That's about it. As a scientist you should always be skeptical.
    – Bitwise
    Commented Jan 25, 2016 at 20:02
  • 1
    Some ideas: Can I get a refund if I bought an article that contains a major flaw? ; Why doesn't arXiv have a comment section?
    – Franck Dernoncourt
    Commented Jan 26, 2016 at 1:55
  • 3
    If it's wrong and you can demonstrate it with science then, by all means, do so and write your own paper.
    – J...
    Commented Jan 26, 2016 at 12:00

2 Answers 2

Reset to default
27

A good way of posting comments, anonymously or not, with a good chance that other people will read them and react is Pubpeer. Although not all fields are covered, you can post your criticism on many articles. Sometimes authors react, sometimes publishers take action, sometimes nothing happens.

what leverage is usually in process for a retraction to happen?

A good list of the cases where Pubpeer comments lead to retraction can be found on the Retraction Watch website:

http://retractionwatch.com/?s=pubpeer

Improve this answer
3
  • Pubpeer is a really great solution to exactly this problem. In the bright future something like pubpeer could be the first place we check just in case whenever we start up a paper we don't know the background of.
    – puslet88
    Commented Jan 25, 2016 at 18:44
  • 1
    @puslet88 I encourage you to write your concerns on this platform if you can spare the time. Also, if you find Pubpeer helpful, you could up-vote its community promotion add.
    – Cape Code
    Commented Jan 25, 2016 at 20:25
  • thanks for the response, I've upvoted the advertisement gladly, didn't post on pubpeer yet on this occasion, but it looks like an excellent site with right goals and good activity. Really great to see how it finally has an effect or when the authors manage to give their own response.
    – puslet88
    Commented Jan 26, 2016 at 18:53
7

IMHO, Ignore it and move on

With the outbreak of many journal publishers and reviewers of diverse levels of experience to review them, you would find many such papers being published today. As a researcher it would be up to you to determine which paper seems verifiable and of proposed results are reproducible.

Improve this answer
4
  • 5
    That would be the most practical advice indeed, and for research purposes, indeed the community should be able to separate the wheat from the chaff. At the same time there is an added complexity recently written about in e.g. andrewgelman.com/2016/01/25/… (has link to slate article) where the popular media consider all that is published as having a true basis in science (also in this case). In this sense, ignoring noticed mistakes as a default strategy, makes the outlook for the future perception of science as rather grim.
    – puslet88
    Commented Jan 25, 2016 at 18:34
  • It then seems that for science, academicians should work really hard to demonstrate to media how getting results in science is really difficult and that not all that is published is gold. Calling things "scientifically" based has a lot of power to it, that may cause mistaken beliefs or people altering their strategies towards poorer states. Bad social science floated in popular media for example can be picked up by people building policy decisions, and at that point it'll be too late (they won't have the time) to point that the basis could have methodological flaws.
    – puslet88
    Commented Jan 25, 2016 at 18:40
  • There may be some complexities here. For practical purposes of managing my own turf and responsibilities and this particular case, I'll probably do exactly that and move on. Thanks! :) Pubpeer offered by @cape-code however seems a perfect solution to, if you want to take a moment and point these things out.
    – puslet88
    Commented Jan 25, 2016 at 18:42
  • 4
    Disagree. This is only true if you indeed consider the "outbreak" of fake science, which actual academics automatically filter. I trust OP is not talking about a fake open access journal.
    – Cape Code
    Commented Jan 25, 2016 at 21:07

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .