A journal for negative results!

A new journal has just been created that publishes papers on negative results (for example, when an experimenter predicts that two variables are correlated and then finds out that they’re not).  This journal, in my mind, will fill a crucial gap.

westernblotThere is a phenomenon in science known as the “filing drawer effect” which refers to the fact that papers that fail to find a correlation where one was expected or where  a popular hypothesis isn’t supported do not get published.  Often, these papers aren’t published because statistical tests reveal that the sample size may not have been large enough to make any strong claims.  In other cases, statistical tests reveal that the sample size was indeed large enough and that the hypothesis really was not supported under that set of conditions.  In these cases, journals often choose to not publish the findings perhaps because these findings are less exciting than positive results. 

It’s really unfortunate that negative results don’t get published more frequently as this can slow the pace of scientific progress.  Knowing that a hypothesis wasn’t supported under a particular set of conditions may allow us to hone our understanding of a phenomenon.  Additionally, some hypotheses don’t get discarded as quickly as they ought to because evidence against them hasn’t been made public.  Finally, I hope that having this kind of journal around will encourage scientists to take chances and allow them to explore more risky but potentially very fruitful hypotheses as they won’t have to worry about being unable to publish their results after the project is completed.  

I’m pleased to see that negative results will now have a home in The Journal of Negative Results in Biomedicine. I hope that other fields will take a tip from the biomedical field and will create similar journals themselves. 


5 thoughts on “A journal for negative results!

  1. That’s actually really excellent.
    I predict it will be negatively read. But if I fail to support that hypothesis, I know where to publish it!

  2. I really hope a Psychology one follows, would certainly stop a lot of undergrads conducting experiments that were doomed to fail from the beginning! Good find.

  3. See, I’m not totally sure this as a good thing. I suppose it depends on the publication standards of your field generally. I know that there are plenty of people trying to get away with publishing null results in behavioral ecology journals–we all so want our results to mean something!

    To me, this just encourages interpretation of null results, something which enters very quickly into speculative territory. I’m all for taking risks, but I’d rather see people do that because we’re funding it more when it merits, not because they know that even if they’re they can over-interpret their results and guess about what might be going on instead of trying to solve problems in another real, falsifiable way.

    And maybe I’m wrong and this journal will maintain high standards for the rigor of statistical tests and their interpretation by its authors, and its field will wind up with a useful compendium of things that have already been tried and failed. But to me, this still won’t solve the underlying problem: that everything must produce some publishable result, and thus that even null results must be squeezed to provide some illusion of new and helpful information.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s