Publishing in the New Millennium: A Forum on Publishing in the Biosciences: topic
This is a public forum
Negative or Null Results: Will Researchers Share?
Hilary Spencer
Wednesday, 14 November 2007 03:13 UTC
A young woman asked a question to the first panel regarding negative or null results in research. I think that Emilie Marcus (editor of Cell) responded by noting that it is very difficult to assess the importance of papers covering NoN-results (Negative or Null results), which is very true. However, as many people have noted, sharing this type of information may prove to be very useful to other researchers. (See Thomas Goetz’s commentary on “dark data” in Wired.)
We’ve been encouraging people to submit papers, posters, and presentations describing NoN-results to Nature Precedings. A recent blog post suggests that at least one researcher found a manuscript on Precedings to be helpful in this regard. However, as Goetz notes, sharing NoN-results “makes many scientists deeply uncomfortable, because it calls for them to reveal their ‘failures’”.
So are scientists likely to share these types of results? We hope that sharing them on a site like Precedings allows authors to turn non-publishable research into at least “citable” research, but is this enough motivation?
-
Replies
Jump to resultsResults
-
This content has been removed by the forum moderators.
-
There have been quite a few discussions and articles in the literature, including in Nature, about this topic. I remember a good article in Bio IT World about a year ago. There is now even a “journal of negative results”. We also have Nature Protocols, in which people can describe protocols that gave negative results as well as positive ones. I think that the Internet is assisting in sharing this kind of information usefully in many ways—and of course projects like open lab notebook/wet ware can also only help, so long as people will share the information. I am sorry I don’t have time to look up all the URLs.
By the way, Sergio is incorrect, I think, to define these results as “errors”. Rather, they are results that failed to show a predicted positive outcome. That doesn’t mean they are wrong, far from it. -
Yes I fully agree with your views Hilary. Knowing about negatuve results does not mean a failure rather this way other researchers will not repeat the same mistakes . After all we learn from our mistakes and even if we consider it as a “failure” , we are inching a step closer to “success”!
-
I agree with what clarke says. I dont consider negative results as ‘errors’. Sometime if your hypothesis is not found to be true but still it is important for others to know this so that they dont spend huge time and money in testing it.
-
I thinkt the problem is lack on incentives. Why would they publish NoN-results? Publishing anything is a difficult process, yet here probability of being cited is very small and there is no financial compensation.
I think the following are ways in which NoN-results might be pushed to be publish:
- Reduce Cost: automated collection of data.
- Money: tie NoN-results to grants and funding.
- Citation: create publishing into verticals so that information is organized in a way that makes it easy to seek out needed information this increasing the probability of citation.
Then, as with most things, it turns into an issue of marketing. There is another possibility – if publishing information brings value back to the scientist, that might also have an effect? Meaning, if I publish my data and someone provides analysis that I don’t have time for? Or something that might make my job easier?
My 2 cents :\
-
All very interesting.
My own personal views are best summarized in a letter I had published almost a year ago to this day in PLoS Clinical Trials. This can be found here
“Search for the truth is the noblest occupation of man; its publication is a duty”.
Madame de Stael (1766 – 1817)
Graham Steel (no relation BTW !!)
-
Although it is true that not all negative results are errors, there are plenty of errors (or weakly supported observations) in ongoing research. If we provide others all of the raw data that we used to reach our conclusions, it will make it much easier to stay close to the truth.
But I think most scientists will not be willing to spend time formatting and submitting “negative results”. The easiest thing to do is simply keep the laboratory notebook online in public – Open Notebook Science.
-
“Open Notebook Science”
Here here I say !!
I think I was coming from quite a specific perspective (clinical trials/data reporting) as that’s an area that I have a lot of first hand experience. I stand by those comments.
—
From a general STM research perspective, I very much applaud your comments Jean-Claude !!
-
Yes, indeed, Jean-Claude. I meant “Open Notebook Science” when I wrote “open lab notebook” in my previous response, sorry for getting the name slightly wrong.
-
I agree with several commentators that the biggest challenge is actually to encourage authors to reach the critical mass of momentum required to actually organize the negative results for public consumption.
I don’t see a lack of citations for negative results as being a problem unless other researchers cannot find the data. That is why it is important to use forums such as Nature Precedings or perhaps PLoS ONE to ensure that these results are made public. They will get cited if they are available.
Results
-