Journal of FUBAR and Negative Results

Last evening I attended a panel discussion entitled, “Making the Web work for Science” hosted by Science Commons. It was held at the Commonwealth Club in San Francisco and moderated by Tim O’Reilly. On the panel were Jimmy Wales, founder of Wikipedia; Stephen Friend, MD, PhD President, CEO and a Co-Founder of Sage; and John Wilbanks, VP of Science at Creative Commons.

While I was hoping more would be discussed on modeling the habits of researchers with web tools, the focus on Open Science was still a good conversation. At one point, Dr. Friend mentioned the need to publish negative results. With the ability to inexpensively self-publish and distribute data on the Web, why then, aren’t we seeing more of this?

Trying to answer from my own experience as a researcher, there are at least three reasons, or rather fears:

  1. Negative results make either you look bad or your model look bad.
  2. The competition will learn what not to do, iterate improvements faster than you, and could end up scooping you.
  3. Negative results do not help you gain tenure or funding. So, why bother with the time involved in the write-up?

For the greater good of science, these points are superficial. As Dr. Friend pointed out, people need to read about negative results, so that the same wheel is not reinvented over and over again. While traditional journals do not have the capacity or resources to include negative results for peer-review and publication, the Web does.

While I was a doctoral candidate I faced this situation. My PI did not want to try to publish some results I had gotten from a project on mouse hematopoietic stem cells and gene therapy. The results were not even entirely negative, but were mild improvements over the existing science. In a thesis committee meeting, I was shocked to hear the response of one of my advisors, a Nobel Laureate, “Publish it. We need that information to build upon it.”

Sadly, those results never did make it into a proper journal, only my dissertation, yet because of the Web, that chapter is now freely available to everyone. I published it on my Mendeley profile. Download the awesomeness ;).

What about those three fears listed above though?

  1. Negative results do not make you look bad. The key, of course, is that not all of your research is negative, but if you are actively self-archiving your negative results, that will get you recognized in a positive light. You will be considered a contributor to science.
  2. Getting scooped. For one, I am not sure that has been proven (citation anyone?). And in fact, there is evidence to suggest that being open with your data will garner more citations. And in any event, if you are the first to publish the negative results that lead to an eventual positive, the credit and recognition, our incentives as researchers, will be given. The Internet and proper citation of previous (negative or positive) research demands it.
  3. The Web is built such that long write-ups are not necessary. Publish the raw data and the materials and methods. How long does that really take? Then let the community come in to draw conclusions in a “living report.” The Wikipedia style. Small chunks of data are better than none, and perhaps better if it comes out faster. And getting back to fear #1, if you are considered the authority in publishing data on the Web in your field, that is going to get recognized by funding and tenure committees.

The Web is a wide-open frontier for the taking. There are very few experts in any field archiving their negative (or positive) data on the Web for others to annotate. Now would be the time to stake a claim on your piece of the frontier. While some are using Mendeley to do that, others are using PLoS ONE. Whichever outlet you choose, the world needs those FUBAR results.

To learn how Mendeley can help you publish and organize research, go here.

Follow Jason Hoyt, PhD on twitter here.