When discussing our recent experiments, I often hear fellow grad students say, “it didn’t work.” But, in fact, the experiment did work; it just wasn’t the needed or expected result. In other words, it was a negative result.
Why do we immediately correlate negative results with failure? It seems to be almost an instinct for grad students. Is it because of the pressure to publish, a process that is notorious for not accepting negative results? Or is it because we are incapable of accepting that our ideas were wrong? Is it that we feel we have wasted time only to find no significance? Additionally, we feel the pressure of our committees, which are unwilling to let us leave without results that make a large impact, results that are most certainly not negative.
It’s important that we, as a scientific community, begin to retrain ourselves and learn to love our negative data for several reasons:
Negative data drive honest science
In 2013, the Economist reported that the publication of negative results has been gradually decreasing over the years, something that shouldn’t surprise you if you’ve ever published a paper. Why is this? It’s simple, reviewers aren’t excited about our insignificant results, they want to see those asterisks. The same article pointed out that negative results are much more reliable. Taken together, this becomes the perfect storm. Imagine: you’re trying to publish a paper, but aren’t getting the results you need to wrap up the story. Maybe you have to adjust several things and keep retrying your experiment until you get the result you want. You blame your negative results on protocol error or unfavorable conditions once you get what you need. Maybe it’s true occasionally, but most often it isn’t, and in reality, you may be falsifying data. If we learn to embrace our negative data, we may be able to avoid this corruptness in our methods.
Negative data are still informative
Many of us work long nights to explore the unexplored, or so we think. We also keep a trove of negative data in our desks that no one else knows about. Do you see the connection here? Wouldn’t it be nice to know if someone else had the same crazy idea, but nothing came out of it?
A vast amount of time and resources are being wasted because we are afraid to publish our negative results, sharing with the world what wasn’t the answer to our questions. If we were more willing to do so, we could save each other a lot of time and grant money.
Embracing your negative data is therapeutic
As mentioned above, we tend to immediately assume our negative data was a failing on our part. We must have taken too much time at a crucial step or forgotten to add something to a buffer. But it’s reasonable to believe that our negative data are reproducible and we know what we’re doing at the bench. In fact, we should be proud of all our data, whether or not it supports our hypothesis, because it was reproducible, and that is something that takes a lot of training and discipline to be able to do. If we can overcome this instant reaction of negative data = I screwed up, then we may feel a little better about ourselves.
Negative data force us to think
As grad students, we tend to be jealous of those who seem never to experience negative data and are able to publish quickly. While that is lucrative, it may not be the best thing for us. During our PhD education we are supposed to learn, to think, and to troubleshoot. If our hypotheses are always supported by our results, then we may never experience having to change directions and completely rethink a project. Struggling with negative data makes us stronger scientists.
In short, negative data are good for us and good for science. Science is all about communicating our results, but it shouldn’t be only about communicating results that support a hypothesis. My plea isn’t that we publish every piece of data we produce, but that we think more about incorporating our negative data so that others can benefit from our work.
What to do with your negative data
Start by discussing your data with your advisor. This may be a difficult conversation, but attempt to stand by your methods if you know you did the experiments well. It’s unlikely that you will be publishing these results by themselves. Discuss the possibility of adding them as supplemental data. Remind your professor that these data points took time and grant money to accumulate and the work should be considered in the publication process. In case you’re wondering, PLOS The Missing Pieces and BMC Research Notes both willingly accept negative results. In addition, online publishers such as F1000 Research and PeerJ are places to publish your negative results standalone, if you choose to do so.
The views and opinions expressed in this blog are the views of the author(s) and do not represent the official policy or position of ASCB.