Negative results

I recently completed a study including several different analyses that met none of my expectations, the results of which were frankly boring. When anticipated outcomes are not met one has obtained “negative results.” For students who are learning how to do research the experience of negative results can be horribly disappointing. The main reason for disappointment is that the researcher does not get to draw the conclusions that they had hoped to make; that is, he/she “does not get to say what they wanted to say.” The truth is researchers put a premium on “having something to say” rather than on “saying it isn’t so.” But that perspective, if applied too broadly, can be misguided.

When I was working on my PhD in anthropology I applied for and received a grant from the National Science Foundation to fund part of my research. I was researching animal remains from a rockshelter site in Missouri that was occupied for much of the last ten thousand years. The money I received was to radiocarbon date bones from the cave to provide a more detailed chronology so that I could more rigorously study change in human diet and in the environment. I received funding for 50 dates, and I eagerly sent off the first batch of 25. It took about a year to get the results back from the independent radiocarbon dating lab, and they were horrifying. Here’s why. Archaeologists rely on the idea that materials near the bottom of deposits tend to be older than those at the top; that being so, a relative ordering of excavated materials can be used to discuss change. If one purchases radiocarbon dates, that’s even better because more precise dating of change can be achieved. I was excited at these prospects for my research.

It turned out that rodent burrowing and other processes had terribly mixed the cave deposit from which I was analyzing animal bones. Not being the original excavator, and given that most of the field notes were missing, I assumed that at least some ordering from top to bottom would exist. Was I wrong! Materials from the bottom of the deposit dated to a couple of hundred years ago, and some of the remains from the top were thousands of years old. The site and the collection of animal remains lacked the integrity needed to accomplish the research I had planned! I was a few years into the dissertation process, and I was demoralized.

Upon receiving the dating results, I sulkily wandered into the office of my major professor. I showed him the results, thinking that he would be very disappointed. To my surprise, he was not; he simply said “people undervalue negative results.” He said, “Don’t worry about it; do the work and report what you find.” What my mentor knew that I did not was that I was learning to do research. One does not get to choose the outcome of research ahead of time; however, a researcher can choose how she/he goes about the process. In the end, I reported what I found, and even though the results were not spectacular, I managed to produce meaningful research, much of which is published. More importantly, I obtained the degree I was seeking.

Today, as a professor who mentors student researchers, I am quite wary of self-proclaimed brilliant students who promise blockbuster results. Instead, I am most impressed with those students who pay attention to learning the research process. The most important products of their time in graduate school will be the skills that they acquire through learning how to manage a large-scale research project. If the results and implications of research were known ahead of time, there would be no need to do research. To use a baseball analogy, it is better to hit a home run while learning to hit singles than it is to swing for the fences at every pitch. To use a different metaphor, the old adage “sometimes you’ve got to kiss a lot of frogs…” applies to research.