Fall Small - Part II: Fail small

In my last post, I wrote about how, even though falling seems like the worst thing that can happen, there is a critical need to fall small in derby to keep yourself safe and to enable yourself to get back up and at it. In science as well as on skates, considering the worst-case scenario, failure, is something I do with every experiment I conduct and I do it before I have pipetted even a single sample. Even at the planning stages, I think about the data I might acquire and what critical control groups would be necessary to ensure that my conclusions are due to my experimental manipulations and not the result of random chance, equipment failure, or alternative explanations. I do pilot studies on small groups to make sure that readout machines are working and that reagents are reacting as expected. In every grant proposal I write, I always include a section ‘Potential Pitfalls and Alternative Approaches” devoted to considering the ways in which my studies might fail and what I am going to do about it if that happens. When properly done, preparing to do an experiment by considering the ways it can go wrong is often more effortful than doing the experiment itself.

Tips for newbie skaters on how to “Fall Small”

Let’s take an example. Let’s say you are a scientist and you have discovered a miracle medicine that you think will dramatically improve scores in mathematics. Let’s say we do an experiment in which you give this medicine to one group of people and give another group of people a sugar pill (a placebo) and measure their ability to complete multiplication tables. And then, just when you think everything is going great, just when all the exams are taken, you look at the data and surprise, the scores of the two groups are nearly identical. What happened? What if the medicine is actually able to improve math scores and something went wrong in the experiment that meant we couldn’t see it? There are about a hundred ways an experiment like this can go wrong so I will focus on just a few. One possibility is that you didn’t give enough medicine to have an effect. A way to address this is to do dose response study where you measure blood levels of various doses of the medicine and selected the dose that produces the highest levels. And what about the math test itself? What if the medicine did help those who took it get 100% but the test was so easy that even the participants who took a sugar pill placebo scored close to perfect, masking the benefit of the medicine? In that case, a pilot study of several test forms of differing complexities would be needed to find a test where placebo treated people scored pretty well (like 80%) but not perfectly so that you could detect an improvement in scores with the medicine. Doing any of these small pre-experiments would have identified some of the potential ways the project could go wrong when the stakes were low, in a sense scientifically falling small. Of course it is possible that the medicine doesn’t impact math ability and if so, then that experiment would not be a failure at all but rather a discovery of the truth. But, preparing for and preventing or at least detecting the worst-case scenario is the way we can be confident in our discoveries.

I have had my fair share of epic fails in the lab and spectacular falls while on track. At first, as a new graduate student and as a newbie skater, I fell a lot. Skates went flying, control groups failed, you name it, it happened to me. And it hurt a lot, both physically or metaphorically due to lost time, wasted money and bruised pride. So, I used to spend a lot of time and energy to avoid it. I was afraid of falling, I was afraid of failing, I was afraid of getting hurt. And that fear held me back as a scientist and a skater. I was afraid to try new things and hesitant to do new experiments to the point of paralysis. But as soon as I accepted that falling and failing was part of the game, I became much more confident and that gave me power to push forward. Well that and a pair of crash pads overnight shipped from Amazon to protect my hips and tailbone from the inevitable.

It took time, and some emotional work to overcome, but eventually, regardless of the tools I used to get there, soon I was beginning each jam and every experiment with this worst-case scenario in mind and that changed everything for me as a skater/scientist. So, don’t avoid falling, equip yourself with the tools to fall safely so that when it happens, and it will happen, you can get back up again and get right back to work.