Checklists in Research – You are not too smart to use them!
In my review of the Checklist Manifesto, I promised to write a post about some checklist applications for scientists and researchers. At first I thought I had exhausted any content with the post about Adam Savage’s tips. However, there are important distinctions between those tips and a good checklist. In the book, Gawande (the author) describes the key elements of the checklist: it must be explicit, clear, and most importantly simple. The Adam Savage steps were explicit, they were hopefully clear, but they were anything but simple – each of them involved multiple questions and many deep lines of thought. Not checklist material. How, then, can we adapt the ideas of a simple checklist to research?
The answer is simple: we apply it to the day-to-day rather than the big picture. Many researchers struggle when trying to organize their research around their life, and have trouble keeping track of all their data, experiments, and ideas. What would the basics of this day-to-day checklist look like, then? Here are a few clear, explicit, simple ideas:
- Were any samples that I started today named properly and catalogued both in my notebook and electronically?
- If I performed any processing steps on my samples today, were all of those steps double-cataloged with ALL of the important variables?
- Have I stored the data that I collected today in a reliable location?
- Is the next processing step that I need to perform ready to go?
- Have I made the reservations and ordered the supplies I need to do the next step?
Quite possibly the most frustrating experience in research (tied perhaps with breaking a sample) is realizing after the fact that my samples were not properly catalogued, and thus I couldn’t match specific pieces of data to the sample. This kind of checklist serves as a reminder – have I kept track of myself? It is easy for us to get ahead of ourselves in research – often our work involves a very large number of serial steps, and if we do not adequately keep track of what we are doing, we could find ourselves in a world of pain (or more accurately, a world of waiting for days and days and days, adviser breathing down our neck..Accckkkkk!)
Ideally, at the end of the day this checklist should be a simple exercise – yes, yes, yes, yes, yes… However, the times where it is not are the times where you will be glad you had it.
We can also create equipment checklists. Many standard operating procedures for research equipment have over 20 steps listed, all verbose with many considerations, cautions, and tips. This can be helpful as a reference, but for the day to day it can be burdensome. What is stressed in the Manifesto are the “kill steps” of your task. Kill steps, in our situation, are the parts of the SOP that, if not followed correctly, will cause your sample to be ruined, the equipment to be ruined, or your physical/mental health to be damaged or destroyed. (Oh dear, I seem to be breathing chlorine gas. / I have opened a portal into a new dimension and have seen what cannot be unseen.) We don’t want any of those things.
For example’s sake, let’s consider a piece of equipment in my laboratory: a tube furnace. There’s not much complicated about this furnace; it’s basically a quartz tube that you stick in a heating chamber. However, when working with certain gases we have to use a “gettering furnace.” This furnace purifies the gas stream of any oxygen present by using hot titanium sponge. One of the most dangerous situations would be if someone flowed oxygen through the gettering furnace – can you guess why? It involves burning and explosions. That is a perfect example of a kill step: “Check that you are not flowing oxygen through the gettering furnace.” A checklist SOP would take all the essentials of the reference SOP, simplify everything to the kill steps, and leave the operator to use his or her judgment on using the reference SOP.
You may wonder why it is worth it to split off these kill steps from the rest of the information. “The more info, the better, right? People need to know all the steps.” It is true that each step is important, but it is doubly true that some steps are more important than others. If we distill to the essentials, people will be more likely to actually follow the checklist without glossing over it. I know because it has happened to me – long SOPs are boring, complicated, and easy to ignore. “I know this,” I say, then later realize that I ruined my sample. Even though I remembered the minutia, I forgot that one important kill step, and now I have to pay for it. Don’t let that happen to you.
Simplify, and stick to the list. You are not too smart for that.