I work in a biomedical engineering lab. With the method I’m establishing, there are hundreds of little steps, repeated 15 times over the course of weeks. For many of these steps, there are no dire consequences for screwing them up. For others, some or all of your work could be ruined if you don’t do them right. There’s nothing intrinsic about the critical steps that scream “PAY ATTENTION RIGHT NOW.”
If your chance of doing any step right is X%, then for some X, you are virtually guaranteed to fail. If in a day, there are 30 critical steps, then you need about a 98% reliability rate on any individual step to achieve a 50% overall chance of success on any particular day. An 80% reliability rate gives you a chance of success virtually indistinguishable from zero.
Normally, habits are what we rely on to make us reliable. When establishing a new method, however, you’re constantly fighting against many of your habits. As you validate, troubleshoot, tweak and optimize, you’re changing bits and pieces of a semi-familiar routine, and it’s a constant struggle to do the modifications right on the first try without letting old habits override them, and without letting the modifications break the old habits. I find more and more that figuring out ways to structure my work to enforce reliability on myself has big payoffs.
Some ways I do this:
I set up my workstation better so that I can design experiments on the computer and print them out for the bench more easily.
I have a rule to always double check my math and do no math in my head
I turn the steps of my method into detailed checklists.
I have a dedicated cardboard box with handles for carrying stuff to other rooms where I do various tests, and I have a label on the cardboard box for what items to bring for specific tests.
I label the reagents according to the order I’ll be using them, if possible. If not, I have a set of bins at my bench. I set the reagents for step 1 in bin 1, the reagents for step 2 in bin 2, etc. That way, I have a physical representation of the task flow that’s set up before me, so that no mental effort need be put into remembering “what next.”
The general rules underpinning these approaches seem to be:
Split apart the steps of thinking and doing as much as possible
Use physical mechanisms you set up during the “thinking stage” to enforce compliance during the “doing” stage
Reliability
I work in a biomedical engineering lab. With the method I’m establishing, there are hundreds of little steps, repeated 15 times over the course of weeks. For many of these steps, there are no dire consequences for screwing them up. For others, some or all of your work could be ruined if you don’t do them right. There’s nothing intrinsic about the critical steps that scream “PAY ATTENTION RIGHT NOW.”
If your chance of doing any step right is X%, then for some X, you are virtually guaranteed to fail. If in a day, there are 30 critical steps, then you need about a 98% reliability rate on any individual step to achieve a 50% overall chance of success on any particular day. An 80% reliability rate gives you a chance of success virtually indistinguishable from zero.
Normally, habits are what we rely on to make us reliable. When establishing a new method, however, you’re constantly fighting against many of your habits. As you validate, troubleshoot, tweak and optimize, you’re changing bits and pieces of a semi-familiar routine, and it’s a constant struggle to do the modifications right on the first try without letting old habits override them, and without letting the modifications break the old habits. I find more and more that figuring out ways to structure my work to enforce reliability on myself has big payoffs.
Some ways I do this:
I set up my workstation better so that I can design experiments on the computer and print them out for the bench more easily.
I have a rule to always double check my math and do no math in my head
I turn the steps of my method into detailed checklists.
I have a dedicated cardboard box with handles for carrying stuff to other rooms where I do various tests, and I have a label on the cardboard box for what items to bring for specific tests.
I label the reagents according to the order I’ll be using them, if possible. If not, I have a set of bins at my bench. I set the reagents for step 1 in bin 1, the reagents for step 2 in bin 2, etc. That way, I have a physical representation of the task flow that’s set up before me, so that no mental effort need be put into remembering “what next.”
The general rules underpinning these approaches seem to be:
Split apart the steps of thinking and doing as much as possible
Use physical mechanisms you set up during the “thinking stage” to enforce compliance during the “doing” stage