I was reading this Nature article by Kim Gardiner and Hugh Kearns on snack writing.  Their argument is that scientists who write in a small, fixed window, 45 minutes a day, every day, write more than “block writers” who blackout entire days for writing.

I’ve been a block writer my entire graduate career, that’s for sure.  But I can see the merit of snack writing, especially when it comes to methodology.  The worst feeling of writing a paper is digging through stacks of ancient data from experiments you don’t remember doing, trying to remember which of the six primer pairs you designed actually worked (let alone what annealing temperature they worked at.)  Writing up results as they are generated is a Godsend, especially when things don’t work out.

I have a tendency to bury things that don’t work: and why shouldn’t I?  Inconsistent, unclear, unexpected: we scientists are storytellers, and “bad” data that doesn’t fit the tale makes us feel uncomfortable.  We don’t know what it means- is our methodology flawed, or is there a fundamental misunderstanding in our thinking?  So instead, we are encouraged to focus on experiments that do work, and build our story from there.

But you never know when you will double back, and find yourself facing the ghosts of unpleasant experiments.  Even if you keep a meticulous notebook, data without the narrative can be confusing.  So spare yourself the trouble, and set aside a time when you generally aren’t very productive, and regularly wrap up what you’ve done and why you did it.