How I read research papers without losing the day

I used to read papers linearly. Abstract, intro, methods, results, discussion. It took me four hours to read one paper, and I would forget half of it by the end of the week.

The fix was to read the figure first. Every paper I care about has one figure that is the paper. The figure is what the authors think is the most important thing they did. Read that figure, read the caption, then go back and read the abstract. By the time you reach the methods you already know what you are looking for.

The second fix was to stop reading every paper. I read the abstract, I read the figure, and I make a decision. If the paper is going to change something I do, I read the whole thing. If it is just interesting, I file it. Most papers are interesting. Most papers are not going to change what I do. Reading them linearly is a waste of a day I am not getting back.

The rule I am still learning: read on paper when it matters. Screens are for triage. When I am reading for keeps — the one paper I will read this month and think about for the next six — I print it. The cost of the print is cheaper than the cost of the distraction.