Some six or so years ago, not long after I had started working as a programmer, I developed a bad case of repetitive strain injury (RSI) and for a while couldn’t use my left hand for typing, but had to rely on my right hand and dictation. This situation being obviously untenable, I took action: I purchased an ergonomic keyboard (Microsoft Sculpt), remapped shift to the space bar when pressed in combination with another key, started doing hand/wrist stretches and switched from typing on the QWERTY layout to Colemak. This grab bag of interventions worked, my RSI got better and, though it still flares up sometimes, I now have it under control.
I have stuck with Colemak since then and have liked it and occasionally even recommended it to friends and colleagues. But beyond reading popular articles about the topic, I never really looked into the evidence on key layouts and ergonomics and productivity – until now, that is!
Before we get to the good stuff, a word of caution. I am in no way an expert on this; I estimate I spent roughly 15-25 hours researching and taking notes, so there is still a lot that I’m uncertain about. My target group here is touch-typing office workers, typing mainly in English, without disabilities that significantly impede their typing (e.g. the loss of an arm or that sort of thing). For people who don’t touch-type, QWERTY is probably fine. For people who have disabilities that affect their typing, I assume special considerations, of which I know little, apply.